Search

Design Parallel-group randomised controlled trial. Methods Participants Adolescents aged 11–17 years, diagnosed with myalgic encephalomyelitis/chronic fatigue syndrome and with no local specialist treatment centre, were referred to a specialist service in South West England. Interventions Fatigue In Teenagers on the interNET in the National Health Service is a web-based myalgic encephalomyelitis/chronic fatigue syndrome-focused cognitive–behavioural therapy programme for adolescents, supported by individualised written, asynchronous electronic consultations with a clinical psychologist/cognitive–behavioural therapy practitioner. The comparator was videocall-delivered activity management with a myalgic encephalomyelitis/chronic fatigue syndrome clinician. Both treatments were intended to last 6 months. Objectives Estimate the effectiveness of Fatigue In Teenagers on the interNET in the National Health Service compared to Activity Management for paediatric myalgic encephalomyelitis/chronic fatigue syndrome. Estimate the effectiveness of Fatigue In Teenagers on the interNET in the National Health Service compared to Activity Management for those with mild/moderate comorbid mood disorders. From a National Health Service perspective, estimate the cost-effectiveness of Fatigue In Teenagers on the interNET in the National Health Service compared to Activity Management over a 12-month horizon. Primary Outcome 36-item Short Form Health Survey Physical Function subscale at 6 months post randomisation. Randomisation Web-based, using minimisation with a random component to balance allocated groups by age and gender. Blinding While the investigators were blinded to group assignment, this was not possible for participants, parents/carers and therapists. Results The treatment of 314 adolescents was randomly allocated, 155 to Fatigue In Teenagers on the interNET in the National Health Service. Mean age was 14 years old and 63% were female. Primary outcome At 6 months, participants allocated to Fatigue In Teenagers on the interNET in the National Health Service were more likely to have improved physical function (mean 60.5, standard deviation 29.5, n = 127) compared to Activity Management (mean 50.3, standard deviation 26.5, n = 138). The mean difference was 8.2 (95% confidence interval 2.7 to 13.6, p = 0.003). The result was similar for participants meeting the National Institute for Health and Care Excellence 2021 diagnostic criteria. Secondary outcomes Fatigue In Teenagers on the interNET in the National Health Service participants attended, on average, half a day more school per week at 6 months than those allocated Activity Management, and this difference was maintained at 12 months. There was no strong evidence that comorbid mood disorder impacted upon the relative effectiveness of the two interventions. Similar improvement was seen in the two groups for pain and the Clinical Global Impression scale, with a mixed picture for fatigue. Both groups continued to improve, and no clear difference in physical function remained at 12 months [difference in means 4.4 (95% confidence interval −1.7 to 10.5)]. One or more of the pre-defined measures of a worsening condition in participants during treatment, combining therapist and patient reports, were met by 39 (25%) participants in the Fatigue In Teenagers on the interNET in the National Health Service group and 42 (26%) participants in the Activity Management group. A small gain was observed for the Fatigue In Teenagers on the interNET in the National Health Service group compared to Activity Management in quality-adjusted life-years (0.002, 95% confidence interval −0.041 to 0.045). From an National Health Service perspective, the costs were £1047.51 greater in the Fatigue In Teenagers on the interNET in the National Health Service group (95% confidence interval £624.61 to £1470.41). At a base cost-effectiveness threshold of £20,000 per quality-adjusted life-year, the incremental cost-effectiveness ratio was £457,721 with incremental net benefit of −£1001 (95% confidence interval −£2041 to £38). Conclusion At 6 months post randomisation, compared with Activity Management, Fatigue In Teenagers on the interNET in the National Health Service improved physical function and school attendance. The additional cost of Fatigue In Teenagers on the interNET in the National Health Service and limited sustained impact mean it is unlikely to be cost-effective. Trial registration This trial is registered as ISRCTN18020851. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 14/192/109) and is published in full in Health Technology Assessment; Vol. 28, No. 70. See the NIHR Funding and Awards website for further award information. Plain language summary Why did we do the study? The best evidence for the treatment of adolescents with myalgic encephalomyelitis/chronic fatigue syndrome is cognitive–behavioural therapy for fatigue delivered in person. In the United Kingdom, most adolescents with myalgic encephalomyelitis/chronic fatigue syndrome cannot get this specialist treatment where they live. Fatigue In Teenagers on the interNET in the National Health Service is an online treatment using cognitive–behavioural therapy designed for myalgic encephalomyelitis/chronic fatigue syndrome, which has been shown to work in the Netherlands. To find out if Fatigue In Teenagers on the interNET in the National Health Service would be beneficial in the United Kingdom, we compared Fatigue In Teenagers on the interNET in the National Health Service to Activity Management. Activity Management is the treatment most often offered to children and young people with myalgic encephalomyelitis/chronic fatigue syndrome in the United Kingdom, and aims to avoid peaks in activity (sometimes called ‘pacing’). What was the question? Does Fatigue In Teenagers on the interNET in the National Health Service lead to greater improvements in children and young people with myalgic encephalomyelitis/chronic fatigue syndrome when compared to Activity Management, when both interventions are delivered remotely? What did we do? We compared Fatigue In Teenagers on the interNET in the National Health Service and Activity Management in two comparable groups of children, and measured physical function at 6 months as the main indication of improvement. We measured how much the treatments cost and we asked children and young people, their parents and treatment providers what they thought about the two interventions. What did we find? At 6 months, adolescents saw greater improvements in physical function, and attended half a day more school per week, with Fatigue In Teenagers on the interNET in the National Health Service compared to Activity Management. Both interventions were associated with improvements over 12 months, with there being no clear difference between them after that time. However, the Fatigue In Teenagers on the interNET in the National Health Service treatment was more expensive. What does this mean? We have shown that cognitive–behavioural therapy for fatigue can be provided online to children as Fatigue In Teenagers on the interNET in the National Health Service, leading to faster improvement in physical function and greater school attendance compared to Activity Management. However, Fatigue In Teenagers on the interNET in the National Health Service is expensive and is unlikely to be good value for money. Scientific summary Objectives Primary objective To investigate whether cognitive–behavioural therapy (CBT) specifically designed for myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) and delivered over the internet [Fatigue In Teenagers on the interNET in the National Health Service (FITNET-NHS)] is effective and cost-effective compared to Activity Management for children with ME/CFS who do not have access to a local specialist ME/CFS service. Secondary objectives (1) Estimate the effectiveness of FITNET-NHS compared to Activity Management for those with mild/moderate comorbid mood disorders (anxiety/depression). (2) Estimate the cost-effectiveness of FITNET-NHS compared to Activity Management. (3) Estimate the cost-effectiveness of FITNET-NHS compared to Activity Management for those with mild/moderate comorbid mood disorders (anxiety/depression). Methods Trial design Randomised controlled trial (RCT) comparing FITNET-NHS with Activity Management. Participants were allocated in a 1 : 1 ratio, minimised by age and gender. We conducted an internal pilot (first 12 months) and integrated qualitative methods to optimise recruitment and retention. Participants Adolescents aged 11–17 years with a diagnosis of ME/CFS [defined using National Institute for Health and Care Excellence (NICE) 2007 Criteria], who did not have a local specialist ME/CFS service. To have a confirmed diagnosis, potential participants were required to have been assessed by a paediatrician (or equivalent specialist doctor), and to have had screening bloods taken. We excluded adolescents if: they were not disabled by fatigue; their fatigue was due to another cause (including primary anxiety/depression); they had access to a local specialist service; they were unable to complete video calls or FITNET-NHS treatment modules; they were pregnant. Setting We identified potentially eligible adolescents at referral to the specialist paediatric ME/CFS service at the Royal United Hospital, Bath. Patients were referred to the service by their general practitioner (GP) or paediatrician (or equivalent specialist). Between September 2018 and March 2020, we offered GP surgeries the opportunity to become patient identification centres. In these sites, database searches were conducted to identify potentially eligible adolescents, who were then offered referral to the Bath Royal United Hospital for eligibility assessment. Recruitment Potentially eligible adolescents were contacted by the clinical team by telephone to discuss the opportunity to take part in the FITNET-NHS trial. Interested adolescents were sent study information, including a link to the Revised Children’s Anxiety and Depression Scale (RCADS) questionnaire. Eligibility assessment Adolescents who were interested were invited to an eligibility assessment conducted by a specialist nurse. The nurse checked if: (1) they had debilitating persistent or relapsing fatigue for at least 3 months, but not life-long; (2) fatigue was not the result of ongoing exertion and not substantially alleviated by rest; (3) they had post-exertional malaise (and increase in fatigue and other symptoms after exertion); and (4) their fatigue was severe enough to cause substantial reduction in previous levels of occupational, educational, social or personal activities. The recruiting nurse checked that the screening blood tests had been done within the previous 12 months and were normal. The nurse then assessed whether the adolescent had comorbid mental health problems or a primary mental health problem. Patients were asked to complete the RCADS. Those scoring within the borderline or above the age-/gender-validated threshold answered further questions to determine if they were at risk of harm and whether their anxiety/depression was sufficiently severe to explain the fatigue. This enabled us to ensure that those recruited had a diagnosis of ME/CFS from a specialist but also had 3 months of disabling fatigue plus one further symptom and did not have an exclusionary diagnosis and therefore had ME/CFS according to the contemporary United Kingdom (UK) guideline. Consent If adolescents were eligible and interested in taking part in the FITNET-NHS trial, consent and assent (patients aged 11–15 years) were obtained using an online form. Randomisation We used an automated web randomisation service operated by Bristol Trials Centre. Participants were allocated in a 1 : 1 ratio. Allocation used minimisation to balance by age and gender with a random component to preserve allocation concealment. Blinding It was not possible to blind the participant, family or the clinical service. The investigators (including the senior statistician who wrote the statistical analysis plan) were blinded to treatment allocation. The study statistician was unblinded as they reported to the Data Safety and Monitoring Committee. Interventions Fatigue In Teenagers on the interNET in the National Health Service (FITNET-NHS) This is a web-based modular specialist CBT programme designed to be used by adolescents with ME/CFS and their parents. We adapted the Dutch FITNET platform for UK adolescents. FITNET-NHS is delivered using asynchronous individualised e-consultations within the web-based platform. Contact is weekly initially and then becomes less frequent. The therapist works with patients and parents separately. FITNET-NHS has psycho-educational and CBT chapters for patients and a parallel programme for parents. There are 19 chapters. Chapters 1–3 introduce ME/CFS, CBT and the role of therapists; Chapter 4 discusses treatment goals; Chapters 5–19 are about cognitive and behavioural strategies. There are diaries for patients that are visible to the therapist. FITNET-NHS is individualised for patients. Chapters 5–19 are unlocked (made available) by the therapist according to clinical presentation, needs and formulation. Fidelity was assessed in clinical supervisions. Activity Management Participants and their families received information on ME/CFS, Activity Management, sleep and symptoms management. Participants could use an online app to record their activity. Activity Management was delivered via videocall. In the initial assessment, participants had a detailed assessment of physical and cognitive activity. The participant and therapist agreed a ‘baseline’ of activity, which is the average level of activity. Participants were asked to record activity, and then, when activity was stable, to increase activity gradually in a flexible and individual way. The initial assessment was 90 minutes (but this could be split into two shorter sessions). Follow-up video calls (60 minutes each) were organised 2–6 weeks apart. In November 2017, the number of follow-up sessions was increased from 3 to 6 in response to participant feedback. Fidelity was assessed using a checklist of mandatory, flexible and prohibited elements. Data sources Adolescent- and parent-completed measures, collected online using research electronic data capture. Therapist reports. Routinely collected data from local systems and from NHS Digital. Qualitative interviews with adolescents, parents and therapists. Measures Primary outcome Disability measured using the 36-item Short Form Health Survey Physical Function subscale (SF-36-PFS) measured 6 months after randomisation. We included outcome data returned in a 5-to-9-month window post randomisation. Secondary outcomes (Measured at 3, 6 and 12 months unless specified): Physical function (3 and 12 months); fatigue [Chalder fatigue scale and checklist individual strength (CIS) fatigue scale]; school attendance (self-reported days per week attending school, or whether receiving home tuition); mental health {RCADS; pain [visual analogue scale (VAS)]; Clinical Global Impression Scale; quality of life [EuroQol-5 Dimensions Youth (EQ-5D-Y)]; parent-completed resource use questionnaire; parent-completed work productivity and activity impairment questionnaire general health (WPAI:GH)}. Demographic data were collected at recruitment and included: age, sex, postcode, ethnicity, symptoms, months since illness onset, presence of comorbid illnesses. Harms/adverse events We prospectively collected the following data: (1) clinician-reported serious deterioration in health, (2) a decrease of ≥ 20 in SF-36-PFS between baseline and 3, 6 or 12 months or scores of ‘much’ or ‘very much’ worse on the Clinical Global Impression scale or (3) withdrawal from treatment because of feeling worse. Safety outcomes were reported to the Data Safety Monitoring Committee. Resource use data In addition to study documentation and patient/parent questionnaires, we used patient-level data recorded on the Royal United Hospitals Bath NHS Trusts electronic patient record system (Millennium). We obtained secondary care data on outpatient visits, inpatient visits and emergency department attendances from NHS Digital. Costs were valued using 2019/20 prices. Analysis Sample size Our sample size of 314 participants (assuming 15% attrition) gives 90% power at 5% significance to detect a 10-point (0.4 standard deviation) difference for the SF-36-PFS for our primary outcome. The original sample size was powered to detect a difference in the subgroup with mental health problems, but this was revised during the study. Statistical methods We used an intention-to-treat analysis in study participants who completed the primary outcome. We used multivariable linear regression analysis adjusting for baseline values of the outcome, age and gender. The treatment effect was estimated as an adjusted difference between sample means. We conducted the following pre-planned sensitivity analyses: we adjusted for variation across participants in the time between randomisation and the 6-month outcome; the primary analysis was repeated with an additional binary covariate distinguishing participants recruited before or after 1 September 2019 (distinguishing those with a 6-month assessment before or during the COVID-19 pandemic); we repeated the analysis in those who completed one or more modules/sessions of their allocated intervention. We conducted a sensitivity analysis for participants with ME/CFS according to the NICE (2021) criteria. We adapted the regression model used for the primary analysis to the secondary outcome variables. We estimated the effectiveness of FITNET-NHS compared with Activity Management in participant subgroups defined by the presence or absence of anxiety or depression (defined by the RCADS). Health economic analysis We used an intention-to-treat approach and multiple imputation by chained equation to minimise bias due to missing data. We combined cost and quality-adjusted life-year (QALY) data to calculate an incremental cost-effectiveness ratioand an incremental net monetary benefit (iNMB) statistic. We performed a subgroup analysis to explore the interaction between comorbid anxiety/depression and cost-effectiveness. Prespecified sensitivity analyses include conducting a complete-case analysis and repeating the primary analysis using the tariff paid by Clinical Commissioning Groups instead of reference costs. Qualitative methods We integrated qualitative methods into the pilot and main phase of the trial to explore trial conduct, recruitment and intervention acceptability. We analysed recruitment to trial consultations and conducted in-depth interviews with recruiters, trial therapists and participants. Results were used to improve recruitment and make small changes to the interventions. Qualitative analysis was ongoing and iterative, commencing soon after data collection. Audio recordings were transcribed, checked and imported into NVivo (QSR International, Warrington, UK). The data were systematically assigned codes and analysed thematically using techniques of constant comparison. Results Of 892 referrals between 1 November 2016 and 31 October 2020, 550 were eligible, of which 155 were allocated to FITNET-NHS and 159 to Activity Management. 265 adolescents were included in the primary analysis (127 for FITNET-NHS and 138 for Activity Management). The baseline characteristics were similar between the treatment arms. 147 participants had either comorbid depression or anxiety (145 depression, 34 anxiety and 32 both). The number completing 80% or more of expected modules/sessions was lower for FITNET-NHS participants (58, 38%) compared to Activity Management participants (124, 78%). Participants in the FITNET-NHS group had a greater improvement in physical function compared to the Activity Management group at 6 months [mean difference 8.2, 95% confidence interval (CI) 2.7 to 13.6]. The sensitivity analyses confirmed the primary analysis. This was true for participants with ME/CFS defined using the NICE (2021) criteria. At 6 months, those allocated to FITNET-NHS were, on average, attending half a day more of school per week compared to Activity Management, and this difference was maintained at 12 months. They experienced less fatigue (with the CIS fatigue measure) at both 6 and 12 months. There was no strong evidence that either treatment was more effective in those with comorbid depression/anxiety, and similar levels of improvement were noted for pain, the Clinical Global Improvement Score, and fatigue measured using the Chalder Fatigue Scale. Participants continued to improve between 6 and 12 months in both treatment arms, and there was little difference in physical function between the two treatment arms at 12 months. Fatigue In Teenagers on the interNET in the National Health Service participants had a small gain in QALYs (0.002, 95% CI −0.041 to 0.045) compared to Activity Management but substantially higher mean costs (£1047.51, 95% CI £624.61 to £1470.41). In the primary analysis, from an NHS perspective, at a threshold of £20,000 per QALY, the iNMB was −£1001.74 (95% CI −£2041.31 to £37.83), indicating that FITNET-NHS is unlikely to be cost-effective. The wide Cis show there is considerable uncertainty in this result. The subgroup analysis suggests that FITNET-NHS is more likely to be cost-effective among those with comorbid anxiety and depression at baseline. In the qualitative interviews, families felt online treatment could fit around everyday life and reduce the increase in symptoms that can accompany travelling to face-to-face appointments. However, some participants still preferred in-person treatment and found it difficult to build a rapport with therapists online. Personalised contact with a therapist was valued in both treatments and remains essential even in online treatment. The reading and writing required on the FITNET-NHS platform was difficult for younger children and those with cognitive symptoms such as brain fog; therefore, adaptations may be needed for these groups. Therapists felt they had to develop a different set of skills to engage patients online. Conclusions Despite the fact that adolescents are more likely to have better physical function at 6 months and attend more school (at 6 and 12 months) after receiving FITNET-NHS compared to Activity Management, FITNET-NHS is unlikely to be cost-effective. This study is consistent with previous RCTs that demonstrated the effectiveness of CBT for children and young people with ME/CFS. It is the first study to attempt to look at cost-effectiveness. The high additional cost of FITNET-NHS and limited substantial sustained impact mean that it may not be a cost-effective use of NHS funds. Alternatively, it is possible that the EQ-5D-Y is not sensitive enough to appropriately reflect the improvements in physical function and school attendance. Most participants in the FITNET-NHS group did not complete 80% or more of the recommended treatment modules. While our qualitative data suggest that FITNET-NHS was acceptable to most patients, it was considered burdensome by some. Implications for health care This study strengthens the evidence that CBT is effective for adolescents with ME/CFS and should be the first line of treatment offered. FITNET-NHS is an intensive treatment approach and is unlikely to be cost-effective. However, the online approach was popular with patients and families. Recommendations for research Would a shorter intervention with less intensive therapist input increase retention and reduce costs without jeopardising effectiveness? We need a validated health economic measure for children and young people that is sensitive to change. Further research needs to be conducted on the best method to deliver remote treatment. A large implementation study in the Netherlands suggested that many patients (and therapists) prefer a mixture of face-to-face and online treatment. Trial registration This trial is registered as ISRCTN18020851. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 14/192/109) and is published in full in Health Technology Assessment; Vol. 28, No. 70. See the NIHR Funding and Awards website for further award information.

Background Stroke often affects recognition and interpretation of information from our senses, resulting in perceptual disorders. Evidence to inform treatment is unclear. Objective To determine the breadth and effectiveness of interventions for stroke-related perceptual disorders and identify priority research questions. Methods We undertook a scoping review and then Cochrane systematic review. Definitions, outcome prioritisation, data interpretation and research prioritisation were coproduced with people who had perceptual disorders post stroke and healthcare professionals. We systematically searched electronic databases (including MEDLINE, EMBASE, inception to August 2021) and grey literature. We included studies (any design) of interventions for people with hearing, smell, somatosensation, taste, touch or visual perception disorders following stroke. Abstracts and full texts were independently dual reviewed. Data were tabulated, synthesised narratively and mapped by availability, sense and interventions. Research quality was not evaluated. Our Cochrane review synthesised the randomised controlled trial data, evaluated risk of bias (including randomisation, blinding, reporting) and meta-analysed intervention comparisons (vs. controls or no treatment) using RevMan 5.4. We judged certainty of evidence using grading of recommendations, assessment, development and evaluation. Activities of daily living after treatment was our primary outcome. Extended activities of daily living, quality of life, mental health and psychological well-being perceptual functional and adverse event data were also extracted. Results Scoping review We included 80 studies (n = 893): case studies (36/80) and randomised controlled trials (22/80). No stroke survivor or family stakeholder involvement was reported. Studies addressed visual (42.5%, 34/80), somatosensation (35%, 28/80), auditory (8.7%, 7/80) and tactile (7.5%, 6/80) perceptual disorders; some studies focused on ‘mixed perceptual disorders’ (6.2%, 5/80 such as taste–smell disorders). We identified 93 pharmacological, non-invasive brain stimulation or rehabilitation (restitution, substitution, compensation or mixed) interventions. Details were limited. Studies commonly measured perceptual (75%, 60/80), motor-sensorimotor (40%, 32/80) activities of daily living (22.5%, 18/80) or sensory function (15%, 12/80) outcomes. Cochrane systematic review We included 18 randomised controlled trials (n = 541) addressing tactile (3 randomised controlled trials; n = 70), somatosensory (7 randomised controlled trials; n = 196), visual (7 randomised controlled trials; n = 225) and mixed tactile-somatosensory (1 randomised controlled trial; n = 50) disorders. None addressed hearing, taste or smell disorders. One non-invasive brain stimulation, one compensation, 25 restitution and 4 mixed interventions were described. Risk of bias was low for random sequence generation (13/18), attrition (14/18) and outcome reporting (16/18). Perception was the most commonly measured outcome (11 randomised controlled trials); only 7 randomised controlled trials measured activities of daily living. Limited data provided insufficient evidence to determine the effectiveness of any intervention. Confidence in the evidence was low–very low. Our clinical (n = 4) and lived experience (n = 5) experts contributed throughout the project, coproducing a list of clinical implications and research priorities. Top research priorities included exploring the impact of, assessment of, and interventions for post-stroke perceptual disorders. Limitations Results are limited by the small number of studies identified and the small sample sizes, with a high proportion of single-participant studies. There was limited description of the perceptual disorders and intervention(s) evaluated. Few studies measured outcomes relating to functional impacts. There was limited investigation of hearing, smell, taste and touch perception disorders. Conclusion Evidence informing interventions for perceptual disorders after stroke is limited for all senses. Future work Further research, including high-quality randomised controlled trials, to inform clinical practice are required. Study registration This study is registered as PROSPERO CRD42019160270. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128829) and is published in full in Health Technology Assessment; Vol. 28, No. 69. See the NIHR Funding and Awards Website for further award information. Plain language summary After a stroke, individuals may have difficulty understanding information gathered through their sense of sight, hearing, smell, taste, touch or somatosensation (body position, temperature, etc.), known as perceptual problems. We estimate perceptual problems affect around 240,000 stroke survivors in the UK, limiting their ability to understand the world around them, affecting everyday activities and reducing quality of life. Healthcare professionals may offer different treatments; medicine, brain stimulation, or rehabilitation activities including puzzles, strategies or physical therapy. We wanted to find the best treatments for stroke-related perceptual problems. We searched for all research on sight, hearing, smell, taste, touch and somatosensation perceptual treatments to find out (1) how well they worked, (2) what the research means for stroke survivors and healthcare professionals and (3) what research is needed next. People with stroke-related perceptual problems and healthcare experts produced this research together. We found 80 studies, involving 893 stroke survivors, describing 93 treatments. Eighteen of these studies used higher-quality randomised controlled trial designs; 535 stroke survivors took part, testing 32 treatments. Randomised controlled trials are important as one-half of those involved receive treatment and one-half do not; they provide the best evidence about whether a treatment works. Most treatments were for visual or somatosensation problems. Each study was small, provided few details about the participants or their treatment, and tested very different treatments. Few measured the effect of treatment on everyday life: only seven measured stroke survivors’ ability to take part in everyday activities. No trial asked stroke survivors about their experiences with the treatments offered. We do not have enough research to identify which treatments benefit the lives of people with stroke-related perceptual problems. We need more research into perceptual problems, especially the impact it has on stroke survivors’ lives, as well as bigger studies into well-described treatments, that measure the impact of the treatment on people’s lives. Scientific summary Background Perception is the synthesis and interpretation of information gathered through the senses: hearing, taste, touch, smell, visual and information on temperature, pressure, vibration and body position, known as somatosensation. Up to a fifth of stroke survivors experience perceptual disorders after stroke, limiting their ability to perceive and process sensory information and reducing their ability to take part in daily activities. To date, the effectiveness of perceptual disorder interventions after stroke is unclear. Clinical guidelines offer limited recommendations. Stroke survivors, carers and healthcare professionals have stated that improving research into perception is important to them. Further, it is important to systematically identify evidence gaps and future research priorities. Objectives We aimed to: Identify all published and unpublished research evaluating interventions for perceptual disorders after stroke, providing a comprehensive report on the scope and nature of the evidence to date and highlighting the research gaps identified. Synthesise and appraise the quality of randomised controlled trial (RCT) evidence of the effectiveness of perceptual disorder interventions after stroke. Understand the implications of our findings for stroke survivors and HCPs working in this area and to determine future research priorities. Methods Our project included a scoping review, the revision and expansion of a Cochrane systematic review and we worked with a Lived Experience Group and a Clinical Expert Group to co-create research recommendations and identify research priorities. Our scoping review of the literature was based on a systematic search of several electronic databases including MEDLINE, EMBASE and CINAHL (inception to August 2021), as well as searches of grey literature, contacting experts and forward citation tracking. We included studies of any design which explored interventions for stroke survivors with hearing, smell, somatosensation, taste, touch or visual perception disorders. Eligible abstracts and full texts were independently reviewed by two reviewers; data were extracted, tabulated and narratively synthesised. Data availability and outcome measures used were mapped. In keeping with scoping review methodology, we did not formally assess research quality. We updated a Cochrane systematic review, including RCTs of adult stroke survivors with perceptual disorders. We assessed the risk of bias, conducted meta-analyses to explore effectiveness of interventions and judged our confidence in the findings using grading quality of evidence and strength of recommendations (GRADE). Outcomes were measured using activities of daily living (ADLs) with extended activities of daily living (EADLs), quality of life, mental health, perceptual function and adverse events data also collated. Using structured involvement and priority setting approaches we worked in partnership with our Lived Experience and Clinical Expert Groups to agree clinical implications and to future research priorities. Results This project was coproduced with people with lived experience of stroke and perceptual disorders (n = 5) and relevant multidisciplinary clinical expertise (n = 4). Working in partnership with the core research team, these groups informed the project throughout, agreeing definitions of perception, relevant outcome measures, clinical implications and priorities for future research. Scoping review Of 91,869 records screened, we included 80 studies (including 36 case reports; 22 RCTs) in the scoping review, most (64%) of which were published in the previous decade. Participants (n = 893) were predominately adults and male; five children were included. Studies generally had small sample sizes, with RCTs accounting for most participants (70.5%; 630/893). The perceptual disorders represented included visual (43%), somatosensory (35%), auditory (9%), tactile (8%) or ‘mixed’ disorders (5%) which included one study on taste–smell disorders. We identified 93 interventions including rehabilitation (84%), pharmacological (6.5%) and non-invasive brain stimulation (NIBS) interventions (7.5%); no surgical or assessment-based interventions were identified. Intervention details were limited. Outcome measures commonly included perceptual function (75%), motor/sensorimotor (40%), ADLs (23%) or sensation (15%). No data on discharge destination, health economic, feasibility, educational (children), psychological well-being and mental health, quality of life, or activity and participation were reported. Time points were typically immediately after the intervention (39%) or within 3 months follow-up with just 15% of studies capturing outcomes beyond that time point. Cochrane systematic review Drawing on the scoping review results, the trials identified in a previous Cochrane Review and an updated search of bibliographic databases, 2575 records were identified. From these, 114 full texts were considered and 18 RCTs (n = 541) were included. All but six were stroke survivors, between 19 days and 4.3 years from onset. The interventions included targeted visual (seven RCTs; n = 225), tactile (three RCTs; n = 70), somatosensory (seven RCTs; n = 196) and one mixed tactile-somatosensory disorders (one RCT; n = 50). No RCT evaluated interventions for stroke-related hearing, taste or smell disorders. Interventions included 1 NIBS, 1 compensatory, 25 restitution, 4 mixed and 1 unclear intervention approach. Seven included RCTs (39%) measured participants’ ADLs though others captured perception (11 RCTs), adverse events (6 RCTs), mobility (4 RCTs) and EADLs (1 RCT). None measured activity and participation, quality of life or psychological well-being and mental health outcomes. We identified 11 ongoing RCTs. The risk of bias of the included RCTs varied, with 72% describing adequate generate of the randomisation sequence and outcome assessor blinding, but concealment of allocation was considered adequate for only a third. Most trials adequately reported participant attrition (78%) and the outcome data gathered (89%). Other sources of bias were noted including an imbalance between the groups at baseline and altered eligibility criteria mid-RCT. With limited data there was insufficient evidence to determine the effectiveness of any one intervention compared to no intervention or an alternative intervention. Based on the small number of RCTs, the small sample sizes and the limited comparisons available, our confidence in the evidence was, using GRADE, judged to be low–very low. Strengths Throughout this project, a Lived Experience and Clinical Expert Group were centrally involved in the development of definitions, categorisation, outcome measurement selection, interpretation of data and research prioritisation, supporting clinical relevance and validity. The consensus working definitions and categorisations developed may support future research on this topic area. Our scoping and systematic reviews were conducted to the highest research conduct and reporting standards. Limitations Despite large numbers of people experiencing one or more perceptual disorder after stroke, there is a striking lack of relevant research to inform interventions. What little has been reported is often based on a single participant or small sample sizes. There is limited description of the perceptual disorder, the intervention(s) evaluated and a focus on perception outcomes rather than measures that reflect the functional impacts described by the Lived Experience Group, for example ADLs. We also found evidence of under-researched subpopulations including children and people with hearing, taste and smell perceptual disorders. The project team, Lived Experience and Clinical Expert Groups are UK-based, and it is unclear whether our priorities (outlined below) capture the wider international picture. Priority setting Our clinical expert (n = 4) and lived experience (n = 5) stakeholders’ input was pivotal throughout the project. Together with the core research team, these groups agreed the clinical implications and research priorities emerging from the findings. Implications for health care Clinical recommendations include the facilitation of improved awareness of stroke-related perceptual disorders, assessment and information provision and holistic intervention approaches and support. While the research evidence was insufficient to support clinical decision-making relating to the choice of intervention approach, the scoping review provides an important information resource for clinicians developing best practice until sufficient evidence becomes available. Recommendations for research The evidence informing interventions for perceptual disorders after stroke is limited, and absent for smell, taste and tactile disorders. Future research should prioritise (1) exploration of the lived experience of people with stroke-related perceptual disorders, (2) improving assessments of stroke-related perceptual disorders, (3) exploring interventions in a way that reflects real-world needs, (4) exploring current clinical practices that address perceptual disorders following stroke and (5) establishing the prevalence of perceptual disorders after stroke. Conclusions Healthcare professionals lack high-quality evidence of effective interventions to inform their provision of advice, treatment and education of stroke survivors with perceptual disorders and their families. Evidence informing these research priority topic areas is urgently required. Study registration This study is registered as PROSPERO CRD42019160270. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128829) and is published in full in Health Technology Assessment; Vol. 28, No. 69. See the NIHR Funding and Awards Website for further award information.

Background Neonates undergoing emergency abdominal surgery frequently require a stoma; closing this stoma with a second operation is an essential part of recovery. Timing of closure varies. Optimal timing is unclear and would be best resolved through a randomised controlled trial; such a trial is likely to be challenging. Aim To determine if it is feasible to conduct a clinical trial comparing ‘early’ versus ‘late’ stoma closure in neonates. Design Mixed methods comprising three parallel workstreams incorporating: a clinician survey, prospective observational cohort study, parent interviews, focus groups, database analyses and consensus meeting. Setting Specialist neonatal surgical centres across the United Kingdom. Participants and data sources Neonatologists, neonatal surgeons, neonatal dietitians and neonatal nurses who care for neonates with stomas. Neonates with recent stoma, their parents and the clinicians looking after them. Three existing, overlapping clinical databases. Results One hundred and sixty-six professionals from all 27 neonatal surgical centres completed the survey: 6 weeks was the most common target time for stoma closure across clinical scenarios, although there was wide variation. Timing of closure was influenced by nutrition, growth and stoma complications. The prospective cohort study enrolled 56 infants from 8 centres. Infants were mostly preterm with necrotising enterocolitis or intestinal perforation. Clinicians identified extreme preterm gestation and clinical conditions as reasons for not randomising babies into a hypothetical trial comparing early and late stoma closure. Parents and healthcare professionals identified that comparator arms needed more clinical flexibility in relation to timing of stoma closure. Analysis of existing databases revealed wide variation in current timing of stoma closure in neonates and identified approximately 300 eligible infants for a trial per annum in the United Kingdom. Conclusions A trial of ‘early’ compared to ‘late’ stoma closure in neonates is feasible and is important to families and health professionals. The population of eligible babies in the United Kingdom is sufficient for such a trial. Challenges centre around lack of equipoise in certain scenarios, specifically: extremely preterm infants; infants waiting too long for stoma closure in the ‘late’ comparator; and logistical issues in closing a stoma at a trial-allocated time. These challenges are addressable by incorporating flexibility based on gestation at birth, communicating that both trial arms are standard practice and valid treatment options, and providing resources, for example, for operating lists. Future work We recommend the following population, intervention, comparator and outcome as a starting point to inform future trial design. Population: neonates with stomas (excluding those with a fixed treatment pathway). Intervention: stoma closure at 6 weeks and after 32 weeks post conceptual age. Comparator: expectant management with stoma closure undertaken when the clinical team determines is best for the infant. Primary outcomes: weight gain/growth or length of hospital stay. Study registration This study is registered as IRAS Project ID 278331, REC Reference 20/LO/1227. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128617) and is published in full in Health Technology Assessment; Vol. 28, No. 71. See the NIHR Funding and Awards website for further award information. Plain language summary Babies who develop a bowel problem soon after birth may need surgical treatment and a stoma (where the bowel is brought out into a bag on the tummy). Stomas can be life-saving but can also cause problems. They must be ‘closed’ with a second operation once the baby has got better. The timing of this second operation varies between hospitals and we do not know the best time to do it. To find out the best time to close a baby’s stoma, we wanted to conduct a clinical trial. This would mean some babies having their stoma closed ‘early’, and other babies ‘later’ and seeing which group did better. Babies would be put into the ‘early’ or ‘late’ group at random so that we could have a fair comparison. Because doing a trial may be challenging, we first set out to find out if a trial would be feasible in the Timing of Stoma Closure in Neonates study. What we did Found out what currently happens to babies with a stoma through a survey of healthcare staff and looking at data from several national databases. Collected data on 56 babies with stomas in 8 hospitals, including asking their doctors whether they would recruit these babies to a trial and when they would close their stoma. Interviewed parents of children who had had a stoma. Held focus groups with staff at the eight hospitals. Held a meeting for National Health Service staff and parents to discuss the way forwards. What we found Parents and doctors think this question is important and that a clinical trial is a good idea. Deciding on the timing of stoma closure in the trial and what we mean by ‘early’ and ‘late’ is critical to the success of a trial. The timing may need to take into account that some babies born extremely preterm may be difficult to include. A trial that compares how well infants grow and/or length of stay in hospital between babies having early or later stoma closure may well be feasible. What happens next The National Institute for Health and Care Research (research arm of the Department of Health and Social Care) will use this to information to help decide if they would like to fund a trial in the future. Scientific summary Background Neonates undergoing emergency abdominal surgery for problems such as necrotising enterocolitis (NEC), spontaneous intestinal perforation (SIP) or a congenital bowel obstruction frequently require a stoma. While stomas can be life-saving, they pose a number of challenges including fluid and electrolyte imbalance, local wound and skin problems, malnutrition and growth failure. Reversing (closing) these stomas with a second operation is therefore an essential part of an infant’s recovery. The timing of stoma closure is known to vary between hospitals and, indeed, clinicians and the best time remains unclear, with conflicting evidence from published studies of low methodological quality. Rationale Determining the best time to close stomas in neonates is imperative, as it has significant implications for: infant health outcomes (e.g. stoma complications and growth) families (e.g. reduced neonatal unit stay, healthcare burden, time off work) healthcare providers [e.g. neonatal unit bed-days, parenteral nutrition (PN) use, surgery, cost]. In addition, reducing unwarranted variability in surgical care is a key priority for the NHS. The most methodologically robust way to determine optimal timing of stoma closure would be through an adequately powered randomised controlled trial (RCT). A trial of different stoma closure times in neonates is likely to be challenging, due to: the patient group who have heterogenous underlying diseases and comorbidities clinician factors such as willingness to recruit parent factors such as trial acceptability. This study aimed to explore these potential challenges and determine if a trial comparing ‘early’ and ‘late’ stoma closure is feasible. Aims The overarching aim of the Timing of Stoma Closure in Neonates (ToSCiN) study was to answer the question: is it feasible to conduct a RCT comparing ‘early’ versus ‘late’ stoma closure in neonates? Objectives The specific objectives of the ToSCiN study were: to establish current UK practice for stoma closure in neonates to determine whether there is equipoise among clinicians (neonatal surgeons and neonatologists) and allied health professionals (specialist nurses and dietitians) over when it is best to close stomas in neonates to define ‘early’ and ‘late’ stoma closure for a potential trial to define a population of neonates for inclusion in a trial (in whom there is significant uncertainty over timing) and determine how many infants are eligible for inclusion to establish the most appropriate design and outcome measures for a trial to determine the willingness of parents, neonatal surgeons and neonatologists to include neonates in a trial that would randomise to ‘early’ or ‘late’ stoma closure and identify potential barriers to recruitment to assess the suitability of using routinely collected data for gathering clinical information for a trial. Methods The ToSCiN study used a mixed-methods approach (qualitative and quantitative methodology) and comprised three parallel workstreams. Workstream 1: Survey of clinician and allied health professional perspectives on neonatal stoma closure An online survey was sent to clinicians and allied health professionals involved in the care of newborn infants with stomas at neonatal surgical units across the UK. The survey questions and a series of clinical scenarios to explore current practice/preferences, perspectives on ‘early’ versus ‘late’ stoma closure, which groups of infants should be included in a trial comparing these, preferred trial design including outcomes and barriers to achieving the optimal timing of closure. Workstream 2: Parent and clinician perspectives regarding a clinical trial of neonatal stoma closure This workstream took place in eight UK neonatal surgical centres and comprised three elements: (2.1) An observational cohort study of neonates who had a stoma formed Key clinical and demographic information was recorded prospectively. The data set comprised factors that could influence the timing of stoma closure and outcomes which are likely to be important in a future trial. (2.2) Questionnaires for the principal clinicians (surgeon and neonatologist) caring for each infant recruited to the cohort study These were completed at three time points and explored the acceptability of a hypothetical trial to clinicians for each infant. Time points: (1) approximately 1 week after stoma formation to explore randomising the infant to a trial, (2) 6 weeks after stoma formation to explore following an allocation to early closure intervention and (3) 12 weeks after stoma formation to explore following an allocation to late closure comparator. (2.3) A qualitative study incorporating: (1) focus groups with clinicians and (2) interviews with parents of neonates who had a stoma Multidisciplinary healthcare professional focus groups explored equipoise, current practice and views on early versus late closure, willingness to recruit to a trial, acceptability of a trial and prioritised outcome measures. Parents of infants in the cohort study and additional parents recruited via social media were invited for a telephone or video interview. Their views were explored on: having a child with a stoma, acceptability of a trial, timing of recruitment, potential barriers to trial participation, trial participant information materials, approach to consent for a trial and prioritised outcome measures. Workstream 3: Analysis of three existing national databases Analyses were carried out on three existing national data sets [from the National Neonatal Research Database (NNRD), The British Association of Paediatric Surgeons Congenital Anomalies Surveillance System (BAPS-CASS) and Hospital Episode Statistics – Admitted Patient Care (HES-APC)]. The analyses aimed to: (1) establish current UK practice, (2) define a population for trial inclusion and provide the number of eligible infants and (3) establish values for key outcome measures. Consensus meeting This was held using an online video platform at the end of the study period when data collection and preliminary analysis had been completed. Professionals and parents previously involved were invited and further parents were recruited via social media. A summary of results from the three workstreams was presented to attendees. Facilitated small group discussions took place, followed by a summary of small group discussions presented by group facilitators to all attendees, and electronic voting was conducted about trial acceptability and design. Results Workstream 1 Results One hundred and sixty-six professionals completed the survey with at least one respondent from each of the 27 UK centres. Six weeks was the most commonly stated target time for closure across all scenarios, although there was a high degree of variability, with intervals of 12 weeks or more frequently advocated. While 70–76% of respondents preferred closure prior to discharge in preterm infants and a term infant with jejunostomy, only 46% preferred this in term infants with ileostomy. A sizeable proportion (41%) use weight, rather than time, to determine when to close a stoma. Thematic analysis of free text identified nine key themes influencing decision-making, mostly related to nutrition, growth and stoma complications. With regard to potential trial design, 86% of respondents indicated they would include preterm infants with NEC/SIP and 72% term infants with other conditions. Closure at 6 weeks was the most commonly cited timing of early intervention; the preferred later time point was 12 weeks. Growth was most commonly selected as the favoured primary outcome for a trial, followed by time to full feeds, length of stay and duration of PN. Workstream 2 Results Cohort study Fifty-six infants were enrolled from eight UK units. Infants were mostly preterm (44/56), with median birthweight 961 g (range 415–3962 g). The cohort comprised 37 infants with NEC/SIP and 19 with other diagnoses. Most stomas were in the small bowel, formed at a median of 8 days (range 1–80 days). At 1 week following stoma formation, 18/56 infants were mechanically ventilated, 8 were receiving inotropes and 13 had received blood products in the preceding 24 hours. At this point, surgeons stated they would be willing to randomise 31 infants (59%) into a hypothetical trial of 6- versus 12-week closure. The commonest reason for not being willing was the infant being ‘too small or premature’. At 6 weeks post stoma, median weight was 2024 g (range 795–4460 g) and nine remained ventilated. Eighty per cent (42/56) were reported to have gained weight in the past week. Median enteral feed volume was 127 ml/kg/day, stoma output was < 20 ml/kg/day in 86% and stoma-related complications were reported in one-third. Surgeons would be willing to follow theoretical trial allocation for ‘early’ closure in 17/56 (33%). In nine infants, the stoma had already been closed. The main reasons for not being willing to follow trial allocation were current clinical status and size. Among infants whom surgeons said they would not be willing to follow trial allocation, most (16/25) were < 28 weeks’ gestation at birth and were lighter (median weight 1609 vs. 2090 g). Twelve weeks after stoma formation, median weight was 2548 g and 10% (4/52) remained ventilated. Most (23/42; 55%) were still receiving PN (a higher proportion than at 6 weeks) and 12% (5/42) were not receiving any enteral feed. Stoma-related problems were present in one-third and one-quarter had stoma output > 20 ml/kg/day. Surgeons stated they would be unwilling to follow trial allocation of closure at 12 weeks in 24 infants: 12 of these had required earlier stoma closure and 9 (mainly premature infants) were considered unsuitable for stoma closure for a variety of mainly clinical reasons (rather than logistical or family factors). Again, most infants for whom the surgeon was unwilling to follow trial allocation were < 28 weeks' gestation at birth (13/24). Overall stoma closure took place before the end of the data collection period in 82% (46/56 infants, 4 infants died with a stoma in situ and 6 infants still had a stoma). Median time to closure was 9.9 weeks (range 4.4–28.3 weeks), at median 88 days of age and weight of 2631 g. Qualitative study We interviewed 24 parents (17 mothers, 7 fathers) of babies who required a stoma in the last 3 years in the UK (July 2021–February 2022). Fifteen parents were recruited via hospital sites and nine via social media. Thirty-six staff (14 surgeons, 10 consultant neonatologists, 6 neonatal surgical nurses, 2 research nurses, 1 dietitian and 3 ‘other’ roles) from five study sites took part in one of six focus groups (November–December 2021). Most parents and staff considered the proposed ToSCiN trial to be acceptable and wanted the research question to be answered. Staff would be willing to randomise babies, except the most extremely preterm, to the trial and parents would hypothetically consent to their baby taking part if the ‘emotive’ ‘early’ (6 weeks) and ‘late’ (12 weeks) terminology was changed. Some parents and staff, however, had clear trial arm preference and concerns about the alternative trial arm. They felt that decisions about when to close a stoma should be led by the overall health and well-being of the baby. Stoma-related factors, logistical and organisational factors and family factors also influenced views on acceptable timing of stoma closure. These findings highlight potential challenges for recruitment and retention and trial success. Workstream 3 Results While direct comparison was limited by differences in data set populations and coverage, there was a reasonably consistent number of potentially eligible infants with a stoma for NEC/SIP across data sets, with 163 in 1 year in the BAPS-CASS cohort and mean annual volumes of 193 (NNRD) and 192 (HES-APC). HES-APC was used to estimate non-NEC infant numbers, as it had the most comprehensive coverage, with a mean of 118 per year. The total UK population of potentially eligible infants would therefore be approximately 300 per annum. Median stoma durations for babies with NEC/SIP were as follows: BAPS-CASS 63 days; NNRD 60 days and HES-APC 78 days. For infants with stomas for other reasons, closure was usually earlier: NNRD 45 days; HES-APC 74 days; and in the BAPS-CASS meconium ileus cohort, 51 days. HES-APC included babies who were discharged home with a stoma in situ, hence longer median stoma durations. Consensus meeting The consensus meeting was attended by 52 individuals from a range of health professional and non-professional backgrounds, including 7 parents. Voting on key aspects of trial design showed: (1) 83% favoured including all infants with a stoma, (2) 58% favoured comparing closure at 6 weeks after stoma formation with expectant management rather than comparison to a specific late timepoint, and (3) the favoured primary outcomes were weight gain/growth (38%) and length of hospital stay (32%). Conclusions We found that a randomised trial of early compared to late stoma closure in neonates is feasible and is important to families and health professionals. We identified the following components as being critical to a successful future trial: comparison of closure at 6 weeks versus expectant management comparison that accounts for completed gestational age, rather than solely duration of stoma. We identified the potential population and outcomes for such a trial and established that a sufficient population exists in the UK. We also identified that a trial comparing ‘early’ and ‘late’ closure at rigidly defined time points (e.g. closure at 6 compared to 12 weeks) would not be feasible. The principal challenge for trial conduct was found to be the ‘baby-led’ narrative that came through strongly from study participants. Parents and professionals appeared to lack equipoise in certain scenarios. However, this lack of personal equipoise is not exclusive to ToSCiN, and similar themes have been overcome in trials in the past (including complex surgical trials) by optimising communication. Our findings provide valuable insight into how best to do this in a future trial. Other challenges identified include: (1) concerns about inclusion of extremely preterm infants, (2) concerns about infants waiting too long for stoma closure if randomised to the ‘late’ comparator arm and (3) logistical arrangements for closing a stoma at the time dictated by trial allocation. These challenges are eminently addressable, by designing the trial to: (1) incorporate a degree of flexibility (e.g. using ‘expectant management’ as the comparator), (2) make allowances for certain groups (e.g. having a higher corrected gestational age limit for extremely preterm infants) and (3) ensure parents and professionals are aware both trial arms are standard practice and valid treatment options. Key research recommendations We recommend the following population, intervention, comparator and outcome (PICO) as a starting point to inform the design of a future trial: Population: neonates with stomas, excluding those with a stoma as part of a fixed treatment pathway, for example, anorectal malformations and Hirschsprung’s disease. Intervention: stoma closure at 6 weeks post stoma formation and where 32 weeks post-conceptual age has been reached. Comparator: expectant management with stoma closure undertaken when the clinical team determines is best for the infant. Outcomes: weight gain/growth or length of hospital stay should be the primary outcome measure. In order to optimise a trial, we recommend the following practical steps: Involve higher volume neonatal surgical centres for efficient recruitment. Ensure trial staff at each centre are highly trained regarding current standard practice and equipoise. Approach parents 1–2 weeks after stoma formation. Provide resources to centres to permit stoma closure as per the trial protocol, for example, for ring-fenced operating theatre time and possibility of extended hospital stay. Study registration This study is registered as IRAS Project ID 278331, REC Reference 20/LO/1227. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128617) and is published in full in Health Technology Assessment; Vol. 28, No. 71. See the NIHR Funding and Awards website for further award information.

Background Interventions for anxiety need to be adapted to meet the needs of autistic people with moderate to severe learning disabilities and successfully modelled before evidence about efficacy can be generated from clinical trials. Objectives The objectives were to: (1) adapt a behavioural intervention for anxiety, develop an intervention fidelity checklist and logic model, and appraise candidate outcome measures, together with carers, autistic people, and clinicians, (2) characterise treatment-as-usual, (3) model the adapted intervention to determine the acceptability and feasibility for all stakeholders, judge the appropriateness of outcome measures, examine the feasibility and acceptability of consent and associated processes and (4) describe factors that facilitate or challenge intervention delivery. Design This study had two phases. Phase 1a: using consensus methods, an intervention adaptation group was formed who met to adapt the intervention, appraise candidate outcome measures, and contribute to the development of the intervention fidelity checklists and logic model. Phase 1b: a national online survey was conducted with professionals to characterise treatment-as-usual. Phase 2: this was a single-group non-randomised feasibility study designed to model the intervention to test intervention feasibility and acceptability, outcome measures, and aspects of the research process. Setting Participants were recruited from National Health Service community adult learning disabilities teams in England. Participants Participants aged 16 and over with a diagnosis of autism, moderate to severe learning disabilities, an anxiety disorder, and a carer who was available to take part in the intervention. For those who lacked capacity to make a decision about taking part, a consultee had to provide advice that the participant should be included in the study. Interventions The intervention comprised 12 sessions alongside treatment-as-usual. Main outcome measures The feasibility and acceptability of the intervention and research processes, outcome measure completion rates, and intervention adherence. Results The intervention was successfully adapted and modelled with 28 autistic participants with moderate to severe learning disabilities. The intervention was judged to be feasible and acceptable by autistic adults with learning disabilities, carers, and therapists. Carers and therapists suggested minor intervention revisions. Carers completed 100% of outcome measures and the missing data rate was low; however, they indicated that some of the questions were repetitive and said they had difficulty responding to some items. The use of the Mental Capacity Act, 2005, led to an average 5-week delay to participant enrolment. The accrual rate was affected by the COVID-19 pandemic and improved during the summer and early autumn of 2022. Limitations Randomisation was not modelled within this feasibility study, although carers and therapists indicated that this would be acceptable. Conclusions The BEAMS-ID intervention and associated study processes were judged to be feasible and acceptable. The intervention required minor revision. Future work The BEAMS-ID intervention should be tested further within a trial. Study registration This study is registered as ISRCTN12637590. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR129804) and is published in full in Health Technology Assessment; Vol. 28, No. 72. See the NIHR Funding and Awards website for further award information. Plain language summary This study was about autistic adults with moderate to severe learning disabilities who have anxiety. There are good interventions for anxiety, but these need to change to meet the needs of autistic adults with moderate to severe learning disabilities. An example of such an intervention is called exposure therapy. This study had two parts. In the first part, we worked with autistic adults, carers and family members, and professionals to adapt an existing anxiety intervention. We also completed a national survey to find out what interventions people are getting now. In the second part, the intervention was tested with 28 autistic adults with moderate to severe learning disabilities. We asked participants, carers and clinicians about their experiences of being part of the study and receiving the intervention. We successfully changed an existing intervention together with autistic adults, carers and family members, and professionals. The intervention included relaxation training and exposure therapy. From our national survey, we found out that autistic people with moderate to severe learning disabilities who have problems with anxiety are most often offered psychological interventions (e.g. relaxation) or medication (e.g. sertraline) for anxiety. Exposure therapy was not offered very frequently. We tried out our intervention and autistic people with moderate to severe learning disabilities told us that they liked coming to the intervention and liked that their carer attended. Carers liked taking part in the research and were positive about the intervention. Therapists told us that the intervention was suitable and were positive about the training they received in delivering the intervention. Both carers and therapists suggested some changes to the intervention to help improve it in the future. We recommended that a larger study should now be completed. Scientific summary Background A large number of people with autism and learning disabilities have problems with anxiety. There is evidence that talking psychological therapies are an effective intervention for anxiety, but many of these interventions have not been tested with people who have both autism and learning disabilities. These interventions need to be adapted before they can be used with this population because of their difficulties with verbal communication, restricted and repetitive behaviours, and behaviours that challenge. Objectives We aimed to (1) work with our patient and public involvement (PPI) partners to adapt an existing intervention manual for anxiety disorders for use with autistic adults with moderate to severe learning disabilities and (2) complete a feasibility study to try out our intervention and seek feedback from participants, families and therapists. In addition, we collected information about what interventions people are currently receiving to characterise treatment-as-usual (TAU) and test out some outcome measures. Phase 1a: adaptation Our objectives were: to establish an intervention adaptation group (IAG) and, during a series of meetings, adapt an existing intervention to treat anxiety symptoms in autistic adults who have moderate to severe learning disabilities to develop an intervention fidelity checklist that can be used alongside the intervention manual to appraise and consider several candidate outcome measures of anxiety-related symptoms and make a recommendation for use within Phase 2 to develop an intervention logic model. Phase 1b: description of treatment-as-usual Our objective was to complete a national survey of existing interventions for adults with anxiety disorders who have moderate to severe learning disabilities. The survey was constructed using items adapted from the Template for Intervention Description and Replication (TIDieR) checklist to ensure a clear description of TAU. This phase ran concurrently with Phase 1a and Phase 2. This allowed for the capture of data to allow for the characterisation of TAU, including any specific interventions offered. We invited participation from all community-based services for people with autism and/or learning disabilities across the United Kingdom. Phase 2: feasibility study Our objectives were: to model the manualised intervention to determine the acceptability and feasibility for all stakeholders, including autistic adults with moderate to severe learning disabilities, carers, and clinicians, and adjust as required to judge the appropriateness, including response rates, of our measures of anxiety-related symptomatology for use within a larger study to examine the feasibility and acceptability of consent and associated processes in the context of the Mental Capacity Act, 2005 and to describe factors that facilitate or challenge the implementation of our intervention. Methods Study design Phase 1a (intervention adaptation) An IAG was established, and during a series of five meetings an existing intervention used to treat anxiety symptoms in adults with autism was adapted for use with people who also have moderate to severe learning disabilities. Phase 1b (treatment-as-usual survey) A national survey of existing community-based interventions for adults with anxiety disorders who have moderate to severe learning disabilities was conducted. The TIDieR checklist was used to inform the development of our survey questions to ensure a clear description of TAU. Phase 2 (feasibility study) This study was a single-arm, non-randomised, feasibility study, with autistic people who have moderate to severe learning disabilities. Participants and their carers received the adapted intervention developed in Phase 1a, in conjunction with any other intervention they were receiving. Both qualitative and quantitative research methods were used to address key components of feasibility. Setting and participants Phase 1a (intervention adaptation) The IAG included eight key stakeholders who were representatives from our PPI partners, carers and family members, people with autism and/or learning disabilities, and clinicians, along with members of the research team. Phase 1b (treatment-as-usual survey) All services for adults with autism and learning disabilities (and learning disabilities services providing support to those who also have autism); this included NHS mental health and learning disabilities services, and the independent and charitable sector, including social enterprises within the UK. The total of number of survey responses was N = 76. Phase 2 (feasibility study) This single-arm non-randomised feasibility study took place within the NHS health services in England. We aimed to recruit up to 30 participants from the following NHS services across England: Coventry and Warwickshire Partnership NHS Trust, Herefordshire and Worcestershire Health and Care NHS Trust, Leicestershire Partnership NHS Trust, Solent NHS Trust, and Mersey Care NHS Foundation Trust. A total of 28 participants were enrolled, and 18% were from a non-white background, while 53.6% had severe learning disabilities. Intervention Participants within this study received up to 12 sessions of individual and manualised behaviour therapy, each lasting between 60 and 90 minutes with support from carers. The intervention was exposure therapy and was adapted for use with this population and manualised. Assessment of feasibility of delivery and acceptability of the intervention We examined the views of participants, parents/carers, and therapists to address (1) intervention accessibility and acceptability, (2) helpful/unhelpful aspects, including barriers to change, (3) the value of our adaptations, (4) relationships with therapists, (5) acceptability of consent processes, (6) acceptability of outcome measures and (7) acceptability of randomisation within a future trial. We completed semistructured interviews with seven parents/carers and eight therapists. Interviews using Talking Mats® were also completed with five autistic adults with moderate to severe learning disabilities who received the intervention to explore their experience of the intervention and their outcomes. Collection of outcome data Phase 1a (intervention adaptation) We held five meetings with the IAG and made use of consensus methods following discussion to make decisions about intervention adaptations, the content of our fidelity checklist, and our selection of candidate outcome measures. Phase 1b Questions for the TAU survey included the type of support/intervention, who, how and where it is delivered, along with dose and any modifications adapted from the TIDieR checklist. The online survey was delivered using Qualtrics (Qualtrics, Provo, UT, USA) or as a telephone interview. This phase lasted 14 months, running concurrently with Phase 1a and Phase 2. Phase 2 Participants were enrolled in the study for approximately 6 months. Participants were assessed at three times points: (1) screening for eligibility, (2) baseline assessment within 4 weeks before commencement of the intervention and (3) within 4 weeks following the completion of the intervention. Eligibility assessments included measures of adaptive behaviour, autistic symptoms, and symptoms of anxiety disorders. Our outcome assessments included measures of anxiety, emotional and behavioural problems, behaviours that challenge, and engagement within the community. Patient and public involvement and engagement Patient and public involvement and engagement was firmly and genuinely an integral part of our methodology. We partnered with the National Autistic Society (NAS), who worked collaboratively with us to adapt our intervention, along with carers, autistic people, and clinicians. They helped prepare study documentation, recruit participants, and collaboratively disseminate information about this study. Autistic people, carers and clinicians were members of the Study Steering Group and shared oversight of our project progress. Our IAG was formed of carers, autistic people and clinicians who are experienced with working with autistic people with moderate to severe learning disabilities. Results Phase 1a We successfully adapted the intervention, developed a logic model and intervention fidelity checklist, and chose outcome measures collaboratively with the IAG. Phase 1b Seventy-eight health and social care professionals responded to our survey and 76 provided data about TAU. The majority worked for the NHS. The most frequently offered anxiety intervention was described as psychological interventions, but exposure therapy was said to be infrequently provided. Respondents also described the adaptations they made to psychological interventions for use with adults with learning disability. These adaptations included: providing communication support, adjusting session content or activities, adjusting the timing, duration, frequency and number of sessions, involving carers within the intervention, or making changes to a person’s environment. The next most offered intervention was medication. Respondents also described less frequently offered interventions including speech and language therapy, occupational therapy, sensory strategies, communication training, increased staff support, touch therapy and music therapy. Four respondents indicated that physical health support was also offered as an intervention for anxiety. Phase 2 (1) Acceptability and feasibility of the intervention: the intervention was feasible to deliver and acceptable to autistic adults with moderate to severe learning disabilities, carers and therapists. Carers and therapists made some suggestions for revisions to the intervention which focused upon reinforcing the importance of consistent carer attendance during the intervention, and the inclusion of further guidance about ensuring that the intervention is person-centred by adapting the timing, frequency, and number of intervention sessions, making use of bespoke or personalised intervention materials, further guidance about delivering relaxation, and strengthening information about the nature and degree of adaptations that can be made by therapists. All suggestions were considered feasible to implement. (2) Appropriateness of outcome measures: percentage of missing data across the outcome measures was low and ranged from 0% to 2.38% and carers completed 100% of all outcome measures. However, carers commented that some of the items were repetitive, and some said they had difficulties answering some questions. However, others said they were acceptable and easy to complete. (3) Feasibility and acceptability of consent and associated processes: carers indicated that the study participant information sheets and associated processes were clear, helpful, and straightforward. Seventy-nine per cent of autistic participants with moderate to severe learning disabilities were judged to lack capacity to make a decision about taking part in this study. On average, it took nearly 5 weeks to seek consultee advice and enrol these participants. (4) Factors that facilitate or challenge the implementation of our intervention: autistic participants with moderate to severe learning disabilities indicated that they liked coming to the intervention sessions and the information booklets that were used during the intervention. They also indicated that they liked some core aspects of the intervention (e.g. fear ladders, relaxation, visual schedules). Carers indicated that the intervention facilitators were adaptations to meet individual needs, the experience of attending therapy and a positive therapeutic relationship with the therapist. Therapists reported that the intervention facilitators were the inclusion of a clear structure, therapist understanding of people with learning disabilities, carer understanding and engagement, effectively meeting carer need, using adapted materials, high-quality training in the intervention, access to supervision, specific aspects of the intervention (e.g. preference assessments, reinforcement, carer-only sessions, positive therapeutic relationship), and intervention flexibility. Carers reported that the key barriers were not directly attributable to the intervention (e.g. illness, holidays, or lack of staff), or were related to challenges accessing intervention material to be used during exposure, or difficulties with using the intervention with some individuals due to additional complexity, the quantity of information provided, and the time commitment required. Therapists suggested that the barriers were challenges adapting relaxation techniques, difficulties with participant engagement and motivation, different carers attending intervention sessions, difficulties with some intervention components with some individuals, quantity of information, and behaviours that challenge. Conclusions The BEAMS-ID intervention was judged to be feasible to deliver and acceptable to autistic adults with moderate to severe learning disabilities, carers and therapists. Carers and therapists made some helpful suggestions for revisions which can be easily incorporated into the existing manualised intervention with minor revision. This study took place during the COVID-19 pandemic and the recruitment of participants during Phase 2 was at a lower rate than anticipated, considering that the study was funded prior to the onset of the pandemic. Nevertheless, 93% of the planned sample size was successfully recruited, and it was noted that the accrual rate improved during the summer of 2022 and was higher within sites who joined towards the end of the study period. The participant attrition rate was low and not attributable to the intervention or study processes. This study benefited from genuine patient and public involvement and engagement during the adaptation of the intervention, development of fidelity checklist and logic model, choice of outcome measures and study management. Following minor revisions to the intervention, and further consideration of outcome measures based upon carer feedback, the BEAMS-ID intervention should be tested in a randomised trial. Protocol The study protocol is available from https://fundingawards.nihr.ac.uk/award/NIHR129804. Study registration This study is registered as ISRCTN12637590. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR129804) and is published in full in Health Technology Assessment; Vol. 28, No. 72. See the NIHR Funding and Awards website for further award information.

Background To limit the use of antimicrobials without disincentivising the development of novel antimicrobials, there is interest in establishing innovative models that fund antimicrobials based on an evaluation of their value as opposed to the volumes used. The aim of this project was to evaluate the population-level health benefit of ceftazidime-avibactam in the NHS in England, for the treatment of severe aerobic Gram-negative bacterial infections when used within its licensed indications. The results were used to inform National Institute for Health and Care Excellence guidance in support of commercial discussions regarding contract value between the manufacturer and NHS England. Methods The health benefit of ceftazidime-avibactam was first derived for a series of high-value clinical scenarios. These represented uses that were expected to have a significant impact on patients’ mortality risks and health-related quality of life. Patient-level costs and health-related quality of life of ceftazidime-avibactam under various usage scenarios compared with alternative management strategies in the high-value clinical scenarios were quantified using decision modelling. Results were reported as incremental net health effects expressed in quality-adjusted life-years, which were scaled to 20-year population in quality-adjusted life-years using infection number forecasts based on data from Public Health England. The outcomes estimated for the high-value clinical scenarios were extrapolated to other expected uses for ceftazidime-avibactam. Results The clinical effectiveness of ceftazidime-avibactam relative to its comparators was estimated by synthesising evidence on susceptibility of the pathogens of interest to the antimicrobials in a network meta-analysis. In the base case, ceftazidime-avibactam was associated with a statistically significantly higher susceptibility relative to colistin (odds ratio 7.24, 95% credible interval 2.58 to 20.94). The remainder of the treatments were associated with lower susceptibility than colistin (odds ratio < 1). The results were sensitive to the definition of resistance and the studies included in the analysis. In the base case, patient-level benefit of ceftazidime-avibactam was between 0.08 and 0.16 quality-adjusted life-years, depending on the site of infection and the usage scenario. There was a high degree of uncertainty surrounding the benefits of ceftazidime-avibactam across all subgroups, and the results were sensitive to assumptions in the meta-analysis used to estimate susceptibility. There was substantial uncertainty in the number of infections that are suitable for treatment with ceftazidime-avibactam, so population-level results are presented for a range of scenarios for the current infection numbers, the expected increases in infections over time, and rates of emergence of resistance. The population-level benefit varied substantially across the scenarios, from 531 to 2342 quality-adjusted life-years over 20 years. Conclusion This work has provided quantitative estimates of the value of ceftazidime-avibactam within its areas of expected usage within the NHS. Limitations Given existing evidence, the estimates of the value of ceftazidime-avibactam are highly uncertain. Future work Future evaluations of antimicrobials would benefit from improvements to NHS data linkages, research to support appropriate synthesis of susceptibility studies, and application of routine data and decision modelling to assess enablement value. Study registration No registration of this study was undertaken. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Policy Research Programme (NIHR award ref: NIHR135592), conducted through the Policy Research Unit in Economic Methods of Evaluation in Health and Social Care Interventions, PR-PRU-1217-20401, and is published in full in Health Technology Assessment; Vol. 28, No. 73. See the NIHR Funding and Awards website for further award information. Plain language summary This project tested new methods for estimating the value to the NHS of an antimicrobial, ceftazidime-avibactam (CAZ-AVI), so its manufacturer could be paid fairly even if very little drug is used in order to reduce the risk of bacteria becoming resistant to the product. Clinicians said that the greatest benefit of CAZ-AVI is when used for complicated urinary tract infections (cUTI) and pneumonia acquired within hospitals caused by bacteria called Enterobacterales, with a resistance mechanism called OXA-48. Because there were no relevant clinical trial data, we estimated how effective CAZ-AVI and alternative treatments were by doing a systematic literature review of studies that grew bacteria from infections in the laboratory and tested the drugs on them. We linked this to data estimating the long-term health and survival of patients. Some evidence was obtained by asking clinicians detailed questions about what they thought the effects would be based on their experience and the available evidence. We included the side effects of the alternative treatments, some of which can cause kidney damage. We estimated how many infections there would be in the UK, whether they would increase over time and how resistance to treatments may change over time. Clinicians told us that they would also use CAZ-AVI to treat intra-abdominal and bloodstream infections. We estimated how many of these infections there would be, and assumed the same health benefits as for cUTI and HAP/VAP, respectively. The total value to the NHS was calculated using these estimates. We also considered whether we had missed any additional elements of value. We estimated that the value to the NHS was £11 million to £47 million over 20 years. This reflects the maximum the NHS could pay for use of CAZ-AVI if the health lost as a result of making these payments rather than funding other NHS services is not to exceed the health benefits of using this antimicrobial. However, these estimates are uncertain due to limitations with the evidence used to produce them and assumptions that had to be made. Scientific summary Background This evaluation is one of two performed as part of a pilot conducted by the National Institute for Health and Care Excellence (NICE), NHS England (NHSE) and NHS Improvement to assess the feasibility of innovative models that pay for antimicrobials (AMs) based on an evaluation of their value to the NHS as opposed to the volumes used. These projects informed commercial discussions regarding contract value for a period of up to 10 years. This report details the evaluation phase for ceftazidime with avibactam (CAZ-AVI, brand name Avycaz), a fixed-dose combination medication composed of ceftazidime, a cephalosporin antibiotic, and avibactam, a beta-lactamase inhibitor manufactured by Pfizer which received a marketing authorisation in June 2016 for treatment in adults and paediatric patients (> 3 months) for complicated intra-abdominal infections (cIAI), complicated UTI (cUTI), hospital-acquired pneumonia, including ventilator-associated pneumonia (HAP/VAP), bacteraemia (adults only) associated with the aforementioned infections and treatment of infections caused by aerobic Gram-negative organisms with limited treatment options. Aim and objectives The aim of this evaluation is to assess the value of CAZ-AVI to the NHS in England when used within its licensed indications. Specific objectives are: To identify two high-value clinical scenarios (HVCSs) for which CAZ-AVI is expected to impact on mortality risks and improve health-related quality of life To undertake an ‘evidence mapping’ exercise and relevant systematic literature reviews to characterise the available clinical effectiveness evidence for the use of CAZ-AVI in the HVCSs. To establish an appropriate decision-analytic model to quantify the costs and health benefits of the use of CAZ-AVI under various usage scenarios compared with usage scenarios involving other available AMs in the HVCSs. The decision-analytic model was required to estimate costs and health effects at both the individual level and the aggregate population level, providing population incremental net health effects (INHEs). Drawing on the systematic reviews and evidence synthesis, national-level data on healthcare-associated infections, and other sources as needed, identify evidence to populate the decision-analytic models. To use structured expert elicitation as necessary to supplement the available evidence to populate the decision-analytic models. To use available evidence and where necessary expert opinion to quantitatively extrapolate estimated population INHEs associated with CAZ-AVI in the HVCSs to other expected uses for the product within the product’s licensed indications. Expected usage and high-value clinical scenarios To control the spread of resistance to CAZ-AVI and to preserve its long-term viability as an effective treatment option against highly resistant pathogens, CAZ-AVI was expected to be used in a narrower group of patients than permitted by its license. This evaluation characterised the value of CAZ-AVI in two steps. First, decision modelling assessed quantitatively the value of CAZ-AVI in a set of ‘high-value clinical scenarios’ (HVCSs). Secondly, rescaling was used to provide a quantitative assessment of value in the overall population expected to receive CAZ-AVI in the English NHS, including scenarios outside the HVCSs. The HVCSs were scenarios where there was currently significant burden from resistant infections, and where CAZ-AVI was expected to offer significant improvements over existing treatments in terms of efficacy and/or safety. The HVCSs focus on the following patient populations: Empiric setting (ES): infections strongly suspected to be caused by oxacillinase-48 (OXA-48) Enterobacterales in patients with hospital-acquired pneumonia or ventilator-associated pneumonia (HAP/VAP). The pathogen, resistance mechanism and antibiotic susceptibility have not yet been established but treatment is initiated immediately due to the severity of the infection. Microbiology-directed setting (MDS): patients with an infection confirmed to be caused by OXA-48 Enterobacterales, where antibiotic susceptibility testing results were available, and where the site of infection was HAP/VAP or cUTI. The resourcing for this project was equivalent to that of a diagnostic assessment review or multiple technology assessment for NICE, but the levels of analysis extend beyond the typical focus of those evaluations, to include population-level health effects now and over time, as well as across several indications and settings. The analyses are inevitably constrained by the time and resources available for the project. Clinical evidence Methods A mapping review was undertaken to establish available evidence and ascertain which approach to estimating comparative effectiveness could inform an economic model. Reviews 1 and 2 considered RCTs and observational studies in patients with HAP/VAP or cUTI infections caused by OXA-48 Enterobacterales or suspected to be caused by carbapenem-resistant Enterobacterales. In Review 3 in vitro susceptibility studies of OXA-48 Enterobacterales isolates which reported the proportion of isolates from any infection site that were susceptible to the HVCS treatment and at least one relevant comparator (colistin, meropenem, tigecycline, aztreonam, fosfomycin, gentamicin, amikacin, tobramycin) were sought. Systematic searches across relevant databases (MEDLINE, EMBASE and Centre for Review and Dissemination (CRD) database) were conducted in March 2021. EEPRU also considered evidence submitted by Pfizer and Public Health England (PHE). After mapping, only Review 3 was pursued, since there was insufficient evidence from Reviews 1 and 2 (see details in Results below). Risk-of-bias assessment was performed using a bespoke tool developed for this evaluation. A random-effects NMA of susceptibility studies was conducted. Subgroup and sensitivity analyses were planned to investigate the impact of clinical sources of heterogeneity including: inclusion criteria (use of resistance to a comparator to select study sample); co-carriage of MBLs; the proportion who were carbapenem-sensitive; whether the sample was recruited consecutively; and what laboratory methods and breakpoints were used to assess susceptibility. Review 4 aimed to provide evidence on the link between susceptibility and clinical outcomes and Review 5 between susceptibility and long-term outcomes in the sites of interest. Review 6 aimed to assess the safety of CAZ-AVI. Results The mapping reviews of RCT and observational studies identified insufficient evidence to inform an assessment of clinical effectiveness since the RCTs included very small numbers of OXA-48 Enterobacterales infections and did not report these data as a subgroup analysis and the observational data were not reported separately for the sites of interest (cUTI and HAP/VAP), the studies were small and highly heterogeneous with respect to prognostic factors, and individual patient data (IPD) were not available within the timeframe of this evaluation. In the review of susceptibility studies, 28 data sources met the initial inclusion criteria. One, from PHE, which included isolates submitted voluntarily to the antimicrobial resistance and healthcare-associated infections (AMRHAI) reference laboratory, had high relevance, but several limitations. EEPRU therefore included a broader set of evidence synthesised using network meta-analysis (NMA). Sixteen studies and data sources met the inclusion criteria. In the synthesis, CAZ-AVI was associated with a statistically significantly higher susceptibility relative to colistin [odds ratio (OR) 7.24, 95% credible interval (CrI) 2.58 to 20.94]. The remainder of the treatments were associated with lower susceptibility than colistin (OR < 1). Heterogeneity was extremely high [standard deviation (SD) 1.56, 95% CrI 1.28 to 1.93] and sensitivity analyses were conducted to investigate the sources of heterogeneity. The evidence network using EUCAST methods and breakpoints to assess susceptibility was selected as the base-case analysis to inform the economic evaluation since heterogeneity was lower and there was a clinical rationale to support restricting to studies that had used EUCAST laboratory methods and breakpoints as these are more commonly used in England. A scenario analysis was planned to include the result from the full analysis set. A further scenario was planned restricted to studies with no-MBLs and that had used EUCAST laboratory methods and breakpoints, which left one study (Vazquez-Ucha et al.). A further scenario analysis was planned using the PHE data alone, due to its high relevance to the evaluation. Review 4 identified two studies that reported mortality or hospital length of stay conditional on susceptibility to empiric treatment and were selected for use in the ES model. No useful evidence relating to the MDS was identified. Review 5 did not identify any relevant literature, but an unpublished study (CARBAR) was submitted by Shionogi during the parallel appraisal of cefiderocol. Review 6 indicated that CAZ-AVI does not appear to increase the risk of acute kidney injury (AKI), C. difficile, or any other serious adverse events (SAE), compared to non-toxic comparators (i.e. comparators that were not colistin or an aminoglycoside). No study reported a comparison of CAZ-AVI exclusively with colistin or aminoglycosides. Event rates were generally very low or zero. Discussion of clinical evidence: there were some limitations to the approach selected and analyses done. Key limitations include: susceptibility could be considered, at best, a surrogate outcome; linking data were limited and not specific to the pathogen–mechanism combination, and expert elicitation had to be relied upon to evidence the link in the MDS; uncertainties relating to the determination and application of breakpoints and the NMA results were associated with high levels of heterogeneity. Economic evidence Methods No published existing economic evaluations that assessed the use of CAZ-AVI in the HVCSs or areas of expected usage were identified. The manufacturer’s cost-effectiveness model estimated that introduction of CAZ-AVI was associated with an expected INHE benefit of approximately 20,000 quality-adjusted life-years (QALYs) over a 10-year time horizon, based on a cost-effectiveness threshold of £30,000 per QALY. EEPRU considered the model to contain several strong assumptions and it assumed a much broader population of patients would receive CAZ-AVI than considered appropriate by clinical advisors to EEPRU. Policy Research Unit in Economic Methods of Evaluation in Health and Social Care Interventions developed a de novo decision-analytic model to predict the cost and health consequences (summarised as population-level INHEs) of introducing CAZ-AVI within the HVCSs. These estimates assumed CAZ-AVI was supplied free of charge to the NHS, and use an estimate of health opportunity cost of £20,000/QALY. This quantitative analysis comprises three components assessing INHEs of the HVCSs at the patient level, INHEs of the HVCSs at the population level, and how INHEs within the HVCSs might appropriately be rescaled to reflect expected usage across the English NHS. The patient-level component characterises the cost, mortality and morbidity consequences of introducing CAZ-AVI over a patient’s lifetime. Separate but related models were developed for the ES and MDS. In the ES, treatment with CAZ-AVI was compared to non-colistin/aminoglycoside-based treatment and colistin/aminoglycoside-based treatment (considered more toxic). In the MDS, outcomes in the overall microbiology-directed cohort using CAZ-AVI were compared to outcomes using existing AMs only. In the ES, outcomes were modelled both for patients in whom OXA-48 Enterobacterales suspicion was confirmed and for those in whom this suspicion turned out to be incorrect. The probability of having the suspected pathogen/resistance mechanism was informed by national surveillance data supplied by PHE. The key driver of effectiveness was in vitro susceptibility (see evidence syntheses above). A combination of evidence from the literature and structured expert elicitation informed the model. Higher rates of AKI for colistin and aminoglycoside-based therapy (compared to other agents, including CAZ-AVI) were informed by published systematic reviews. Lifetime costs, quality of life and mortality were predicted accounting for the highly comorbid nature of the patient population and the increased risk of chronic kidney disease resulting from AKI. The population-level component used a forecast-based approach to aggregate the patient-level predictions to the population-level accounting for the size of, and growth over time in, the eligible patient population within each HVCS. This component also reflects how resistance is likely to develop to CAZ-AVI and existing AMs over time. Predictions were presented for patients initiated on treatment over the next 20 years. National surveillance data were used to estimate current numbers of patients within the HVCS and to forecast growth in patient numbers over time. A series of scenarios reflect the potential emergence of resistance to CAZ-AVI. Changes in resistance to existing AMs over time were not modelled due to the sparsity of evidence available to inform these forecasts. Predicted overall population INHEs corresponding to the expected use of CAZ-AVI in the English NHS were generated by rescaling the population INHEs from the HVCSs to reflect additional areas of expected usage (known or suspected OXA-48 Enterobacterales bloodstream and intra-abdominal infections). Rescaling was based on surveillance data on population size and expert assessments of the similarity in per-patient INHEs between the HVCSs and these additional sites of interest. The literature on the economic evaluation of AMs has described a range of elements of value associated with these products that are not relevant to evaluations of other healthcare interventions. We summarised the extent to which these elements of value were captured within the quantitative estimates and, where this has not been possible, whether they were likely to substantively modify the quantitative estimates of value presented. Results Patient-level INHEs in the HVCS The benefits of CAZ-AVI are driven by similar susceptibility but improved safety compared to colistin/aminoglycoside-based treatments, and, in the ES, by higher susceptibility than non-colistin/aminoglycoside-based treatment. The two most significant sources of uncertainty relate to the ES and are (1) the susceptibility evidence and (2) the proportion of patients in the ES who are suspected of having OXA-48 Enterobacterales who are later confirmed to have this resistant pathogen. CAZ-AVI generated INHEs of 0.16 QALYs per patient in the ES (range across scenarios 0.00–0.26) and 0.08 (0.05–0.12) in the microbiology-directed settings. Policy Research Unit in Economic Methods of Evaluation in Health and Social Care Interventions were unable to select a base case for the population-level results. Population-level results are, therefore, presented for two different approaches to estimating current infection numbers (based on different methods to classify infections from clinical specimen sites), two alternative approaches to forecasting infections over time (based on whether observed trends are assumed to persist indefinitely or not) and three different trajectories with respect to CAZ-AVI resistance emergence (1%, 5% and 10% at 20 years). Across these scenarios, population INHEs varied from 531 to 2342 QALYs. The population size estimates are subject to additional uncertainties relating to the challenges of inferring patient population size from microbiology data and in forecasting population size in the future. Estimates were generated using a number of strong assumptions about the generalisability of patient-level INHEs between settings, for example, that patient-level INHEs for bloodstream infections can be generalised to HAP/VAP infections. These assumptions were based on discussions with clinical experts. Policy Research Unit in Economic Methods of Evaluation in Health and Social Care Interventions considers that it is possible that CAZ-AVI use will facilitate additional or at least more prompt receipt of required treatments/procedures for certain groups and therefore provide additional enablement value not captured within the model estimates of INHEs. The magnitude of these enablement benefits remains highly uncertain. Conclusion Policy Research Unit in Economic Methods of Evaluation in Health and Social Care Interventions’s quantitative assessment of value of CAZ-AVI was associated with a base-case population INHE across its areas of expected usage of 531–2342 QALYs over 20 years. These values were informed by interlinked decision-analytic models informed by systematic reviews of the literature and evidence synthesis, additional national data provided by PHE, structured expert elicitation and, where necessary, assumptions informed by clinical opinion. The quantitative estimates of the value for CAZ-AVI within the areas of expected usage within the NHS indicate that the maximum amount the NHS could pay for CAZ-AVI was £11 million to £47 million if the health lost as a result of making these payments rather than funding other NHS services is not to exceed the health benefits of using CAZ-AVI. The high level of uncertainty could be addressed via further research; however, the appropriateness of amending payments to reflect revised estimates of value requires further consideration. A broader and important question is ‘what would represent the “optimal” scope of usage for CAZ-AVI?’ Further methodological and quantitative work is required to address this question. Study registration No registration of this study was undertaken. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Policy Research Programme (NIHR award ref: NIHR135592), conducted through the Policy Research Unit in Economic Methods of Evaluation in Health and Social Care Interventions, PR-PRU-1217-20401, and is published in full in Health Technology Assessment; Vol. 12, No. 73. See the NIHR Funding and Awards website for further award information.

Background Opioids kill more people than any other drug. Naloxone is an opioid antagonist which can be distributed in take-home ‘kits’ for peer administration (take-home naloxone). Aim To determine the feasibility of carrying out a definitive randomised controlled trial of take-home naloxone in emergency settings. Design We used Welsh routine data (2015–21) to test the feasibility of developing a discriminant function to identify people at high risk of fatal opioid overdose. We carried out a cluster randomised controlled trial and qualitative study to examine experiences of service users and providers. We assessed feasibility of intervention and trial methods against predetermined progression criteria related to: site sign-up, staff trained, identification of eligible patients, proportion given kits, identification of people who died of opioid poisoning, data linkage and retrieval of outcomes. Setting This study was carried out in the emergency environment; sites comprised an emergency department and associated ambulance service catchment area. Participants At intervention sites, we invited emergency department clinicians and paramedics to participate. We recruited adult patients who arrived at the emergency department or were attended to by ambulance paramedics for a problem related to opioid use with capacity to consent to receiving the take-home naloxone and related training. Interventions Usual care comprised basic life support plus naloxone by paramedics or emergency department staff. The take-home naloxone intervention was offered in addition to usual care, with guidance for recipients on basic life support, the importance of calling the emergency services, duration of effect, safety and legality of naloxone administration. Results Discriminant function With low numbers of opioid-related deaths (1105/3,227,396) and a high proportion having no contact with health services in the year before death, the predictive link between death and opioid-related healthcare events was weak. Logistic regression models indicated we would need to monitor one-third of the population to capture 75% of the decedents from opioid overdose in 1-year follow-up. Randomised controlled trial Four sites participated in the trial and 299 of 687 (44%) eligible clinical staff were trained. Sixty take-home naloxone kits were supplied to patients during 1-year recruitment. Eligible patients were not offered take-home naloxone kits 164 times: ‘forgot’ (n = 136); ‘too busy’ (n = 15); suspected intentional overdose (n = 3). Qualitative interviews Service users had high levels of knowledge about take-home naloxone. They were supportive of the intervention but noted concerns about opioid withdrawal and resistance to attending hospital for an overdose. Service providers were positive about the intervention but reported barriers including difficulty with consenting and training high-risk opioid users. Health economics We were able to calculate costs to train staff at three sites (£40 per AS and £17 in Site 1 ED). No adverse events were reported. Progression criteria were not met – fewer than 50% of eligible staff were trained, fewer than 50% of eligible patients received the intervention and outcomes were not retrieved within reasonable timescales. Future work The take-home naloxone intervention needs to be developed and evaluated in emergency care settings, with appropriate methods. Limitations The Take-home naloxone Intervention Multicentre Emergency setting study was interrupted by coronavirus disease. Conclusions This study did not meet progression criteria for intervention or trial methods feasibility, so outcomes were not followed up and a fully powered trial is not planned. Trial registration This trial is registered as ISRCTN13232859. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 16/91/04) and is published in full in Health Technology Assessment; Vol. 28, No. 74. See the NIHR Funding and Awards website for further award information. Plain language summary Emergency ambulance staff and doctors in the emergency department regularly administer naloxone to people who have overdosed, reversing effects of the opioid – so-called ‘take-home naloxone’ – to administer to others in an emergency. We don’t know whether take-home naloxone saves lives. We carried out this feasibility study to see whether: we could identify a high-risk population to include in a trial to determine whether take-home naloxone reduces deaths from overdose ambulance paramedics and emergency department staff could be trained, and would then give out take-home naloxone kits to drug users they see. We included four areas in the study. We randomly selected two for distribution of take-home naloxone to patients at risk of or following an overdose by emergency department clinicians and local ambulance paramedics. We attempted to identify people at high risk of death from opioid overdose. We collected data about patients that were eligible and received take-home naloxone in the two intervention areas. We carried out interviews to find out about the views of patients and staff who gave out the kits. At the start, we agreed criteria that should be met for us to recommend that a full evaluation be carried out. We could not reliably identify people at high risk of death from opioid overdose, as deaths were rare in the population, and previous health service usage was variable. Less than half of eligible staff were trained to supply take-home naloxone (299/687, 44%), and less than half of eligible patients were given take-home naloxone (60/277, 21.7%) over 1 year. Patients were not offered take-home naloxone because staff forgot, were too busy or suspected an intentional overdose. Other patients left before receiving a kit. Service users liked the idea of take-home naloxone kits being provided in the emergency setting but reported resistance to attending hospital following an overdose. Service providers were supportive of providing take-home naloxone in the emergency setting but reported challenges in consent and training. Conclusion This study found that it was not feasible to deliver or evaluate this form of take-home naloxone, using this study design, in emergency care. Scientific summary Parts of this summary have been reproduced from Jones M, Bell F, Benger J, Black S, Buykx P, Dixon S, et al. Protocol for Take-home naloxone In Multicentre Emergency (TIME) settings: feasibility study. Pilot Feasib Stud 2020;6(1):1–10. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text. Background Opioids, such as heroin, kill more people worldwide by overdose than any other type of drug, and death rates associated with opioid poisoning in the United Kingdom (UK) are at record levels. Naloxone is an opioid antagonist which can be distributed in ‘kits’ for administration by witnesses in an overdose emergency. This intervention is known as take-home naloxone (THN). We know that THN can save lives on an individual level, but there is currently limited evidence about the effectiveness of THN distribution on an aggregate level, in specialist drug service settings or in emergency service (ES) settings. Notably, we do not know whether THN kits reduce deaths from opioid overdose in at-risk populations, if there are unforeseen harms associated with THN distribution or if THN is cost-effective. To address this research gap, we aimed to determine the feasibility of a fully powered cluster randomised controlled trial (RCT) of THN distribution in emergency settings. Aim To determine the feasibility of carrying out a definitive RCT of THN in emergency settings. Objectives To determine: the best form of THN kit, training and delivery whether a trial clustered by emergency department (ED) catchment area and the associated ambulance service (AS) is deliverable, as assessed against predefined progression criteria related to intervention, trial design and methods. Design We assessed feasibility of intervention and trial methods based upon the following predetermined progression criteria. Intervention feasibility Sign-up of four sites, including ≥ 50% eligible staff to complete training in delivering the intervention at each intervention site. Identification of ≥ 75% of people who have presented to ED or AS with opioid overdose or an opioid use-related problem over a 12-month period. THN kits issued to ≥ 50% eligible patients over a 12-month period at intervention sites. Serious adverse event rate [to be defined in agreement with Data Monitoring and Ethics Committee (DMEC)] of no more than 10% difference in intervention sites to control sites at the conclusion of recruitment. Trial methods feasibility Identification and inclusion for follow-up of ≥ 75% of people who died of opioid poisoning in the following year in the study areas according to Office of National Statistics (ONS) mortality data (previous ONS data suggest between 140 and 180 such deaths across the 4 participating sites during the study period). Matching and data linkage in ≥ 90% of cases not dissented at the conclusion of quantitative data collection. Retrieval of primary and secondary outcomes from National Health Service (NHS) Digital and National Welsh Informatics Service within 6 months of projected timeline. As the intervention tested is for administration to recipients of the THN kits and peers who may suffer an overdose, we needed to find a way to identify cohorts to include in outcome comparisons. We therefore analysed Welsh routine data to test the feasibility of developing a discriminant function to identify a high-risk population for fatal opioid overdose. We scoped anonymised routine retrospective data from 1 January 2015 to 30 November 2021, sourced from the Welsh Demographic Service Dataset (WDSD) to define the study population. To categorise death associated with an opioid overdose, the annual district death extract (ADDE) dataset was used in conjunction with the WDSD to calculate an individual-level study end date. Finally, we considered critical care, ED and hospital admissions as well as substance misuse treatment for the 36 months up to the end of the study period. We carried out a RCT clustered by site in the emergency environment with a qualitative study to examine processes of implementation, patient safety, costs of training NHS staff and experiences of service users and providers. Two intervention sites (paired ED and local AS catchment area) were randomly selected from the four participating sites. Usual practice was continued in the other two sites, acting as controls. Alongside the RCT, we collected qualitative data via semistructured interviews with service users from substance use treatment centres and third-sector organisations. The interview questions were guided by literature around opioid overdose experience and emergency naloxone use, with the aim to explore how opioid users interact with the knowledge, behaviour and attitudes towards the use of THN kits and training to use the kits. Focus groups and interviews with service providers (paramedics and ED clinical staff) were conducted to discuss barriers in the provision of THN in the emergency setting as well as facilitators to this implementation. We assessed the feasibility of collecting costs associated with THN provision in the emergency setting by measuring the health service contacts and incorporating healthcare resource groups (HRGs) into the analysis to produce an overall cost. Setting This feasibility study was carried out in the emergency care environment, across study sites each centred on a receiving ED and defined geographically as the local AS catchment area for that receiving ED. Participants At intervention sites, we invited ED clinicians and paramedics to participate in the trial and recruited adult patients who arrived at the ED or were attended by ambulance paramedics for a problem related to opioid use with capacity to consent to receiving the THN and related training. Participants were to be identified for outcome comparison by application of the discriminant function, if completed, to the study site general populations. Interventions Usual care comprised administration of basic life support plus naloxone by paramedics or ED staff. The THN intervention was offered in addition to usual care and included a multi dose THN kit (Prenoxad) containing 2 mg naloxone hydrochloride 1 mg/1 ml solution for intramuscular (IM) injection, and instructions on the correct administration of the naloxone dose. Recipients also received guidance on: BLS; the importance of calling the ES; duration of effect; the safety of naloxone in terms of adverse events and overdose; and the legality of bystander administration of naloxone. Results TABLE S1Assessment against preset progression criteria Criteria Achieved Criteria met Sign-up of four sites, including ≥ 50% eligible staff to complete training in delivering the intervention at each intervention site Site 1: ED trained 81.1%, AS trained 54% of eligible staffSite 2: ED trained 8.1%, AS trained 33.8% of eligible staff No Identification of ≥ 75% of people who have presented to ED or AS with opioid overdose or an opioid use-related problem over a 12-month period Unable to assess Not known THN kits issued to ≥ 50% eligible patients over a 12-month period at intervention sites 21.7% of eligible patients were given kits No Serious adverse event rate (to be defined in agreement with the DMEC) of no more than 10% difference in intervention sites to control sites at the conclusion of recruitment No serious adverse events were reported Yes Identification and inclusion for follow-up of ≥ 75% of people who died of opioid poisoning in the following year in the study areas according to ONS mortality data (previous ONS data suggest between 140 and 180 such deaths across the four participating sites during the study period) We were able to identify decedents from opioid poisoning in Wales but were unable to produce a discriminant function which included this group in a sufficiently small section of the general population, or to test these methods in a second population No Matching and data linkage in ≥ 90% of cases not dissented at the conclusion of quantitative data collection Due to significant delays in permissions processes for routine- linked data retrieval from NHS Digital, and low administration of THN kits, we did not attempt to match and link records for patients recruited to the trial No Retrieval of routinely recorded primary and secondary outcomes from national repositories within six-months of projected timeline Again, due to significant delays in permissions processes for routine- linked data retrieval, and low administration of THN kits, we did not attempt to retrieve routinely recorded primary and secondary outcomes No Discriminant function With low numbers of opioid-related deaths (1105/3,227,396) and a high proportion of them having no contact with health services in the year before death, the predictive link between death and opioid-related healthcare events was weak. Logistic regression models indicated we would need to monitor one-third of the population to capture 75% of the decedents from opioid overdose in 1-year follow-up. RCT In total, 299 of 687 (43.5%) eligible staff were trained to supply THN kits to eligible patients at the two sites (Site 1: ED n = 107, AS n = 121; Site 2: ED n = 25, AS n = 46). Sixty THN kits were supplied to eligible patients during the recruitment period (Site 1: ED n = 36, AS n = 4; Site 2: ED n = 16, AS n = 4). Eligible patients were recorded as not being offered THN kits 164 times, with reasons reported for not offering eligible patients kits: staff forgot (n = 136); staff too busy (n = 15); and suspected intentional overdose (n = 3). Staff recorded 626 people as being considered for inclusion but found not to be eligible, with reasons listed as: uncooperative including being abusive towards staff (n = 55); lack of capacity (n = 35); reduced consciousness level (n = 41); patient in custody (n = 21); and patient absconded (n = 161). Qualitative interviews Service users had high levels of knowledge about THN, with variable previous access to kits. They generally supported the provision of THN kits and training in the emergency setting and felt that it should be expanded further to chemists and needle exchanges. They also noted the importance of including loved ones in training and felt that this gave them a sense of empowerment and motivation to help others in an overdose situation. They noted concerns with regards to opioid withdrawal and resistance to attending hospital for an overdose. The service users reported that the provision of THN kits and training to friends and family of opioid users would possibly be more beneficial and believed that incorporating THN provision into normal practice would help mitigate some of these barriers. Interviews and focus groups with service providers found that they were supportive about the provision of THN kits and training in the emergency setting. However, they also reported barriers including difficulties consenting and training opioid users, a high turnover of staff impacting the cascade of the intervention as well as negative attitudes towards the patient group and the coronavirus disease 2019 (COVID-19) pandemic. No adverse events were reported. Conclusion This study did not meet progression criteria for intervention or trial methods feasibility, so outcomes were not followed up and a fully powered trial is not planned. There does appear to be appetite for THN kit provision and training in the emergency setting. We conclude that the THN intervention as defined and administered in the Take-home naloxone Intervention Multicentre Emergency setting (TIME) study was not feasible and should not therefore go forward to full trial. However, there may be space for further development of this complex intervention in emergency care – for example, for protocols to allow administration to family and friends of opioid users; as well as methods for definition and identification of study cohorts for outcome comparisons. Trial registration This trial is registered as ISRCTN13232859. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 16/91/04) and is published in full in Health Technology Assessment; Vol. 28, No. 74. See the NIHR Funding and Awards website for further award information.

Background Neonates with suspected sepsis are commonly treated with gentamicin, an aminoglycoside. These antibiotics are associated with high risk of ototoxicity, including profound bilateral deafness, in people with the m.1555A>G mitochondrial genetic variant. Objective This early value assessment summarised and critically assessed the clinical effectiveness and cost-effectiveness of the Genedrive MT-RNR1 ID Kit for identifying the gene m.1555A>G variant in neonates and mothers of neonates needing antibiotics or anticipated to need antibiotics. Following feedback from the scoping workshop and specialist assessment subgroup meeting, we also considered the Genedrive MT-RNR1 ID Kit for identifying the m.1555A>G variant in mothers prior to giving birth. Data sources For clinical effectiveness, we searched three major databases in October 2022: MEDLINE, EMBASE and CINAHL (Cumulative Index to Nursing and Allied Health Literature). For cost-effectiveness, in addition to the three mentioned databases we searched Cochrane and RePEc-IDEAS. Study selection Study selection and risk-of-bias assessment were conducted by two independent reviewers (Ryan PW Kenny and Akvile Stoniute for clinical effectiveness and Hosein Shabaninejad and Tomos Robinson for cost-effectiveness). Any differences were resolved through discussion, or by a third reviewer (Nick Meader). Study appraisal Risk of bias was assessed using Quality Assessment of Diagnostic Accuracy Studies-2. One study (n = 751 neonates recruited) was included in the clinical effectiveness review and no studies were included in the cost-effectiveness review. All except one outcome (test failure rate: low risk of bias) were rated as being at moderate risk of bias. The study reported accuracy of the test (sensitivity 100%, 95% confidence interval 29.2% to 100%; specificity 99.2%, 95% confidence interval 98% to 99.7%), number of neonates successfully tested (n = 424/526 admissions), test failure rate (17.1%, although this was reduced to 5.7%), impact on antibiotic use (all those with a m.1555A>G genotype avoided aminoglycosides), time taken to obtain a sample (6 minutes), time to genotyping (26 minutes), time to antibiotic treatment (55.18 minutes) and the number of neonates with m.1555A>G (n = 3). Limitations The economic component of this work identified key evidence gaps for which further data are required before a robust economic evaluation can be conducted. These include the sensitivity of the Genedrive MT-RNR1 ID Kit for identifying the gene m.1555A>G variant in neonates, the magnitude of risk for aminoglycoside-induced hearing loss in neonates with m.1555A>G, and the prevalence of the m.1555A>G variant. Other potentially important gaps include how data regarding maternal inheritance may potentially be used in the clinical pathway. Conclusions This early value assessment suggests that the Genedrive MT-RNR1 ID Kit has the potential to identify the m.1555A>G variant and to be cost-effective. The Genedrive MT-RNR1 ID Kit dominates the current standard of care over the lifetime, as it is less costly and more effective. For a 50-year time horizon, the Genedrive MT-RNR1 ID Kit was also the dominant strategy. For a 10-year time horizon, the incremental cost-effectiveness ratio was estimated to be £103 per quality-adjusted life-year gained. Nevertheless, as anticipated, there is insufficient evidence to conduct a full diagnostic assessment of the clinical effectiveness and cost-effectiveness of the Genedrive MT-RNR1 ID Kit in neonates directly or in their mothers. This report includes a list of research priorities to reduce the uncertainty around this early value assessment and to provide the additional data needed to inform a full diagnostic assessment, including cost-effectiveness modelling. Study registration This study is registered as PROSPERO (CRD42022364770). Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135636) and is published in full in Health Technology Assessment; Vol. 28, No. 75. See the NIHR Funding and Awards website for further award information. Plain language summary Our immune system usually fights off invading germs, such as bacteria, viruses, fungi or parasites, in order to prevent infection. Sometimes the immune system stops fighting the ‘invaders’ and begins to turn in on itself. This life-threatening reaction is known as sepsis. Bacterial infections and sepsis are significant causes of death and illness in newborns. Newborns with suspected bacterial infection or sepsis are normally treated with an aminoglycoside antibiotic called gentamicin (a type of medicine that is meant to kill bacteria). These antibiotics are associated with a very high risk of ototoxicity (damage to the ear, including deafness) among people with the m.1555A>G MT-RNR1 gene variant [a specific change to the small section of deoxyribonucleic acid storing biological information] within their mitochondrial deoxyribonucleic acid (small circles of deoxyribonucleic acid located in the mitochondria, the cell’s energy producer). The aim of this review was to summarise and critically evaluate existing evidence on how effective (the degree to which a test does more harm than good) and cost-effective (how effective a test is in relation to its cost) the Genedrive MT-RNR1 ID Kit is for identifying the m.1555A>G gene variant in newborns or in their mothers. We collected and analysed all relevant research studies, one moderate quality study was included in the clinical effectiveness review and no studies were included in the cost-effectiveness review. The quality of the included study was assessed as moderate for most of the outcomes (things measured to monitor the degree to which the test does more good than harm) reported due to uncertainty regarding the failure rate of the test. The results suggested that the test was capable of identifying newborns with the m.1555A>G variant. This was accomplished by successfully testing 424 out of 526 patients, with three newborns identified as carrying the gene and avoiding aminoglycoside treatment. Because of these small numbers, there does remain some uncertainty regarding the accuracy of the test. Additionally, time to antibiotic administration was not negatively impacted by the test. Similar time for treatment initiation was taken for those tested as for those not tested. This review shows that the Genedrive MT-RNR1 ID Kit has the potential to identify the m.1555A>G variant and the potential to provide value for money for the National Health Service. However, as expected, there is not enough evidence to conduct a full assessment of the clinical effectiveness and cost-effectiveness of Genedrive MT-RNR1 ID Kit in newborns directly, or their mothers. This could be addressed by generating further evidence. The risk and severity of hearing loss from aminoglycoside use is of particular interest, as is further testing of the Genedrive MT-RNR1 ID Kit in both neonates and mothers or neonates who need treatment. Such testing conducted in other settings would be of great importance. Scientific summary Background Sepsis and bacterial infections are a significant cause of mortality and morbidity in neonates (up to and including a corrected gestational age of 28 days). Expert opinion suggests that the incidence of culture-confirmed neonatal infection is around 1 in 2000 deliveries. But a larger proportion of babies will go on to receive precautionary antibiotic treatment for suspected infection [e.g. 30–60 in 1000 for those admitted to neonatal intensive care units (NICUs)]. Treatment for suspected infection or sepsis is commonly conducted using gentamicin, an antibiotic of the aminoglycoside family. This antibiotic is associated with a high risk of ototoxicity in those with a genetic variation of the mitochondrial MT-RNR1 gene, specifically m.1555A>G. The purpose of this assessment was to investigate the use of the Genedrive MT-RNR1 ID Kit for identifying the m.1555A>G variant in neonates with suspected infection or sepsis. This technology has the potential to identify those at most risk of ototoxicity from aminoglycoside antibiotics and inform treatment decisions within the time frame recommended by National Institute for Health and Care Excellence (NICE) guidance. Aim The overall aim of this early value assessment was to summarise and critically appraise existing evidence on the clinical effectiveness and cost-effectiveness of the Genedrive MT-RNR1 ID Kit for identifying the m.1555A>G gene variant in neonates or their mothers. Methods A rapid review methodology was used to identify eligible studies for clinical effectiveness and cost-effectiveness. Databases searches were conducted on MEDLINE, EMBASE and CINAHL (Cumulative Index to Nursing and Allied Health Literature) for both aspects of the review; additionally, the cost-effectiveness review searched Cochrane and RePEc-IDEAS from 2010 to November 2022. Search results were screened by two independent reviewers. Only one study met the inclusion criteria for the clinical effectiveness rapid review, and no studies met the eligibility criteria for the cost-effectiveness rapid review. Data extraction and quality appraisal of the clinical effectiveness study were completed by one reviewer and checked for accuracy by another. Quality appraisal was conducted per outcome, the QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies-2) tool was used to assess diagnostic test accuracy outcomes, and the ROBINS-I (Risk Of Bias In Non-randomized Studies – of Interventions) tool was used for all other outcomes. Meta-analyses were not possible as only one study was included in the clinical effectiveness rapid review. Care pathways with and without the use of the Genedrive MT-RNR1 ID Kit were developed and from these a conceptual economic evaluation model was developed. This was used to identify the information required to parameterise the model. Attempts were then made to identify relevant parameter values and evidence gaps where no or few data were identified. Using available information, an early health economic model was developed to provide initial estimates of the incremental cost per quality-adjusted life-year (QALY) gained for the comparison of the use of Genedrive MT-RNR1 ID Kit with current standard care. Results The evidence to inform this early value assessment was extremely limited. Only one study was included in the clinical effectiveness rapid review, for which risk of bias was rated as being moderate for most of the outcomes measured. The included study suggested high diagnostic test accuracy (sensitivity 100%, specificity 99.2%). Estimates of sensitivity were very uncertain due to a small number of true-positive cases (i.e. people with the m.1555A>G variant), but no false negatives were identified. However, there were some false positives (n = 5 of 8), and the specificity estimate was very high with sufficient precision. This was established from 424 successful tests, with a test failure rate of 17.1% (90 patients). The failure rate was reduced to 5.1% in repeated testing of samples after modifications were made to the assay buffer and the test cartridge was redesigned. Overall, three neonates were identified with the genetic variant. The trial research team were able to genotype the m.1555A>G variant using the Genedrive MT-RNR1 ID Kit in 26 minutes. Time to antibiotics when using the Genedrive MT-RNR1 ID Kit did not differ from normal practice (i.e. not using the test kit). Difference between groups was not statistically significant (mean difference –0.87 minutes, 95% confidence interval –5.96 to 4.23 minutes) and the 95% confidence interval was within the predefined boundary for statistical equivalence. We did not identify any studies that reported on the following intermediate, clinical or patient-related outcomes: impact of test implementation and use on healthcare resources, usability of the test, mortality and morbidity. Additionally, no studies assessed the use of the point-of-care test in mothers. No relevant economic evaluations were identified. From the conceptual economic model, key evidence gaps were identified. These include the sensitivity of the Genedrive MT-RNR1 ID Kit for identifying the m.1555A>G gene variant in neonates, the magnitude of risk for aminoglycoside-induced hearing loss (AIHL) in neonates and mothers with m.1555A>G, and the prevalence of the gene m.1555A>G variant. Other potential important gaps include how data regarding maternal inheritance may potentially be used in the clinical pathway. The early health economic model focused on some of those parameters where, on consideration of the available data, the estimates of cost-effectiveness would be most sensitive to changes. The results of this model showed that the use of the Genedrive MT-RNR1 ID Kit for identification of the m.1555A>G genetic variant could potentially be cost-effective, with lower costs (£58.48) and higher effectiveness in terms of QALYs (0.01) over the patient lifetime. In a deterministic sensitivity analysis, the results were shown to be most sensitive to changes in the time horizon, the sensitivity of the Genedrive MT-RNR1 ID Kit system, the proportion of neonates with m.1555A>G variant suffering from AIHL after being exposed to aminoglycosides and the prevalence of the m.1555A>G variant in the UK population. Conclusions There is limited evidence for the assessment of the Genedrive MT-RNR1 ID Kit for identification of the m.1555A>G genetic variant. The test was conducted in two large NICUs and thus may not be generalisable to smaller NICUs or other hospitals. Therefore, the use of the Genedrive MT-RNR1 Kit should be investigated further in varying settings. Furthermore, although modifications were made to the kit to reduce its failure rate, when it was used in the clinical setting this was not completely eradicated. However, there is evidence to suggest that the use of the kit did not substantially impact on time to antibiotics and has the potential to identify the m.1555A>G variant. There were no existing economic evaluations that addressed this topic. The total cost per test to the NHS was estimated to be £130; however, there is uncertainty surrounding this estimate given that this cost is likely to vary by size and type of site. The results of the early economic evaluation model suggest that the use of the Genedrive MT-RNR1 ID Kit to identify the m.1555A>G genetic variant could potentially be cost-effective. Once evidence regarding the reported evidence gaps has been identified, a full diagnostic assessment should be undertaken to establish the cost-effectiveness of the Genedrive MT-RNR1 ID Kit. Suggested priorities for further research This report identifies two key priorities for research required to reduce the uncertainty around this early value assessment and to provide the additional data needed to inform a full diagnostic assessment, including cost-effectiveness modelling. The risk and the severity of AIHL in neonates with the m.1555A>G variant was identified as key uncertainties in the economic model. Limitations of the current literature, which is primarily based on case–control studies in hearing-impaired populations with the m.1555A>G variant, are provided in more detail below. Future studies, perhaps including existing cohorts in the UK, are required to identify sufficient numbers of people with the m.1555A>G variant who have been exposed to aminoglycosides in a sample that includes participants with and participants without hearing impairment. A second priority for research is further validation of the Genedrive MT-RNR1 ID Kit in both neonates and mothers of neonates who need or may need aminoglycoside treatment. The sensitivity of the test was an important uncertainty in the economic model. Further studies including more people with the m.1555A>G variant will increase the precision of the estimated sensitivity of the test. In addition, only the pharmacogenetics to avoid loss of hearing (PALOH) study has investigated the validity of the Genedrive MT-RNR1 ID Kit. This study was conducted in two large NICUs, and further research is needed to assess if the findings of the PALOH study generalise to smaller NICUs and other relevant hospital settings. In addition, our focus group with parents and a review of parents’ comments on internet forums identified that further work may be required to obtain informed consent. A final area for further research is to provide updated and more comprehensive estimates of health state utility values. Data that are currently available are restricted in terms of health states considered or use health-related quality-of-life tools whose relevance to UK decision-makers may be limited. Study registration This study is registered as PROSPERO (CRD42022364770). Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135636) and is published in full in Health Technology Assessment; Vol. 28, No. 75. See the NIHR Funding and Awards website for further award information.

Background Urinary tract infection is the most diagnosed infection in older people. It accounts for more than 50% of antibiotic prescriptions in care homes and is a frequent reason for care home residents being hospitalised. Objective This realist review developed and refined programme theories for preventing and recognising urinary tract infection, exploring what works, for whom and in what circumstances. Design The review used realist synthesis to explore existing literature on the detection and prevention of urinary tract infection, complemented by stakeholder consultation. It applies to the UK context, although other healthcare systems may identify synergies in our findings. Data sources Bibliographic databases searched included MEDLINE, CINAHL, EMBASE, Cochrane Library, Web of Science Core Collection (including the Social Sciences Citation Index), Sociological Abstracts, Bibliomap and National Institute for Health and Care Research Journals Library. Data selection and extraction Title and abstract screening were undertaken by two researchers independently of each other. Selection and assessment were based on relevance and rigour and cross-checked by a second researcher. Data extracted from the included studies were explored for explanations about how the interventions were considered to work (or not). Evidence tables were constructed to enable identification of patterns across studies that offered insight about the features of successful interventions. Data analysis and synthesis Programme theories were constructed through a four-stage process involving scoping workshops, examination of relevant extant theory, analysis and synthesis of primary research, teacher–learner interviews and a cross-system stakeholder event. A process of abductive and retroductive reasoning was used to construct context–mechanism–outcome configurations to inform programme theory. Results The scoping review and stakeholder engagement identified three theory areas that address the prevention and recognition of urinary tract infection and show what is needed to implement best practice. Nine context–mechanism–outcome configurations provided an explanation of how interventions to prevent and recognise urinary tract infection might work in care homes. These were (1) recognition of urinary tract infection is informed by skills in clinical reasoning, (2) decision-support tools enable a whole care team approach to communication, (3) active monitoring is recognised as a legitimate care routine, (4) hydration is recognised as a care priority for all residents, (5) systems are in place to drive action that helps residents to drink more, (6) good infection prevention practice is applied to indwelling urinary catheters, (7) proactive strategies are in place to prevent recurrent urinary tract infection, (8) care home leadership and culture fosters safe fundamental care and (9) developing knowledgeable care teams. Limitations We adapted our approach and work to online interactions with stakeholders and as a research team because of COVID-19. This also had an impact on bringing stakeholders together at a face-to-face event at the end of the project. Studies focusing on the prevention of urinary tract infection in care home settings were predominantly from the USA and Europe where the regulatory and funding systems for the long-term care of the elderly have some differences, particularly in the USA where national reporting plays a significant role in driving improvements in care. Conclusions Care home staff have a vital role in the prevention and recognition of urinary tract infection, which can be enabled through integration and prioritisation within the systems and routines of care homes and delivery of person-centred care. Promoting fundamental care as a means of facilitating a holistic approach to prevention and recognition of urinary tract infection helps staff to recognise how they can contribute to antimicrobial stewardship and recognition of sepsis. Challenging assumptions made by staff about the presentation of urinary tract infection is complex and requires education that facilitates ‘unlearning’ and questioning of low-value practices. Programmes to prevent urinary tract infection need to be co-designed and supported through active and visible leadership by care home managers with support from specialist practitioners. Future work We will focus on co-designing tools that facilitate implementation of our findings to ensure they fit with the care home context and address some of the challenges faced by care home leaders. This will underpin action at care home and system levels. Further research is needed to better understand the perspectives of residents and family carers, the effectiveness of non-pharmacological, pharmacological and specialist practitioner interventions and non-traditional approaches to training and educating the workforce in care home settings. Study registration This study is registered as PROSPERO CRD42020201782. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR130396) and is published in full in Health Technology Assessment; Vol. 28, No. 68. See the NIHR Funding and Awards website for further award information. Plain language summary This study was about how interventions to prevent and recognise urinary tract infections might work in care homes to reduce urinary tract infection in older people. We used an approach called realist synthesis. This aims to understand everything that influences how care is delivered to find out what works in particular situations and settings. We did this by talking to care home staff, residents, family carers and other experts about how to reduce urinary tract infection in older people living in care homes. We combined their experiences and ideas with the evidence from the research literature. From this, we developed three areas of focus: care approaches to support accurate recognition of urinary tract infection care approaches to prevent urinary tract infection/catheter-associated urinary tract infection making best practice happen. Our research shows that care staff are best placed to recognise subtle changes in a resident’s behaviour or well-being which might indicate an infection. There are several things care staff can do to proactively help residents from developing a urinary tract infection such as prioritising residents’ hydration needs and addressing poor fluid intake. Putting in place infection prevention measures such as caring for, or where possible, removing a urinary catheter can also help. Actively monitoring residents to determine reasons for changes in behaviour can increase the focus on preventative activities and help avoid inappropriate treatment. The detection and prevention of urinary tract infection in older people could be improved in several ways, including: education tailored to the roles and work of care home staff leaders of care homes providing a culture of safety and improvement where urinary tract infection is recognised as something preventable rather than unavoidable having a safe and supportive environment where staff are confident to communicate their concerns all care team members, residents and their families and other professionals linked with the home having a common language and shared goals in the recognition, prevention and diagnosis of urinary tract infection. The COVID-19 pandemic required us to adapt our approach and work mainly online both in our interactions with stakeholders and as a research team. While this provided opportunities to extend our reach to a wider group of stakeholders, it also limited some of our engagement work. Scientific summary Background Urinary tract infection (UTI) is the most diagnosed infection in older people, accounting for more than 50% of antibiotic prescriptions in long-term care settings. It is a frequent reason for hospitalisation from care homes. Guidance about strategies for UTI prevention and recognition in care homes is limited and focuses on indwelling urinary catheters and catheter-associated urinary tract infections (CAUTIs). It does not account for the varying contexts in which care is delivered, the challenges presented by residents with complex health needs or the demands of care delivery by unqualified care staff with limited supervision from registered nurses. As little is known about the practicality of implementing interventions to prevent UTI in care homes in the UK, this review aimed to create an evidence-informed theoretical explanation of which strategies are effective (or not) in the prevention and recognition of UTI and CAUTI in older people in care homes. Objectives To identify which interventions could be effective, the mechanisms by which these strategies work (or why they fail), for whom and under what circumstances. To understand what needs to be in place for the implementation of programmes to support the prevention of UTI and its recurrence in older people with and without a urinary catheter living in care homes in the UK. Through identifying the active components of complex interventions designed to prevent UTI in older people living in care homes, we sought to help guide the development and successful delivery of future programmes. More specifically, we planned to produce actionable recommendations to underpin the prevention and recognition of UTI in older people of relevance to UK care homes. Review methods We used realist synthesis as its explanatory focus and emphasis on understanding how complex interventions work in context fitted with our aim and objectives. Multiple types of evidence were sought to construct programme theory through a four-stage process that involved a high degree of stakeholder engagement throughout, including scoping workshops with care home staff, residents and carers, teacher–learner interviews and a cross-system stakeholder event. Context–mechanism–outcome configurations (CMOcs) were developed to test initial programme theory propositions and explain how they might or might not work and in which contexts and circumstances. This provided a clear account of the mechanisms of action that were considered essential to understanding how interventions to prevent UTI can and should be delivered in UK care homes. Evidence sources The review drew on evidence from health and social care, including primary research relating to UTI prevention and recognition in older people in care homes and improvement project reports. In September 2020, during stage 1 of the research, a generic topic-based multipurpose search was undertaken and focused on evidence that directly addressed the prevention and recognition of UTI in older people in long-term care facilities. Bibliographic databases searched were MEDLINE, CINAHL, EMBASE, Cochrane Library, Web of Science Core Collection (including the Social Sciences Citation Index), Sociological Abstracts, Bibliomap and National Institute for Health and Care Research (NIHR) Journals Library. Further searches were undertaken in October 2020 using key index studies (highly cited) to find ‘sibling’ studies/papers, for example, contemporaneous papers or studies that share a context by means of Google Scholar and ‘Publish or Perish’ software. Additional supplementary searches were undertaken in stage 2 (July–December 2021) to address gaps in evidence and inform the realist synthesis. These were informed by the tentative programme theories with the aim of refining them. A targeted approach with purposive searches of academic and grey literature focused specifically on continence and UTI, hydration and UTI, non-antimicrobial strategies for recurrent UTI, recognition of soft signs and family involvement in older people’s care in care homes. Data extraction Full-text papers were screened and reviewed to determine their relevance and rigour. Relevance was defined as the extent to which evidence contributed to theory building, testing and refinement and rigour as the extent to which methods used were credible and trustworthy. Data were extracted from included studies using a bespoke data extraction form to record explanations about how the interventions were considered to work (or not). The data were organised into evidence tables to enable comparison of findings and identification of patterns across studies that offered insight about the components of successful interventions. In stage 3, we tested and refined the programme theory through nine ‘teacher–learner’ interviews with a range of participants. A mixture of purposive and convenience sampling was used to gain perspectives from individuals reflective of the different audiences likely to be interested in acting on the findings from this review. Semistructured interview schedules, guided by the content of the CMOc, were used to ensure the interviews focused on participants’ perspectives of the theories as relevant to their role and expertise. In stage 4, further input was gained from an online stakeholder event that included participants from across the care system, commissioners and regulators of care. Results Fifty-six papers were included in the review. The scoping review and stakeholder engagement identified three theory areas that address the prevention and recognition of UTI and show what is needed to implement best practice. Nine CMOcs provided an explanation of how interventions to prevent and recognise UTI might work in care homes to reduce UTI: (1) recognition of UTI is informed by skills in clinical reasoning, (2) decision-support tools enable a whole care team approach to communication, (3) active monitoring is recognised as a legitimate care routine, (4) hydration is recognised as a care priority for all residents, (5) systems are in place to drive action that helps residents to drink, (6) good infection prevention practice is applied to indwelling urinary catheters, (7) proactive strategies are in place to prevent recurrent UTI, (8) care home leadership and culture foster safe fundamental care and (9) developing knowledgeable care teams. The prevention of UTI in older people living in care homes requires attention to fundamental aspects of their physical and psychosocial needs with clear communication across the whole care team that is trusted and valued by residents and families. This necessitates a context of care with a culture of safety and person-centredness, promoted by commissioners, regulators and providers, where leadership and resources are committed to support preventative action by knowledgeable care staff. For the prevention of UTI to be fully realised in care homes, it is crucial to attend to the hydration needs of all residents, adhere to infection prevention measures in the care of indwelling urinary catheters and increase access to specialists to identify appropriate treatment options for residents with recurrent UTI. Care staff need knowledge and skills to differentiate between UTI and other diagnoses using their knowledge of the resident and what is normal for them. Assumptions by staff about the inevitability of infection and tendencies to consider non-evidence-based signs and symptoms as being a UTI by default need to be challenged. Use of decision-support tools that enable staff to gather and convey accurate and relevant information about a resident’s condition using a shared language helps them to feel that their knowledge of the resident is valued by healthcare professionals (HCPs) and that their concerns will be listened to. Protocols for active monitoring provide the opportunity to focus on preventative measures and permit a more reflective approach to the recognition and diagnosis of UTI, reducing the potential for inappropriate treatment. Conclusion At the outset of this review, we identified that the coherence and detail of what works for providers to prevent UTI in older people living in UK care homes was lacking. What we now know is that care home staff have a vital role in the prevention and recognition of UTI, which can be enabled through improved integration and prioritisation within the systems and routines of care homes and delivery of person-centred care. Promoting fundamental care as a means of facilitating a more holistic approach to prevention and recognition of UTI helps staff to recognise how they can contribute to antimicrobial stewardship and the recognition of sepsis. Challenging assumptions about UTI presentation is complex and requires education that facilitates ‘unlearning’ and questioning of low-value practices. Programmes to prevent UTI need to be co-designed and supported through active and visible leadership by care home managers. The involvement of specialist practitioners such as community matrons, specialist pharmacists, continence advisors and infection prevention specialists may help to create a network of practitioners that provides peer support for change. Limitations The COVID-19 pandemic required us to adapt our approach and work mainly online both in our interactions with stakeholders and as a research team. While this provided opportunities to extend our reach to a wider group of stakeholders, it also limited some of our engagement work. We wanted to have greater input from residents and their families, but this was not possible in the circumstances. The move to virtual meetings also had an impact on our intention to bring a wide range of stakeholders together, including representatives of care home residents, at a face-to-face event towards the end of the project with many preferring to meet virtually. We acknowledge that these adaptations may have limited the generation of insights and discussion to inform both the review and our strategies for development and dissemination of outputs. For this reason, we recognise the importance of continuing to engage with stakeholders as part of our ongoing work, reporting this process and any outcomes. Studies focusing on the prevention of UTI and CAUTI in care home settings were predominantly from the United States of America (USA) and Europe where the regulatory and funding systems for the long-term care of older people has some differences, particularly in the USA where national reporting plays a significant role in driving improvements in care. Furthermore, care homes (also known as nursing homes) in the USA provide a range of medical services including post-acute care, rehabilitation, palliative and hospice care, as well as long-term care. The studies undertaken in the UK and Europe were primarily focused on interventions to reduce antimicrobial resistance through stewardship but had significant learning that was transferable to the prevention and recognition of UTI. Our synthesis tried to take account of these differences, but we are aware that we will not have reflected all realities. Through the review, we identified several tools and resources that were being used by staff in care homes in addition to those reflected in the literature. It became clear from our teacher–learner interviews and stakeholder event that adding another set of tools would be unhelpful and that any resource we developed would need to follow the principles of co-design to address some of the challenges faced by care home leaders. The difficulty in bringing stakeholders to achieve this was insurmountable within the period for this review. Implications for practice The review findings point towards actionable recommendations for UTI prevention and recognition in the care of older people living in care homes, which we describe in relation to organisational- and system-level actions. These centre on the need to align UTI prevention and recognition to the goals of person-centred, fundamental care and prioritise this in routine daily care to improve quality and safety. For example, understanding and targeting personal barriers to drinking more fluids, such as fears about incontinence and getting to the toilet, may assist in addressing poor fluid intake. This in turn can help to reduce falls, confusion and drowsiness, as well as UTI. Care home providers Best practice to prevent and recognise UTI in care home residents requires focusing on a set of evidence-informed actions as part of routine daily care with the involvement of the whole care team, including individual residents, their family carers and care home staff. Preventative actions include: supporting each resident’s hydration preferences and needs use of fluid intake monitoring systems that enable realistic targets and actions to be agreed for residents with poor fluid consumption accessing specialists who can support the care of residents with recurrent UTI applying infection prevention practice to the care and management of indwelling urinary catheters. Accurate recognition of UTI requires: knowledge of the individual resident and what is normal for them understanding of evidence-based signs and symptoms of UTI use of structured tools that align with existing care processes and a shared language to convey accurate and relevant information to healthcare professionals. Care home staff need opportunities to: develop knowledge and skills so they can interpret a resident’s signs and symptoms and consider possible explanations for generalised changes in their condition reflect on practice and learn from each other about how to recognise a UTI and support preventative actions. System level A system-wide approach with regulatory and inspection frameworks aligned to evidence on prevention and recognition of UTI is vital to ensure that resources and infrastructure are available to enable care home managers and their staff to prioritise this as part of person-centred care. There is a need to: integrate prevention of UTI with diagnostic and antimicrobial stewardship in the care of older people living in care homes harmonise the prevention and recognition of UTI decision and communication tools with those focused on recognising deterioration to facilitate adoption and integration in care homes use cocreative approaches to develop and implement resources and improvement initiatives that involve the whole care team, residents and family carers build a knowledgeable workforce of care home support workers and registered nurses who can deliver evidence-informed care and communicate their observations in a way that enables care to be reviewed before escalation improve access to expert practitioners and services to support the provision of personalised, multidisciplinary assessment and treatment plans for residents with recurrent UTI who have the greatest potential to benefit from effective treatment. Research recommendations Well-designed research to improve the prevention and recognition of UTI in older people living in care homes should address the following: perspectives and beliefs of residents and family carers relating to the prevention and recognition of UTI and the concept of active monitoring to avoid unnecessary treatment the effectiveness of specialist practitioners in supporting initiatives to recognise and prevent UTI, including expertise in facilitating improvement the effectiveness of preventative pharmacological and non-pharmacological interventions to manage recurrent UTI in care homes the effectiveness of non-traditional education interventions such as huddles or structured reflection to facilitate decision-making in care homes in cocreating interventions research should be explicit about the elements of an intervention that can be tailored to individual care homes and those which are important to deliver as intended. Study registration This study is registered as PROSPERO 2020 CRD42020201782. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR130396) and is published in full in Health Technology Assessment; Vol. 28, No. 68. See the NIHR Funding and Awards website for further award information.

Background Irritable bowel syndrome, characterised by abdominal pain and a change in stool form or frequency, is most often managed in primary care. When first-line therapies are ineffective, National Institute for Health and Care Excellence guidelines suggest considering low-dose tricyclic antidepressants as second-line treatment, but their effectiveness in primary care is unknown and they are infrequently prescribed by general practitioners. Objective To evaluate the clinical and cost-effectiveness of low-dose titrated amitriptyline as a second-line treatment for irritable bowel syndrome in primary care. Design A pragmatic, randomised, multicentre, two-arm, double-blind, placebo-controlled trial. A nested, qualitative study explored participant and general practitioner experiences of treatments and trial participation, and implications for wider use of amitriptyline for irritable bowel syndrome in primary care. Participants, clinicians, investigators and analysts were masked to allocation. Setting Fifty-five general practices in three regions in England (Wessex, West of England, West Yorkshire). Participants Patients aged ≥ 18 years meeting Rome IV criteria for irritable bowel syndrome with ongoing symptoms after trying first-line treatments and no contraindications to TCAs. Intervention Amitriptyline 10 mg once-daily, self-titrated by participants to a maximum of 30 mg once-daily or matched placebo for 6 months. Participants randomised 1 : 1 with most having the option to continue blinded treatment for a further 6 months. Main outcome measures The primary participant-reported outcome was the effect of amitriptyline on global irritable bowel syndrome symptoms at 6 months, measured using the irritable bowel syndrome Severity Scoring System, with a 35-point between-group difference defined as the minimum clinically important difference. The key secondary outcome was the proportion of participants reporting subjective global assessment of relief at 6 months, defined as somewhat, considerable, or complete relief of symptoms. Other secondary outcomes included: effect on global symptoms, via the irritable bowel syndrome Severity Scoring System, and subjective global assessment of relief of irritable bowel syndrome symptoms at 3 and 12 months; effect on somatic symptom-reporting at 6 months; anxiety an–d depression scores; ability to work and participate in other activities at 3, 6 and 12 months; acceptability, tolerability and adherence to trial medication. Results Four hundred and sixty-three participants were randomised to amitriptyline (232) or placebo (231). An intention-to-treat analysis of the primary outcome showed a significant difference in favour of amitriptyline for irritable bowel syndrome Severity Scoring System score between arms at 6 months [−27.0, 95% confidence interval (CI) −46.9 to −7.10; p = 0.008]. For the key secondary outcome of subjective global assessment of relief of irritable bowel syndrome symptoms, amitriptyline was superior to placebo at 6 months (odds ratio 1.78, 95% CI 1.19 to 2.66; p = 0.005). Amitriptyline was superior to placebo across a range of other irritable bowel syndrome symptom measures but had no impact on somatoform symptom-reporting, anxiety, depression, or work and social adjustment scores. Adverse event trial withdrawals were more common with amitriptyline (12.9% vs. 8.7% for placebo) but most adverse events were mild. The qualitative study thematically analysed 77 semistructured interviews with 42 participants and 16 GPs. Most participants found the self-titration process acceptable and empowering. Conclusions General practitioners should offer low-dose amitriptyline to patients with irritable bowel syndrome whose symptoms do not improve with first-line therapies. Guidance and resources should support GP–patient communication to distinguish amitriptyline for irritable bowel syndrome from use as an antidepressant and to support patients managing their own dose titration. Study registration This trial is registered as ISRCTN48075063. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 16/162/01) and is published in full in Health Technology Assessment Vol. 28, No. 66. See the NIHR Funding and Awards website for further award information. Plain language summary Background People with irritable bowel syndrome experience stomach (abdominal) pain and changes to their bowel movements. Irritable bowel syndrome can have a serious impact on people’s lives. Previous small trials suggest that a drug called amitriptyline used at a low dose may help irritable bowel syndrome. Amitriptyline is already used to treat other conditions. It is available for irritable bowel syndrome but is not used much by general practitioners. Methods We recruited adults aged ≥ 18 years with irritable bowel syndrome from UK general practices who did not have any issues preventing the use of amitriptyline. Patients received either low-dose amitriptyline or placebo (a dummy tablet) for 6 months. Patients could adjust the dose according to symptoms and side effects. Neither the researchers nor the patients knew which treatment they were getting. Participants recorded symptoms using a questionnaire containing an irritable bowel syndrome severity score. We looked at the difference in average irritable bowel syndrome severity score between patients receiving amitriptyline and placebo. We also looked at effects of amitriptyline on mood, ability to work, and non-gut symptoms related to irritable bowel syndrome, as well as safety and acceptability. Some patients and general practitioners were interviewed about their experiences. Results Four hundred and sixty-three patients took part. Participants receiving amitriptyline reported a bigger improvement in their irritable bowel syndrome severity scores at 6 months, compared with patients on placebo. Amitriptyline was better across a range of irritable bowel syndrome symptom measures but did not impact anxiety, depression or ability to work. Forty-six people (19.8%) stopped taking amitriptyline and 59 (25.5%) stopped the placebo before 6 months. Patients liked being able to adjust their dose and valued contact with the research team. Conclusion This study showed that amitriptyline is more effective than a placebo and is safe. General practitioners should offer low-dose amitriptyline to people with irritable bowel syndrome if symptoms do not improve with other standard treatments. Patients should be supported and helped to adjust their dose as needed. The dose adjustment sheet used in this trial will be made available. Scientific summary Background Irritable bowel syndrome (IBS) affects 5% of the population, accounting for > 3% of all consultations in primary care in England and Wales. Symptoms include abdominal pain in association with a change in stool form or frequency. The condition impacts on quality of life and ability to work and limits social activities. The medical management of IBS is unsatisfactory, with no therapy proven to alter the long-term natural history and, at best, modest symptom reduction. Previous meta-analyses of trials conducted in secondary and tertiary care suggest low-dose tricyclic antidepressants (TCAs) may be efficacious, probably because of their pain-modifying properties, as well as their influence on gut motility, rather than any effects on mood. Although National Institute for Health and Care Excellence guidelines for the management of IBS in primary care suggest considering low-dose TCAs as second-line treatment, their effectiveness in this setting is unknown and they are infrequently prescribed by general practitioners (GPs). Objectives Our objective was to determine the clinical and cost-effectiveness of low-dose titrated amitriptyline compared with placebo for 6 months as a second-line treatment in adults with IBS in primary care. Methods ATLANTIS was a pragmatic, randomised, multicentre, parallel-group, two-arm, double-blind, placebo-controlled trial. A nested, qualitative study explored participant and GP experiences of treatments and trial participation. A within-study cost-effectiveness analysis was planned but, due to the coronavirus disease discovered in 2019 (COVID-19) pandemic, health economic analyses were removed after obtaining additional funding to complete the trial to prioritise funds for participant recruitment. These will be subject to further funding. Participants, their GPs, investigators, the research team, and the analysis team were all masked to treatment allocation throughout the trial. Patients meeting Rome IV criteria for IBS who had tried first-line treatments and with ongoing IBS symptoms [score of ≥ 75 on the IBS Severity Scoring System (IBS-SSS)] were recruited via mail-out from 55 general practices in three regions in England. Participants were randomised 1 : 1 to receive either low-dose titrated amitriptyline or placebo. Both treatments were supplied for 6 months, with the dose commenced at 10 mg o.d. and titrated to a maximum of 30 mg o.d. or a minimum of 10 mg alternate days. Dose titration was participant-led according to IBS symptoms and side effects, with support from the trial team and a dose titration document developed with input via patient and public involvement. Participants recruited earlier to the trial had the option to continue blinded treatment for an additional 6 months. The primary outcome was the effect of amitriptyline on global IBS symptom scores at 6 months. The key secondary outcome was the proportion of participants with relief of IBS symptoms at 6 months. Other secondary outcomes included effect on global IBS symptoms and relief of IBS symptoms at 3 and 12 months, effect on IBS-associated somatic symptoms at 6 months, effect on quality of life, anxiety, and depression scores, and ability to work and participate in other activities at 3, 6 and 12 months, as well as acceptability and tolerability of, and adherence to, treatment. Patient-reported questionnaires at baseline and 3, 6 and 12 months post randomisation (unless otherwise indicated) were used to assess IBS symptom severity (measured via the IBS-SSS), relief of IBS symptoms [measured by subjective global assessment (SGA) of relief], adequate relief of IBS symptoms (measured by a weekly response to the question ‘Have you had adequate relief of your IBS symptoms?’), IBS-associated somatic symptoms [using the Patient Health Questionnaire-12 (PHQ-12)], mood [using the Hospital Anxiety and Depression Scale (HADS)], ability to work and participate in other activities [using the Work and Social Adjustment Scale (WSAS)], quality of life (using the EQ-5D-3L), healthcare use (using a bespoke health resource use questionnaire), and tolerability [using the Antidepressant Side-Effect Checklist (ASEC)]. Numbers of participants reporting serious adverse events (SAEs), including serious adverse reactions (SARs), were reported for each treatment group. An evaluable sample size of 414 participants would provide 90% power to detect a minimum clinically important difference of 35 points between amitriptyline and placebo at 6 months on the IBS-SSS. This sample size provided at least 85% power to detect a 15% absolute difference in the key secondary outcome of SGA of relief of IBS symptoms at 6 months. We planned to recruit 518 participants, allowing for 20% loss to follow-up. Effectiveness outcomes were analysed in the intention-to-treat population, defined as all participants randomised, regardless of adherence. All statistical testing used two-sided 5% significance levels. The primary outcome was analysed using a linear regression model, adjusted for minimisation variables and baseline IBS-SSS score. Missing data were imputed by treatment arm, via multiple imputation, and results were expressed as point estimates with 95% confidence intervals (CIs). Secondary binary outcomes were analysed in logistic or ordinal regression models, with results expressed as odds ratios (ORs) with 95% CIs. Continuous secondary outcomes, including PHQ-12, HADS and WSAS scores, were analysed as for the primary outcome, adjusted for the respective baseline score. All participants receiving at least one dose of trial medication, according to medication received, were included in the safety analysis. The nested, qualitative study aimed to identify factors that would facilitate or impede prescribing of, acceptability of, and adherence to, low-dose amitriptyline in IBS, to identify participants’ and GPs’ perspectives on the broader impact of the trial, and to explore psychosocial and contextual factors that might shape wider use of amitriptyline for IBS. Familiarity with amitriptyline may both hinder uptake, given its association with depression, and facilitate it, given that it is a known drug, taken in a low dose distinct from the antidepressant dose, already used for a range of other painful conditions and has comparatively mild, and in some cases potentially beneficial, side effects such as on sleep. Semi-structured audio-recorded telephone interviews were conducted with a diverse sub-sample of trial participants and GPs involved in the trial and transcribed verbatim. The final sample size was dependent on saturation, to achieve a rigorous, credible analysis in relation to the aims. Topic guides allowed flexible exploration of all required topics, while remaining open to participants’ individual experiences and perspectives. To enhance trustworthiness of the analysis, all qualitative study team members contributed to avoid producing idiosyncratic interpretations, a negative case analysis was undertaken, and an audit trail was produced to enhance transparency, including detailed coding manuals and interviewer field notes. Reflexive thematic analysis, incorporating techniques from grounded theory, was used to analyse the qualitative data. Data collection and initial analyses proceeded iteratively, and informed subsequent interviews. Analysis was primarily inductive, with researchers identifying themes in the data rather than imposing any pre-existing interpretive framework. Qualitative findings were related to the main trial findings by comparing themes across subgroups and against the quantitative data. Clinical results In total, 15,672 potentially eligible patients were invited to take part, of whom 1253 were interested and were screened. Of those screened, 463 (37.0%) were randomised {mean age 48.5 years [standard deviation (SD) 16.1 years], 315 (68.0%) female}, to amitriptyline (n = 232) or placebo (n = 231). Six-month follow-up was achieved for 401 (86.6%) participants, 204 (87.9%) in the amitriptyline arm, and 197 (85.3%) in the placebo arm. Participants were well balanced between treatment arms according to demographics and baseline characteristics, IBS symptom severity, PHQ-12 scores, HADS-depression and HADS-anxiety scores, and previous first-line treatments. Among participants, 80.4% had IBS-D or IBS-M, 84.2% had a normal HADS-depression score, and 84.7% had moderate to severe scores on the IBS-SSS, with a mean IBS-SSS score in all participants of 272.8 (SD 90.3). The median duration of IBS was 10 years. In total, 338 (73.0%) participants completed 6 months of treatment, 173 (74.6%) randomised to amitriptyline and 165 (71.4%) to placebo. Discontinuation of trial medication before 6 months occurred in 105 (22.7%) participants, 46 (19.8%) allocated to amitriptyline and 59 (25.5%) to placebo. The most common reason for discontinuing trial medication was adverse events (AEs) in 30 (12.9%) participants allocated to amitriptyline and 20 (8.7%) to placebo, followed by lack of benefit in 7 (3.0%) randomised to amitriptyline and 18 (7.8%) to placebo. There were a further 17 (3.7%) participants lost to follow-up and 3 (0.6%) who did not commence trial medication. By 3 months, similar proportions of participants randomised to amitriptyline had titrated their dose to 20 mg o.d. (35.2%) or 30 mg o.d. (37.8%), although by 6 months this had increased to 42.8% taking 30 mg o.d. However, in the placebo arm, 57.0% of participants titrated their dose to 30 mg o.d. within 3 months and this proportion was similar at 6 months. For the primary outcome, amitriptyline was superior to placebo at 6 months in the intention-to-treat analysis, with a significant difference in mean IBS-SSS score between arms (−27.0, 95% CI −46.9 to −7.1; p = 0.008). For the key secondary outcome, SGA of relief of IBS symptoms, amitriptyline was also superior to placebo (OR for relief of IBS symptoms = 1.78, 95% CI 1.19 to 2.66; p = 0.005). At 3 months, the difference in mean change in IBS-SSS score also favoured amitriptyline (−23.3, 95% CI −42.0 to −4.6; p = 0.014), as did the SGA of relief of IBS symptoms (OR = 1.70, 95% CI 1.15 to 2.53; p = 0.008). In a sensitivity analysis using an alternative definition of SGA of relief of IBS symptoms, where only those reporting considerable or complete relief of IBS symptoms at 3 or 6 months were classed as responders, the effect size in the amitriptyline arm increased at both 3 (OR = 1.81, 95% CI 1.17 to 2.79) and 6 months (OR = 1.88, 95% CI 1.20 to 2.95). Other sensitivity analyses on the per-protocol population for the primary outcome and on participants with complete data for the primary and key secondary outcomes gave consistent results, albeit with larger estimated treatment effects. In terms of adequate relief of IBS symptoms, amitriptyline was also superior to placebo with increased odds of adequate relief across all 25 weeks (OR = 1.56, 95% CI 1.20 to 2.03; p < 0.001), and a higher proportion of participants reporting adequate relief for ≥ 13 of 25 weeks [90/222 (40.5%) vs. 67/221 (30.3%)]. Significantly higher numbers of participants taking amitriptyline reported the drug to be acceptable and would have been willing to continue taking it at 6 months (OR = 1.60, 95% CI 1.08 to 2.35; p = 0.018). Adherence at 3 months was identical in the two treatment arms, but it was higher in the amitriptyline arm at 6 months [172/232 (74.1%) vs. 155/228 (68.0%)]. Amitriptyline had no significant effect on PHQ-12 scores at 6 months, or HADS-anxiety, HADS-depression or WSAS scores at either 3 or 6 months. In terms of treatment-emergent AEs, there was a statistically significant increase in the total ASEC score in those receiving amitriptyline compared with placebo at 3 months (1.39, 95% CI 0.29 to 2.50; p = 0.013) but not at 6 months (0.26, 95% CI −0.98 to 1.51; p = 0.681). The AEs reported in participants receiving amitriptyline in excess of those reported by the placebo arm mainly related to its known anticholinergic effects, including dry mouth, drowsiness, blurred vision and problems with urination. However, rates of treatment-emergent AEs fell between 3 and 6 months and few were severe. The commonest AEs leading to discontinuation in the amitriptyline arm were drowsiness and deterioration of mood. In total, there were five SARs, two in the amitriptyline arm and three in the placebo arm. There were five SAEs unrelated to trial medication, of which four occurred in the amitriptyline arm and one in the placebo arm. In the subset of participants recruited to 12 months’ follow-up and with the choice to continue treatment beyond 6 months, 44% of participants completed 12 months’ treatment. Despite the mixed sample, in the 12-month ITT population, weak evidence of a significant effect in favour of low-dose amitriptyline remained on the mean IBS-SSS (−22.6, 95% CI −49.35 to −4.16; p = 0.098) and the SGA of relief of global IBS symptoms (OR = 1.58, 95% CI 0.94 to 2.64; p = 0.083). In contrast to 6-month results, there was a statistically significant effect on the HADS-depression (−0.88, 95% CI −1.71 to 0.06; p = 0.036) and WSAS (−2.14, 95% CI −3.80 to −0.49; p = 0.011) scores in favour of low-dose amitriptyline. Qualitative results The qualitative study conducted and thematically analysed 77 semistructured interviews with 42 participants and 16 GPs. A multidisciplinary team including patient collaborators explored multiple aspects of participants’ and GPs’ experiences of treatments and participating in the ATLANTIS trial. The qualitative analysis of barriers and facilitators suggests that low-dose amitriptyline for IBS is acceptable to, and is often welcomed by, GPs and patients as an additional treatment option. Addressing concerns and promoting facilitators could facilitate wider use of low-dose amitriptyline for IBS which may be achieved through: Clear communication to clinicians, for example in clinical guidelines, that distinguishes low-dose amitriptyline for IBS from amitriptyline use for other conditions (especially depression). Resources to support GP–patient communication to distinguish low-dose amitriptyline for IBS from amitriptyline for other conditions (especially depression). This might include, for example, tips for GPs when discussing amitriptyline for IBS with patients, online materials to support or reinforce messages given during consultations, tailored packaging and patient inserts, and education for pharmacists. Clear guidance about low-dose amitriptyline for IBS and anticholinergic burden. This should highlight that low-dose amitriptyline has lower potential risk and that currently anticholinergic burden risk scores do not account for dose, so can overinterpret risk with low-dose amitriptyline. Guidance and resources for GPs and patients to support patients managing their own dose titration. The dose-titration document used in ATLANTIS was well received by GPs and patients. Conclusions In the largest trial of a TCA in IBS ever conducted, titrated low-dose amitriptyline was superior to placebo as a second-line treatment for IBS in primary care across multiple outcomes and was safe. The results of ATLANTIS strongly support use of titrated low-dose amitriptyline in this setting. GPs should offer low-dose amitriptyline to patients with IBS whose symptoms do not improve with first-line therapies, with appropriate support to guide patient-led dose titration, such as the self-titration document developed for this trial. Trial registration This trial is registered as ISRCTN48075063. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 16/162/01) and is published in full in Health Technology Assessment; Vol. 28, No. 66. See the NIHR Funding and Awards website for further award information.

Background Chronic limb-threatening ischaemia with ischaemic pain and/or tissue loss. Objective To examine the clinical and cost-effectiveness of a vein bypass-first compared to a best endovascular treatment-first revascularisation strategy in preventing major amputation or death. Design Superiority, open, pragmatic, multicentre, phase III randomised trial. Setting Thirty-nine vascular surgery units in the United Kingdom, and one each in Sweden and Denmark. Participants Patients with chronic limb-threatening ischaemia due to atherosclerotic peripheral arterial disease who required an infra-popliteal revascularisation, with or without an additional more proximal infra-inguinal revascularisation procedure, to restore limb perfusion. Interventions A vein bypass-first or a best endovascular treatment-first infra-popliteal, with or without an additional more proximal infra-inguinal revascularisation strategy. Main outcome measures The primary outcome was amputation-free survival. Secondary outcomes included overall survival, major amputation, further revascularisation interventions, major adverse limb event, health-related quality of life and serious adverse events. Methods Participants were randomised to a vein bypass-first or a best endovascular treatment-first revascularisation strategy. The original sample size of 600 participants (247 events) was based on a hazard ratio of 0.66 with amputation-free survival rates of 0.72, 0.62, 0.53, 0.47 and 0.35 in years 1–5 in the best endovascular treatment-first group with 90% power and alpha at p = 0.05. The sample size was revised to an event-based approach as a result of increased follow-up time due to slower than anticipated recruitment rates. Participants were followed up for a minimum of 2 years. A cost-effectiveness analysis was employed to estimate differences in total hospital costs and amputation-free survival between the groups. Additionally, a cost–utility analysis was carried out and the total cost and quality-adjusted life-years, 2 and 3 years after randomisation were used. Results Between 22 July 2014 and 30 November 2020, 345 participants were randomised, 172 to vein bypass-first and 173 to best endovascular treatment-first. Non-amputation-free survival occurred in 108 (63%) of 172 patients in the vein bypass-first group and 92 (53%) of 173 patients in the best endovascular treatment-first group [adjusted hazard ratio 1.35 (95% confidence interval 1.02 to 1.80); p = 0.037]. Ninety-one (53%) of 172 patients in the vein bypass-first group and 77 (45%) of 173 patients in the best endovascular treatment-first group died [adjusted hazard ratio 1.37 (95% confidence interval 1.00 to 1.87)]. Over follow-up, the economic evaluation discounted results showed that best endovascular treatment-first was associated with £1690 less hospital costs compared to vein bypass-first. The cost utility analysis showed that compared to vein bypass-first, best endovascular treatment-first was associated with £224 and £2233 less discounted hospital costs and 0.016 and 0.085 discounted quality-adjusted life-year gain after 2 and 3 years from randomisation. Limitations Recruiting patients to the Bypass versus Angioplasty in Severe Ischaemia of the Leg Trial-2 trial was difficult and the target number of events was not achieved. Conclusions A best endovascular treatment-first revascularisation strategy was associated with better amputation-free survival, which was largely driven by fewer deaths. Overall, the economic evaluation results suggest that best endovascular treatment-first dominates vein bypass-first in the cost-effectiveness analysis and cost–utility analysis as it was less costly and more effective than a vein bypass-first strategy. Future work The Bypass versus Angioplasty in Severe Ischaemia of the Leg Trial-2 investigators have a data sharing agreement with the BEst Surgical Therapy in patients with Chronic Limb threatening Ischaemia investigators. One output of this collaboration will be an individual patient data meta-analysis. Study registration Current Controlled Trials ISRCTN27728689. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 12/35/45) and is published in full in Health Technology Assessment; Vol. 28, No. 65. See the NIHR Funding and Awards website for further award information. Plain language summary Atherosclerosis, or narrowing of the arteries, can occur as a result of smoking, high blood pressure, diabetes, or high cholesterol in the blood. Atherosclerosis can affect any artery, including those supplying the legs, where the condition is called peripheral arterial disease. The most severe form of peripheral arterial disease is chronic limb-threatening ischaemia which can cause severe pain in the foot as well as ulcers and gangrene. Unless the blood supply to the leg and foot is improved, by a process called revascularisation, people with chronic limb-threatening ischaemia are at high risk of amputation and death. The blood supply can be improved by using a vein from the leg to bypass around the blockages (vein bypass) or by using a balloon (angioplasty) or small metal tubes (stents) to reopen the blocked arteries (best endovascular treatment). There is debate about which type of revascularisation is best in terms of preventing amputation and death, especially in people who need revascularisation of the arteries below the knee. Bypass versus Angioplasty in Severe Ischaemia of the Leg Trial-2 is the first randomised controlled trial to compare vein bypass-first and best endovascular treatment-first in this group of patients. Bypass versus Angioplasty in Severe Ischaemia of the Leg Trial-2 found that people randomised to a vein bypass-first revascularisation strategy were 35% more likely to require a major amputation or die than those randomised to a best endovascular treatment-first strategy. Most of this difference in favour of best endovascular treatment-first was due to a higher number of patients dying in the vein bypass-first group. Best endovascular treatment-first was also cheaper for the National Health Service. The results of this study suggest that in patients with chronic limb-threatening ischaemia due to peripheral arterial disease in the arteries below the knee, who are suitable for both vein bypass and best endovascular treatment and where there is uncertainty as to which is best, best endovascular treatment should be offered first rather than vein bypass. Scientific summary Background Chronic limb-threatening ischaemia (CLTI) is the severest manifestation of peripheral arterial disease (PAD) and presents with ischaemic pain at rest or tissue loss (ulceration, gangrene or both) or both. The researchers compared the effectiveness of a vein bypass (VB)-first with a best endovascular treatment (BET)-first revascularisation strategy in terms of preventing major amputation and death in patients with CLTI who required an infra-popliteal (IP) revascularisation, with or without an additional more proximal infra-inguinal revascularisation procedure, to restore limb perfusion. Objectives The primary objective of the study was to examine the effectiveness and cost-effectiveness of using a VB-first compared to a BET-first revascularisation strategy in terms of preventing major (above the ankle) amputation or death from any cause [amputation-free survival (AFS)] in patients with CLTI who required an IP, with or without an additional more proximal infra-inguinal revascularisation procedure, to restore limb perfusion. Design Superiority, open-label, pragmatic, multicentre, phase III randomised trial. Setting Thirty-nine vascular surgery units in the United Kingdom, and one each in Sweden and Denmark. Participants Those with CLTI due to atherosclerotic PAD and who required an IP revascularisation, with or without an additional more proximal infra-inguinal revascularisation procedure, to restore limb perfusion. Interventions A VB-first or a BET-first IP, with or without an additional more proximal infra-inguinal revascularisation strategy, to restore limb perfusion. Most VBs were constructed with great saphenous vein and originated from the common or superficial femoral arteries. Most endovascular interventions comprised plain balloon angioplasty with selective use of plain or drug-eluting stents. Outcome measures The primary outcome was AFS defined as time to first major (above the ankle) amputation or death from any cause. Secondary outcomes included: time to death from any cause (overall survival); time to major amputation of the trial leg; major adverse limb event (defined as major amputation of the trial leg, or any further major revascularisation intervention to the trial leg, following the first revascularisation intervention); major adverse cardiac event (defined as chronic limb-threatening ischaemia and/or major amputation affecting the non-trial leg, myocardial infarction, transient ischaemic attack or stroke); 30-day morbidity and mortality; relief of ischaemic pain as determined by visual analogue scale and opiate usage; health- related quality of life using generic [EuroQol-5 Dimensions, five-level version (EQ-5D-5L), Short Form questionnaire-12 items, ICEpop CAPability measure for Older people] and disease-specific (the Vascular Quality of Life Questionnaire) tools; further major revascularisation intervention to the trial leg (following the first revascularisation intervention); re-intervention and crossover intervention (where re-intervention is defined as the same, and a crossover procedure is defined as an alternative, revascularisation procedure to the first revascularisation procedure post-randomisation); healing of tissue loss (ulcers, gangrene) at or below the ankle presumed to be caused by PAD as assessed by the perfusion, extent, depth, infection and sensation score, the Wound Ischaemia and foot Infection tool; and haemodynamic measurements (ankle–brachial pressure index, and toe brachial pressure index). Serious adverse events were collected up to 30 days post first revascularisation. Economic evaluation analyses in the form of cost-effectiveness and cost–utility analysis (CUA) were conducted from the perspective of the UK NHS alongside the trial. The base case analyses considered only hospital costs and an additional scenario and subgroup analyses were carried out. Sample size The original sample size was based on a time-to-event analysis to be undertaken 2 years after completion of recruitment. It was anticipated that recruitment would take place over 3 years: 20% of patients recruited in year 1, 40% in year 2 and 40% in year 3. Based on the Bypass versus Angioplasty in Severe Ischaemia of the Leg Trial (BASIL-1) trial. AFS rates were assumed to be 0.72 in year 1, 0.62 in year 2, 0.53 in year 3, 0.47 in year 4 and 0.35 in year 5. Allowing for a 10% attrition rate and based on the survival estimates calculated using the BASIL-1 data, a population of 600 participants (247 primary outcome events) would have 90% power to detect a reduction in AFS of one-third [hazard ratio (HR) 0.66] at the 5% significance level. The initial assumptions made in this trial concerning recruitment rates were not achieved; therefore, recruitment continued beyond year 3. As a result, the median length of follow-up was longer than originally planned. Therefore, the number of randomised participants required to observe 247 events (as per the original sample size target) was reduced due to the increased exposure time. With support of the funder and independent oversight from the Data Monitoring Committee, recruitment rates, length of follow-up, and pooled event rates over time were modelled to predict the number of participants needed to reach 247 events, with 24 months minimum follow-up in each participant. The modelling was updated approximately every 6 months based on emerging data. Results Between 22 July 2014 and 30 November 2020, 345 participants were randomised, 172 to the VB-first group and 173 to the BET-first group. The baseline characteristics of the two groups were similar. Major amputation or death occurred in 108 (63%) of 172 patients in the VB-first group and 92 (53%) of 173 patients in the BET-first group {adjusted HR 1.35 [95% confidence interval (CI) 1.02 to 1.80]; p = 0.037}. Ninety-one (53%) of 172 patients in the VB-first group and 77 (45%) of 173 patients in the BET-first group died [adjusted HR 1.37 (95% CI 1.00 to 1.87)]. In both groups the commonest causes of morbidity and mortality, including those occurring within 30 days of their first revascularisation, were cardiovascular and respiratory events. Over follow-up, the economic evaluation results showed that BET-first was associated with £1690 less hospital costs compared to VB-first with additional survival without amputation. The cost utility analysis showed that compared to VB-first, BET-first was associated with £2524 and £2233 less hospital costs and 0.016 and 0.085 QALY gain after 2 and 3 years from randomisation. Conclusions In the BASIL-2 trial, a BET-first revascularisation strategy was associated with a better AFS, which was largely driven by fewer deaths in the BET-first group. In both the CUA and cost-effectiveness analysis (CEA) analyses and from a NHS and societal perspective BET-first dominated VB-first and is therefore a highly cost-effective intervention for the NHS. These data suggest a greater role for BET in the management of patients with CLTI who require an IP revascularisation to restore limb perfusion. In patients with CLTI who required an IP, with or without an additional more proximal infra-inguinal revascularisation procedure, to restore limb perfusion, a BET-first revascularisation strategy was associated with reduced hospital costs (£1690) and improved AFS (0.429 years), out to 7 years following randomisation. BET-first therefore dominated VB-first in the CEA. Similarly, in the health-related quality of life the CUA, BET-first was cost-saving with improved QALYs (£2524 and £2233 less hospital costs and 0.016 and 0.085 more QALYs at 2 and 3 years, respectively) and so dominated VB-first. This economic analysis therefore shows that BET-first is a cost-effective option from an NHS and societal perspective. The sensitivity analysis supported the base-case analysis and BET-first was found to be cost-effective at different willingness-to-pay thresholds. Similar findings were found in all other scenario analyses when considering costs of primary and other hospital healthcare services, taking a broader societal perspective, which includes out-of-pocket expenditure and the costs associated with productivity loss, patient’s adherence to study protocol, the impact of imputation by focusing on participants with complete hospital cost and EQ-5D-5L data only. However, these findings should be interpreted cautiously given the large number of imputed cost values and the substantial probability of a very small QALY difference. Study registration Current Controlled Trials ISRCTN27728689. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 12/35/45) and is published in full in Health Technology Assessment; Vol. 28, No. 65. See the NIHR Funding and Awards website for further award information.

Background People with severe mental illness who experience co-occurring substance use experience poor outcome including suicide, violence, relapses and use of crisis services. They struggle to access care and treatment due to a lack of an integrated and co-ordinated approach which means that some people can fall between services. Despite these concerns, there is limited evidence as to what works for this population. Objectives To undertake a realist evaluation of service models in order to identify and refine programme theories of what works under what contexts for this population. Design Realist synthesis and evaluation using published literature and case study data. Setting Mental health, substance use and related services that had some form of service provision in six locations in the United Kingdom (five in England and one in Northern Ireland). Participants People with lived experience of severe mental illness and co-occurring substance use, carers and staff who work in the specialist roles as well as staff in mental health and substance use services. Results Eleven initial programme theories were generated by the evidence synthesis and in conjunction with stakeholders. These theories were refined through focus groups and interviews with 58 staff, 25 service users and 12 carers across the 6 case study areas. We identified three forms of service provision (network, consultancy and lead and link worker); however, all offered broadly similar interventions. Evidence was identified to support most of the 11 programme theories. Theories clustered around effective leadership, workforce development and collaborative integrated care pathways. Outcomes that are meaningful for service users and staff were identified, including the importance of engagement. Limitations The requirement for online data collection (due to the COVID-19 pandemic) worked well for staff data but worked less well for service users and carers. Consequently, this may have reduced the involvement of those without access to information technology equipment. Conclusion The realist evaluation co-occurring study provides details on how and in what circumstances integrated care can work better for people with co-occurring severe mental health and alcohol/drug conditions. This requires joined-up policy at government level and local integration of services. We have also identified the value of expert clinicians who can support the workforce in sustaining this programme of work. People with co-occurring severe mental health and alcohol/drug conditions have complex and multifaceted needs which require a comprehensive and long-term integrated approach. The shift to integrated health and social care is promising but will require local support (local expert leaders, network opportunities and clarity of roles). Future work Further work should evaluate the effectiveness and cost-effectiveness of service models for this group. Study registration This study is registered as PROSPERO CRD42020168667. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128128) and is published in full in Health Technology Assessment; Vol. 28, No. 67. See the NIHR Funding and Awards website for further award information. Plain language summary People who experience serious mental illness (the kind that affects people’s daily life and needs long-term support) can also have problems with drugs and alcohol. This can affect access to care and treatment and result in a crisis. We wanted to develop a better understanding of how services could help those with co-occurring serious mental health and alcohol/drug conditions (or COSMHAD for short). To answer these questions, we carried out the following: We gathered information from publications to identify what treatments are available for co-occurring severe mental health and alcohol/drug conditions and which aspects of these were useful (evidence synthesis). We asked services across the United Kingdom about the local provision for people with co-occurring severe mental health and alcohol/drug conditions. We conducted focus groups in six locations with service users, carers and staff about their experiences of care for co-occurring severe mental health and alcohol/drug conditions. What we found There are very few services in the United Kingdom that currently provide a service for people with co-occurring severe mental health and alcohol/drug conditions. We identified a set of factors that are likely to promote better outcomes for people with co-occurring severe mental health and alcohol/drug conditions including commitment from leaders across organisations to address this issue and support staff training. A local expert clinical leader was seen to be important in enabling closer working between mental health and substance use. Service users and carers recognised that when care was co-ordinated, and staff demonstrated empathy and compassion, they were more likely to engage in treatment. Further work will be needed to evaluate how helpful some of the aspects of the models of care are in helping people in their recovery goals. Scientific summary Background Approximately 30–50% of people with serious mental health illness (SMI) have a coexisting alcohol/drug condition, leading to significant negative health and social outcomes. Despite the scale of these co-occurring conditions, there is limited evidence to inform treatment, with the evidence that is available failing to provide a definitive answer as to how services and treatments should be best delivered to improve health and other outcomes for this diverse group. Objectives The aim of this project was to use a realist approach to understand what works, how, for whom and in what circumstances by synthesising data from published and grey literature, mapping and describing the characteristics of UK services and service provision, and undertaking in-depth focus groups and interviews in locations picked to be representative of the range of provision identified in the mapping and review of the literature. The outcome was a set of refined programme theories (PTs), which underpin an explanatory framework that can be used to inform future research, policy and practice. Methods We conducted a series of distinct, yet interrelated work packages (WPs) to achieve our research objectives. Work package 1: development of programme theories The aim of WP1 was to (1) map the literature to provide a systematic overview of the nature of the published and grey literature on types of service provision for people with co-occurring severe mental health and alcohol/drug conditions (COSMHAD) and (2) develop realist PTs for interventions and service models for COSMHAD. An a priori protocol was registered with PROSPERO. In phase 1, in consultation with stakeholders (including clinical experts in COSMHAD and those with lived experience), we elicited a set of initial PTs in a workshop and analysed policy documents and articles describing COSHMAD services in practice in the UK. In phase 2, we followed the five stages for realist synthesis. A total of 172 papers were included in the synthesis. Work package 2: service mapping Work package 2a: mapping of United Kingdom co-occurring severe mental health and alcohol/drug conditions services The aim of WP2 was to gather information on the availability of COSMHAD treatment across the UK. To achieve this, information was gathered by direct requests to relevant health and social care organisations, either as speculative e-mails or as Freedom of Information requests. In addition, internet searches for relevant services were also conducted. The initial information requests were sent by e-mail in March 2020 and had to be halted due to the start of the coronavirus disease 2019 (COVID-19) pandemic. A second wave of requests was sent in October 2020 to organisations that had not responded. The information requested included the details of the approach/treatment pathway for COSMHAD that each organisation commissioned/provided, and whether treatment was commissioned/provided specifically for COSMHAD, knowledge of any other COSMHAD service being provided in the same location, and name and contact details of the organisation’s COSMHAD lead (if there was one). Work package 2b: service audit/survey Using the information gathered during the national mapping, 16 organisations were identified as providing COSMHAD services. This was on the basis that their response indicated that there was some form of dedicated resource (typically in the form of specifically funded staff roles) to deliver the COSMHAD model. The 16 services were asked to provide more detail using an online audit form. This included items on the approach to treatment, the range of treatments offered, staff training/supervision, commissioning and funding of COSMHAD services, and health economic data. This survey was sent to a key member of staff in each of the organisations to complete, with the option of a follow-up telephone call to aid response rates. The data were used to identify models of service delivery. Work package 3: refining programme theories The purpose of WP3 was to test and refine the PTs developed in WP1 in real-world settings. Six case study sites were selected that represented examples from the three types of service models identified in WP2. Staff were recruited and consented to participate in online focus groups. Service users and carers were approached by clinical staff in the service to inform them of the study and to pass on contact details to the realist evaluation co-occurring (RECO) researcher. The RECO researcher would then contact and discuss the project, and if they were interested, they would give informed consent. Service users and carers participated in online focus groups and individual interviews. One carer focus group was face to face. Topic guides were developed from the 11 PTs that were developed in the realist synthesis, and all the interviews and focus groups were conducted using realist interviewing style. All interviews were recorded using Microsoft Teams and the auto-transcription facility. The transcripts were anonymised and analysed in NVivo. Results Work package 1: realist synthesis A set of 11 refined PTs were identified from the literature and these broadly fell into three interconnecting categories: committed leadership; clear expectations regarding COSMHAD from mental health and substance use workforces; and clear processes to co-ordinate care. Work package 2: service mapping The initial mapping identified that most areas of the UK offer some form of integrated care or treatment pathway for people with COSMHAD. However, only 16 services indicated that they offered more than an agreed pathway and offered something more tangible. The information provided from the audit of the services was used to broadly divide into three main models of delivery: Network – this is a broad collection of staff from a range of local services that share agreed care pathways and come together for network meetings and shared training. Consultancy – a specialist team that provides consultancy to the mainstream mental health services, including joint assessment, advice on care planning and clinical management and continuous professional development (CPD) training. Lead and link – this model is led by a clinical expert leader (often a consultant nurse or equivalent) with ‘link workers’ or ‘champions’ based across the mental health service to support this role. However, there was significant variation in the level of investment in terms of dedicated staff. Six sites were selected to be case studies. Work package 3: case studies A total of 58 staff, 25 service users and 12 carers participated in the focus groups and interviews. Data were obtained that supported 10 out of 11 PTs. There were limited data regarding pre-registration/undergraduate training (despite prompts) because the participants had no connection to the curriculum development and delivery for nursing, allied health, social care or medical training. Some of the PTs had more supporting evidence than others. The main findings were that positive attitudes including empathy and compassion were seen as a critical context for engaging people in services. Positive attitudes were influenced by access to CPD training, and also ongoing support and supervision in practice. It was important that people could see that the clinical skills taught in CPD worked effectively in practice. This was also promoted by working alongside (shadowing) a clinical expert in COSMHAD. The role of a clinical expert (most commonly a consultant nurse in COSMHAD) was rated highly by the staff for a number of reasons, including keeping COSMHAD on the agenda both clinically and strategically; having an oversight of the whole care pathway; and fostering relationships between key agencies (such as substance use, housing, etc.). The expert lead also offered a training and supervision programme within the organisation, and in some cases facilitated local special interest networks which included peer support, sharing expertise across disciplines, and building interagency and interdisciplinary relationships. The quality of these inter-agency relationships was seen to be critical for integrating care for COSMHAD. Clear and collaborative care pathways that were needs led, person-centred and holistic were important. The service users and carers reported that fragmented local services were really difficult to navigate, and therefore service users often ended up falling through the gaps, increasing the need for crisis care including the police, ambulance service and emergency rooms, none of which is satisfactory in helping people to move forward with their recovery. For effective integrated care to occur, there was a clear need for organisational commitment to this agenda at a senior strategic level (including local commissioning) as well as commitment in operational managers. This commitment ensures investment in posts (such as the expert leaders); protected time for staff to access training, supervision and network meetings; and mandating standards of care for people with COSMHAD. In addition, there needs to be a commitment to recruiting and retaining the workforce that have the requisite knowledge, skills and values. This also links to the need to ensure that pre-registration education for nurses and allied health professionals as well as psychiatry and psychology includes content in working with COSMHAD, as well as offering placements where health and social care students get positive experience of assessing and planning care for people with COSMHAD. Carers felt that they carried a lot of the burden of caring for their loved one, and often felt excluded from the care team. They found that peer support was very helpful, but they too would benefit from more clarity in the care pathways and consistency of approach. Discussion Despite over two decades of attention and policy guidance to improve responses to people with COSMHAD, the mapping revealed a lack of investment in specific roles to support care pathways, workforce development and quality of care. Most of the services we identified had been in existence in some form or another for many years and were driven by enthusiastic champions with fragile funding, and often limited strategic or organisational support. Models of care were generally based on an expert lead role plus link workers (dedicated roles) or local champions (not dedicated roles, but supplemental to their role) to support the work within the service as well as co-ordinating services across several agencies. In addition to leadership and supporting care pathways, all of the case studies offered in-house training and consultation. One of the case studies involved a dedicated team of clinicians who offered consultation and training only, as opposed to carrying out any direct clinical work. One model operated as a network of link workers but was missing the strategic lead as that post had not been filled following the departure of the previous lead. The RECO study has been able to identify the important contexts under which integrated care can be provided (mechanism) which produces positive outcomes for service users (and their carers). It is clear that COSMHAD needs to be prioritised as a service development issue and this requires drivers from national and local policy-makers and commissioners. NHS mental health providers should be the main provider with key responsibilities for people with serious and enduring mental health problems irrespective of other comorbidities (in this case alcohol and or drugs). Staff in mental health services should have a minimum level of knowledge and skills regarding the interface between mental health and substance use. Training alone, without clinical supervision and/or other practice-based learning opportunities, is likely to be insufficient. The RECO study demonstrates how valued and multifaceted the role of the clinical expert is in promoting good practice for COSMHAD. They need senior organisational level support for this work. Strengths This is the first study to use realist methods to identify the contexts under which mechanisms work and identify the types of outcomes that are meaningful for staff, service users and carers. The PTs were generated from a number of sources (stakeholder consultation, literature and document synthesis and primary data) using robust methodology. This is also one of the first studies to include the lived experience of being a service user and carer and triangulate with the staff data. We sampled case studies across three types identified from mapping the broader UK service provision. Limitations The study was impacted by COVID-19. The national response to COVID-19 in health and social care impacted on responses to requests for information in the UK mapping exercise. Therefore, there may have been additional services in existence that we did not identify. Due to social distancing and various national and regional lockdowns, the data collection was almost entirely conducted online. While we collected our target sample of staff, we did not recruit to target for the service user and carer focus groups and interviews. Online methods were a barrier to some people. However, the data from the service user and carers were obtained across most of the sites and were sufficient to answer the research questions. Conclusion Co-occurring serious mental illness and substance use is a common phenomenon in mental health care, yet the mental health, substance use and related services remain ill-equipped to meet the needs of people with COSMHAD. The RECO study provides details on how and in what circumstances integrated care can work better for people with COSMHAD. This requires joined-up policy at government level and local integration of services. We have also identified the value of expert clinicians who can support the workforce in sustaining this programme of work. People with COSMHAD have complex and multifaceted needs which require a comprehensive and long-term integrated approach. The shift to integrated health and social care is promising but will require local support (local expert leaders, network opportunities and clarity of roles). Future research should focus on evaluating whole system approaches as opposed to reductionist individual interventions. This includes establishing the cost-effectiveness of key components of the COSMHAD models. Study registration This study is registered as PROSPERO CRD42020168667. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128128) and is published in full in Health Technology Assessment; Vol. 28, No. 67. See the NIHR Funding and Awards website for further award information.

Background Hand eczema is common and a cause of morbidity and occupational disability. When education, irritant/contact allergen avoidance, moisturisation and topical corticosteroids are insufficient to control chronic hand eczema, ultraviolet therapy or systemic immune-modifying drugs are used. There is no treatment pathway generally accepted by UK dermatologists. Primary objective Compare alitretinoin and ultraviolet therapy as first-line therapy in terms of disease activity at 12 weeks post planned start of treatment. Design Prospective, multicentre, open-label, two-arm parallel group, adaptive randomised controlled trial with one planned interim analysis, and an economic evaluation. Setting UK secondary care dermatology outpatient clinics. Participants Patients with severe chronic hand eczema unresponsive to at least 4 weeks of treatment with potent topical corticosteroids. Primary end point Natural logarithm of the Hand Eczema Severity Index + 1, 12 weeks post planned start of treatment. Randomisation Participants randomised 1 : 1 by minimisation to alitretinoin or ultraviolet therapy for 12 to 24 weeks. Blinding Blinded primary end-point assessor. Results Intention-to-treat population: 441 (100.0%) participants; 220 (49.9%) alitretinoin and 221 (50.1%) ultraviolet therapy. At least one dose was received by 212 (96.4%) alitretinoin and 196 (88.7%) ultraviolet therapy participants. Primary outcome The unadjusted median (interquartile range) relative change in hand eczema severity index at 12 weeks was 30% (10–70%) of that at baseline for alitretinoin compared with 50% (20–100%) for ultraviolet therapy. There was a statistically significant benefit of alitretinoin compared with ultraviolet therapy at 12 weeks, with an estimated fold change or relative difference (95% confidence interval) = 0.66 (0.52 to 0.82), p = 0.0003 at 12 weeks. There was no evidence of a difference at 24 or 52 weeks, with the estimated fold change (95% confidence interval) equal to 0.92 (0.798 to 1.08) and 1.27 (0.97 to 1.67), respectively. Primary analysis results were consistent for secondary end points Fifty-nine per cent allocated to alitretinoin and 61% allocated to ultraviolet therapy achieved a clear/almost clear assessment during the trial period. Differential treatment compliance observed: 145 (65.9%) alitretinoin and 53 (24.0%) ultraviolet therapy participants confirmed compliance (≥ 80% received, no treatment breaks > 7 days during first 12 weeks). High levels of missing data were observed. Safety One hundred and thirty-five reportable adverse events across 79 participants, 55 (25.0%) alitretinoin and 24 (10.9%) ultraviolet therapy. Four serious adverse events (two alitretinoin, two ultraviolet therapy). Four pregnancies reported (three alitretinoin, one ultraviolet therapy). No new safety signals were detected. Conclusion As a first-line therapy, alitretinoin showed more rapid improvement and superiority to ultraviolet therapy at week 12. This difference was not observed at later time points. Alitretinoin is cost-effective at weeks 12 and 52. Ultraviolet therapy is cost-effective after 10 years, with a high degree of uncertainty. Hand eczema severity index may be a useful primary outcome measure for hand eczema trials; ALPHA results will inform future trials. Limitations Treatment compliance was poor for ultraviolet therapy. Regular twice weekly treatment was not received by most patients. Assessment of long-term effects of randomised treatments was complicated by use of second-line treatments post treatment phase. Further work Further analysis of substudies and pilot data will provide valuable information for future studies. A clear need for better therapeutic approaches for severe chronic hand eczema remains. Future studies will need to further address long-term benefits of treatments given. Trial registration This trial is registered as ISRCTN80206075. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 12/186/01) and is published in full in Health Technology Assessment; Vol. 28, No. 59. See the NIHR Funding and Awards website for further award information. Plain language summary What was the question? The main question was which treatment was better at easing symptoms of severe hand eczema after 12 weeks. The two treatments compared were ones used most often by UK dermatologists. The first is a tablet called alitretinoin, which is taken once a day. The second is called ultraviolet therapy, where hands are soaked in a special liquid and placed under ultraviolet light twice a week at a hospital. What did we do? We treated 220 patients with alitretinoin and 221 patients with ultraviolet therapy. Patients received treatment for 12 to 24 weeks depending on how well their hand eczema responded. Patients could have different treatments afterwards, and we collected information on their hand eczema symptoms for up to 1 year. What did we find? After 12 weeks, severe hand eczema symptoms improved for both groups of patients but improved most for patients who took alitretinoin. However, 1 year after joining the trial, there was no evidence of a difference between alitretinoin and ultraviolet therapy as a first-line treatment. More patients stopped ultraviolet therapy early compared with patients who received alitretinoin. Different treatments may have been prescribed after the first treatment. What does this mean? Alitretinoin provides a convenient, instant relief or a ‘quick fix’ for patients with severe hand eczema. Alitretinoin is more convenient for lots of people, but it is important to have other options available for people who would prefer not to, or are unable to, take alitretinoin. For example, people who take alitretinoin can experience unwanted side effects, and people who are able to become pregnant must also use contraception. Long-term control of severe hand eczema is important. Individual discussions on the pros and cons of each treatment for hand eczema symptoms is needed. Providing flexible options to attend ultraviolet therapy appointments could be helpful (e.g. weekend/evenings). Scientific summary Background Hand eczema (HE) is common and an important cause of morbidity and occupational disability. One-year prevalence of HE is up to 10% in the general population, with 5–7% developing chronic hand eczema (CHE). Current evidence is not compelling enough to guide clinical practice. When patient education, irritant/contact allergen avoidance, moisturisation and topical corticosteroids are insufficient to control CHE, ultraviolet therapy (PUVA) or systemic immune-modifying drugs are used. There is no treatment pathway generally accepted among UK dermatologists. Most UK dermatology centres use phototherapy (mostly Immersion PUVA) or alitretinoin as first-line treatment for uncontrolled CHE. ALPHA is the first trial comparing alitretinoin with Immersion PUVA as a first-line therapy for patients with uncontrolled severe CHE. Objectives Primary objective Compare alitretinoin and Immersion PUVA as first-line therapy in terms of disease activity at 12 weeks post planned start of treatment. Secondary objectives Compare alitretinoin and Immersion PUVA in terms of disease activity over time (focusing on 24 and 52 weeks post planned start of treatment). Compare alitretinoin and Immersion PUVA in terms of time to relapse. Compare alitretinoin and Immersion PUVA in terms of quality of life (QoL) and patient benefit over 52 weeks post planned start of treatment. Determine cost-effectiveness of alitretinoin compared with Immersion PUVA over the short and longer term. Determine the educational need for patients. Compare alitretinoin and Immersion PUVA in terms of safety. Exploratory objectives included subgroup analyses to assess whether treatment response is affected by pre-specified baseline characteristics. Methods Design Prospective, multicentre, open-label, two-arm parallel group, adaptive randomised controlled trial with one planned interim analysis and an economic evaluation. Participants Patients with severe CHE unresponsive to at least 4 weeks of treatment with potent topical corticosteroids. Primary end point Natural logarithm of the Hand Eczema Severity Index (HECSI) + 1 at 12 weeks post planned start of treatment. Secondary end points Disease activity, measured by HECSI, modified Total Lesion Scoring System (mTLSS) and Physician’s Global Assessment (PGA) at 12, 24, 52 weeks post planned start of treatment. Time to relapse, defined as time between achieving clear/almost clear overall on the blinded assessor PGA and scoring 75% of their baseline HECSI, with sensitivity assessed by redefining relapse as 50% of their baseline HECSI. Dermatology Life Quality Index (DLQI) and Patient Benefit Index for Hand Eczema (PBI-HE) at 12, 24 and 52 weeks post planned start of treatment. Person-Centred Dermatology Self-Care Index (PeDeSi) at 12 and 52 (or 24) weeks post planned start of treatment. Reported adverse events (AEs) and serious adverse events (SAEs) over 52 weeks post planned start of treatment. Cost-effectiveness of alitretinoin compared with Immersion PUVA at week 12, 52 (short term) and over 10 years (long term). Randomisation Participants were randomised using minimisation 1 : 1 to alitretinoin 30 mg/day or Immersion PUVA for 12 weeks to 24 weeks. Randomisation factors were: randomising site, disease duration, clinical phenotype, atopy status, DLQI and skin type. Analysis Linear mixed models accounting for the longitudinal data structure were fitted to primary (HECSI) and secondary end points (mTLSS, DLQI) on the intention-to-treat (ITT) population, with adjustment for: smoking history, body mass index (BMI), foot involvement, baseline score, time since planned start of treatment and treatment group. Participant and participant–time interaction were fitted as random effects. An ordinal logistic mixed model was fitted to the PGA, and an ordinal logistic model without random effects was fitted to the PeDeSI at 12 weeks. Exploratory analyses explored differential treatment effects in pre-planned subgroups, correlation of scoring systems and second-line therapies, and safety data were summarised. Total cost and quality-adjusted life-years (QALYs) were estimated at weeks 12, 52, and 10 years after treatment initiation. The latter were combined to estimate short- and long-term cost-effectiveness. Short-term cost-effectiveness was estimated via a within-trial analysis at weeks 12 and 52 and via Markov model to estimate 10-year cost-effectiveness of both interventions. A resource use questionnaire was used to estimate short- and long-term cost of interventions and standard care, while EuroQol-5 Dimensions, three-level version (EQ-5D-3L) responses were used to estimate utility values and generate QALYs. This analysis followed the NHS and personal and social perspective (PSS). Cost-effectiveness was determined using National Institute for Health and Care Excellence (NICE) guidelines. Secondary analysis included societal perspective and QALYs determined by the DLQI. Data collection Data were collected at baseline, 4-weekly to week 36 and 8-weekly to week 52. Sample size A minimum of 500 and maximum of 780 participants were required to detect a relative difference, or fold change, of 1.3 in HECSI score between treatment arms at 12 weeks post planned start of treatment (80% power; two-sided 5% significance level) assuming a coefficient of variation (CV) between 1.175 and 1.7 and 20% attrition. Note that a fold change of 1.3 is equivalent to a fold change of 0.77 if there was a benefit in the opposite direction. A sample size review was planned after 364 participants reached 12 weeks post planned start of treatment but was conducted early in August 2017. Following Data Monitoring and Ethics Committee (DMEC) review, a sample size of 514 participants was recommended based on a CV of 1.2. Results Screening and recruitment Thirty-one NHS hospitals in England, Scotland and Wales assessed 1557 patients for eligibility, registered 582 (37.4%) and randomised 441 (75.8%) participants between October 2015 and June 2021. Of 1557 patients screened, 642 (41.2%) were ineligible, with 347 (54.0%) not having a severe CHE diagnosis. Of 915 eligible patients, 582 (63.6%) were consented and registered. Of 333 patients who did not consent, 74 (22.2%) thought the Immersion PUVA schedule or travel was inconvenient, and 28 (8.4%) did not want alitretinoin. Of 141 patients not randomised, the main reasons were not meeting the eligibility criteria [n = 69 (48.9%)] or patient choice [n = 29 (20.6%)]. Of patients randomised, 220 (49.9%) were allocated to alitretinoin and 221 (50.1%) to Immersion PUVA. Of those allocated to alitretinoin, 201 (91.4%) started treatment within 7 days post randomisation compared with 165 (74.7%) of those allocated to Immersion PUVA. In total, 132 (29.9%) participants withdrew from, or were lost to, follow-up. The ITT population included 441 (100.0%) participants. Primary outcome Hand eczema severity index at 12 weeks In terms of relative change, the median [interquartile range (IQR)] score at 12 weeks was equal to 30% (10–70%) of that at baseline for the alitretinoin group compared with 50% (20–100%) in the Immersion PUVA group. There was a statistically significant benefit of alitretinoin compared with Immersion PUVA at 12 weeks, with an estimated fold change of 0.66 (0.52, 0.82), p = 0.0003 at 12 weeks. Secondary end points Hand eczema severity index over 52 weeks There was no evidence of a difference between alitretinoin and Immersion PUVA at 24 weeks or 52 weeks, with the estimated fold change [95% confidence interval (CI)] equal to 0.92 (0.798 to 1.08) and 1.27 (0.97 to 1.67), respectively. Modified total lesion symptom score over 52 weeks There was no evidence of a difference between treatment groups over 52 weeks, with the estimated difference (95% CI) equal to −0.37 (−1.23 to 0.48) at 12 weeks, −0.14 (−0.84 to 0.56) at 24 weeks and 0.41 (−0.59 to 1.40) at 52 weeks. Physician’s Global Assessment over 52 weeks At 12 weeks, the proportion of participants with available data achieving clear or almost clear (via blinded assessment) was 27.6% (47/170) for alitretinoin and 23.6% (35/148) for Immersion PUVA. Over 52 weeks, 59.4% (123/207) of alitretinoin participants with available data achieved at least one clear/almost clear assessment compared with 61.5% (118/192) of Immersion PUVA participants. There was no evidence of a difference between treatment groups over 52 weeks, with the estimated odds ratios (alitretinoin vs. Immersion PUVA) for achieving lower PGA scores and 95% CI equal to 1.22 (0.90 to 1.64) at 12 weeks, 1.18 (0.89 to 1.56) at 24 weeks and 1.10 (0.64 to 1.89) at 52 weeks. Dermatology quality of life index over 52 weeks There was a statistically significant benefit of alitretinoin compared with Immersion PUVA, with mean scores (95% CI) estimated to be 0.95 (0.09 to 1.82) lower at 12 weeks for the alitretinoin group compared with Immersion PUVA. There was no statistically significant treatment effect at 24 weeks [estimated difference in mean scores = −0.18 (−0.92 to 0.56)], but at 52 weeks, there was a statistically significant treatment effect with increased scores in the alitretinoin group compared with the Immersion PUVA group [estimated difference in mean scores = −1.62 (−2.62 to −0.62)]. The differences observed in the DLQI were not clinically significant. Person-Centred Dermatology Self-Care Index at 12 weeks At baseline, 16.1% (n = 71) were assessed as having sufficient knowledge, ability and confidence to self-manage their condition, with a similar distribution of education needs across treatment groups. At 12 weeks, the proportion of participants with available data who had sufficient knowledge and education was 26.2% (117/324), 28.0% (49/175) in the alitretinoin group and 24.2% (36/149) in the Immersion PUVA group. There was no evidence of a difference in terms of educational need between treatment groups, with an estimated odds ratio (alitretinoin vs. Immersion PUVA) of 0.65 (0.39 to 1.08). Patient Benefit Index – hand eczema over 52 weeks At 12 weeks, the median (IQR) score was equal to 2.3 (1.4–3.2) in the alitretinoin group and 1.9 (0.9–2.8) in the Immersion PUVA group. At 24 weeks, the median (IQR) score was unchanged at 2.3 (1.2–3.5) in the alitretinoin group but had increased to 2.8 (1.4–3.5) in the Immersion PUVA group. At 52 weeks, the median (IQR) score was 2.6 (1.6–3.3) in the alitretinoin group and 3.0 (1.7–3.6) in the Immersion PUVA group. Time to relapse At treatment phase completion, 34.1% (N = 75) of participants allocated to alitretinoin achieved a clear or almost clear response, compared with 25.8% (N = 57) allocated to Immersion PUVA. When relapse was defined as 75% of the baseline HECSI score, 20% (n = 15) of alitretinoin responders relapsed compared with 15.8% (n = 9) of Immersion PUVA responders. When relapse was defined as 50% of the baseline HECSI score, 37.3% (n = 28) of alitretinoin responders and 36.8% (n = 21) of Immersion PUVA responders relapsed. Health economic analysis Within-trial estimated costs indicate that Immersion PUVA is more costly than alitretinoin at weeks 12 and 52 (week 12: £3236 vs. £1904; week 52: £4424 vs. £3336, respectively). Treatment costs are the main drivers of Immersion PUVA costs (including therapy cost, medication and follow-up). Out-of-pocket expenditures are also higher for patients assigned to Immersion PUVA. These increase the cost difference at weeks 12 and 52 (from £1333 to £1650 at week 12 and from £1081 to £1841 at week 52). Immersion PUVA’s QALYs are higher at both time points (week 12: 0.165 vs. 0.159 and week 52: 0.798 vs. 0.761). When using DLQI, the estimated QALYs were slightly higher for alitretinoin compared with Immersion PUVA (week 12: 0.158 vs. 0.154; week 52: 0.784 vs. 0.782, respectively). When combined, the incremental cost-effectiveness ratios (ICERs) indicate alitretinoin is cost-effective at weeks 12 and 52 in all scenarios (primary analysis, societal perspective, and QALYs derived from DLQI). These results are robust, as the probability of cost-effectiveness is between 96% and 100%. Long-term cost-effectiveness The 10-year cost estimates indicate that patients allocated to alitretinoin are slightly more costly than Immersion PUVA (£5432 vs. £5361, respectively). In terms of QALYs, Immersion PUVA is slightly more effective (6.535 vs. 6.530, respectively). The latter suggests Immersion PUVA is the cost-effective strategy; however, results are uncertain, as only 50% of the probabilistic analysis iterations indicate this is the case. Safety data Seventy-nine participants had 135 reportable AEs, 25.0% (n = 55) participants allocated to alitretinoin and 10.9% (n = 24) participants allocated to Immersion PUVA. There were four serious AEs (SAEs; two alitretinoin, two Immersion PUVA); one suspected/expected to be related to treatment, and three related to underlying CHE. There were four pregnancies (three alitretinoin, one Immersion PUVA), including one alitretinoin patient in active treatment, who stopped treatment immediately. Treatment compliance In total, 212 (96.4%) participants randomised to alitretinoin and 196 (88.7%) randomised to Immersion PUVA received treatment. Full ‘trial’ compliance (≥ 80% received and no treatment breaks > 7 days during the first 12 weeks) was observed in 65.9% (N = 145) participants allocated to alitretinoin and 24.0% (N = 53) allocated to Immersion PUVA. Subgroup analysis No differential treatment effects were observed within subgroups defined by disease duration, clinical phenotype, disease severity, presence of atopy, filaggrin loss of function mutation, smoking history, BMI, foot involvement or biomarkers identified through tape stripping. Conclusion Alitretinoin and Immersion PUVA both led to reductions in symptoms. As a first-line therapy, alitretinoin showed a more rapid improvement and was superior to Immersion PUVA in terms of primary end point at 12 weeks. This difference was not observed at later time points. Alitretinoin was more cost-effective over the short term (12 and 52 weeks) with robust results. Long-term analysis indicates that Immersion PUVA is cost-effective in the long term; these results, however, are uncertain. These findings will inform clinical management of patients with severe CHE. Alitretinoin may be considered as the recommended first-line therapy; however, Immersion PUVA may also be considered for longer-term outcomes, particularly in patients where alitretinoin is not an appropriate treatment. Recruitment was difficult, with perceived treatment pathway challenges. The trial implemented a self-referral service, through which 13.8% of randomised participants were recruited. The Immersion PUVA schedule was problematic, with more than 20% of participants declining participation citing this as their main reason. Immersion PUVA treatment compliance was poor, and regular twice-weekly treatment was not achieved by most patients. Hand eczema severity index (HECSI) was the chosen primary outcome measure because it incorporated more details about the condition compared with PGA, which had been used previously. Furthermore, HECSI is continuous rather than discrete, which meant a smaller sample size was required for a desired level of power. The results from ALPHA will help inform future CHE trial designs. Most participants had some educational need for how to manage their condition, and overall, there remains a need for better therapeutic approaches for severe CHE. Just 59% and 61% of patients allocated to alitretinoin and Immersion PUVA, respectively, were known to achieve a clear/almost clear assessment at any time point during the trial period. Further work includes deeper analysis of substudies and pilot data to inform future research, which is needed to understand the long-term effects of treatments. Trial registration This trial is registered as ISRCTN80206075. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 12/186/01) and is published in full in Health Technology Assessment; Vol. 28, No. 59. See the NIHR Funding and Awards website for further award information.

Background Speech impairments are common with Parkinson’s disease (reported prevalence 68%), increasing conversational demands, reliance on family and social withdrawal. Objective(s) The PD COMM trial compared the clinical and cost-effectiveness of two speech and language therapy approaches: Lee Silverman Voice Treatment LOUD and National Health Service speech and language therapy for the treatment of speech or voice problems in people with Parkinson’s disease to no speech and language therapy (control) and against each other. Design PD COMM is a phase III, multicentre, three-arm, unblinded, randomised controlled trial. Participants were randomised in a 1 : 1 : 1 ratio to control, National Health Service speech and language therapy or Lee Silverman Voice Treatment LOUD via a central computer-generated programme, using a minimisation procedure with a random element, to ensure allocation concealment. Mixed-methods process and health economic evaluations were conducted. Setting United Kingdom outpatient and home settings. Participants People with idiopathic Parkinson’s disease, with self-reported or carer-reported speech or voice problems. We excluded people with dementia, laryngeal pathology and those within 24 months of previous speech and language therapy. Interventions The Lee Silverman Voice Treatment LOUD intervention included maximum effort drills and high-effort speech production tasks delivered over four 50-minute therapist-led personalised sessions per week, for 4 weeks with prescribed daily home practice. National Health Service speech and language therapy content and dosage reflected local non-Lee Silverman Voice Treatment speech and language therapy practices, usually 1 hour, once weekly, for 6 weeks. Trained, experienced speech and language therapists or assistants provided interventions. The control was no speech and language therapy until the trial was completed. Main outcome measures Primary outcome: Voice Handicap Index total score at 3 months. Secondary outcomes: Voice Handicap Index subscales, Parkinson’s Disease Questionnaire-39; Questionnaire on Acquired Speech Disorders; EuroQol-5D-5L; ICEpop Capabilities Measure for Older Adults; Parkinson’s Disease Questionnaire – Carers; resource utilisation; and adverse events. Assessments were completed pre-randomisation and at 3, 6 and 12 months post randomisation. Results Three hundred and eighty-eight participants were randomised to Lee Silverman Voice Treatment LOUD (n = 130), National Health Service speech and language therapy (n = 129) and control (n = 129). The impact of voice problems at 3 months after randomisation was lower for Lee Silverman Voice Treatment LOUD participants than control [−8.0 (99% confidence interval: −13.3, −2.6); p = 0.001]. There was no evidence of improvement for those with access to National Health Service speech and language therapy when compared to control [1.7 (99% confidence interval: −3.8, 7.1); p = 0.4]. Participants randomised to Lee Silverman Voice Treatment LOUD reported a lower impact of their voice problems than participants randomised to National Health Service speech and language therapy [99% confidence interval: −9.6 (−14.9, −4.4); p 70 years); disease severity measured using the Hoehn and Yahr staging (1.0–2.5, 3.0–5.0) and severity of speech measured using the Voice Handicap Index (VHI) total score (≤ 33, mild 34–44, moderate 45–61, severe > 61). To avoid any possibility of the treatment allocation becoming predictable, a random element was included within the randomisation process. Once the participant was randomised into the trial, they were given a unique trial identifier and the treatment allocation was confirmed by e-mail to the site. The primary outcome measure for the trial was patient-reported VHI total score at 3 months. The VHI measures the psychosocial consequences of voice disorders and provides an overall perception of effectiveness of voice-related communication. The VHI comprises of 30 questions (0–120 negatively scored) divided into emotional, functional and physical subscales (0–40 subscale score). Secondary outcomes were: the VHI subscales; the Parkinson’s Disease Questionnaire-39, a validated, health-related quality-of-life measure specific to PD; the Questionnaire on Acquired Speech Disorders, self-reported participation restriction related to speech and communication; the participant-rated EuroQol5D (5-level version) which provides a simple descriptive profile and a single index value for health status; the ICEpop Capabilities Measure for Older Adults (ICECAP-O), a measure of capability in older people for use in economic evaluations; a disease-specific participant resource usage questionnaire; carer quality of life, measured using the PD Questionnaire – carer; and adverse events. For participants in either SLT arm, any vocal strain or abuse believed to be associated with treatment was identified by the therapists at the participants’ SLT session and was reported in the adverse event log. All participant-reported resource usage forms completed for the trial were checked to ensure that no vocal strain or abuse had occurred in outpatient appointments with ENT specialists. At the 12-month clinical visit, the medical professional also checked whether any adverse events had occurred since entering the trial. There were two nested studies within this trial: a process evaluation and an economic evaluation. A process evaluation of the content of the interventions examined participants’ and therapists’ reported content of a subset of the two SLT interventions. Individual participant therapy data recorded by participants and therapists during the PD COMM trial were extracted from treatment record forms and therapy notes. Extraction category headings were piloted and reflected the Template for Intervention Description and Replication intervention reporting guidelines, with ongoing refinement as extraction progressed and components of the interventions were identified which were inadequately accounted for in the initial framework. The process evaluation of the implementation of the interventions used normalisation process theory which is highly attuned to the challenges of complex interventions as it encourages looking at systems as a whole. We considered that its conceptual framework would assist the interpretation and synthesis of data and analysis to explain what implementation processes took place and the interactions and gaps between the PD COMM interventions, the changing context, speech and language therapists and their practice. A within-trial economic evaluation in the form of cost–utility analysis was conducted from the perspective of the NHS. The primary results were expressed in terms of cost per quality-adjusted life-year (QALY) gained at 12 months for the three trial comparisons. Additional secondary analyses were performed from the NHS, personal social services perspective, broader societal perspective and using the capability approach, which uses broader measures of capability well-being. Results The mean age of participants in PD COMM was approximately 70 years old, and 74% were male (286/388). The mean duration of PD was between 5 and 6 years, with a broad range from newly diagnosed to over 30 years, and the majority (61%) were in Hoehn and Yahr stage of 2 or less (

Background People with suspected prostate cancer are usually offered either a local anaesthetic transrectal ultrasound-guided prostate biopsy or a general anaesthetic transperineal prostate biopsy. Transperineal prostate biopsy is often carried out under general anaesthetic due to pain caused by the procedure. However, recent studies suggest that performing local anaesthetic transperineal prostate biopsy may better identify cancer in particular regions of the prostate and reduce infection rates, while being carried out in an outpatient setting. Devices to assist with freehand methods of local anaesthetic transperineal prostate may also help practitioners performing prostate biopsies. Objectives To evaluate the clinical effectiveness and cost-effectiveness of local anaesthetic transperineal prostate compared to local anaesthetic transrectal ultrasound-guided prostate and general anaesthetic transperineal prostate biopsy for people with suspected prostate cancer, and local anaesthetic transperineal prostate with specific freehand devices in comparison with local anaesthetic transrectal ultrasound-guided prostate and transperineal prostate biopsy conducted with a grid and stepping device conducted under local or general anaesthetic. Data sources and methods We conducted a systematic review of studies comparing the diagnostic yield and clinical effectiveness of different methods for performing prostate biopsies. We used pairwise and network meta-analyses to pool evidence on cancer detection rates and structured narrative synthesis for other outcomes. For the economic evaluation, we reviewed published and submitted evidence and developed a model to assess the cost-effectiveness of the different biopsy methods. Results We included 19 comparative studies (6 randomised controlled trials and 13 observational comparative studies) and 4 single-arm studies of freehand devices. There were no statistically significant differences in cancer detection rates for local anaesthetic transperineal prostate (any method) compared to local anaesthetic transrectal ultrasound-guided prostate (relative risk 1.00, 95% confidence interval 0.85 to 1.18) (n = 5 randomised controlled trials), as was the case for local anaesthetic transperineal prostate with a freehand device compared to local anaesthetic transrectal ultrasound-guided prostate (relative risk 1.40, 95% confidence interval 0.96 to 2.04) (n = 1 randomised controlled trial). Results of meta-analyses of observational studies were similar. The economic analysis indicated that local anaesthetic transperineal prostate is likely to be cost-effective compared with local anaesthetic transrectal ultrasound-guided prostate (incremental cost below £20,000 per quality-adjusted life-year gained) and less costly and no less effective than general anaesthetic transperineal prostate. local anaesthetic transperineal prostate with a freehand device is likely to be the most cost-effective strategy: incremental cost versus local anaesthetic transrectal ultrasound-guided prostate of £743 per quality-adjusted life-year for people with magnetic resonance imaging Likert score of 3 or more at first biopsy. Limitations There is limited evidence for efficacy in detecting clinically significant prostate cancer. There is comparative evidence for the PrecisionPoint™ Transperineal Access System (BXTAccelyon Ltd, Burnham, UK) but limited or no evidence for the other freehand devices. Evidence for other outcomes is sparse. The cost-effectiveness results are sensitive to uncertainty over cancer detection rates, complication rates and the numbers of core samples taken with the different biopsy methods and the costs of processing them. Conclusions Transperineal prostate biopsy under local anaesthetic is equally efficient at detecting prostate cancer as transrectal ultrasound-guided prostate biopsy under local anaesthetic but it may be better with a freehand device. local anaesthetic transperineal prostate is associated with urinary retention type complications, whereas local anaesthetic transrectal ultrasound-guided prostate has a higher infection rate. local anaesthetic transperineal prostate biopsy with a freehand device appears to meet conventional levels of costeffectiveness compared with local anaesthetic transrectal ultrasound-guided prostate. Study registration This study is registered as PROSPERO CRD42021266443. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR134220) and is published in full in Health Technology Assessment Vol. 28, No. 60. See the NIHR Funding and Awards website for further award information. Plain language summary A prostate biopsy can help determine if a person has prostate cancer. The main ways of performing a prostate biopsy involve taking small samples of the prostate out through the rectum (back passage) or through the perineum – the skin area between the anus and the scrotum (testicles). Both methods use ultrasound images from a probe inserted into the rectum to help the clinician see what they are doing. Taking samples through the rectum is usually carried out under local anaesthetic, whereas taking samples through the perineum is usually carried out under general anaesthetic. We wanted to find out if taking samples through the perineum under local anaesthetic (instead of general anaesthetic) would be equally effective at detecting prostate cancer as the other biopsy methods and whether there was any improvement or change in the sorts of side effects people may have. We also wanted to know if people found the biopsy painful or not. We carried out searches of computer research databases to find relevant clinical and cost-effectiveness studies and compared the effectiveness of the different biopsy methods they used. We read and summarised the results of the studies we found in our search. Our findings showed that taking biopsy samples through the perineum under local anaesthetic had rates of detecting prostate cancer similar to those of the other biopsy methods. But if the clinician also used a freehand device that helps guide the biopsy needle as part of the procedure, then this may be a better method for detecting cancer. The studies we found agreed that performing this prostate biopsy under local anaesthetic was not too painful for most people. Our economic estimates suggest that using a freehand device for local anaesthetic perineal (through the skin of the perineum) biopsy may be a cost-effective use of National Health Service resources. Scientific summary Background Prostate cancer accounts for 30% of all cancers diagnosed in men in the UK and the incidence is rising. It is more common in men over 45 years of age. Symptoms that cannot be attributed to other health conditions include lower back or bone pain, lethargy, erectile dysfunction, haematuria, weight loss and lower urinary tract symptoms. National Institute for Health and Care Excellence (NICE) guideline NG12 advises on recognition and referral of people presenting with possible prostate cancer. A prostate-specific antigen (PSA) test and digital rectal examination should be performed. If PSA levels are raised above normal or if the prostate feels malignant, then the person should be referred for suspected cancer. NICE guideline NG131 advises on diagnosis and management. It recommends a multiparametric magnetic resonance imaging (mpMRI) test with the results reported using a five-point Likert scale to indicate how likely the presence of prostate cancer is. The Likert scale score, or alternatively the Prostate Imaging Reporting and Data System (PI-RADS score, not mentioned in the NICE guideline), is used to assess whether the person is offered a prostate biopsy. People with a score of 3 or above should be offered a multiparametric magnetic resonance imaging (mpMRI)-influenced prostate biopsy. People with a score of 1 or 2 will discuss risks and benefits with a clinician and if a prostate biopsy goes ahead, it should be a systematic biopsy. Two main options for biopsy are transrectal ultrasound prostate biopsy under local anaesthetic (LATRUS) and transperineal prostate biopsy under general anaesthetic (GATP). Biopsies can be either targeted (based on mpMRI findings) or systematic (samples are taken according to a predefined scheme) or both. Recent studies suggest that performing transperineal prostate biopsy under local anaesthetic (LATP) could better identify cancer in particular regions of the prostate and could have lower infection rates than transrectal biopsies while also being able to be carried out in an outpatient setting. Transperineal prostate biopsy is usually carried out under general anaesthetic due to pain caused by the procedure and tolerability is a key issue. Various freehand devices to assist with LATP prostate biopsy are being introduced to the market. The six specific freehand devices specified in the NICE scope for this review are: Cambridge Prostate Biopsy Device (CamPROBE) (JEB Technologies Ltd, Suffolk, UK); EZU-PA3U (Hitachi Ltd, Tokyo, Japan); PrecisionPoint™ Transperineal Access System (BXTAccelyon Ltd, Burnham, UK); SureFire Guide (LeapMed, Jiangsu, China); Trinity® Perine Grid (KOELIS®, NJ, USA); UA1232 puncture attachment (BK Medical, MA, USA). Objectives The aim of this review is to evaluate the diagnostic yield, clinical effectiveness and cost-effectiveness of LATP prostate biopsies performed with or without available specialist devices and equipment, in people with suspected prostate cancer. Two decision questions were prioritised by NICE for this assessment, with input from relevant stakeholders: Decision question 1. Do LATP prostate biopsies in patients with suspected prostate cancer represent a clinically and cost-effective use of National Health Service (NHS) resources? Decision question 2. Do freehand transperineal biopsy devices for LATP prostate biopsies in patients with suspected prostate cancer represent a clinically effective and cost-effective use of NHS resources? There are five comparisons required to address the two decision questions in the NICE scope: LATP-any (using coaxial needle or grid and stepping device or freehand device) versus LATRUS LATP-any (using coaxial needle or grid and stepping device or freehand device) versus GATP LATP-freehand (freehand device only) versus LATRUS LATP-freehand (freehand device only) versus GATP LATP-freehand (freehand device only) versus LATP-grid and stepping device. Methods Systematic review of diagnostic test evaluation and clinical effectiveness A systematic review of diagnostic and clinical effectiveness evidence was conducted following a peer-reviewed protocol. Searches were based on a comprehensive search strategy. Bibliographic databases, including MEDLINE, EMBASE, Web of Science, The Cochrane Library and the International HTA database, were searched for English-language references in July 2021, and these searches were updated at the end of October 2021. Urology conferences and freehand-device company submissions were hand-searched, and reference lists of identified systematic reviews and meta-analyses were checked. Relevant studies were sought through contact with study authors and NICE Specialist Committee members. Studies were eligible if they included people with suspected prostate cancer with an indication for prostate biopsy and reported diagnostic yield, for example, cancer detection rates, or other clinical or patient-reported outcomes. The eligible interventions were any LATP biopsy (of which LATP-freehand biopsy is a subset) and the eligible comparators were LATRUS and GATP; the LATP-grid and stepping device was an eligible comparator when compared with the LATP-freehand intervention. The Cochrane risk of bias tool (version 1) was used to assess risk of bias for the included randomised controlled trials (RCTs) and The Joanna Briggs Institute critical appraisal checklists were used to assess the included observational studies. Two reviewers carried out study selection, data extraction and critical appraisal, with any disagreements resolved through discussion and referred to a third reviewer for resolution as necessary. We conducted meta-analysis of the cancer detection rate outcomes for which sufficient comparative data were available. Pairwise meta-analysis was conducted for the above comparisons, with randomised and non-randomised studies analysed separately. Network meta-analysis was conducted for the two decision questions specified in the NICE scope. We synthesised the data for other outcomes narratively, as evidence was too sparse for meta-analysis. Review of economic evaluations We conducted a systematic review of the cost-effectiveness of the prostate biopsy methods in scope. The search strategies were based on an early version of the clinical effectiveness searches with the addition of an economics search filter. Included studies were full economic evaluations that assessed both costs and consequences for the different prostate biopsy methods. Outcomes included measures of resource use and costs and health outcomes: life-years or quality-adjusted life-years (QALYs) gained. Economic evaluations not meeting the inclusion criteria and studies that reported on resource use and costs, and health-related quality of life (utilities) were assessed as potential sources of information for the economic model. External Assessment Group independent economic assessment We developed a decision model to estimate the cost-effectiveness of alternative biopsy methods for people referred for biopsy with suspected prostate cancer. The model includes a decision tree to estimate diagnostic outcomes and biopsy-related complications, and a Markov model that predicts the long-term costs and consequences of false-negative biopsy results. We assessed cost-effectiveness for four subgroups at different prior levels of risk, based on previous mpMRI results (Likert 1 or 2; or Likert 3 or more) and history of prostate biopsy (none; previous negative biopsy). The decision tree used published results from the economic evaluation of the Prostate MR imaging study (PROMIS) to estimate baseline prevalence in the subgroups of interest, and diagnostic yield of LATRUS biopsy. Cancer detection rates were adjusted for the other biopsy methods using relative risks (RRs) from our network meta-analyses, and evidence from the literature on biopsy complication rates and the probability of repeat biopsy. Costs of the biopsy methods were estimated in a microcosting analysis, as well as from submitted evidence and published sources. The Markov model was a replicated version of a model developed for the 2019 update of the NICE guideline (NG131). Model parameters were based on those in the NG131 model, with some adjustments to costs and utilities from more recent published sources. Results Systematic review of diagnostic test evaluation and clinical effectiveness The literature searches identified a total of 1969 references of which 111 references were subjected to full-text screening. Twenty-seven publications reported 23 studies meeting the inclusion criteria for this review: 19 comparative studies of which 6 were RCTs and 13 were observational studies (1 of which is unpublished); and 4 single-arm studies for LATP-freehand devices where no comparative evidence was identified. There were no statistically significant differences in cancer detection rates for LATP (any method) compared to LATRUS with RR = 1.00 [95% confidence interval (CI) 0.85 to 1.18] (n = 5 RCTs). A single randomised trial estimated a non-significant difference in cancer detection rates in favour of LATP using a freehand device (PrecisionPoint) compared to LATRUS RR = 1.40, 95% CI 0.96 to 2.04. This finding was supported by meta-analysis of observational comparative studies RR = 1.21 (95% CI 1.08 to 1.34) (n = 4 studies). There were no statistically significant differences in cancer detection rates in meta-analyses comparing other biopsy methods. Evidence from the systematic review for other clinical outcomes, biopsy-related complications and patient-reported outcomes was sparse. Review of economic evaluations One economic evaluation was eligible for inclusion in the economic review out of 725 results from the original and update searches. This study evaluated the CamPROBE (LATP-freehand) device versus LATRUS for use in diagnosing prostate cancer from the perspective of the UK NHS. It used a decision-tree model with a Markov model at the terminal nodes and was informed by a prospective case series for the CamPROBE device and data from the PROMIS study. The study suggested that compared with LATRUS, LATP using the CamPROBE freehand device would be cost saving at a device price below £41 per procedure, or more cost-effective at the £20,000 per QALY threshold with a price below £81 per procedure. These calculations assume a zero rate of infection for LATP and equal diagnostic accuracy for LATP using CamPROBE and LATRUS. We considered 13 excluded economic studies as sources to inform our model structure and inputs, including the cost-effectiveness analysis for the PROMIS study and the analysis for the update of the NICE guideline on prostate cancer published in May 2019 (NG131). Evidence from the BXTAccelyon company submission included a cost minimisation study developed in 2020 by the York Health Consortium that compared the costs of LATP (with the PrecisionPoint freehand device) with different combinations of LATRUS and GATP for UK NHS Trusts. The study suggests that LATP using the PrecisionPoint freehand device is cost saving, assuming equal diagnostic yield of the different biopsy methods. Independent economic assessment The base-case economic analysis comparing LATP (all methods) with LATRUS and GATP indicated that LATP is likely to be the most cost-effective option in all four subgroups, with incremental cost-effectiveness ratio (ICER) estimates for LATP compared with LATRUS below £20,000 per QALY gained, and GATP estimated as more expensive and less effective than LATP. These conclusions were supported by probabilistic sensitivity analysis and a wide range of scenario analyses, although the results for LATP compared with LATRUS were sensitive to some alternative sources of cancer detection rates, rates of biopsy-related hospital admissions, numbers of core samples and histopathology costs. The economic analysis including LATP-freehand compared with other LATP methods, as well as LATRUS and GATP, indicated that LATP with a freehand device was the most cost-effective strategy, with an ICER of £743 per QALY for the highest-risk subgroup with MRI Likert score of 3 or more at first biopsy, and £4595 per QALY for the subgroup with a MRI Likert score 1 or 2 at first biopsy. For the subgroups with a previous negative biopsy, the ICER remained below £20,000 per QALY. Again, probabilistic sensitivity analysis supported these results, but scenario analysis highlighted uncertainty related to the cost of the devices, the number of core samples and costs of processing them, and the use of other sources of evidence for cancer detection and biopsy-related complication rates. The more favourable cost-effectiveness estimates for LATP with a freehand device are mostly driven by the cancer detection rates, which rest on a single RCT for LATP with a freehand device (PrecisionPoint). In the scenario based on observational evidence of cancer detection rates, the ICERs for LATP with a freehand device were less favourable, although still well below £20,000 per QALY. Increasing the cost of LATP with a freehand device by assuming the cost of the most expensive device, the ICER remained below £20,000 per QALY for the highest-risk subgroup but not for the other subgroups. Conclusions Transperineal prostate biopsy under local anaesthetic is equally efficient at detecting prostate cancer as transrectal ultrasound-guided prostate biopsy under local anaesthetic but evidence from one RCT, supported by observational studies, suggests that it might be better when using a freehand device. Local anaesthetic transperineal prostate biopsy is associated with urinary retention-type complications, whereas local anaesthetic transrectal ultrasound-guided prostate biopsy has a higher infection rate. Economic evaluation suggests that LATP with a freehand device is likely to be cost-effective compared with LATP with other methods, LATRUS and GATP for patients with no previous biopsy at high risk of having prostate cancer indicated by previous MRI results. This result is sensitive to the estimated cost of the freehand device, the number of and cost of core samples taken, and the sources for biopsy complication rates. Recommendations for research Evidence for freehand devices. There was no comparative evidence for several of the freehand devices in the NICE scope. The TRANSLATE study is expected to help address this question, as it is evaluating the PrecisionPoint, UA1232 and ‘any ultrasound probe-mounted needle guidance device’. Outcomes not covered in included available evidence. We suggest that incidence of defined complications (standardised for grading of severity and length of follow-up), health-related quality of life and longer-term clinical outcomes could be defined in a core outcome set. LATP versus GATP. Evidence for this comparison is sparse (we identified one RCT reporting cancer detection rates). Repeat biopsy population. There is a need for separate reporting of results for this subgroup, or a separate prospective RCT. UK NHS setting. The three UK studies included in our review were single-centre observational studies with a limited set of outcomes. The TRANSLATE study is expected to remedy this; it is a multicentre randomised study across nine NHS Trusts in England. Study registration This study is registered as PROSPERO CRD42021266443. Funding This awardproject was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR134220) and is published in full in Health Technology Assessment Vol. 28, No. 60. See the NIHR Funding and Awards website for further award information.

Background Chronic migraine is a disabling condition, affecting 2–4% of adults globally. With the introduction of expensive calcitonin gene-related peptide monoclonal antibodies, it is timely to compare the clinical effectiveness and cost-effectiveness of preventive drugs for chronic migraine. Objective To assess the clinical effectiveness and cost-effectiveness of medications used for chronic migraine through systematic reviews and economic modelling. Eligibility criteria Randomised controlled trials of drug treatments for efficacy with > 100 participants with chronic migraine per arm; for adverse events > 100 participants with episodic or chronic migraine per arm. Previous economic analyses of preventive drugs for chronic migraine. Data sources Eight databases. Reviews methods Systematic reviews, network meta-analysis and economic modelling. Outcomes Monthly headache days, monthly migraine days, headache-related quality of life, cost-effectiveness. Results We found 51 individual articles, reporting 11 randomised controlled trials, testing 6 drugs (topiramate, Botox, eptinezumab, erenumab, fremanezumab, galcanezumab), versus placebo, on 7352 adults with chronic migraine. Calcitonin gene-related peptide monoclonal antibodies, Botox and topiramate reduced headache/migraine days by 2.0–2.5, just under two, or by less than 1.5 days per month, respectively. In the network meta-analysis, eptinezumab 300 mg and fremanezumab monthly ranked in first place in both monthly headache day and monthly migraine day analyses. The calcitonin gene-related peptide monoclonal antibodies were consistently the best choices for headache/migraine days and headache-related quality of life. Topiramate was very unlikely to be the best choice for headache/migraine days and headache-related quality of life when compared to calcitonin gene-related peptide monoclonal antibodies or Botox. We found no trials of the commonly used drugs, such as propranolol or amitriptyline, to include in the analysis. The adverse events review included 40 randomised controlled trials with 25,891 participants; 3 additional drugs, amitriptyline, atogepant and rimegepant, were included. There were very few serious adverse events – none of which were linked to the use of these medications. Adverse events were common. Most people using some calcitonin gene-related peptide monoclonal antibodies reported injection site issues; and people using topiramate or amitriptyline had nervous system or gastrointestinal issues. The cost-effectiveness review identified 16 studies evaluating chronic migraine medications in adults. The newer, injected drugs are more costly than the oral preventatives, but they were cost-effective. Our economic model showed that topiramate was the least costly option and had the fewest quality-adjusted life-year gains, whereas eptinezumab 300 mg was more costly but generated the most quality-adjusted life-year gains. The cost-effectiveness acceptability frontier showed that topiramate was the most cost-effective medication if the decision maker is willing to pay up to £50,000 per quality-adjusted life-year. Our consensus workshop brought together people with chronic migraine and headache experts. Consensus was reached on the top three recommendations for future research on medications to prevent chronic migraine: (1) calcitonin gene-related peptide monoclonal antibodies and Botox versus calcitonin gene-related peptide monoclonal antibodies, (2) candesartan versus placebo and (3) flunarizine versus placebo. Limitations Topiramate was the only oral drug for which we were able to include data. We did not find sufficient quality evidence to support the use of other oral drugs. Conclusions We did not find evidence that the calcitonin gene-related peptide monoclonal antibodies are more clinically and cost-effective when compared to topiramate or Botox. We identified directions for future research these drugs might take. Study registration This study is registered as PROSPERO CRD42021265990, CRD42021265993 and CRD42021265995. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR132803) and is published in full in Health Technology Assessment; Vol. 28, No. 63. See the NIHR Funding and Awards website for further award information. Plain language summary What is the problem? Chronic migraine is a disabling condition that can destroy work and family life. Treatments include cheap tablets (e.g. amitriptyline, propranolol and topiramate), Botox and expensive new drugs (the calcitonin gene-related peptide monoclonal antibodies). It is not known which of these drugs is the best choice. What did we want to find out? We wanted to find out which of these drugs works best. We wanted to know if they reduced the number of headache/migraine days and improved headache-related quality of life, how many side effects people experienced, and if they provided good value for the National Health Service. How did we do this? We first looked for research comparing these drugs to placebo (fake) drugs, and to each other. We then worked out which provide best value for money. What did we find out? Calcitonin gene-related peptide monoclonal antibodies reduced headache/migraine days by 2.0–2.5 days per month; Botox reduced headache/migraine days per month by around 1.9; and topiramate reduced headache/migraine days by 1.1–1.5 days per month. Many people taking topiramate or amitriptyline have nervous system and/or stomach/bowel side effects. Some people using calcitonin gene-related peptide monoclonal antibodies reported side effects associated with injections. Some calcitonin gene-related peptide monoclonal antibodies and Botox provide worthwhile benefits on headache-related quality of life. We were not able to identify any studies of sufficient quality to assess the effectiveness of other oral drugs. The best value drug was topiramate which gave better health outcomes at a lower cost than the placebos. What does this mean? After sharing the results with a panel of people with chronic migraine and headache experts, we identified a need for new studies comparing commonly used cheap oral drugs with placebo, Botox and calcitonin gene-related peptide monoclonal antibodies. Scientific summary Background Chronic migraine is a profoundly disabling condition and affects 2–4% of the world’s adult population. It is defined as headaches on 15 days or more a month with features of migraine on at least 8 of those days. Since 2020, expensive calcitonin gene-related peptide (CGRP) monoclonal antibodies (MAbs) have become established as specific treatments for people with chronic migraine who have failed to improve with other medications. Little is known about the effectiveness of these drugs when compared with each other, or with other well-established, cheaper, oral drugs used to treat chronic migraine. Therefore, it is timely to compare the clinical effectiveness and cost-effectiveness of these medications to treat chronic migraine. We set out to address the following research question: What is the clinical effectiveness and cost-effectiveness of prophylactic drug treatments for people with chronic migraine? Objectives Our overall aim is to produce evidence needed for people with chronic migraines and their doctors to make more informed decisions about prophylactic medications for chronic migraine. Our objectives were: What is the comparative effectiveness of prophylactic drugs for chronic migraine? What are the comparative incidences of adverse events (AEs) of prophylactic drugs used for migraine? What is known about the cost-effectiveness of prophylactic drugs for chronic migraine? Which prophylactic drugs for the management of chronic migraine are the most cost-effective? Based on our findings, what should the research recommendations be? Methods Systematic reviews of trial evidence on: The clinical effectiveness of prophylactic medications for chronic migraine; analyses included headache days, migraine days and headache-related quality of life: migraine- specific quality of life (MSQ); headache impact test-6 (HIT-6). Only randomised controlled trials (RCTs) with at least 100 people per arm were included. We report the comparative effectiveness using a network meta-analysis (NMA) for these different outcomes to see which drug was the most ‘effective’. To identify the comparative incidence of AEs of prophylactic drugs used for chronic or episodic migraine. RCTs with at least 100 people per arm were included. The cost-effectiveness studies of prophylactic drugs used for treatment of chronic migraine. We developed an economic model comparing the cost-effectiveness of prophylactic drugs for chronic migraine for the adult population from a National Health Service (NHS) and personal social services (PSS) perspective. The base-case analysis used a 2-year time horizon, with a starting age of 30 years for the patient cohort. Health states in the model were based on effectiveness data [reduction in the mean difference (MD) in monthly headache days (MHDs)] from the NMA. Costs are in 2021–2 prices and utilities were estimated based on EuroQol-5 Dimensions, five-level version (EQ-5D-5L) scores from the CHESS trial using the Hernandez-Alava crosswalk algorithm. Cost-effectiveness was measured in terms of an incremental cost per quality-adjusted life-year (QALY) gained [Institute for Clinical and Economic Review (ICER)]. Probabilistic sensitivity analysis (PSA) was undertaken to account for uncertainty in model parameters. Uncertainty around the cost-effectiveness of the various medications showing which is the preferred strategy is presented using a cost-effectiveness acceptability frontier (CEAF). At the end of the project, we held a consensus workshop bringing together people with chronic migraine and clinicians and other health professionals who are experts in chronic migraine. We presented the findings from our reviews, NMA, the economic model and some potential recommendations. We then split into groups (mixed with health professionals and participants) and asked them to discuss our suggested research recommendations, identify any other recommendations, and rank these recommendations in terms of priority. We then had another breakaway session, where all participants with chronic migraine met and all health professionals met. Finally, everyone was bought back together to discuss their rankings as a wider group and to reach a consensus using anonymous polling. Results The clinical effectiveness review focused on prophylactic medications which might be used in the UK for the prevention of chronic migraine. We found 11 RCTs reported across 51 individual publications, involving 7352 adult participants with chronic migraine, which showed that all pharmacological treatments for all outcomes of interest were beneficial in preventing migraine when compared to placebo. There were no trials of sufficient quality of the commonly used drugs, such as amitriptyline, candesartan, flunarizine or propranolol. Overall, the CGRP MAbs reduced headache/migraine days by 2.0 to 2.5 days per month. The most effective medication in reducing MHDs was eptinezumab 300 mg which reduced MHDs by 2.46 [95% credible interval (CrI) 3.24 to −1.67] days. The most effective medication in reducing monthly migraine days (MMDs) was fremanezumab monthly which reduced MMDs by 2.76 (95% CrI −3.36 to −2.15) days. Botox (BTA) reduced MHDs by 1.87 (95% CrI −2.55 to −1.18) days per month and MMDs by 1.96 (95% CrI −2.69 to −1.24) days per month. Topiramate was the least effective, prescribable drug and only reduced headache/migraine days by less than 1.5 fewer headache/migraine days per month. The NMA results showed that eptinezumab 300 mg had the highest probability ranking to reduce MHDs and MMDs – Surface Under the Cumulative Ranking Area (SUCRA) was 0.88 and 0.77, respectively. The CGRP MAbs provided a worthwhile improvement on the HIT-6 measure of headache-related quality of life (eptinezumab 300 mg reducing the HIT-6 by a score of 3.22 points); BTA had a worthwhile effect on the HIT-6 measure, reducing the HIT-6 score by 2.10 points; and there was no convincing benefit of topiramate on the MSQ measure. Galcanezumab 120 mg provided the best improvement in quality of life for the preventative role dimension of migraine-specific quality of life (MSQ-PR) (MD 6.97, 95% CrI 3.79 to 10.24, SUCRA 0.88), but for two other dimensions of the MSQ, erenumab 140 mg was superior to other treatments: for migraine-specific quality of life-restrictive role (MSQ-RR) – MD: 7.28, 95% CrI: 3.05 to 11.65, SUCRA 0.75, and for migraine- specific quality of life-emotional function (MSQ-EF) – MD: 8.89, 95% CrI: 3.20 to 14.55, SUCRA 0.79. The results from the quality assessment using the revised Cochrane risk-of-bias (RoB 2) tool for RCTs found that approximately 46% of the included RCTs in this review had low RoB and 36% of the RCTs had some concerns of bias. The incidence of AEs and serious adverse events (SAEs) review used evidence from 40 RCTs reported across 67 articles, which investigated pharmacological interventions to manage chronic or episodic migraine. These trials included 25,891 participants and 3 additional drugs were included – amitriptyline, atogepant and rimegepant. There were very few SAEs – none of which were linked to the use of these drugs. Non-SAEs were common, and results suggested that all the pharmacological medications included in this review were found to be tolerable. There were differences in the incidence of AEs between the CGRP MAbs, with most people using fremanezumab and one in four people using galcanezumab reporting injection site issues. These issues were much less common in people using eptinezumab or erenumab. Most people using topiramate or amitriptyline had nervous system or gastrointestinal side effects; topiramate was also linked to a higher prevalence of psychiatric disorders; and AEs related to BTA were uncommon. The cost-effectiveness review identified nine peer-reviewed journal articles and seven published reports of chronic migraine prophylactic medications in the adult population. All articles were model-based evaluations, and none were trial-based economic evaluations. We found that although these newer drugs (BTA and CGRP MAbs) were more costly than the oral preventatives, they were however deemed cost-effective. Generally, the articles were classed as high quality when appraised by the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) reporting tool. We developed a Markov (state-transition) model to assess the cost-effectiveness of different pharmacological medications to treat or prevent chronic migraine in the adult population. Our base-case deterministic results showed when comparing each of the medications separately against placebo, topiramate dominated placebo (cheaper and more effective); and each of the other medications, when compared separately, were more expensive than placebo; however, they generated more QALYs than placebo. The best value medication when compared with placebo was BTA, with the cost per QALY around £25,000. When comparing all medications together, the deterministic results showed that topiramate was the least costly option and had slightly more QALY gains than placebo, whereas eptinezumab 300 mg was the more costly option but generated the most QALY gains. Most medications were eliminated due to dominance. The ICER between BTA and topiramate and the ICER between eptinezumab 300 mg and BTA were not within plausible cost-effectiveness thresholds. Probabilistic results were similar to deterministic results. The CEAF showed that when comparing all medications topiramate was the most cost-effective medication if the decision maker is willing to pay up to £50,000 per QALY. None of the CGRP MAbs represented good value for money in this comparative analysis. Extensive sensitivity analyses showed that when MHDs is used as an outcome measure, the results were generally in line with the base-case results. The main exception was when using MMDs as an outcome measure instead of MHDs, fremanezumab monthly generated more QALY gains than eptinezumab 300 mg; the ICERs between the plausible options, once any dominated options were removed, were not within plausible cost-effectiveness thresholds. Our consensus workshop brought together 8 participants with chronic migraine and 11 health professionals with expertise in chronic migraine to set research priorities for preventive drugs for chronic migraine. Each of the small groups found that the need for trials of cheaper, oral medications, tricyclic antidepressants, serotonin and norepinephrine reuptake inhibitors (SNRIs) when compared with placebo were ranked highly; and for trials comparing the medications with each other, the CGRP MAbs and BTA separately or in combination with each other were ranked highly. The final (anonymised) rankings showed that the top three research priorities versus placebo were: (1) candesartan, (2) flunarizine and (3) melatonin; and for medications compared with each other were: (1) CGRP MAbs and BTA versus CGRP MAbs, (2) CGRP MAbs versus BTA and (3) a multi-arm trial of CGRPs MAbs receptor (erenumab) versus CGRP MAbs ligand (eptinezumab, fremanezumab and galcanezumab). In terms of priority, a consensus was established regarding the three most recommended medication comparisons for treating chronic migraine: (1) CGRP MAbs and BTA versus CGRP MAbs, (2) candesartan compared to placebo and (3) flunarizine in comparison to placebo. Discussion and conclusions Of the treatments included in the NMA, the CGRP MAbs overall were consistently the best choices for headache days, migraine days and headache-related quality of life. BTA was less likely than CGRP MAbs to be the best choice for headache days, migraine days and headache-related quality of life. Topiramate was very unlikely to be the best choice for headache days, migraine days and headache-related quality of life when compared to CGRP MAbs or BTA. The economic model found that topiramate was the best value drug if you are prepared to pay up to £50,000 per QALY. It is likely that CGRP MAbs are likely to be cost-effective in people who have failed treatment with BTA. At the workshop, general consensus was agreed on the top three choices of medication for chronic migraine. Topiramate was the only established oral drug for which we were able to include data. It is disappointing that we did not find a sufficient quality evidence base to support the use of drugs, such as amitriptyline, candesartan, flunarizine and propranolol that are recommended by National Institute for Health and Care Excellence (NICE) and/or Scottish Intercollegiate Guidelines Network (SIGN). Our consensus meeting identified the need for trials comparing candesartan and flunarizine with placebo as the top priorities for placebo-controlled trials. Only for topiramate can we make any observations for how this may compare with CGRP MAbs. The CGRP MAbs appear to be clinically superior, but even so topiramate, in spite of its high incidence of AEs, represents the best value for money. Within the current care pathway, it is unlikely that CGRP MAbs will be recommended ahead of topiramate without a very substantial reduction on price. What is perhaps a more critical decision point is whether BTA or CGRP MAbs might be preferred as the first choice after failure of oral medication. Our findings support continuing with the current care pathway since our CEAF found that only topiramate met an acceptable threshold. Data from our health economics review, however, do support the use of CGRP MAbs after failure of BTA for chronic migraine. Our consensus group identified the direct comparison of BTA and CGRP MAbs as a key research question. They also identified the question of whether these drug effects might be additive. The effect sizes, in terms of mean monthly migraine/headache days for each of these drugs, are at best modest, the largest being 2.76 days for fremanezumab monthly dose. As these drugs work through different pathways, it might be that more substantial effects are possible. Adding together the effects of BTA and a CGRP MAb, assuming no negative interaction, might have a mean effect size of 4–5 days that would be transformative for many people with chronic migraine. Our consensus group identified the comparative, and additive, effects of BTA and CGRP MAbs as high priority research questions. In conclusion, we have summarised the existing clinical and cost-effectiveness data on preventive drugs for chronic migraine and identified which directions future research on these drugs might take. We did not find convincing evidence that the CGRP MAbs are more clinically effective and cost-effective compared to topiramate or BTA. Study registration This study is registered as PROSPERO CRD42021265990, CRD42021265993 and CRD42021265995. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR132803) and is published in full in Health Technology Assessment; Vol. 28, No. 63. See the NIHR Funding and Awards website for further award information.

Background Chronic pain is common and costly. Antidepressants are prescribed to reduce pain. However, there has not been a network meta-analysis examining all antidepressants across all chronic pain conditions, so effectiveness and safety for most antidepressants for pain conditions remain unknown. Objective To assess the efficacy and safety of antidepressants for chronic pain (except headache) in adults. Our primary outcomes were as follows: substantial pain relief (50%), pain intensity, mood and adverse events. Our secondary outcomes were as follows: moderate pain relief (30%), physical function, sleep, quality of life, Patient Global Impression of Change, serious adverse events and withdrawal. Design This was a systematic review with a network meta-analysis. We searched CENTRAL, MEDLINE, EMBASE, CINAHL, LILACS, AMED and PsycINFO databases for randomised controlled trials of antidepressants for chronic pain conditions up until 4 January 2022. The review was registered in PROSPERO (CRD42020171855), and the protocol was published in the Cochrane Library (https://doi.org/10.1002/14651858.CD014682). Setting We analysed trials from all settings. Participants We included trials in which participants had chronic pain, defined as longer than 3 months, from any condition excluding headache. Interventions We included all antidepressants. Main outcome measures Our primary outcome was substantial pain relief, defined as a reduction ˃ 50%. We also measured pain intensity, mood and adverse events. Secondary measures included moderate pain relief (above 30% reduction), physical function, sleep, quality of life, Global Impression of Change, serious adverse events, and withdrawal from trial. Results We identified 176 studies with a total of 28,664 participants. Most studies were placebo-controlled (n = 83) and parallel armed (n = 141). The most common pain conditions examined were fibromyalgia (59 studies), neuropathic pain (49 studies) and musculoskeletal pain (40 studies). The average length of randomised controlled trials was 10 weeks. Most studies measured short-term outcomes only and excluded people with low mood and other mental health conditions. Across efficacy outcomes, duloxetine was consistently the highest-ranked antidepressant with moderate- to high-certainty evidence. Standard dose was equally efficacious as high dose for the majority of outcomes. Milnacipran was often ranked as the next most efficacious antidepressant, although the certainty of evidence was lower than that for duloxetine. There was insufficient evidence to draw robust conclusions for the efficacy and safety of any other antidepressant for chronic pain. Limitations The evidence for antidepressants other than duloxetine is poor. For duloxetine, it is not clear whether the effect applies to groups with both pain and low mood, since these groups were excluded from trials. There is also insufficient evidence on long-term outcomes and on adverse effects. Conclusions There is only reliable evidence for duloxetine in the treatment of chronic pain. Duloxetine was moderately efficacious across all outcomes at standard dose. There is also promising evidence for milnacipran, although further high-quality research is needed to be confident in these conclusions. Data for all other antidepressants were of low certainty. However, the findings should not be read as an encouragement to prescribe antidepressants where other non-pharmacological intervention could be equally effective, especially in the absence of good evidence on side effects and safety. Future work There is a need for large, methodologically sound trials testing the effectiveness of antidepressants for chronic pain. These trials should examine long-term outcomes (> 6 months) and include people with low mood. There should also be better reporting of adverse events, tolerance of drugs, and long-term compliance. Study registration This study is registered as PROSPERO CRD42020171855. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128782) and is published in full in Health Technology Assessment; Vol. 28, No. 62. See the NIHR Funding and Awards website for further award information. Plain language summary What was the question? Chronic pain is pain that lasts for more than 3 months. Over one-third of people across the world experience chronic pain. This often has a detrimental impact on people’s mood, disability and well-being. Antidepressants are often prescribed to reduce pain, but we are not sure which antidepressants work best for different types of pain, or whether they are safe. We wanted to find out whether antidepressants were effective and safe for management of chronic pain. What did we do? We searched for studies that had compared any antidepressant with any other treatment for any type of chronic pain (except headache). We compared all the treatments against each other using a statistical method called network meta-analysis. This method allows us to rank the treatments in order of best to worst for each outcome. What did we find? We found 176 studies that included a total of 28,664 people with chronic pain. Most of the studies (83/176) compared an antidepressant with a placebo (which looks like the real medicine but does not have any medicine in it). The evidence from our analysis suggests that: Duloxetine is the antidepressant that we have the most confidence in. It was the best antidepressant for reducing pain and improving physical function. A standard dose of duloxetine was equally as effective for reducing pain as a high dose of duloxetine. Milnacipran was also effective at reducing pain, but we are not as confident in this result as in the one for duloxetine because there were fewer studies with fewer people involved. Aside from duloxetine and milnacipran, we do not have confidence in the results from any other antidepressant included in this review, and even for duloxetine and milnacipran, we do not know the long-term effects. It is important to recognise that the lack of evidence for the majority of antidepressants in this review does not necessarily equal a lack of benefit. Rather, this means that the large, high-quality trials required for us to be certain of an antidepressant’s effectiveness have not been undertaken. Altogether, although duloxetine and milnacipran are effective, the results of this review should not be read as an encouragement to prescribe antidepressants where other non-pharmacological intervention could be equally effective, especially in the absence of good evidence on side effects and safety. These conclusions were informed by our patient and public involvement group. Scientific summary Background Chronic pain is common in adults, and often has a detrimental impact upon physical ability, well-being, and quality of life. Previous reviews have shown that certain antidepressants may be effective in reducing pain with some benefit in improving patients’ global impression of change for certain chronic pain conditions. However, there has not been a network meta-analysis examining all antidepressants across all chronic pain conditions. Objectives Our objective was to assess the efficacy and safety of antidepressants for chronic pain (except headache) in adults. Our primary outcomes were as follows: substantial pain relief (50%), pain intensity, mood and adverse events. Our secondary outcomes were as follows: moderate pain relief (30%), physical function, sleep, quality of life, Patient Global Impression of Change (PGIC), serious adverse events and withdrawal. Search methods We searched CENTRAL, MEDLINE, EMBASE, CINAHL, LILACS, AMED and PsycINFO databases for randomised controlled trials (RCTs) of antidepressants for chronic pain conditions up until 4 January 2022. Selection criteria We included RCTs that examined antidepressants for chronic pain against any comparator. If the comparator was placebo, another medication, another antidepressant or the same antidepressant at different doses, then the study was required to be double-blind. RCTs with active comparators that were unable to be double-blinded (e.g. psychotherapy) were included but rated as at high risk of bias. We excluded RCTs where the follow-up was < 2 weeks and those with < 10 participants in each trial arm. We included any antidepressant at any dose, for any indication but used primarily for treatment of people with chronic pain and compared to placebo or active intervention. Participants We included adults (aged 18 years or older) reporting primary or secondary pain in any part of their body (except headache) as their primary complaint, that matched the International Association for the Study of Pain definition of chronic pain (i.e. at least 3 months’ duration). We included all trials regardless of the severity of participants’ chronic pain, although we extracted whether severity was part of the inclusion criteria of the individual studies. We excluded studies where the participants’ primary pain condition was headache or migraine. Data collection and analysis Two authors separately screened, extracted data and judged risk of bias. We synthesised the data using Bayesian network meta-analysis (NMA) and pairwise meta-analyses for each outcome and ranked the antidepressants in terms of their effectiveness using the surface under the cumulative ranking curve. We primarily used the Confidence in Network Meta-Analysis (CINeMA) framework and ‘Risk Of Bias due to Missing Evidence in Network meta-analysis’ (ROB-MEN) tool to assess the certainty of the evidence. Where it was not possible to use CINeMA and ROB-MEN due to the complexity of the networks, we used Grading of Recommendations, Assessment, Development and Evaluations (GRADE) to assess the certainty of the evidence. Main results This review and NMA included 176 studies with a total of 28,664 participants. The majority of studies were placebo-controlled (n = 83) and parallel-armed (n = 141). The most common pain conditions examined were fibromyalgia (59 studies), neuropathic pain (49 studies) and musculoskeletal pain (40 studies). The average length of RCTs was 10 weeks; seven studies provided no useable data and were omitted from the NMAs. The majority of studies measured short-term outcomes only and excluded people with low mood and other mental health conditions. Across efficacy outcomes, duloxetine was consistently the highest-ranked antidepressant with moderate- to high-certainty evidence. In duloxetine trials, standard dose was equally efficacious as high dose for the majority of outcomes. Milnacipran was often ranked as the next most efficacious antidepressant, although the certainty of evidence was lower than that of duloxetine. There was insufficient evidence to draw robust conclusions for the efficacy and safety of any other antidepressant for chronic pain. Primary efficacy outcomes For pain relief, duloxetine standard dose showed a small to moderate effect for substantial pain relief [odds ratio 0.91, 95% confidence interval (CI) 0.56 to 0.84] and continuous pain intensity [standardised mean difference (SMD) −0.31, 95% CI −0.39 to −0.24]. For pain intensity, milnacipran standard dose also showed a small effect (SMD −0.22, 95% CI −0.39 to 0.06) with moderate-certainty evidence. For mood, mirtazapine had a moderate effect (SMD −0.5, 95% CI −0.78 to −0.22), while duloxetine showed a small effect (−0.16, 95% CI −0.22 to −0.1); however, it is important to note that most trials excluded participants with mental health conditions, and so average anxiety and depression scores tended to be in the ‘normal’ or ‘subclinical’ ranges at baseline already. Secondary efficacy outcomes Across all secondary efficacy outcomes (moderate pain relief, physical function, sleep, quality of life and PGIC), duloxetine and milnacipran were the highest-ranked antidepressants with moderate-certainty evidence, although effects were small. For both duloxetine and milnacipran, standard doses were equally as efficacious as high doses. Safety There was very low-certainty evidence for all safety outcomes (adverse events, serious adverse events and withdrawal) across all antidepressants. We cannot draw any reliable conclusions from the NMAs for these outcomes. Authors’ conclusions Our review and NMAs show that despite studies investigating 25 different antidepressants, there is reliable evidence for only duloxetine in the treatment of chronic pain. Duloxetine was moderately efficacious across all outcomes at standard dose. There is also promising evidence for milnacipran, although further high-quality research is needed to be confident in these conclusions. Data for all other antidepressants were of low certainty. As RCTs excluded people with low mood, we were unable to establish the effects of antidepressants for people with chronic pain and depression. There is currently no reliable evidence for the long-term efficacy and safety of any antidepressant. Study registration This study is registered as PROSPERO CRD42020171855. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128782) and is published in full in Health Technology Assessment; Vol. 28, No. 62. See the NIHR Funding and Awards website for further award information.

Background Magnetic resonance imaging localises cancer in the prostate, allowing for a targeted biopsy with or without transrectal ultrasound-guided systematic biopsy. Targeted biopsy methods include cognitive fusion, where prostate lesions suspicious on magnetic resonance imaging are targeted visually during live ultrasound, and software fusion, where computer software overlays the magnetic resonance imaging image onto the ultrasound in real time. The effectiveness and cost-effectiveness of software fusion technologies compared with cognitive fusion biopsy are uncertain. Objectives To assess the clinical and cost-effectiveness of software fusion biopsy technologies in people with suspected localised and locally advanced prostate cancer. A systematic review was conducted to evaluate the diagnostic accuracy, clinical efficacy and practical implementation of nine software fusion devices compared to cognitive fusion biopsies, and with each other, in people with suspected prostate cancer. Comprehensive searches including MEDLINE, and Embase were conducted up to August 2022 to identify studies which compared software fusion and cognitive fusion biopsies in people with suspected prostate cancer. Risk of bias was assessed with quality assessment of diagnostic accuracy studies-comparative tool. A network meta-analysis comparing software and cognitive fusion with or without concomitant systematic biopsy, and systematic biopsy alone was conducted. Additional outcomes, including safety and usability, were synthesised narratively. A de novo decision model was developed to estimate the cost-effectiveness of targeted software fusion biopsy relative to cognitive fusion biopsy with or without concomitant systematic biopsy for prostate cancer identification in biopsy-naive people. Scenario analyses were undertaken to explore the robustness of the results to variation in the model data sources and alternative assumptions. Results Twenty-three studies (3773 patients with software fusion, 2154 cognitive fusion) were included, of which 13 informed the main meta-analyses. Evidence was available for seven of the nine fusion devices specified in the protocol and at high risk of bias. The meta-analyses show that patients undergoing software fusion biopsy may have: (1) a lower probability of being classified as not having cancer, (2) similar probability of being classified as having non-clinically significant cancer (International Society of Urological Pathology grade 1) and (3) higher probability of being classified at higher International Society of Urological Pathology grades, particularly International Society of Urological Pathology 2. Similar results were obtained when comparing between same biopsy methods where both were combined with systematic biopsy. Evidence was insufficient to conclude whether any individual devices were superior to cognitive fusion, or whether some software fusion technologies were superior to others. Uncertainty in the relative diagnostic accuracy of software fusion versus cognitive fusion reduce the strength of any statements on its cost-effectiveness. The economic analysis suggests incremental cost-effectiveness ratios for software fusion biopsy versus cognitive fusion are within the bounds of cost-effectiveness (£1826 and £5623 per additional quality-adjusted life-year with or with concomitant systematic biopsy, respectively), but this finding needs cautious interpretation. Limitations There was insufficient evidence to explore the impact of effect modifiers. Conclusions Software fusion biopsies may be associated with increased cancer detection in relation to cognitive fusion biopsies, but the evidence is at high risk of bias. Sufficiently powered, high-quality studies are required. Cost-effectiveness results should be interpreted with caution given the limitations of the diagnostic accuracy evidence. Study registration This trial is registered as PROSPERO CRD42022329259. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: 135477) and is published in full in Health Technology Assessment; Vol. 28, No. 61. See the NIHR Funding and Awards website for further information. Plain language summary Men with an magnetic resonance imaging scan that shows possible prostate cancer (PCa) are offered prostate biopsies, where samples of the prostate tissue are collected with a needle, to confirm the presence and severity of cancer. Different biopsy methods exist. In a cognitive fusion biopsy, clinicians will target abnormal looking parts of the prostate by looking at the magnetic resonance imaging scan alongside ‘live’ ultrasound images. During a software fusion (SF) biopsy, a computer software is used to overlay the magnetic resonance imaging scan onto the ultrasound image. This study evaluated whether SF is better at detecting cancer compared with cognitive fusion biopsy, and whether it represents value for money for the National Health Service. We did a comprehensive review of the literature. We combined and re-analysed the evidence, and assessed its quality. We investigated whether SF biopsies are sufficient value for money. Compared with cognitive fusion, patients receiving a SF biopsy may have: (1) a lower probability of having a ‘no cancer’ result, (2) similar probability of having a benign, non-clinically significant (CS) cancer result and (3) higher probability of detecting CS cancer. However, it is uncertain to what extent SF is more accurate than cognitive fusion, because of concerns about the quality of the evidence. We found no evidence that any SF devices were superior to others. Using additional, random biopsies alongside software or cognitive fusion would increase the detection of PCa. We also looked for evidence on the value for money of the SF biopsies to detect PCa and found no relevant studies. We weighed the costs and the benefits of SF biopsy compared to cognitive fusion to determine whether it could be a good use of National Health Service money. The poor quality of information makes the value of the technologies largely unknown. Scientific summary Background Prostate cancer (PCa) is the most commonly diagnosed cancer in men in the UK. In the NHS people with suspected PCa are offered multiparametric magnetic resonance imaging (mpMRI). People with suspected PCa, according to MRI, are offered a biopsy procedure to confirm the presence and severity of cancer. Traditionally patients were offered a systematic transrectal, ultrasound-guided prostate biopsy (or systematic biopsy). Since the introduction of mpMRI, specific areas of abnormal tissue can be targeted, by combining (or fusing) the results of mpMRI and ultrasound imaging. Several methods for fusing MRI and ultrasound images exist, including cognitive fusion (CF), in which a region of interest is identified prior to biopsy and the biopsy operator estimates where it might be on an ultrasound image, and software fusion (SF), where regions of interest on magnetic resonace images are identified and contoured before biopsy and overlayed with the prostate contours on ultrasound images during the biopsy. Systematic biopsy may be used in addition to targeted biopsy. A number of SF technologies are available. However, the effectiveness and cost-effectiveness of SF compared with CF is uncertain. Objectives This study aimed to assess the clinical and cost-effectiveness of SF biopsy systems in people with suspected localised and locally advanced PCa. Methods Systematic review A systematic review of the diagnostic accuracy, clinical effectiveness, safety and practical implementation of nine SF systems compared with CF and with each other, in people suspected PCa according to MRI was conducted. Comprehensive bibliographic searches, including MEDLINE and EMBASE and supplementary sources, were conducted up to 2 August 2022 for published and unpublished literature. Studies of people with suspected PCa who have had a MRI scan that indicates a significant lesion [Likert or prostate imaging – reporting and data system (PI-RADS) score of 3 or more], including biopsy-naive and repeat biopsy patients with a previous negative prostate biopsy, and comparing SF with CF or with another SF device, were included. The following SF technologies were included: ARTEMIS (InnoMedicus ARTEMIS), BioJet (Healthcare Supply Solutions Ltd), BiopSee (Medcom), bkFusion (BK Medical UK Ltd and MIM Software Inc.), Fusion Bx 2.0 (Focal Healthcare), FusionVu (Exact Imaging), iSR’obot Mona LisaTM (Biobot iSR’obot), KOELIS Trinity (KOELIS and Kebomed) and UroNav Fusion Biopsy System (Phillips). Previous versions were also eligible. In-bore (or in-gantry) biopsies were excluded. Prospective, randomised and non-randomised comparative studies were included, and retrospective evidence where no prospective evidence could be found for an eligible SF device. To provide sufficient evidence for a network meta-analysis (NMA), within-patient comparisons or randomised controlled trials (RCTs) between SF and systematic biopsy, and between CF and systematic biopsy, were also eligible to inform indirect comparisons of diagnostic accuracy. Two researchers independently screened the titles and abstracts of all reports identified by the bibliographic searches and of all full-text papers subsequently obtained. Data extraction and quality assessment were conducted by at least one researcher and checked by a second. Risk of bias of diagnostic accuracy studies was assessed using quality assessment of diagnostic accuracy studies-comparative (QUADAS-C). For diagnostic accuracy outcomes, studies reporting sufficient data were included in network meta-analyses comparing SF and CF with or without concomitant systematic biopsy, and systematic biopsy alone, where odds of being categorised in each of different cancer grades were allowed to vary by biopsy type. Results were reported as odds ratios with 95% credible intervals (CrIs). Additional diagnostic accuracy results that could not be pooled in a meta-analysis and clinical effectiveness, safety and implementation outcomes were synthesised narratively. Economic analysis Cost-effectiveness evidence comparing SF biopsy systems with CF for targeted prostate biopsy in men with suspected PCa was identified by the previously mentioned searches, with evidence narratively summarised and tabulated. Studies were appraised for their quality, generalisability and appropriateness to inform the decision problem as defined by the National Institute for Health and Care Excellence Diagnostics Assessment Report (NICE DAR) scope. A targeted search was conducted to identify evidence to support the development of a de novo decision model. The searches aimed to identify cost-effectiveness evidence of diagnostic strategies at the point of biopsy to support the model conceptualisation. Evidence was reviewed to (1) identify value components of the biopsy approaches, (2) characterise alternative mechanisms of evidence linkage from disease prevalence, diagnostic accuracy, choice of treatment to final outcomes, and (3) identify any UK-relevant sources of evidence. A de novo decision analytic model was developed to estimate the cost-effectiveness of SF compared to CF. The model evaluated two strategies for two alternative comparisons: (1) targeted SF biopsy versus targeted cognitive biopsy and (2) combined (targeted and systematic) SF biopsy versus combined cognitive biopsy. The four strategies could not be incrementally compared due to the mechanism of evidence generation for the diagnostic accuracy, which relied on separate evidence networks. The de novo model consisted of two components: (1) a decision tree, which captured biopsy adverse events (AEs), repeated biopsies and classified individuals according to their biopsy results and underlying true disease status, and (2) long-term model to link classification to clinical management decisions and this to longer-term costs and consequences (e.g. disease progression and PCa mortality) so that differences in costs, life-year gains and quality-adjusted life-years (QALYs) were quantified over a lifetime horizon. The model required the development of (1) an extension to the evidence synthesis to allow quantifying the extension of test misclassification in the diagnostic model with SF biopsy and CF biopsy, and (2) an inference model to derive unobservable transition probabilities for the long-term model. Results The systematic review of clinical evidence included a total of 3733 patients who received SF and 2154 individuals with CF from 23 studies. Evidence was included for all devices specified in the protocol, except for Fusion Bx 2.0 and FusionVu. Overall, the evidence for all devices was at high risk of bias. Overall, biopsy-naive patients were under-represented. Fourteen studies were included in the meta-analyses. Diagnostic accuracy Across all analyses results must be interpreted with caution due to the high risk of bias in the evidence base and wide uncertainty over the results. The meta-analyses show that patients undergoing SF biopsy may have: (1) a lower probability of being classified as not having cancer, (2) similar probability of being classified as having non-clinically significant cancer [International Society of Urological Pathology (ISUP) grade 1], and (3) higher probability of being classified at higher ISUP grades, particularly ISUP 2. Similar results were obtained where both biopsy methods were combined with systematic biopsy. Additional meta-analyses of cancer detection rates suggest that, compared with CF biopsy, SF may identify more PCa (any grade) (OR 1.30; 95% CrI 1.06, 1.61). Adding systematic biopsy to cognitive or SF may increase the detection of all PCa and of clinically significant (CS) cancer, and from this evidence there is no suggestion that SF with concomitant systematic biopsy is superior to CF with systematic biopsy. Meta-analyses of cancer detection rates, by individual device, showed that compared with CF biopsy, BioJet and Urostation are associated with a higher detection of PCa overall. There was no evidence that any of the SF devices increased detection of CS cancer (except for BioJet, although this is based on one low-quality study), and overall, the evidence was insufficient to conclude whether any individual devices were superior to CF, or whether some SF technologies are more accurate than others. Clinical effectiveness There is no evidence that biopsy positivity rates and safety outcomes differ significantly between SF and CF, or between SF devices. There was some evidence that systems with rigid registration (BioJet or UroNav) are easier and faster to use than elastic registration (KOELIS Trinity), although this is informed by a single, small study and is not conclusive. Cost-effectiveness One full cost-effectiveness study of SF compared targeted SF to targeted CF. However, the findings of the study were not considered generalisable to the decision problem under assessment. Sixteen studies were identified of which nine were selected to inform the conceptualisation and parameterisation of the de novo decision model. The base-case cost-effectiveness analysis suggests for the targeted biopsy and the combined biopsy comparisons, that SF strategy is on average costlier and yields greater QALYs than the CF strategy, resulting in a probabilistic incremental cost-effectiveness ratio (ICER) of £6197 and £2199 per additional QALY for each comparison, respectively. These ICERs are below the lower bound of the cost-effectiveness threshold range recommended by NICE, suggesting that SF may be cost-effective compared to CFs in both the targeted and the combined comparisons. However, these results should be interpreted cautiously given the uncertainties in the relative diagnostic accuracy evidence which informs the model. The probabilistic analysis suggests a higher probability of cost-effectiveness for SF versus CF at the range of cost-effectiveness thresholds recommended by NICE (0.64 and 0.68 at £20,000 and £30,000 per additional QALY for targeted SF biopsy). Discussion This assessment includes a broad, comprehensive literature search for software and CF technologies and has been conducted following recognised guidelines to ensure high quality. The review identified evidence on the diagnostic accuracy of nine SF technologies, and is the first systematic review to formally compare the relative accuracy of SF and CF, with and without systematic biopsy, as well as different SF devices, using both direct and indirect evidence in a NMA. Unlike recent systematic review evidence, our review found that SF increased detection of clinically insignificant cancer compared with CF. Our review has a number of limitations. The evidence included in the systematic review is at high risk of bias overall. There was variation in patient and study characteristics. Biopsy-naive patients, who form the large majority of patients eligible for targeted biopsy, were under-represented, although there was insufficient evidence to evaluate whether the relative accuracy of software and CF differed between biopsy-naive and repeat biopsy patients. There was insufficient evidence to explore the impact of a number of other potential effect modifiers, including lesion location, operator experience, biopsy routes and anaesthesia methods. There were few studies per comparison, not all studies reported outcomes by all cancer grades, and most estimates from the meta-analyses were imprecise, particularly at higher cancer grades where data were most sparse. The network meta-analyses relied on the assumption that CF was equivalent across different centres, which is uncertain. No evidence was found for most of this assessment’s prespecified outcomes: biopsy sample suitability/quality, number of repeat biopsies performed, procedure completion rates, software failure rate, time to diagnosis, length of hospital stay, time taken for MR image preparation, subsequent PCa management, re-biopsy rate, hospitalisation, overall survival, progression-free survival (PFS), patient- and carer- reported outcomes [including tolerability and health-related quality of life (HRQoL)], barriers and facilitators to implementations. The cost-effectiveness results are driven by the modelled differences in diagnostic accuracy between software and CF, particularly the increased correct detection of Cambridge Prognostic Group 1 (CPG 1) (resulting in net losses for SF) and CPG 2 (resulting in net gains for SF). The External Assessment Group (EAG)’s NMA and its extension underpinned the economic model, so its limitations apply to the cost-effectiveness estimates. The magnitude of value realised for SF, compared with CF, depends on the balance between different degrees of misclassification and correct classification with the two technologies and on the prevalence of disease at each cancer grade. The value of SF is thus driven by comparative diagnostic accuracy (compared to ‘gold standard’) derived where evidence is particularly sparse (cancer grades above 2), and by prevalence, which is also affected by evidence sparsity. Therefore, the estimates of cost-effectiveness are affected by unquantified uncertainty and should be interpreted with caution. Conclusions Compared to CF biopsy, patients undergoing SF biopsy may show a lower probability of being classified as not having cancer, similar probability of being classified as having non-CS cancer, and a higher probability of being classified at higher ISUPs, particularly ISUP 2. Both SF and CF biopsy can miss CS cancer lesions, and the addition of standard-systematic biopsy increases the detection of all PCa and CS cancer for both fusion methods. There is insufficient evidence to conclude on the relative accuracy and clinical effectiveness of different software devices. Cost-effectiveness estimates comparing software to CF were generally favourable to SF, except where the technologies were assumed to have the same diagnostic accuracy. The drivers of economic value of SF, comparative diagnostic accuracy and prevalence, are affected by unquantified uncertainty. Judgements on the economic value of SF require integration of the uncertainties over the clinical evidence with the overall cost-effectiveness. Recommendations for further research High-quality, sufficiently powered RCT evidence comparing SF biopsy with CF biopsy is required to address limitations from the existing evidence. Improved reporting of diagnostic accuracy outcomes would enable future syntheses to make use of a larger body of evidence. Study registration This trial is registered as PROSPERO CRD42022329259. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: 135477) and is published in full in Health Technology Assessment; Vol. 28, No. 61. See the NIHR Funding and Awards website for further information.

Background There is a lack of interventions for specific phobia in children and adolescents with moderate to severe intellectual disabilities. Objectives The objectives were to: (a) develop an intervention for specific phobia, together with an intervention fidelity checklist and logic model, and evaluate candidate outcome measures, together with parents/carers and clinicians; (b) describe treatment as usual; (c) model the intervention to determine the acceptability and feasibility for all stakeholders, judge the appropriateness of outcome measures, explore recruitment pathways, and examine the feasibility and acceptability of consent and associated processes; and (d) describe factors that facilitate or challenge the intervention. Design Phase 1a: using consensus methods, an Intervention Development Group was established who met to develop the intervention, review candidate outcome measures and contribute to the development of the intervention fidelity checklists and logic model. Phase 1b: a national online survey was conducted with parents and professionals to describe treatment as usual. Phase 2: a single-group non-randomised feasibility study was designed to model the intervention and to test intervention feasibility and acceptability, outcome measures and aspects of the research process. Setting Phase 2: participants were recruited from National Health Service community child learning disabilities teams and special schools in England. Treatment was delivered in the child learning disabilities teams. Participants Children aged 5–15 years with moderate to severe learning disability and specific phobia, and their parents/carers. Interventions The SPIRIT intervention comprised two half-day workshops and eight support sessions plus treatment as usual. Main outcomes The feasibility and acceptability of the intervention and research processes, recruitment, outcome measure completion rates and acceptability, and intervention adherence. Parents completed all of the outcome measures, with very low rates of missing data. The recruitment of sites and participants was impacted by the COVID-19 pandemic. Results The intervention was successfully developed and modelled with 15 participants with moderate to severe learning disabilities and their parents. The intervention was judged to be feasible and acceptable by parents/carers and therapists. Parents/carers and therapists suggested minor intervention revisions. Limitations Randomisation was not modelled within this feasibility study, although the majority of parents and therapists indicated that this would be acceptable. Conclusions The SPIRIT intervention and associated study processes were judged to be feasible and acceptable. The intervention requires minor revisions. Future work The SPIRIT intervention should be tested further within a clinical trial. Study registration Current Controlled Trials ISRCTN34766613. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR130177) and is published in full in Health Technology Assessment; Vol. 28, No. 64. See the NIHR Funding and Awards website for further award information. Plain language summary This study was about children and adolescents who have moderate to severe learning disabilities and specific phobia. This study had two parts. In the first part, we worked with parents of young people with learning disabilities and therapists to develop a treatment for specific phobia in children and adolescents with moderate to severe learning disabilities. We also collected information about what treatment young people were currently getting. To do this, we conducted a national (United Kingdom) survey of parents/carers who have a child with a learning disability and a phobia, along with a survey of health professionals who work with children with learning disabilities. Together with parents and therapists, we developed a treatment for specific phobia in children and adolescents with moderate to severe learning disabilities. We collected information about what treatments young people received for specific phobia and found that many do not receive any treatment for their specific phobia. In the second part, we wanted to find out whether the treatment was acceptable to parents and therapists. To do this, we tried out the treatment with 15 children and adolescents. We had difficulties getting people involved in the study due to the COVID-19 pandemic. We got enough people involved to help us to work out whether the treatment was acceptable to parents and therapists. We interviewed parents and therapists to find out how they felt about the treatment and being part of the study. We also talked to therapists to ask them what they thought about the treatment. Parents told us that they liked being involved in the study and found the treatment helped them to help their children. Parents and therapists suggested some changes to the treatment to help improve it in the future. It was recommended that a larger study should be completed. Scientific summary Background Many children with learning disabilities have significant fears or phobias. These can, for example, include a severe fear of dogs or other animals, visiting the dentist, or having an injection. Children and adolescents with learning disabilities are at least twice as likely to experience specific phobia than their typically developing peers. There is good evidence that psychological therapies, particularly exposure-based therapies, are an effective treatment for phobias, but these treatments have not been evaluated for use with people with learning disabilities, in particular for children and adolescents with moderate to severe learning disabilities. Due to difficulties with verbal communication, understanding, restricted and repetitive behaviours, and challenging behaviour, these treatments need to be adapted before they can be used. Objectives The aim of this study was to, using coproduction with our patient and public involvement (PPI) partners, develop and evaluate the feasibility of an exposure-based intervention for specific phobia in children and adolescents with moderate to severe learning disabilities. This work was undertaken in two phases: (1a) development of the intervention and (1b) description of treatment as usual (TAU); and (2) evaluation of the feasibility of the proposed intervention. Phase 1a: development The objectives were to: establish an Intervention Development Group (IDG), and using coproduction over a series of meetings, develop an intervention for specific phobia for use with children and adolescents who have moderate to severe learning disabilities with and without autism develop a treatment fidelity checklist to be used alongside the intervention manual appraise and consider several candidate outcome measures of anxiety-related symptoms, and secondary outcomes, and make a recommendation for use within phase 2. Phase 1b: description of treatment as usual The objective was to describe the current standard treatment provided for children and adolescents with moderate to severe learning disabilities and specific phobia within the UK. Phase 2: feasibility study The objectives were to: evaluate the manualised intervention to determine the acceptability and feasibility for all stakeholders, including children and young people, carers and therapists judge the appropriateness of the measures of anxiety-related symptomatology, and secondary outcomes, for use within a larger study explore recruitment pathways describe factors that challenge or facilitate the implementation of the intervention (e.g. comorbid behaviour problems, other mental health problems, community resources to support exposure) determine the feasibility and acceptability of consent and associated processes determine the acceptability of randomisation in a future trial describe the parameters of a future study to examine the effectiveness of exposure-based therapy to treat phobias in this population. Methods Study design Phase 1a (intervention development): an IDG was established. Informed by co-applicant Williams’s existing intervention developed for dog phobia in adolescents with severe learning disabilities and little to no speech, we developed an intervention that aimed to be developmentally appropriate for use with both children and adolescents with moderate to severe learning disabilities, and with phobia related to any specific stimulus, as defined by the DSM-5 [American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (DSM-5). 5th edn. Washington, DC: American Psychiatric Association; 2013] (animal, natural environment, blood-injection-injury, situational, other). Phase 1b (TAU survey): to determine current community-based TAU, an online survey (UK-wide) was conducted of parents/carers who identified their child (aged 5–15 years) with moderate to severe learning disabilities as having a specific phobia, together with interviews/online survey of professionals. Phase 2 (feasibility study): this study was a single-arm, non-randomised feasibility study, with participants receiving the intervention developed in phase 1a, in conjunction with any other treatment they were receiving. Overall, the study ran from January 2021 to June 2023. Setting and participants Phase 1a (intervention development) The IDG recruited six key stakeholders who were representatives from our PPI partners, carers and family members and clinicians. The Principal Investigator, Study Manager and Research Assistant attended all of the IDG meetings. Other members of the research team attended the IDG sessions as observers only. Phase 1b (TAU survey) The study aimed to recruit 50 parents/carers who identified their child with moderate to severe learning disability as having a specific phobia and 25 learning disability professionals (health professionals, service providers and commissioners). We utilised our existing Midlands and wider UK networks of schools, support groups, charities and our PPI partner (the Foundation for People with Learning Disabilities) to disseminate the online survey to parents/carers of children and adolescents throughout England, Wales, Northern Ireland and Scotland. We also used our existing learning disabilities health professional networks, together with the local National Institute for Health and Care Research Clinical Research Network. The survey included questions informed by the Template for Intervention Description and Replication (TIDieR) checklist. The TIDieR checklist is used to provide a description of an intervention, including the use of any associated materials. Phase 2 (feasibility study) This single-arm, non-randomised feasibility study took place within the NHS – either specialist learning disabilities or mainstream child and adolescent mental health services in England. Five NHS services across England were recruited for this study: Cambridgeshire Community Services NHS Trust, Dorset Healthcare University NHS Foundation Trust, Avon and Wiltshire Mental Health Partnership NHS Trust, Norfolk Community Health and Care NHS Trust and Hertfordshire Community NHS Trust. Children currently receiving treatment for specific phobia or psychological intervention for other anxiety disorders were not eligible to participate. A total of 15 participants were recruited. Intervention During phase 1a, the exposure-based intervention for children with severe learning disabilities and limited communication skills developed by co-applicant Williams informed the development of an intervention for both children and adolescents with moderate to severe learning disabilities and with specific phobias. The intervention consisted of two parts: (1) a parents/carers skills training group workshop (two half-days), and (2) weekly therapist support telephone calls with individual parents/carers over 8 weeks, lasting approximately 30 minutes each, with an additional 30 minutes of therapist time to prepare and write notes after the session. Assessment of feasibility of delivery and acceptability of the intervention We examined the views of parents/carers and therapists to address: (a) intervention accessibility and acceptability; (b) helpful/unhelpful aspects, including barriers to change; (c) the value of our adaptations; (d) relationships with therapists within the intervention; (e) acceptability of consent processes; (f) acceptability of outcome measures; and (g) acceptability of randomisation within a future trial. We completed semistructured interviews with five parents/carers and five therapists. We aimed to complete interviews with the young people who received the intervention to explore their experience of the intervention and the outcomes for them. Although we planned to use augmented communication methods to aid our interview as much as possible, all parents indicated that their child would not be able to participate in an interview due to limited communication skills. Recruitment Phase 1b The online survey was delivered using Qualtrics® (Qualtrics, Provo, UT, USA). This phase lasted 14 months, running concurrently with phase 1a and phase 2. Fifty-two parents completed the survey on TAU. The health professionals survey covered the same content as the parent survey, in relation to support/treatment proved for specific phobia. Although originally conceived as an interview, the survey was also offered as an online survey to facilitate recruitment. Twenty-five professionals completed the online survey. Phase 2 Young people were enrolled in the study for approximately 6 months and were assessed at three time points: (1) eligibility assessment; (2) baseline assessment within 4 weeks prior to commencement of the intervention; and (3) assessment at completion of the intervention. The primary outcome measure was a parent/carer-completed checklist of symptoms of phobia and their severity. As there were no measures of specific phobia available, we modified existing measures to assess symptoms and their impact. Together with the IDG, in phase 1a of the project, the child version of the Severity Measure for Specific Phobia was adapted, modifying it consistent with the recommended adaptations in the Diagnostic Manual – Intellectual Disability (DM-ID-2) (Fletcher RJ, Barnhill J, Cooper SA, editors. Diagnostic Manual – Intellectual Disability 2: A Textbook of Diagnosis of Mental Disorders in Persons with Intellectual Disability. 2nd edn. Kingston, NY: National Association for the Dually Diagnosed; 2017), and adapting it to be completed by a parent/carer. The impact of the phobia was also considered using an adapted version of the Strengths and Difficulties Questionnaire impact supplement (Goodman R. The extended version of the Strengths and Difficulties Questionnaire as a guide to child psychiatric caseness and consequent burden. J Child Psychol Psychiatry 1999;40:791–801), modified to be focused on specific phobia and to be appropriate for children with moderate to severe learning disabilities (Impact of Phobia measure). The IDG also considered a range of secondary outcomes, including: (a) specific phobia diagnosis {e.g. diagnostic checklist using DM-ID-2 or clinical interview [Anxiety Disorders Interview Schedule (Silverman W, Albano A. Anxiety Disorders Interview Schedule. Oxford: Oxford University Press; 1996)]}; (b) emotional and behaviour problems {Developmental Behavior Checklist-2 [Gray KM, Tonge B, Einfeld S, Gruber C, Klein A. Developmental Behavior Checklist (DBC2). 2nd edn. Torrance, California: Western Psychological Services; 2018]}; (c) challenging behaviour [Behavior Problems Inventory (Rojahn J, Rowe E, Sharber A, Hastings R, Matson J, Didden R, et al. The behavior problems inventory‐short form for individuals with intellectual disabilities: Part I: development and provisional clinical reference data. J Intellect Disabil Res 2012;56:527–45)]; and (d) physiological measures (heart rate). These measures were completed prior to commencement of the intervention, and within 4 weeks after the completion of the intervention. Patient and public involvement Patient and public involvement was a key part of our methods. We partnered with the Foundation for People with Learning Disabilities, who worked with us collaboratively to support coproduction (with family carers) to develop the intervention. Service users, carers and clinicians were members of our Study Steering Group and shared oversight of the progress of the project. PPI partners played a key role in contributing to the preparation of study documents, provided advice on recruitment, and helped to collaboratively disseminate information about the study findings. Results Phase 1a We successfully adapted the intervention, developed a logic model and intervention fidelity checklist and selected outcome measures collaboratively with the IDG. Phase 1b A national survey of TAU was undertaken to describe interventions for specific phobia in children and adolescents with moderate to severe learning disabilities. Parents/carers (n = 52) of children and adolescents with specific phobia and moderate to severe learning disabilities were surveyed as well as professionals (n = 25) working in services providing care to children and adolescents with moderate to severe learning disabilities. A key finding from the survey was that a significant proportion of parents (73%) reported not being offered any treatment for their child’s specific phobia. Of those who did receive treatment for their child, a range of treatments were offered, with the most frequent being medication. Other treatments were psychological, and included exposure therapy, sensory integration therapy and counselling. While the majority of treatments provided were in community-based health and social care settings, 28% were school based. Of the professionals who completed the TAU survey, the majority worked in health and care services (95%), and one was based in a school. Just over half (54%) indicated that their service offered treatment for specific phobia. Of these, 50% offered exposure therapy. Other therapies were also offered including cognitive behavioural therapy (CBT), medication, acceptance and commitment therapy, primary care support, systemic intervention and psychoeducation. With the exception of CBT, the other therapies did not include a graded exposure component. Phase 2 Acceptability and feasibility of the intervention The intervention was feasible to deliver and was acceptable to the parents of children with moderate to severe learning disabilities and to therapists. A number of revisions were suggested by the parents to improve clarity of some of the materials. Parents and therapists felt that some flexibility in the delivery of the support sessions would be useful. A number of the challenges identified by the therapists could be addressed in minor revisions to the therapist training workshop. Appropriateness of the outcome measures The study outcome measures were judged to be appropriate. With the exception of one parent, all measures were completed by parents/carers who remained in the study at both time points. The percentage of missing data on completed measures was extremely low. Recruitment pathways Recruitment of sites was challenging, with the two original planned sites withdrawing from the study due to capacity issues. The study team discussed participation in the study with 22 sites in order to recruit 5. Sites often declined to be involved on the basis of staff capacity. Barriers to taking part in the study were primarily COVID-19 related. Recruitment of participants was also challenging. Sites reported finding it challenging to identify potential participants from caseloads, as information on systems did not tend to record specific phobia as a primary problem. Three of the five sites were only able to recruit through current caseloads, while two sites were able to recruit externally (recruiting through local special schools and support/advocacy organisations in the region). In total, 93 potential participants were identified and contacted about the study; 47 of these were identified by NHS sites (caseloads) and 46 through external recruitment, highlighting the importance of being able to recruit from organisations external to the NHS sites. Factors that facilitate or challenge the implementation of the intervention For parents, logistical issues around finding time to do the tasks involved in the intervention presented a key challenge. Other challenges included sharing the data sheets with the therapists and the need for further support with understanding reinforcement. Accessing the feared stimulus (e.g. dogs) was a challenge for some. Therapists felt that the structure of the intervention and the troubleshooting component in particular facilitated the implementation of the intervention, and that placing parents as the experts on their child and their needs was a strength. Challenges included some difficulties with implementing the relaxation strategies, ensuring the exposure steps were sufficiently small and steps were not skipped, motivating parents, managing negative experiences during exposure, and accessing a dog for exposure steps. A number felt that more support/time was needed for parents. Feasibility and acceptability of consent and associated processes Parents/carers reported no difficulties with the participant information sheets and consent forms. Acceptability of randomisation in a future trial The majority (60%) of parents felt that participating in a future trial with randomisation was acceptable; however, 40% were concerned they may not be able to access the intervention. Therapists felt that it would be acceptable if all children were able to be offered the intervention at the end of the trial. Conclusions The SPIRIT intervention was judged to be feasible to deliver and acceptable to parents of children and adolescents with moderate to severe intellectual disabilities and therapists. Carers and therapists made some helpful suggestions for revisions which can be easily incorporated into the existing manualised intervention with minor revisions. This study took place during the COVID-19 pandemic, and the recruitment of sites and participants during phase 2 was at a lower rate than anticipated. The study aimed to recruit up to 20 participants and recruited 15. The participant attrition rate was low and not attributable to the intervention or study processes. This study benefited from genuine PPI during the adaptation of the intervention, development of the fidelity checklist and logic model, choice of outcome measures and study management. Following minor revisions to the intervention, the SPIRIT intervention should be tested in a randomised trial. Study registration Current Controlled Trials ISRCTN34766613. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR130177) and is published in full in Health Technology Assessment; Vol. 28, No. 64. See the NIHR Funding and Awards website for further award information.

Background Acute respiratory infections are a common reason for consultation with primary and emergency healthcare services. Identifying individuals with a bacterial infection is crucial to ensure appropriate treatment. However, it is also important to avoid overprescription of antibiotics, to prevent unnecessary side effects and antimicrobial resistance. We conducted a systematic review to summarise evidence on the diagnostic accuracy of symptoms, signs and point-of-care tests to diagnose bacterial respiratory tract infection in adults, and to diagnose two common respiratory viruses, influenza and respiratory syncytial virus. Methods The primary approach was an overview of existing systematic reviews. We conducted literature searches (22 May 2023) to identify systematic reviews of the diagnostic accuracy of point-of-care tests. Where multiple reviews were identified, we selected the most recent and comprehensive review, with the greatest overlap in scope with our review question. Methodological quality was assessed using the Risk of Bias in Systematic Reviews tool. Summary estimates of diagnostic accuracy (sensitivity, specificity or area under the curve) were extracted. Where no systematic review was identified, we searched for primary studies. We extracted sufficient data to construct a 2 × 2 table of diagnostic accuracy, to calculate sensitivity and specificity. Methodological quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies version 2 tool. Where possible, meta-analyses were conducted. We used GRADE to assess the certainty of the evidence from existing reviews and new analyses. Results We identified 23 reviews which addressed our review question; 6 were selected as the most comprehensive and similar in scope to our review protocol. These systematic reviews considered the following tests for bacterial respiratory infection: individual symptoms and signs; combinations of symptoms and signs (in clinical prediction models); clinical prediction models incorporating C-reactive protein; and biological markers related to infection (including C-reactive protein, procalcitonin and others). We also identified systematic reviews that reported the accuracy of specific tests for influenza and respiratory syncytial virus. No reviews were found that assessed the diagnostic accuracy of white cell count for bacterial respiratory infection, or multiplex tests for influenza and respiratory syncytial virus. We therefore conducted searches for primary studies, and carried out meta-analyses for these index tests. Overall, we found that symptoms and signs have poor diagnostic accuracy for bacterial respiratory infection (sensitivity ranging from 9.6% to 89.1%; specificity ranging from 13.4% to 95%). Accuracy of biomarkers was slightly better, particularly when combinations of biomarkers were used (sensitivity 80–90%, specificity 82–93%). The sensitivity and specificity for influenza or respiratory syncytial virus varied considerably across the different types of tests. Tests involving nucleic acid amplification techniques (either single pathogen or multiplex tests) had the highest diagnostic accuracy for influenza (sensitivity 91–99.8%, specificity 96.8–99.4%). Limitations Most of the evidence was considered low or very low certainty when assessed with GRADE, due to imprecision in effect estimates, the potential for bias and the inclusion of participants outside the scope of this review (children, or people in hospital). Future work Currently evidence is insufficient to support routine use of point-of-care tests in primary and emergency care. Further work must establish whether the introduction of point-of-care tests adds value, or simply increases healthcare costs. Funding This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number NIHR159948. Plain language summary What is the problem? Respiratory infections are a common cause of illness. Currently, healthcare professionals use clinical experience to decide whether an infection is caused by a virus or bacteria, and whether antibiotics are needed. However, this is not always easy to establish. We tried to identify the effectiveness of rapid tests (with results in under 45 minutes) at distinguishing between viral and bacterial respiratory infections. What did we do? We identified and summarised all the existing reviews and studies in this area. We looked at many different tests which aim to distinguish between bacterial and viral causes of respiratory infections. In particular, we assessed: individual symptoms and signs (such as the presence of cough, or a fever) combinations of symptoms and signs (the presence or absence of multiple symptoms) various ‘biomarker’ tests (blood tests for evidence that the body has used its defence mechanisms) We also looked at specific tests for flu and respiratory syncytial virus, which are common causes of viral infection. What did we find? The reviews we found showed symptoms and signs were not able to identify bacterial infections in people accurately. The accuracy of biomarker tests was slightly better, particularly when multiple markers were used. The accuracy of rapid tests for flu and respiratory syncytial virus varied; the most accurate tests were those that detect viral genetic material. We also found studies showing that genetic tests that identify many viruses at once (multiplex tests) were very accurate. However, most of the evidence we identified was not robust. There were concerns about the conduct of some of the studies. In some cases there was uncertainty whether a test was really accurate enough to be useful. Therefore there is still doubt about whether any of these tests will be useful additions to current clinical care.

Background The BioImpedance Spectroscopy to maintain Renal Output randomised controlled trial investigated the effect of bioimpedance spectroscopy added to a standardised fluid management protocol on the risk of anuria and preservation of residual kidney function (primary trial outcomes) in incident haemodialysis patients. Despite the economic burden of kidney disease, the cost-effectiveness of using bioimpedance measurements to guide fluid management in haemodialysis is not known. Objectives To assess the cost-effectiveness of bioimpedance-guided fluid management against current fluid management without bioimpedance. Design Within-trial economic evaluation (cost–utility analysis) carried out alongside the open-label, multicentre BioImpedance Spectroscopy to maintain Renal Output randomised controlled trial. Setting Thirty-four United Kingdom outpatient haemodialysis centres, both main and satellite units, and their associated inpatient hospitals. Participants Four hundred and thirty-nine adult haemodialysis patients with > 500 ml urine/day or residual glomerular filtration rate > 3 ml/minute/1.73 m2. Intervention The study intervention was the incorporation of bioimpedance technology-derived information about body composition into the clinical assessment of fluid status in patients with residual kidney function undergoing haemodialysis. Bioimpedance measurements were used in conjunction with usual clinical judgement to set a target weight that would avoid excessive fluid depletion at the end of a dialysis session. Main outcome measures The primary outcome measure of the BioImpedance Spectroscopy to maintain Renal Output economic evaluation was incremental cost per additional quality-adjusted life-year gained over 24 months following randomisation. In the main (base-case) analysis, this was calculated from the perspective of the National Health Service and Personal Social Services. Sensitivity analyses explored the impact of different scenarios, sources of resource use data and value sets. Results The bioimpedance-guided fluid management group was associated with £382 lower average cost per patient (95% CI −£3319 to £2556) and 0.043 more quality-adjusted life-years (95% CI −0.019 to 0.105) compared with the current fluid management group, with neither values being statistically significant. The probability of bioimpedance-guided fluid management being cost-effective was 76% and 83% at commonly cited willingness-to-pay threshold of £20,000 and £30,000 per quality-adjusted life-year gained, respectively. The results remained robust to a series of sensitivity analyses. Limitations The missing data level was high for some resource use categories collected through case report forms, due to COVID-19 disruptions and a significant dropout rate in the informing BioImpedance Spectroscopy to maintain Renal Output trial. Conclusions Compared with current fluid management, bioimpedance-guided fluid management produced a marginal reduction in costs and a small improvement in quality-adjusted life-years. Results from both the base-case and sensitivity analyses suggested that use of bioimpedance is likely to be cost-effective. Future work Future work exploring the association between primary outcomes and longer-term survival would be useful. Should an important link be established, and relevant evidence becomes available, it would be informative to determine whether and how this might affect longer-term costs and benefits associated with bioimpedance-guided fluid management. Funding details This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number HTA 14/216/01 (NIHR136142). Plain language summary ‘Bioimpedance’ is a measure of how difficult it is for an electric current to pass through a biological object. Bioimpedance is used in devices that assess fluid status (over- or under-hydration) because it is very sensitive to the amount of water in tissue. Bioimpedance can be used in addition to clinical judgement when deciding how much water should be removed from someone with kidney failure during a dialysis treatment session. This is the first study to examine whether using this treatment represents a cost-effective use of National Health Service resources. We carried out an economic evaluation within a large randomised controlled trial in patients with kidney disease undergoing haemodialysis. We calculated the additional costs and quality-adjusted life-years (a measure that combines quality and quantity of life) using established methods. Over 2 years, our study showed that taking into account bioimpedance measurements about target weight resulted in slightly lower costs and marginally more quality-adjusted life-years, although there is uncertainty around these findings.

Background Acne is common, can cause significant impact on quality of life and is a frequent reason for long-term antibiotic use. Spironolactone has been prescribed for acne in women for many years, but robust evidence is lacking. Objective To evaluate whether spironolactone is clinically effective and cost-effective in treating acne in women. Design Pragmatic, parallel, double-blind, randomised superiority trial. Setting Primary and secondary healthcare and community settings (community and social media advertising). Participants Women aged 18 years and older with facial acne persisting for at least 6 months, judged to potentially warrant oral antibiotic treatment. Interventions Participants were randomised 1 : 1, using an independent web-based procedure, to either 50 mg/day spironolactone or matched placebo until week 6, increasing to 100 mg/day spironolactone or matched placebo until week 24. Participants continued usual topical treatment. Main outcome measures Primary outcome was the adjusted mean difference in Acne-Specific Quality of Life symptom subscale score at 12 weeks. Secondary outcomes included Acne-Specific Quality of Life total and subscales; participant self-assessed improvement; Investigator’s Global Assessment; Participant’s Global Assessment; satisfaction; adverse effects and cost-effectiveness. Results Of 1267 women assessed for eligibility, 410 were randomised (201 intervention, 209 control), 342 in the primary analysis (176 intervention, 166 control). Mean age was 29.2 years (standard deviation 7.2) and 7.9% (28/356) were from non-white backgrounds. At baseline, Investigator’s Global Assessment classified acne as mild in 46%, moderate in 40% and severe in 13%. At baseline, 82.9% were using topical treatments. Over 95% of participants in both groups tolerated the treatment and increased their dose. Mean baseline Acne-Specific Quality of Life symptom subscale was 13.0 (standard deviation 4.7) across both groups. Mean scores at week 12 were 19.2 (standard deviation 6.1) for spironolactone and 17.8 (standard deviation 5.6) for placebo [difference favouring spironolactone 1.27 (95% confidence interval 0.07 to 2.46) adjusting for baseline variables]. Mean scores at week 24 were 21.2 (standard deviation 5.9) in spironolactone group and 17.4 (standard deviation 5.8) in placebo group [adjusted difference 3.77 (95% confidence interval 2.50 to 5.03) adjusted]. Secondary outcomes also favoured spironolactone at 12 weeks with greater differences at 24 weeks. Participants taking spironolactone were more likely than those taking placebo to report overall acne improvement at 12 weeks {72.2% vs. 67.9% [adjusted odds ratio 1.16 (95% confidence interval 0.70 to 1.91)]} and at 24 weeks {81.9% vs. 63.3% [adjusted odds ratio 2.72 (95% confidence interval 1.50 to 4.93)]}. Investigator’s Global Assessment was judged successful at week 12 for 31/201 (18.5%) taking spironolactone and 9/209 (5.6%) taking placebo [adjusted odds ratio 5.18 (95% confidence interval 2.18 to 12.28)]. Satisfaction with treatment improved in 70.6% of participants taking spironolactone compared with 43.1% taking placebo [adjusted odds ratio 3.12 (95% confidence interval 1.80 to 5.41)]. Adverse reactions were similar between groups, but headaches were reported more commonly on spironolactone (20.4% vs. 12.0%). No serious adverse reactions were reported. Taking account for missing data through multiple imputation gave an incremental cost per quality-adjusted life-year of £27,879 (adjusted) compared to placebo or £2683 per quality-adjusted life-year compared to oral antibiotics. Conclusions Spironolactone resulted in better participant-reported and investigator-reported outcomes than placebo, with greater differences at week 24 than week 12. Trial registration This trial is registered as ISRCTN12892056 and EudraCT (2018-003630-33). Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 16/13/02) and is published in full in Health Technology Assessment; Vol. 28, No. 56. See the NIHR Funding and Awards website for further award information. Plain language summary What was the question? Acne (or spots) is common and often persists into adulthood. Many people take long courses of antibiotic tablets, but concerns about antibiotic resistance mean alternatives are needed. Spironolactone is a medicine that is sometimes used for acne in women. However, we do not know whether it works. This trial aimed to answer this question. What did we do? We invited women aged over 18 who had acne on their face for at least 6 months to take part via their general practitioner surgery, hospital or advertising. Women were randomly assigned to two groups: one group was given spironolactone and the other group was given identical-looking placebo (‘dummy pill’) daily for 24 weeks. Women in both groups could continue using acne treatments applied to the skin (gels/creams/lotions). We asked participants to rate their acne using a questionnaire called Acne-Specific Quality of Life, asked whether they felt their skin had improved and asked skin specialists to assess their skin. What did we find? Four hundred and ten women took part, many of whom had had acne for a long time. Acne-Specific Quality of Life scores improved in both groups by 12 weeks but improved more in the spironolactone group at 12 and 24 weeks. When asked directly whether their skin had improved, 71% of participants in the spironolactone group said it had, compared with 43% on placebo. Skin specialists were also more likely to report that the acne had improved in the spironolactone group. Side effects were mild and similar in both groups but there were slightly more headaches on spironolactone (20% compared with 12%). Spironolactone is likely to represent value for money for the National Health Service, though this depends on a number of factors including what it is compared to. What does this mean? This trial suggests that spironolactone is a useful additional treatment for women with persistent acne. Scientific summary Background Acne vulgaris (hereon ‘acne’) is common, can cause significant psychosocial impact and risks permanent scarring. Topical treatments are first line, but people commonly receive long courses of oral antibiotics, raising concerns regarding antimicrobial resistance. Spironolactone, a potassium-sparing diuretic, is widely used for other conditions, such as hypertension. Spironolactone has anti-androgenic properties and is prescribed by dermatologists to treat acne in women, but robust evidence of effectiveness is lacking. Objective To evaluate whether spironolactone is clinically effective and cost-effective in treating persistent facial acne in women. Methods Design This was a pragmatic, multicentre, participant-blinded and clinician-blinded, placebo-controlled randomised trial. Participants were recruited through primary care (search and mail-out or opportunistic recruitment), secondary care (opportunistic recruitment) and advertising, including community and social media advertising. Trial participants were randomised to receive either 50 mg spironolactone or matched placebo one tablet daily for the first 6 weeks and then two tablets daily (total 100 mg spironolactone or matched placebo) at (or after) week 6, providing the participant was tolerating any side effects. Treatment continued for 24 weeks in both groups. Participants in both groups could continue to use usual topical treatments throughout the trial but adherence to topicals was not promoted beyond usual care. Between baseline and week 12, participants were asked not to change their topical treatments and not to take oral treatments for acne such as oral antibiotics, hormonal treatments or isotretinoin. After week 12, participants could change usual acne care, including oral treatments, if needed. In both groups, spironolactone or placebo was stopped at week 24, participants were informed of their treatment allocation and entered an unblinded follow-up period up to week 52. After week 24, participants could seek any treatment from their usual clinical team, including spironolactone. Baseline assessment was carried out in secondary care to ensure standardisation of clinical assessments, as the Investigator’s Global Assessment (IGA) for acne is not commonly used in primary care and was an important secondary outcome. The baseline appointment included pregnancy test, blood test (to exclude renal impairment or raised serum potassium), participant photo to aid recall about changes in acne, and contraceptive counselling. Baseline visits were conducted by research nurses and/or dermatologists. Participants were followed up face-to-face (or by video call or telephone due to the COVID-19 pandemic) in secondary care at week 6 and week 12, with primary outcome assessment at week 12, and longer-term follow-up by questionnaires at week 24 and up to week 52. Participants Participants eligible for inclusion were women aged 18 years or over: with facial acne vulgaris for at least 6 months acne of sufficient severity to warrant oral antibiotics, as judged by trial clinician; and with IGA ≥ 2 (mild or worse) women of childbearing potential at risk of pregnancy had to be willing to use their usual hormonal or barrier method of contraception for the first 6 months of the trial and for at least 4 weeks afterwards willing to be randomised sufficient English to self-complete acne-specific quality of life (Acne-QoL). Potential participants were excluded if: IGA acne grade was 0–1 (clear or almost clear) ever taken spironolactone taken oral isotretinoin within past 6 months taken oral antibiotics (lasting longer than 1 week) for acne within previous month started, stopped or changed hormonal contraception, co-cyprindiol or other hormonal treatment within past 3 months or planning to start, stop or change within the next 3 months intending to become pregnant in next 6 months spironolactone contraindicated: currently taking potassium-sparing diuretic, angiotensin-converting enzyme inhibitor, angiotensin II receptor blocker or digoxin hereditary problems of galactose intolerance, lactase deficiency or glucose–galactose malabsorption (as the spironolactone and placebo tablets contain lactose) androgen-secreting adrenal or ovarian tumour Cushing syndrome congenital adrenal hyperplasia estimated glomerular filtration rate below 60 ml/minute/1.73 m2 serum potassium level above upper limit of reference range for laboratory. Outcomes Primary outcome was comparison of mean Acne-QoL symptom subscale scores between groups at week 12. Acne-QoL contains 19 questions with seven response categories, each referring to the past week, reported in four domains (acne symptoms, self-perception, role-social, role-emotional). Secondary outcomes included: Acne-QoL symptom subscale score at week 24 and up to week 52 Acne-QoL other subscales (self-perception, role-emotional and role-social) at week 12, week 24 and up to week 52 participant self-assessed overall improvement at week 12 recorded on six-point Likert scale (with baseline photo to aid recall) IGA change from baseline to week 12 Participant’s Global Assessment (PGA) change at week 12 and week 24 generic health-related quality of life at week 6, week 12, week 24 and up to week 52 [EuroQol-5 Dimensions, five-level version (EQ-5D-5L)] adverse reactions use of oral acne treatment during follow-up (e.g. antibiotics, isotretinoin) resource use. Sample size Sample size calculation was based on Acne-QoL symptom subscale scores as recommended by the questionnaire developer. Based on comparison Acne-QoL symptom subscale scores between groups at week 12, power 90%, alpha 0.05 and seeking a difference of 2 points between groups (effect size 0.35), a target sample size of 346 participants was initially estimated, or 434 participants allowing for 20% loss to follow-up. During the trial, the target sample size was revised to allow for correlation between baseline and follow-up measures, following discussion with trial monitoring committees and funder. Allowing for a correlation with baseline of 0.293 and a deflation factor of 1-ρ2, gave a revised target sample size of 398 participants (including allowing for 20% loss to follow-up). Randomisation and blinding Participants were randomised in a 1 : 1 ratio to either spironolactone or matched placebo using an independent web-based system (block sizes 2 and 4). Participants were stratified by recruitment centre and baseline acne severity (IGA

Background Stroke or transient ischaemic attack patients are at increased risk of secondary vascular events. Antiplatelet medications, most commonly clopidogrel, are prescribed to reduce this risk. Factors including CYP2C19 genetic variants can hinder clopidogrel metabolism. Laboratory-based or point-of-care tests can detect these variants, enabling targeted treatment. Objective To assess the effectiveness of genetic testing to identify clopidogrel resistance in people with ischaemic stroke or transient ischaemic attack. Specific objectives: Do people tested for clopidogrel resistance, and treated accordingly, have a reduced risk of secondary vascular events? Do people with loss-of-function alleles associated with clopidogrel resistance have a reduced risk of secondary vascular events if treated with alternative interventions compared to clopidogrel? Do people with loss-of-function alleles associated with clopidogrel resistance have an increased risk of secondary vascular events when treated with clopidogrel? What is the accuracy of point-of-care tests for detecting variants associated with clopidogrel resistance? What is the technical performance and cost of CYP2C19 genetic tests? Is genetic testing for clopidogrel resistance cost-effective compared with no testing? Design Systematic review and economic model. Results Objective 1: Two studies assessed secondary vascular events in patients tested for loss-of-function alleles and treated accordingly. They found a reduced risk, but confidence intervals were wide (hazard ratio 0.50, 95% confidence interval 0.09 to 2.74 and hazard ratio 0.53, 95% confidence interval 0.24 to 1.18). Objective 2: Seven randomised controlled trials compared clopidogrel with alternative treatment in people with genetic variants. Ticagrelor was associated with a lower risk of secondary vascular events than clopidogrel (summary hazard ratio 0.76, 95% confidence interval 0.65 to 0.90; two studies). Objective 3: Twenty-five studies compared outcomes in people with and without genetic variants treated with clopidogrel. People with genetic variants were at an increased risk of secondary vascular events (hazard ratio 1.72, 95% confidence interval 1.43 to 2.08; 18 studies). There was no difference in bleeding risk (hazard ratio 0.98, 95% confidence interval 0.68 to 1.40; five studies). Objective 4: Eleven studies evaluated Genomadix Cube accuracy; no studies evaluated Genedrive. Summary sensitivity and specificity against laboratory reference standards were both 100% (95% confidence interval 94% to 100% and 99% to 100%). Objective 5: Seventeen studies evaluated technical performance of point-of-care tests. Test failure rate ranged from 0.4% to 19% for Genomadix Cube. A survey of 8/10 genomic laboratory hubs revealed variation in preferred technologies for testing, and cost per test ranging from £15 to £250. Most laboratories expected test failure rate to be

Background This work was undertaken to inform a National Institute for Health and Care Excellence guideline on the initial assessment of adults with suspected acute respiratory infection. Objective To undertake a rapid evidence synthesis of systematic reviews and cost-effectiveness studies of signs, symptoms and early warning scores for the initial assessment of adults with suspected acute respiratory infection. Methods MEDLINE, EMBASE and Cochrane Database of Systematic Reviews were searched for systematic reviews and MEDLINE, EMBASE, EconLit and National Health Service Economic Evaluation Database were searched for cost-effectiveness studies in May 2023. References of relevant studies were checked. Clinical outcomes of interest included escalation of care, antibiotic/antiviral use, time to resolution of symptoms, mortality and health-related quality of life. Risk of bias was assessed using the Risk of Bias in Systematic Reviews tool or the National Institute for Health and Care Excellence economic evaluations checklist. Results were summarised using narrative synthesis. Results Nine systematic reviews and one cost-effectiveness study met eligibility criteria. Seven reviews assessed several early warning scores for patients with community- acquired pneumonia, one assessed early warning scores for nursing home-acquired pneumonia and one assessed individual signs/symptoms and the Centor score for patients with sore throat symptoms; all in face-to-face settings. Two good-quality reviews concluded that further research is needed to validate the CRB-65 in primary care/community settings. One also concluded that further research is needed on the Pneumonia Severity Index in community settings; however, the Pneumonia Severity Index requires data from tests not routinely conducted in community settings. One good-quality review concluded that National Early Warning Score appears to be useful in an emergency department/acute medical setting. One review (unclear quality) concluded that the Pneumonia Severity Index and CURB-65 appear useful in an emergency department setting. Two poor-quality reviews concluded that early warning scores can support clinical judgement and one poor-quality review found numerous problems with using early warning scores in a nursing home setting. A good-quality review concluded that individual signs and symptoms have a modest ability to diagnose streptococcal pharyngitis, and that the Centor score can enhance appropriate prescribing of antibiotics. The cost-effectiveness study assessed clinical scores and rapid antigen detection tests for sore throat, compared to delayed antibiotic prescribing. The study concluded that the clinical score is a cost-effective approach when compared to delayed prescribing and rapid antigen testing. Conclusions Several early warning scores have been evaluated in adults with suspected acute respiratory infection, mainly the CRB-65, CURB-65 and Pneumonia Severity Index in patients with community-acquired pneumonia. The evidence was insufficient to determine what triage strategies avoid serious illness. Some early warning scores (CURB-65, Pneumonia Severity Index and National Early Warning Score) appear to be useful in an emergency department/acute medical setting; however, further research is required to validate the CRB-65 and Pneumonia Severity Index in primary care/community settings. The economic evidence indicated that clinical scores may be a cost-effective approach to triage patients compared with delayed prescribing. Future work and limitations Only systematic reviews were eligible for inclusion in the synthesis of clinical evidence. There was a great deal of overlap in the primary studies included in the reviews, many of which had significant limitations. No studies were undertaken in remote settings (e.g. NHS 111). Only one cost-effectiveness study was identified, with limited applicability to the review question. Funding This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number NIHR159945. Plain language summary Acute respiratory infections, such as cold and flu, are common and can be caused by viruses or bacteria. People with symptoms of acute respiratory infection often go to their general practitioner, who may advise them to stay at home (with or without antibiotics or antivirals) or might refer them to hospital if the infection is serious. Doctors assess the patient’s symptoms or may use a tool called an ‘early warning score’ to judge whether the infection is serious. A systematic review is a research method where all relevant studies assessing a specific question are found and summarised. We aimed to summarise all systematic reviews and cost-effectiveness studies that assessed signs, symptoms and ‘early warning scores’ in adults with suspected acute respiratory infections in the community (i.e. not hospitalised patients). We found nine systematic reviews and one cost-effectiveness study. Several different early warning scores for acute respiratory infection have been assessed in systematic reviews. Seven of the reviews assessed early warning scores in patients with community-acquired pneumonia. Good-quality reviews concluded that further research is needed to see how useful the ‘CRB-65’ and ‘Pneumonia Severity Index’ early warning scores are for assessing pneumonia severity in the community. Another good-quality review concluded that the ‘National Early Warning Score’ early warning score appears to be useful in an emergency department setting. A good-quality review found that individual symptoms are not very reliable for diagnosing pharyngitis caused by streptococcal bacteria in patients with sore throat; the review also found that the ‘Centor score’ can help doctors decide whether to prescribe antibiotics for pharyngitis. The cost-effectiveness study assessed clinical scores and rapid antigen detection tests (which test for substances that increase in our blood when we have certain infections) in patients with sore throat, and found that clinical scores may be cost-effective compared to delaying prescribing antibiotics.

Background The presence of dental caries impacts on children’s daily lives, particularly among those living in deprived areas. There are successful interventions across the United Kingdom for young children based on toothbrushing with fluoride toothpaste. However, evidence is lacking for oral health improvement programmes in secondary-school pupils to reduce dental caries and its sequelae. Objectives To determine the clinical and cost effectiveness of a behaviour change intervention promoting toothbrushing for preventing dental caries in secondary-school pupils. Design A multicentre, school-based, assessor-blinded, two-arm cluster randomised controlled trial with an internal pilot and embedded health economic and process evaluations. Setting Secondary schools in Scotland, England and Wales with above-average proportion of pupils eligible for free school meals. Randomisation occurred within schools (year-group level), using block randomisation stratified by school. Participants Pupils aged 11–13 years at recruitment, who have their own mobile telephone. Interventions Two-component intervention based on behaviour change theory: (1) 50-minute lesson delivered by teachers, and (2) twice-daily text messages to pupils’ mobile phones about toothbrushing, compared with routine education. Main outcome measures Primary outcome: presence of at least one treated or untreated carious lesion using DICDAS4–6MFT (Decayed, Missing and Filled Teeth) in any permanent tooth, measured at pupil level at 2.5 years. Secondary outcomes included: number of DICDAS4–6MFT; presence and number of DICDAS1–6MFT; plaque; bleeding; twice-daily toothbrushing; health-related quality of life (Child Health Utility 9D); and oral health-related quality of life (Caries Impacts and Experiences Questionnaire for Children). Results Four thousand six hundred and eighty pupils (intervention, n = 2262; control, n = 2418) from 42 schools were randomised. The primary analysis on 2383 pupils (50.9%; intervention 1153, 51.0%; control 1230, 50.9%) with valid data at baseline and 2.5 years found 44.6% in the intervention group and 43.0% in control had obvious decay experience in at least one permanent tooth. There was no evidence of a difference (odds ratio 1.04, 95% confidence interval 0.85 to 1.26, p = 0.72) and no statistically significant differences in secondary outcomes except for twice-daily toothbrushing at 6 months (odds ratio 1.30, 95% confidence interval 1.03 to 1.63, p = 0.03) and gingival bleeding score (borderline) at 2.5 years (geometric mean difference 0.92, 95% confidence interval 0.85 to 1.00, p = 0.05). The intervention had higher incremental mean costs (£1.02, 95% confidence interval −1.29 to 3.23) and lower incremental mean quality-adjusted life-years (−0.003, 95% confidence interval −0.009 to 0.002). The probability of the intervention being cost-effective was 7% at 2.5 years. However, in two subgroups, pilot trial schools and schools with higher proportions of pupils eligible for free school meals, there was an 84% and 60% chance of cost effectiveness, respectively, although their incremental costs and quality-adjusted life-years remained small and not statistically significant. The process evaluation revealed that the intervention was generally acceptable, although the implementation of text messages proved challenging. The COVID-19 pandemic hampered data collection. High rates of missing economic data mean findings should be interpreted with caution. Conclusions Engagement with the intervention and evidence of 6-month change in toothbrushing behaviour was positive but did not translate into a reduction of caries. Future work should include work with secondary-school pupils to develop an understanding of the determinants of oral health behaviours, including toothbrushing and sugar consumption, particularly according to free school meal eligibility. Trial registration This trial is registered as ISRCTN12139369. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 15/166/08) and is published in full in Health Technology Assessment; Vol. 28, No. 52. See the NIHR Funding and Awards website for further award information. Plain language summary What was the problem? Tooth decay has an impact on children and young people’s daily lives, particularly those living in deprived areas. For young children, programmes to improve toothbrushing with fluoride toothpaste help prevent tooth decay. The Brushing RemInder 4 Good oral HealTh trial (BRIGHT) investigated whether a secondary-school-based toothbrushing programme would work. What did we do? We developed a new programme which included a lesson and twice-daily text messages sent to pupils’ phones. In total, 4680 pupils, aged 11–13 years, from 42 secondary schools in the United Kingdom took part in the trial. At each school, one year group was randomly selected to receive the programme, while the other year group did not receive it. All pupils were followed up for 2.5 years to see whether there were any differences in levels of tooth decay, frequency of toothbrushing, plaque or quality of life. We also considered the programme’s value for money and the views of pupils and school staff. What did we find? We followed up 2383 pupils and found no difference in tooth decay, plaque or quality of life. We found those who had the programme were more likely to brush their teeth twice daily after 6 months than those who did not. The programme was not good value for money overall. However, the programme appeared to be of more benefit at preventing tooth decay in pupils eligible for free school meals compared to those not eligible. In the schools with more pupils eligible for free school meals, the chance of the programme representing good value for money increased. The programme was generally liked by the pupils and school staff. Some pupils found the text messages useful, although others said they were annoying. What does this mean? The programme helped pupils brush their teeth more frequently in the short term, but this did not lead to less tooth decay. Further research is needed to understand how to prevent tooth decay in secondary-school pupils. Scientific summary Background Reducing the high prevalence and severity of dental caries in the UK is a public health priority. Dental caries is largely preventable with several evidence-based child oral health promotion interventions including toothbrushing with fluoride toothpaste. However, evidence is lacking for interventions targeting toothbrushing practices to reduce dental caries and its sequelae (pain, infection and tooth loss) in secondary-school pupils. The Brushing RemInder 4 Good oral HealTh (BRIGHT) trial investigated a two-component behaviour change intervention (lesson and text messages) to prevent dental caries in secondary-school pupils. Objectives Conduct an internal pilot phase with feasibility components to: Tailor the intervention to young people. Test trial processes. Assess feasibility of within-school cluster randomisation (by year group). Investigate the intervention’s effect on caries prevalence, twice-daily toothbrushing, oral health-related quality of life and oral health behaviours. Investigate the intervention’s cost effectiveness. Explore implementation, mechanisms of impact and context through process evaluation. Methods Design A multicentre, school-based, assessor-blinded, two-arm cluster randomised controlled trial with an internal pilot phase, and embedded health economic and process evaluations. Criteria determining progression to the main phase were pre-specified and reviewed within the internal pilot phase. Setting Secondary schools in Scotland, England and Wales with above national average percentage of pupils eligible for free school meals (FSM) were recruited. Schools had to have pupils aged 11–16 years and at least 60 pupils per year group. Participant recruitment Pupils aged 11–13 years (Years 7 and 8 in England and Wales; S1 and S2 in Scotland) in participating schools were eligible to take part and received an information session about BRIGHT. Parents/carers had 2 weeks to decline their child’s participation. Pupils who had not been opted out were invited to participate by completing a consent form and providing their mobile telephone number. If they did not own a mobile telephone or could not provide their number, they were ineligible. Sample size We proposed to recruit 5040 pupils from 42 schools (84 year groups) to give 90% power to detect a reduction in the proportion of pupils with obvious decay experience from 34% to 26%, assuming: within-school (year-group level) randomisation, partial contamination effects (i.e. those contaminated gain half the treatment benefits) for 27% of the control group, recruiting an average of 60 pupils per year group, an intracluster correlation coefficient of 0.02% and 20% attrition at follow-up. A sample size of 10 schools was considered sufficient to test internal pilot objectives. Randomisation Year groups within schools were randomised 1 : 1 to the intervention or control arm. An allocation sequence, stratified by school using blocks of size two, was generated by an independent statistician. Once each school’s baseline assessments were completed, the year groups were randomised by allocation to the next available block. Trial interventions The intervention consisted of two components: (1) a 50-minute classroom-based session (CBS) about dental health delivered by teachers using a lesson plan and pupil-facing materials followed by (2) a series of twice-daily text messages to participants’ mobile phones about toothbrushing. Year groups allocated to the control arm received routine education only. Follow-up Assessment time points were baseline, after the lesson (internal pilot only), 12 weeks (internal pilot only), 6 months, 1, 2 (internal pilot only) and 2.5 years following lesson delivery. Pupils completed self-report questionnaires at all these time points. Dental examinations at baseline, 2 and 2.5 years included assessment of caries, plaque and gingival bleeding. Parents/carers were sent a resource use questionnaire at baseline, and 1, 2 and 2.5 years. Note, the descriptors of the time points for the assessments (e.g. 2.5 years) reflect the planned follow-up schedule and are used throughout the report; however, some of the actual average time intervals of follow-up varied from that planned (e.g. the actual average length of follow-up at the 2.5-year assessment was closer to 3 years as this was delayed due to disruptions caused by the COVID-19 pandemic). Primary outcome The primary outcome was the presence of at least one treated or untreated carious lesion in any permanent tooth, measured at the pupil level during the 2.5-year dental assessment using DICDAS4–6MFT (Decayed, Missing and Filled Teeth). Secondary outcomes Secondary outcomes assessed during the dental assessments were DICDAS4–6MFT at 2 years, and the following at 2 and 2.5 years: the presence of at least one treated or untreated carious lesion in any permanent tooth measured using DICDAS1–6MFT, the number of permanent DICDAS4–6MFT and DICDAS1–6MFT, plaque score (modified gingival index of Löe), bleeding score (gingival index) and number of teeth with bleeding gingivae. Participant-reported secondary outcome measures were: twice-daily toothbrushing, health-related quality of life measured by the Child Health Utility 9D (CHU9D) and oral health-related quality of life assessed using Caries Impacts and Experiences Questionnaire for Children (CARIES-QC). Statistical methods Both the clinical effectiveness and economic analyses were conducted to pre-specified and externally endorsed analysis plans. Analyses were conducted in STATA v17 (StataCorp LP, College Station, TX, USA) following the principles of intention to treat using two-sided statistical tests assessed at the 5% significance level. Baseline data were summarised descriptively overall and by randomised arm. The primary outcome was analysed using mixed-effect logistic regression, adjusting for the number of DICDAS4–6MFT and school year at baseline as fixed effects, and school as a random effect. Sensitivity analyses were conducted in which year group, nested within school, was additionally included as a random effect, and including additional covariates that were significantly associated with missing primary outcome data. A further sensitivity analysis excluded pupils whose dental assessments were completed outside of 3 months either side of the average length of follow-up for the 2.5-year time point. A complier-average causal effect (CACE) analysis, using a two-stage instrumental variable regression approach with the randomised group as the instrumental variable, was implemented to assess the impact of: attending the lesson; attending the lesson and receiving at least seven text messages a week for the first 12 weeks; and number of text messages sent. Subgroup analyses considered whether or not the intervention effect differed by FSM status, number of carious teeth at baseline and whether the school was recruited during the pilot or main trial phase. The secondary outcome of presence of at least one DICDAS1–6MFT was analysed as described for the primary outcome. The number of permanent DICDAS4–6MFT and DICDAS1–6MFT and number of teeth with bleeding gingivae were analysed using mixed-effect negative binomial regression, self-reported twice-daily toothbrushing via mixed-effect logistic regression and plaque, bleeding and CARIES-QC scores via mixed-effect linear regression. Models were adjusted for school year and associated baseline value of the outcome as fixed effects and school as a random effect. Attendance data, adverse events and safeguarding issues are summarised descriptively. Economic analysis The economic evaluation was undertaken using individual-level trial data, took the NHS perspective and generated quality-adjusted life-years (QALYs) using the CHU9D. The time horizon of the analysis was set to cover all relevant costs and outcomes. Parental questionnaires were used to collect data on dental treatments and child health-related quality of life. Other sources of cost data were used in relation to the text messages and the lesson. Results Clinical effectiveness results Of the 14,083 pupils approached in the 42 recruited schools, 4699 (33.4%) were eligible and consented and were asked to complete baseline data collection; however, 19 withdrew pre randomisation leaving 4680 pupils (92.9% of our target of 5040) included in the randomised sample (intervention, n = 2262; control, n = 2418). The average number of pupils recruited per school was 111.4 [standard deviation (SD) 35.9, median 107, range 46–189] and per year group was 55.7 (SD 21.6, median 53, range 13–119). The average age of pupils at recruitment was 12.7 years (SD 0.6), 54.2% were female and 21.9% were eligible for FSM. Over three-quarters (77.6%) reported brushing their teeth at least twice a day. There was a valid baseline dental assessment for 4625 pupils; 1603 (34.7%) had evidence of obvious decay experience (presence of DICDAS4–6MFT in at least one permanent tooth) and 2929 (63.3%) had at least one treated or untreated carious lesion in any permanent tooth (DICDAS1–6MFT). Baseline data were similar between the intervention and control groups. Confirmation of lesson delivery was received from 39 of the 42 schools, with an estimated 2016 (89.1%) of 2262 pupils randomised to the intervention group attending. Text messages were sent to 2258 (99.8%) intervention pupils. The other four withdrew shortly after receiving the lesson and so their messages were not commenced. Participants were sent text messages until they requested them to stop or until 12 July 2020 (when a technical error occurred with the text provider that meant texts stopped being sent). A total of 962 intervention participants (42.5%) withdrew from receiving the text messages, a median of 2.8 months after they commenced (range 1 day to 30 months). Participants were sent messages for between 0 and 127 weeks (approximately 30 months, mean 53.4 weeks, SD 35.4, median 62). This equated to between 1 and 1708 text messages (mean 694.5, SD 468.9, median 789). On average, 71.4% of the text messages sent to a participant were successfully delivered. At 2.5 years, 1043 out of 2383 pupils (43.8%; intervention 514, 44.6%; control 529, 43.0%) had obvious decay experience in at least one permanent tooth. There was no evidence of a difference between the intervention and control groups [odds ratio (OR) 1.04, 95% confidence interval (CI) 0.85 to 1.26, p = 0.72]. The sensitivity analyses produced similar results to the primary analysis. The CACE estimates of the treatment effect based on attending the CBS session and on attending the CBS session and receiving at least 50% of their messages per week for the first 12 weeks were similar to the intention to treat estimate (OR 1.05, 95% CI 0.85 to 1.31, p = 0.64, and 1.07, 95% CI 0.72 to 1.59, p = 0.74, respectively). The CACE estimate associated with the number of texts sent was OR 1.00 (95% CI 0.999 to 1.001, p = 0.93), which indicates that for every additional text message sent, there was no evidence of a decrease in likelihood of having a carious lesion. There was no significant interaction between treatment allocation and either number of carious teeth at baseline or pilot/main trial schools, but there was evidence of a qualitative interaction for FSM status, with a benefit of the intervention seen among FSM pupils (OR 0.69, 95% CI 0.44 to 1.08, p = 0.10) but not among non-FSM pupils (OR 1.17, 95% CI 0.93 to 1.46, p = 0.18). There was evidence of a statistically significant difference for twice-daily toothbrushing at 6 months (OR 1.30, 95% CI 1.03 to 1.63, p = 0.03) and borderline evidence of a difference in gingival index score between the two groups (geometric mean difference 0.92, 95% CI 0.85 to 1.00, p = 0.05) at 2.5 years. One non-serious adverse event was recorded during the trial, which was deemed possibly related and unexpected, for a pupil in the control group. No suspected serious pathologies were identified during dental assessments. Fifteen safeguarding issues arose during the course of the trial. All were dealt with according to the trial safeguarding procedure. Cost-effectiveness results Clinical results suggested that there was unlikely to be effects beyond 2.5 years, which was therefore set as the time horizon. Due to high rates of missing data, treatment costs were estimated from dental assessment data using multiple imputation. Text messages cost £32.53 per pupil on average, while mean dental costs over the 2.5-year follow-up were £20.73 and £21.02 for control and intervention groups, respectively. QALYs were similar in the control and intervention groups (2.196 and 2.193, respectively). Regression analysis estimated incremental costs and QALYs of the intervention, relative to the control group, to be £1.02 (95% CI −1.29 to 3.23) and −0.003 (95% CI −0.009 to 0.002), respectively, with a 7% chance that the intervention is cost-effective using a £20,000 per QALY gained threshold. The results were robust to the changes explored in the sensitivity analyses, except for an alternative approach to estimating QALYs using the CARIES-QC. While this only had a very small impact on incremental costs and QALYs, it was enough to generate an incremental QALY gain, which, in combination with a very small incremental cost, generates an incremental cost-effectiveness ratio (ICER) of £79 per QALY gained and an associated probability of the intervention being cost-effective of 96%. The subgroup analyses suggest that there is no evidence of a difference in the probability of the intervention being cost-effective within schools in Scotland versus England/Wales (6% in both cases). A positive QALY gain is observed in those schools with higher levels of pupils eligible for FSM than schools with lower levels of FSM eligibility. This produces an ICER of £2254 per QALY gained and a probability of the intervention being cost-effective of 60%. We also observed a QALY gain for pilot schools which leads to the intervention having an ICER of £3049 per QALY gained (and an 84% chance of it being cost-effective). Process evaluation The process evaluation showed that the BRIGHT intervention was generally implemented as intended, although there were challenges establishing the dose delivered and received for both components and technical difficulties delivering the text messages. Overall, pupils, staff members and stakeholders felt the intervention was acceptable. Conclusions At the 2.5-year follow-up, there was no evidence of a difference between the control and intervention groups in the prevalence of caries extending to dentine (primary outcome) or including enamel and dentine lesions. The proportion of participants with caries into dentine was high (intervention group 34.4% at baseline and 44.7% at the 2.5-year follow-up, and control group 34.9% at baseline and 43.1% at follow-up). There was an indication of a positive effect on short-term toothbrushing behaviour at 6 months. The subgroup analysis of participants eligible for FSM suggests a significant, qualitative interaction effect whereby the intervention appeared to be beneficial in terms of caries prevalence within pupils eligible for FSM but not for those not eligible for FSM. The process evaluation found the intervention to be broadly acceptable with some technical issues of text message delivery. The primary economic analysis shows that the intervention is not likely to be cost-effective. Further research is needed to understand how to prevent dental caries in secondary-school pupils. Trial registration This trial is registered as ISRCTN12139369. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 15/166/08) and is published in full in Health Technology Assessment; Vol. 28, No. 52. See the NIHR Funding and Awards website for further award information.

Background Older adults with multimorbidity experience impaired health-related quality of life and treatment burden. Yoga has the potential to improve several aspects of health and well-being. The British Wheel of Yoga’s Gentle Years Yoga© programme was developed specifically for older adults, including those with chronic conditions. A pilot trial demonstrated feasibility of using Gentle Years Yoga in this population, but there was limited evidence of its effectiveness and cost-effectiveness. Objective To determine the effectiveness and cost-effectiveness of the Gentle Years Yoga programme in addition to usual care versus usual care alone in older adults with multimorbidity. Design Pragmatic, multisite, individually randomised controlled trial with embedded economic and process evaluations. Setting Participants were recruited from 15 general practices in England and Wales from July 2019 with final follow-up in October 2022. Participants Community-dwelling adults aged 65 years and over with multimorbidity, defined as two or more chronic health conditions from a predefined list. Interventions All participants continued with any usual care provided by primary, secondary, community and social services. The intervention group was offered a 12-week programme of Gentle Years Yoga. Main outcome measures The primary outcome and end point were health-related quality of life measured using the EuroQol-5 Dimensions, five-level version utility index score over 12 months. Secondary outcomes were health-related quality of life, depression, anxiety, loneliness, incidence of falls, adverse events and healthcare resource use. Results The mean age of the 454 randomised participants was 73.5 years; 60.6% were female, and participants had a median of three chronic conditions. The primary analysis included 422 participants (intervention, n = 227 of 240, 94.6%; usual care, n = 195 of 214, 91.1%). There was no statistically or clinically significant difference in the EuroQol-5 Dimensions, five-level version utility index score over 12 months: the predicted mean score for the intervention group was 0.729 (95% confidence interval 0.712 to 0.747) and for usual care it was 0.710 [95% confidence interval (CI) 0.691 to 0.729], with an adjusted mean difference of 0.020 favouring intervention (95% CI −0.006 to 0.045, p = 0.14). No statistically significant differences were observed in secondary outcomes, except for the pain items of the Patient-Reported Outcomes Measurement Information System-29. No serious, related adverse events were reported. The intervention cost £80.85 more per participant (95% CI £76.73 to £84.97) than usual care, generated an additional 0.0178 quality-adjusted life-years per participant (95% CI 0.0175 to 0.0180) and had a 79% probability of being cost-effective at the National Institute for Health and Care Excellence threshold of £20,000 per quality-adjusted life-year gained. The intervention was acceptable to participants, with seven courses delivered face to face and 12 online. Limitations Self-reported outcome data raise the potential for bias in an unblinded trial. The COVID-19 pandemic affected recruitment, follow-up and the mode of intervention delivery. Conclusions Although the Gentle Years Yoga programme was not associated with any statistically significant benefits in terms of health-related quality of life, mental health, loneliness or falls, the intervention was safe, acceptable to most participants and highly valued by some. The economic evaluation suggests that the intervention could be cost-effective. Future work Longer-term cost-effectiveness modelling and identifying subgroups of people who are most likely to benefit from this type of intervention. Trial registration This trial is registered as ISRCTN13567538. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/94/36) and is published in full in Health Technology Assessment; Vol. 28, No. 53. See the NIHR Funding and Awards website for further award information. Plain language summary It is common for older adults to have two or more long-term health conditions. These conditions affect quality of life differently, with some people feeling well and others needing healthcare support. The Gentle Years Yoga programme was developed to improve quality of life for older adults, including those with long-term health conditions. We wanted to see how well the programme worked and if it offered good value for money for the NHS. We tested whether offering a 12-week course of Gentle Years Yoga improved the quality of life and reduced anxiety, depression, loneliness and falls for people aged 65 years and over who had two or more long-term health conditions. We recruited 454 people through general practices across England and Wales, with 240 people selected at random to be invited to take part in the Gentle Years Yoga programme and the other 214 to continue with their usual care and not be offered Gentle Years Yoga. The average age of participants was 74 years, nearly two-thirds were female and the number of long-term health conditions participants had ranged from two to nine (average was three). They completed four questionnaires over a 12-month period. We also interviewed some of the participants and the yoga teachers to find out how the approach worked in practice. The yoga was delivered either face to face or online. We did not find any significant benefits in terms of quality of life, anxiety, depression, loneliness or falls. At interview, some yoga participants noted no or a modest impact on their health or lifestyle, while others described Gentle Years Yoga as transformative, having substantial impacts and improvements on their physical health and emotional well-being. Because running the yoga classes was relatively inexpensive and some insignificant benefits were seen, the Gentle Years Yoga programme may be good value for money. Scientific summary Background Multimorbidity, having two or more chronic health conditions, is a major challenge for older adults and increases healthcare utilisation and associated costs. Multimorbidity is associated with reduced health-related quality of life (HRQoL), impaired functional status, worse physical and mental health and premature death. In 2015, 54% of people aged 65 years and over in England had multimorbidity. There has been limited exploration of the effectiveness of interventions to improve outcomes for people with multimorbidity. There is some evidence to suggest that yoga may help to prevent and treat various physical and mental illnesses and improve HRQoL. The available data offer support for the beneficial effects of yoga in older adults and for several chronic conditions. However, robust evidence of clinical and cost effectiveness is limited, and little research has specifically focused on older adults with multimorbidity. The Gentle Years Yoga (GYY) programme was developed to cater specifically to the needs of older adults, including those with health conditions common to an older cohort such as osteoarthritis, hypertension and cognitive impairment. A pilot randomised trial of the GYY programme (n = 52 adults, mean age 75 years) demonstrated feasibility, with a potential for a clinically important benefit on health status [EuroQol-5 Dimensions, five-level version (EQ-5D-5L) utility index score] at 3 months after randomisation [mean difference 0.12, 95% confidence interval (CI) 0.03 to 0.21]. Consequently, we conducted this larger trial, as if shown to be clinically and cost-effective, GYY could be widely implemented, leading to improved outcomes for this population. Objectives The primary objective was to establish if the offer of a 12-week GYY programme in addition to usual care is more effective compared with usual care alone in improving HRQoL (EQ-5D-5L utility index score) over 12 months in people aged 65 years or over with multimorbidity. Secondary objectives were to: explore the effect of the GYY programme on HRQoL, depression, anxiety, loneliness and incidence of falls explore the safety of the GYY programme in terms of the occurrence of adverse events assess the cost-effectiveness of the GYY programme undertake a qualitative process evaluation to explore the acceptability of the intervention and the experience of participants and teachers, explain the determinants of delivery and identify the optimal implementation strategies. Methods Design This was a multisite, two-arm, parallel-group, superiority, individually randomised controlled trial comparing an experimental strategy of offering a 12-week GYY programme against a control strategy of no offer of GYY in community-dwelling people aged 65 years or over who had multimorbidity. Both trial arms continued with any usual care provided by primary, secondary, community and social services. The study also included cost-effectiveness and qualitative process evaluations. Setting Participants were recruited from primary care general practices in the UK. General practices interested in taking part in the trial were identified with help from the NHS Clinical Research Networks and the Health and Care Research Wales Support and Delivery Centre. General practices were selected based on their proximity to yoga class venues and local transport routes. The yoga courses were delivered either face-to-face in a yoga studio, community hall or leisure centre or online via video conferencing during periods of social distancing restrictions. The trial’s yoga consultants identified yoga teachers who were eligible and potentially interested in taking part in the trial. Yoga teachers needed to have completed the British Wheel of Yoga (BWY) Qualification Level 4 Teaching GYY and have valid BWY membership and insurance. For online courses, teachers also needed to be proficient in remote teaching. Participant recruitment and consent Patients were eligible to join the study if they were aged 65 years or older, community-dwelling and had two or more of the predefined chronic health conditions derived from the NHS Quality and Outcomes Framework. Exclusion criteria were: inability to attend one of the GYY courses on offer; yoga practice in the previous 6 months; contraindications to yoga participation; severe mental health problem; learning disability; or being unable to provide consent and/or return the baseline questionnaire. For online classes, ineligibility also included no internet access; inability to use the internet; no suitable device; insufficient space at home; and/or no sturdy chair for use during the classes. Potential participants were identified by searching general practitioner (GP) electronic patient databases. Participating practices ran a custom-built search, based on predefined read codes, which identified patients with eligible health conditions. Potentially eligible patients were sent a recruitment pack including an information sheet and, if interested, were invited to sign and return a consent form. After the research team checked their suitability for the trial with their GP, eligible patients were asked to provide baseline data on sociodemographic measures, primary and secondary outcome measures and preferences/beliefs for the treatments on offer in the trial. Participants indicated on the consent form if they also wanted to be considered to take part in the process evaluation interviews. Intervention and comparison Gentle Years Yoga is designed for older adults, including those with chronic conditions. Based on standard Hatha Yoga, it incorporates physical postures and transitions as well as breathing, concentration and relaxation activities. The aims of GYY are to improve muscle strength, flexibility, balance, mobility and mental and social well-being. Chairs are used for seated exercise and support when standing. The yoga practices are modified so individuals with varying medical conditions and functional abilities can participate safely. Props are used to modify some of the postures and concentration activities. The physical challenge of each posture can be progressed throughout the course as participants become more able and confident. Participants randomised to the intervention were invited to take part in a free GYY course. Each course involved 12, 75-minute sessions of group-based yoga, usually delivered 1 week apart, either face to face or online. Each class included: ‘housekeeping’ activities (5 minutes); an introduction to the theme and practices of the class, basic breathing and focusing activities (5 minutes); an extended warmup/mobilisation and preparatory postures (30–35 minutes); focused postures and restorative activities (10–15 minutes); breathing exercises (5–10 minutes); and relaxation and concentration activities (5–10 minutes), followed by optional after-class social time (15–30 minutes). Throughout the trial, both trial arms continued with any usual care provided by primary care, secondary care, community and social services. Sample size We proposed to randomise 586 participants in a 1:1 ratio to be able to detect a difference of 0.06 in EQ-5D-5L utility index score, assuming a standard deviation (SD) of 0.20, with 90% power, a two-sided alpha of 0.05 and 20% attrition. In October 2021, an interim calculation of the correlation between baseline and 12-month EQ-5D-5L utility index score indicated we would be able to detect a clinically important difference with close to or greater than 90% power with 454 participants, since the primary analysis adjusted for baseline score, which affords gains in power. Randomisation Participants were randomised via a central, computer-based randomisation system designed and managed by York Trials Unit (YTU), University of York. Randomisation was stratified by site using varying block sizes and allocation ratios. Blinding of participants or the yoga teachers was not possible. Outcome measures Outcomes were self-reported by the participant and collected using questionnaires at baseline and 3, 6 and 12 months after randomisation. The primary outcome and end point was the EQ-5D-5L utility index score over the 12-month follow-up period. Utility index scores were calculated following current National Institute for Health and Care Excellence guidance. Secondary outcomes were: EQ-5D-5L utility index score at 3, 6 and 12 months after randomisation EQ-5D-5L visual analogue scale (VAS) score at 3, 6 and 12 months and overall HRQoL at 3, 6 and 12 months and overall using PROMIS-29 (v2.1) depression severity at 3, 6 and 12 months and overall using the Patient Health Questionnaire-8 (PHQ-8) anxiety severity at 3, 6 and 12 months and overall using the Generalised Anxiety Disorder-7 (GAD-7) loneliness at 3, 6 and 12 months and overall. Questions used to capture loneliness were taken from the English Longitudinal Study of Ageing (ELSA), based on the University of California, Los Angeles 3-item (UCLA-3) loneliness scale and a direct question about how often the respondent felt lonely the incidence of falls, adverse events and healthcare resource use over 12 months. Analysis All analyses were conducted in Stata version 17 (StataCorp LP, College Station, TX, USA) following the principles of intention-to-treat using two-sided statistical tests at the 5% significance level. The primary outcome was analysed using a linear mixed model, including data at all available follow-up time points, adjusting for baseline EQ-5D-5L utility index score, time point, trial arm and an arm-by-time interaction as fixed effects. Participant (to account for the repeated measures) and site were included as random effects. In sensitivity analyses, age, gender and adapted Bayliss (severity-adjusted count of health conditions) score were added as covariates to the primary analysis model, and the site was substituted for yoga teacher as a random effect. Complier-average causal effect (CACE) analyses, using a two-stage instrumental variable regression approach with randomised group as the instrumental variable, were implemented to assess the impact of receiving GYY on the primary treatment estimate. A subgroup analysis was conducted to assess for differential effects of the intervention based on mode of delivery (face to face or online). The secondary outcomes of EQ-5D VAS, GAD-7, PHQ-8, T-scores from each of the seven subscales of the PROMIS-29 and the physical and mental health component scores and the global item score, UCLA-3 score and ELSA single-item direct loneliness question were analysed using the same methods as for the primary outcome, with baseline EQ-5D-5L utility index score swapped as a covariate for baseline value of the outcome. The incidence of falls during the 12-month follow-up period was analysed using a mixed-effect negative binomial regression model, adjusting for the number of falls in the 3 months prior to baseline and site as a random effect. Adverse events are summarised descriptively. Economic analysis The economic analysis assessed the relative cost-effectiveness of the GYY programme in addition to continued access to usual care compared with usual care alone. Costs and health outcomes were evaluated from the perspective of the NHS and Personal Social Services using a within-trial economic analysis and a cost–consequence analysis, both over a 12-month time horizon; hence, discounting was not required. Health outcomes were assessed in terms of quality-adjusted life-years (QALYs) using EQ-5D-5L data, with costs collected for healthcare resource use, medications and the intervention. Findings were presented in terms of the incremental cost-effectiveness ratio for the intervention versus usual care and net monetary benefit. The base-case analysis was undertaken on an intention-to-treat basis, with multiple imputation used to deal with missing data and sensitivity analyses conducted to explore uncertainty around the cost-effectiveness findings. Process evaluation The process evaluation was informed by qualitative interviews with trial participants, trial decliners, trial yoga teachers and stakeholders, as well as by observations of standardisation training sessions and yoga classes. Intervention fidelity was assessed by class observation and via interviews with teachers as part of the process evaluation. Interviews were audio-recorded, transcribed verbatim and edited to ensure respondent anonymity. Data analysis was iterative throughout the trial and conducted according to standard procedures of qualitative analysis. Results Between July 2019 and August 2021, 13,070 invitation packs were sent to potentially eligible participants, of which 1297 responded. Of these, 454 (35.0%) participants were randomised: 240 to the intervention and 214 to usual care. The mean age of participants was 73.5 years (range 65–99); 60.6% were female, and participants had a median of three chronic conditions. Among the intervention group, the mean number of GYY sessions attended was 8.8 (SD 3.7, median 10, range 0–12). Two hundred and twenty-two (92.5%) participants attended at least one session, and fifty-three (22.1%) attended all twelve. Eighty per cent (n = 192) attended at least three of the first six sessions and at least three other sessions. One participant in the usual care group was invited to attend classes in error; they attended eight sessions, including five of the first six. The primary analysis included 422 participants with valid EQ-5D-5L data at baseline and at least one post-randomisation time point (intervention n = 227 of 240, 94.6%; usual care n = 195 of 214, 91.1%). There was no statistically or clinically significant difference in the EQ-5D-5L utility index score over 12 months: the predicted mean score for the intervention group was 0.729 (95% CI 0.712 to 0.747) and for usual care was 0.710 (95% CI 0.691 to 0.729), with an adjusted mean difference of 0.020 favouring intervention (95% CI −0.006 to 0.045; p = 0.14). The sensitivity analyses produced very similar results. The CACE analyses, which considered compliance as attending (1) ≥1 GYY session and (2) ≥6 including 3 of the first 6, produced slightly greater, but not clinically relevant, treatment estimates (0.025, 95% CI −0.002 to 0.052; p = 0.07; and 0.029, 95% CI −0.002 to 0.059; p = 0.06, respectively). There was no evidence of an interaction between trial arm and intended mode of delivery (interaction effect 0.007, 95% CI −0.042 to 0.057; p = 0.77). No statistically significant differences were observed in secondary outcomes, except in the T-score for the pain interference subscale of the PROMIS-29 at 3 months (−1.44, 95% CI −2.63 to −0.26; p = 0.02) and over the 12 months (−1.14, 95% CI −2.24 to −0.04; p = 0.04), and in the global (pain intensity) PROMIS-29 item at 12 months (−0.45, 95% CI −0.83 to −0.08; p = 0.02) and over the 12 months (−0.32, 95% CI −0.61 to −0.04; p = 0.03), favouring intervention. No serious, related adverse events were reported. Economic evaluation The base-case economic evaluation found that the intervention cost £80.85 more per participant (95% CI £76.73 to £84.97) than usual care, generated an additional 0.0178 QALYs per participant (95% CI 0.0175 to 0.0180) and had a 79% probability of being cost-effective at the willingness-to-pay threshold of £20,000 per QALY gained. Process evaluation Participants found both face-to-face and online courses acceptable. Participants were highly motivated; most viewed their health as good and engaged well with the GYY classes over time. The majority viewed GYY as a form of gentle exercise with mindful breathing. Some participants noted no or only a modest impact of GYY on their health or lifestyle, including physical, psychological and self-management benefits, while others described GYY as transformative, having substantial impacts and improvements on their physical health and emotional well-being. Conclusions The offer of a 12-week GYY programme was not associated with any statistically significant benefits in terms of HRQoL, mental health, loneliness or falls in older adults with multimorbidity. However, the intervention was safe, acceptable to most participants and highly valued by some. The economic evaluation suggests that the intervention could be cost-effective. Trial registration This trial is registered as ISRCTN13567538. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/94/36) and is published in full in Health Technology Assessment; Vol. 28, No. 53. See the NIHR Funding and Awards website for further award information.

Background The most common cause of preventable death after injury is haemorrhage. Resuscitative endovascular balloon occlusion of the aorta is intended to provide earlier, temporary haemorrhage control, to facilitate transfer to an operating theatre or interventional radiology suite for definitive haemostasis. Objective To compare standard care plus resuscitative endovascular balloon occlusion of the aorta versus standard care in patients with exsanguinating haemorrhage in the emergency department. Design Pragmatic, multicentre, Bayesian, group-sequential, registry-enabled, open-label, parallel-group randomised controlled trial to determine the clinical and cost-effectiveness of standard care plus resuscitative endovascular balloon occlusion of the aorta, compared to standard care alone. Setting United Kingdom Major Trauma Centres. Participants Trauma patients aged 16 years or older with confirmed or suspected life-threatening torso haemorrhage deemed amenable to adjunctive treatment with resuscitative endovascular balloon occlusion of the aorta. Interventions Participants were randomly assigned 1 : 1 to: standard care, as expected in a major trauma centre standard care plus resuscitative endovascular balloon occlusion of the aorta. Main outcome measures Primary: Mortality at 90 days. Secondary: Mortality at 6 months, while in hospital, and within 24, 6 and 3 hours; need for haemorrhage control procedures, time to commencement of haemorrhage procedure, complications, length of stay (hospital and intensive care unit-free days), blood product use. Health economic: Expected United Kingdom National Health Service perspective costs, life-years and quality-adjusted life-years, modelled over a lifetime horizon. Data sources Case report forms, Trauma Audit and Research Network registry, NHS Digital (Hospital Episode Statistics and Office of National Statistics data). Results Ninety patients were enrolled: 46 were randomised to standard care plus resuscitative endovascular balloon occlusion of the aorta and 44 to standard care. Mortality at 90 days was higher in the standard care plus resuscitative endovascular balloon occlusion of the aorta group (54%) compared to the standard care group (42%). The odds ratio was 1.58 (95% credible interval 0.72 to 3.52). The posterior probability of an odds ratio > 1 (indicating increased odds of death with resuscitative endovascular balloon occlusion of the aorta) was 86.9%. The overall effect did not change when an enthusiastic prior was used or when the estimate was adjusted for baseline characteristics. For the secondary outcomes (3, 6 and 24 hours mortality), the posterior probability that standard care plus resuscitative endovascular balloon occlusion of the aorta was harmful was higher than for the primary outcome. Additional analyses to account for intercurrent events did not change the direction of the estimate for mortality at any time point. Death due to haemorrhage was more common in the standard care plus resuscitative endovascular balloon occlusion of the aorta group than in the standard care group. There were no serious adverse device effects. Resuscitative endovascular balloon occlusion of the aorta is less costly (probability 99%), due to the competing mortality risk but also substantially less effective in terms of lifetime quality-adjusted life-years (probability 91%). Limitations The size of the study reflects the relative infrequency of exsanguinating traumatic haemorrhage in the United Kingdom. There were some baseline imbalances between groups, but adjusted analyses had little effect on the estimates. Conclusions This is the first randomised trial of the addition of resuscitative endovascular balloon occlusion of the aorta to standard care in the management of exsanguinating haemorrhage. All the analyses suggest that a strategy of standard care plus resuscitative endovascular balloon occlusion of the aorta is potentially harmful. Future work The role (if any) of resuscitative endovascular balloon occlusion of the aorta in the pre-hospital setting remains unclear. Further research to clarify its potential (or not) may be required. Trial registration This trial is registered as ISRCTN16184981. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 14/199/09) and is published in full in Health Technology Assessment; Vol. 28, No. 54. See the NIHR Funding and Awards website for further award information. Plain language summary Trauma (physical injury) is a major cause of death and disability. The most common cause of preventable death after injury is uncontrolled bleeding. Resuscitative endovascular balloon occlusion of the aorta is a technique whereby a small balloon is inflated in the aorta (main blood vessel) which aims to limit blood loss until an operation can be done to stop the bleeding. In this study, which is the first randomised trial in the world of this technique, we investigated whether adding resuscitative endovascular balloon occlusion of the aorta to the standard care received in a major trauma centre reduced the risk of death in trauma patients who had life-threatening uncontrolled bleeding. The study took place in 16 major trauma centres in the United Kingdom. Ninety adult trauma patients with confirmed or suspected uncontrolled bleeding took part and were randomly divided into two groups: (1) those who received standard care and (2) those who received standard care plus resuscitative endovascular balloon occlusion of the aorta. We followed participants for 6 months using routinely collected data from the National Health Service and from the Trauma Audit Research Network registry. We also contacted surviving patients at 6 months to ask about their quality of life. In the standard care group, 42% of participants died within 90 days of their injury compared to 54% of participants in the standard care plus resuscitative endovascular balloon occlusion of the aorta group. Risk of death was also higher in the standard care plus resuscitative endovascular balloon occlusion of the aorta group at all other time points (3, 6 and 24 hours, in hospital and at 6 months). Overall, the study showed that the use of resuscitative endovascular balloon occlusion of the aorta in hospital increased the risk of death. Scientific summary Background Trauma is a major cause of death and disability. Trauma (physical injury) disproportionately affects the young, killing those who might otherwise have lived long and productive lives. The most common cause of preventable death after injury is haemorrhage. The addition of resuscitative endovascular balloon occlusion of the aorta (REBOA) to current standard care is intended to provide earlier, temporary haemorrhage control, to facilitate transfer to an operating theatre or interventional radiology suite, for definitive haemostasis. Objectives The UK-REBOA trial was a pragmatic, multicentre, Bayesian, open-label, group-sequential, parallel-group randomised controlled trial comparing standard care plus REBOA versus standard care in patients with exsanguinating haemorrhage in the emergency room. The study included an elicitation exercise, an embedded mixed-methods process evaluation and a health economic evaluation. The primary clinical outcome was 90-day mortality (defined as death within 90 days of injury, before or after discharge from hospital). Secondary clinical outcomes included 3-, 6- and 24-hour mortality, in-hospital mortality, 6-month mortality, length of stay (in hospital and intensive care unit), 24-hour blood product use, need for haemorrhage control procedures (operation or angioembolisation), time to commencement of haemorrhage control procedure, complications/safety data and functional outcome [measured using the extended Glasgow Outcome Scale (GOS-E)] at discharge. Economic outcomes were 6-month (within trial) and lifetime (modelled) UK NHS perspective costs, life-years and quality-adjusted life-years (QALYs) [calculated using EuroQol Group’s 5-dimension health status 5-level questionnaire (EQ-5D-5L)], 6-month quality of life (measured using EQ-5D-5L). Methods Trauma patients were recruited in 16 UK major trauma centres. Trauma patients aged (or believed to be aged) 16 years or older, with confirmed or suspected life-threatening torso haemorrhage thought to be amenable to adjunctive treatment with REBOA were eligible. Women known (or thought to be) pregnant and those with injuries deemed unsurvivable were excluded. The trauma team leader assessed the patients for eligibility. Patients who were eligible for inclusion in the trial were incapacitated and unable to give consent at the time of eligibility assessment and randomisation. There was also not sufficient time to consult a surrogate decision-maker, or even an independent medical practitioner, for advice about including the patient. Enrolment therefore took place without prior consent following Research Ethics Committee approval for this approach. Consent for continuing participation (i.e. data collection) was sought by a member of the UK-REBOA trial team once patients were no longer in a critical condition (defined as being cared for in a ward area rather than an intensive care unit or high-dependency unit) or from a personal (or nominated professional) consultee. The trauma team leader enrolled the participant using a dedicated, secure website, available on a handheld device (smartphone, tablet) or desktop computer which is linked directly to the 24-hour randomisation system at the Centre for Healthcare Randomised Trials, based in the Health Services Research Unit, University of Aberdeen. Patients were randomised into one of the two intervention arms, in a 1 : 1 allocation ratio, in randomly generated blocks of two or four. Standard care: Patients allocated to the control group received ‘standard care’, as expected in a specialist major trauma centre. Such treatment typically included intubation, blood transfusion including blood products in a 1 : 1 : 1 ratio, interventions such as tourniquet application, and early operative or endovascular haemorrhage control. Treatment could also have included open aortic occlusion of the thoracic or abdominal aorta. Standard care plus REBOA: Patients allocated to this arm of the trial additionally received the technique of endovascular aortic occlusion, for the purpose of resuscitation, as part of an overall treatment strategy. The addition of REBOA to current standard care was intended to provide earlier, temporary haemorrhage control, to facilitate transfer to an operating theatre or interventional radiology suite for definitive haemostasis. The trial sought to evaluate the technique of REBOA rather than a specific brand of device, and therefore permitted the use of any licensed occlusion balloon, and did not prescribe or mandate a particular product. The trial had an integrated training programme to ensure familiarity with the REBOA procedure. In patients who had been randomised to the standard care plus REBOA arm of the trial, clinicians could decide not to insert the balloon occlusion device if: the patient’s haemodynamic status improved (as a result of other resuscitative measures), if they were deemed to no longer have life-threatening torso haemorrhage requiring adjunctive treatment with REBOA; they deteriorated (to the point of imminent death); or there was technical difficulty in obtaining arterial access, and it was felt that operative control of haemorrhage could be obtained more quickly. The data collection strategy for the UK-REBOA trial was designed to minimise the burden on participants and clinicians, and for the avoidance of duplication. The randomisation system collected balloon inflation/deflation times. The trial drew on routinely collected data, primarily from the Trauma Audit and Research Network (TARN) registry which includes demographic, injury, treatment and outcome data (including the GOS-E and EQ-5D-5L). Mortality and hospital resource use data were also sought from NHS Digital. The main analysis was based on the intention-to-treat principle. There were two planned interim analyses of survival and a final analysis of all outcomes after follow-up was complete. Baseline and follow-up data were summarised using descriptive statistics and graphical summaries. Treatment effects are presented with 95% credible intervals for the primary and secondary outcomes. Elicitation exercise An elicitation exercise involving 20 subject matter experts (12 emergency medicine physicians, 3 pre-hospital care doctors, 4 surgeons and 1 intensivist) was undertaken to derive prior probability distributions to help contextualise the interpretation for the primary and secondary outcomes of the trial. Subject matter experts, on average, estimated in-hospital and 90-day mortality in this patient group, without the use of REBOA, to be in excess of 50%. Mortality at earlier time points (6 and 24 hours) was estimated to be closer to 25%. The elicited data, and the resulting prior probability distributions, indicate that the experts, on average, had a favourable opinion of REBOA, that is they expect the addition of REBOA to standard care to improve mortality at all time points. Clinical results Sixteen recruitment sites were opened in a staggered manner. Recruitment commenced in October 2017, was halted in March 2020 due to COVID-19, and restarted in July 2020. The second interim analysis (including 80 participants) triggered one of the pre-specified stopping rules, and recruitment closed in March 2022, by which time 90 participants had been recruited. Forty-four participants were randomised to standard care (2 of whom did receive REBOA) and 46 were randomised to standard care plus REBOA (19 of whom had the catheter inserted and balloon inserted and the remaining 27 progressed to different time points along this pathway). The groups were well-matched in terms of age, gender, comorbidities, mechanism of injury and injury severity. In the standard care arm, the median age was 39 years (interquartile range 30–56 years) and 77% were male. In the standard care plus REBOA arm, the median age was 46 years (interquartile range 33–62) and 61% were male. The median Injury Severity Score in both arms was 41 (interquartile range 29–50), with the majority classed as having very severe injury. Of the 46 patients allocated to standard care plus REBOA treatment, 25 (54%) died within 90 days. Of the 43 standard care patients for whom primary outcome data are available, 18 (42%) died within 90 days. Using the minimally informative prior, the odds of 90-day mortality were 1.58 for patients allocated to the standard care plus REBOA arm (95% credible interval 0.72 to 3.52). The posterior probability of an odds ratio > 1 (i.e. that REBOA was harmful) was 86.9%. The direction of the estimate did not change when an enthusiastic (the elicited) prior was used or when the estimate was adjusted for baseline characteristics. For the secondary outcomes (3-, 6- and 24-hour mortality), the posterior probability that REBOA was harmful was higher than for the primary outcome. Additional analyses to account for intercurrent events did not change the direction of the estimate for mortality at 3, 6 and 24 hours, at 90 days or 6 months, or in-hospital mortality. Death due to haemorrhage was more common in the standard care plus REBOA group than in the standard care group. The mean number of days spent in intensive care and in hospital were lower in the standard care plus REBOA group compared to the standard care group, partly because of the higher number of earlier deaths in the standard care plus REBOA arm. There were no serious adverse device effects. Health economics We costed individual components of resources and summed these to generate a total cost for the whole initial hospitalisation admission period. Total NHS resource use for the index hospitalisation was obtained from patient-level data in TARN and the key resource use variables for costing included time of arrival, time of emergency department departure, time of first operation, time of death/discharge, number and type of operative procedures and volume of blood transfusions that were required. Secondary care contacts and episodes of care that were commenced between the date of discharge from the index hospitalisation through 6 months post randomisation were sourced, where available, through linkage of patient records to the Hospital Episode Statistics database. All costs are reported from a UK NHS perspective in Great British pounds (GBP) (year 2020-1). Quality of life was measured using the EQ-5D-5L prior to patient’s discharge from their index hospitalisation and at 6 months post admission. EQ-5D-5L asks respondents to report any problems on a given day across five dimensions of mobility, self-care, usual activities, anxiety/depression and pain. The data were available from TARN and supplemented with data collected by the local trial teams. EQ-5D-5L data were cross-walked to the 3L version and valued using UK general population preference tariffs. Baseline utility was set equal to the unconscious state (−0.402) and utility following death was set to 0. QALYs were calculated using an area under the curve approach assuming linear extrapolation between time points. From the within-trial health economic analysis, participants in the standard care plus REBOA arm of the study incurred lower costs {index hospital admission: mean cost £57,384 [standard deviation (SD) £62,863]} compared to those in standard care [mean cost £116,064 (SD £128,957)]. Lower costs in the standard care plus REBOA arm of the study were mainly due to lower use of hospital resources (length of stay, etc.) due to the competing risk of death (i.e. a higher number of deaths in the REBOA plus standard care group). Similarly, life-years accrued and QALYs over 6 months post randomisation were also lower in standard care plus REBOA compared to standard care due to a greater proportion of trial participants dying, with mortality also occurring earlier in the follow-up period for the REBOA arm. The mean life-years gained in the standard care plus REBOA arm was 0.232 (SD 0.247) compared to 0.305 (SD 0.236) in the standard care arm. When modelled over a full lifetime horizon, standard care plus REBOA is less costly (probability 99%), due to the competing risk of mortality but is also substantially less effective in terms of QALYs accrued over a lifetime horizon (probability 91%). The findings are robust to a range of scenario analyses undertaken, with the probability of standard care being the optimal treatment strategy ranging from 66% to 81% at a threshold value of a QALY = £50,000. Process evaluation The process evaluation was conducted in two phases; both phases involved interviews with clinical and research staff based at recruitment sites. Phase 1 was designed to identify barriers during trial initial and set-up; Phase 2 focused on exploring barriers and facilitators of recruitment into the trial and intervention delivery. A behavioural framework was used in Phase 2 to direct analysis and generate solutions designed to enhance trial practices, which included regular online meetings between the principal investigators from each site, updates to training materials and delivery, and e-mail/Twitter feedback on recruitment activity. Conclusions This is the first randomised trial ever to be conducted examining the potential clinical effectiveness of REBOA for the management of exsanguinating haemorrhage. All the analyses conducted suggest with high probability that a strategy of standard care plus REBOA is harmful. Implications for health care: The continuing use of REBOA, at least in the UK in-hospital setting, should be re-evaluated. Implications for research: The role (if any) of REBOA in the pre-hospital setting remains unclear. Further research to clarify its potential (or not) may be required. Trial registration This trial is registered as ISRCTN16184981. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 14/199/09) and will be published in full in Health Technology Assessment; Vol. 28, No. 54. See the NIHR Funding and Awards website for further award information.

Abstract Background Endometriosis affects 1 in 10 women, many of whom have surgery for persistent pain. Recurrence of symptoms following an operation is common. Although hormonal treatment can reduce this risk, there is uncertainty about the best option. Objectives To evaluate the clinical and cost-effectiveness of long-acting progestogen therapy compared with the combined oral contraceptive pill in preventing recurrence of endometriosis-related pain and quality of life. Design A multicentre, open, randomised trial with parallel economic evaluation. The final design was informed by a pilot study, qualitative exploration of women’s lived experience of endometriosis and a pretrial economic model. Setting Thirty-four United Kingdom hospitals. Participants Women of reproductive age undergoing conservative surgery for endometriosis. Interventions Long-acting progestogen reversible contraceptive (either 150 mg depot medroxyprogesterone acetate or 52 mg levonorgestrel-releasing intrauterine system) or combined oral contraceptive pill (30 µg ethinylestradiol, 150 µg levonorgestrel). Main outcome measures The primary outcome was the pain domain of the Endometriosis Health Profile-30 questionnaire at 36 months post randomisation. The economic evaluation estimated the cost per quality-adjusted life-years gained. Results Four hundred and five women were randomised to receive either long-acting reversible contraceptive (N = 205) or combined oral contraceptive pill (N = 200). Pain scores improved in both groups (24 and 23 points on average) compared with preoperative values but there was no difference between the two (adjusted mean difference: −0.8, 95% confidence interval −5.7 to 4.2; p = 0.76). The long-acting reversible contraceptive group underwent fewer surgical procedures or second-line treatments compared with the combined oral contraceptive group (73 vs. 97; hazard ratio 0.67, 95% confidence interval 0.44 to 1.00). The mean adjusted quality-adjusted life-year difference between two arms was 0.043 (95% confidence interval −0.069 to 0.152) in favour of the combined oral contraceptive pill, although this cost an additional £533 (95% confidence interval 52 to 983) per woman. Limitations Limitations include the absence of a no-treatment group and the fact that many women changed treatments over the 3 years of follow-up. Use of telephone follow-up to collect primary outcome data in those who failed to return questionnaires resulted in missing data for secondary outcomes. The COVID pandemic may have affected rates of further surgical treatment. Conclusions At 36 months, women allocated to either intervention had comparable levels of pain, with both groups showing around a 40% improvement from presurgical levels. Although the combined oral contraceptive was cost-effective at a threshold of £20,000 per quality-adjusted life-year, the difference between the two was marginal and lower rates of repeat surgery might make long-acting reversible contraceptives preferable to some women. Future work Future research needs to focus on evaluating newer hormonal preparations, a more holistic approach to symptom suppression and identification of biomarkers to diagnose endometriosis and its recurrence. Trial registration This trial is registered as ISRCTN97865475. https://doi.org/10.1186/ISRCTN97865475. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 11/114/01) and is published in full in Health Technology Assessment; Vol. 28, No. 55. See the NIHR Funding and Awards website for further award information. The NIHR recognises that people have diverse gender identities, and in this report, the word ‘woman’ is used to describe patients or individuals whose sex assigned at birth was female, whether they identify as female, male or non-binary. Plain language summary What was the question? Endometriosis is a condition where cells similar to ones that line the womb are found elsewhere in the body. Endometriosis affects 1 in 10 women, many of whom have surgery for persistent pain. Unfortunately, symptoms often return and some women will need repeat operations. Hormonal contraceptives can prevent the return of endometriosis-related pain: either long-acting reversible contraceptives (injections or a coil, fitted inside the womb) or the combined oral contraceptive pill (often called ‘the pill’). We do not know which is the best option. The aim of this trial was to find out which of these two hormone treatments was more effective in terms of symptom relief, avoidance of further surgery and costs. What did we do? Four hundred and five women with endometriosis, who were not intending to get pregnant, participated in a clinical trial. Half of the participants took long-acting reversible contraceptives, and the other half took the pill for 3 years following endometriosis surgery. The choice of treatment was made at random by a computer to ensure a fair comparison, although those allocated to the long-acting contraceptive could choose between injections or the coil. Participants completed questionnaires about their symptoms and life quality at intervals up to 3 years. What did we find? Both treatments were equally good at reducing pain but more women using the pill had repeat operations. The pill was a little more costly overall but associated with a slightly higher quality of life. What does this mean? Both treatments are equally effective in reducing pain up to 3 years after surgery for endometriosis. The differences in costs are small and the choice of treatment should be based on personal preference. Scientific summary Background Endometriosis, which affects up to 1 in 10 women, is characterised by the proliferation of endometrial cells outside the uterus, usually within the pelvis. These endometriotic deposits undergo cyclical proliferation in response to ovarian oestrogen, resulting in internal bleeding, scarring and adhesion formation, which causes pain and has a serious impact on quality of life in affected women. Surgical removal or destruction of endometriotic tissue is currently the preferred treatment but the risk of recurrence is high. Recurrence can be controlled by post-surgical hormonal treatment to reduce circulating levels of oestrogen but there is uncertainty as to the clinical and cost-effectiveness of two commonly used modalities: long-acting reversible contraceptive (LARC) and the combined oral contraceptive pill (COCP). Progestogen-based LARCs used in the trial were the levonorgestrel-releasing intrauterine system (LNG-IUS) or depot medroxyprogesterone acetate injection (DMPA). Objectives To evaluate the clinical and cost-effectiveness of LARCs compared with COCP in preventing recurrence of endometriosis-related pain and quality of life. Design A multicentre, open, randomised trial with parallel economic evaluation. The final design was informed by a pilot study, qualitative exploration of women’s lived experience of endometriosis and a pretrial economic model. Setting Thirty-four NHS hospitals within the UK. Participants Women of reproductive age undergoing laparoscopic surgery for pelvic pain due to endometriosis were eligible if they fulfilled the following criteria: Inclusion criteria: Aged 16–45 years. No immediate plans to conceive. Scheduled for laparoscopic conservative surgery, or diagnostic laparoscopy with concurrent surgery if endometriosis is found, for pelvic pain associated with endometriosis. Willing to be randomised to one long-acting progestogen (LNG-IUS or DMPA) and COCP. The following women were also eligible if they had recurrent pain and were to have conservative surgery for endometriosis: Had one or more previous diagnostic laparoscopies. Had previous laparoscopic conservative surgery for endometriosis, provided that this did not involve rectovaginal dissection or bowel resection. Used postoperative medical treatment, including the treatment options included in the trial. Previous use of treatment options included in the trial as contraceptives. Use of preoperative gonadotropin-releasing hormone analogues (GnRHa), provided that this was stopped at least 4 weeks prior to laparoscopy. Exclusion criteria: No endometriosis identified at diagnostic laparoscopy. Infertility. Any plans for further elective endometriosis surgery (for deep disease or endometrioma). Contraindications to the use of hormonal treatment with oestrogen or progestogens. Suspicion of malignancy. Interventions Four hundred and five women were randomised in a one-to-one ratio via secure internet facility to either LARCs or COCP. The LARC was either 150 mg DMPA or 52 mg LNG-IUS. The COCP formulation contained 30 µg ethinylestradiol and 150 µg levonorgestrel. The LARC (LNG-IUS or DMPA) was selected before randomisation by the patient if a preference was apparent (or alternatively allocated randomly if there was no opinion). Outcome measures Primary outcome Pain as evaluated by the pain domain of the Endometriosis Health Profile – 30-item (EHP-30) questionnaire at 36 months post randomisation. Secondary outcomes (evaluated at 6 months, 1, 2 and 3 years): The four core domains of the EHP-30 questionnaire (control and powerlessness; emotional well-being; social support; self-image). The six modular domains of the EHP-30 (work; relationships with family; sexual relationships; feelings about medical profession; feelings about treatment; feelings about infertility). Pelvic pain measured by visual analogue scale (VAS) during periods; during intercourse; at any time. Responses to the question ‘compared to 1 month ago, would you say your pelvic pain has “Got much better”, “Got a little better”, “Not changed much”, “Got worse”’. Fatigue, as measured by Fatigue Severity Scale (FSS) score. Menstrual regularity. Generic quality of life questionnaire [EuroQol-5 Dimensions, five-level version (EQ-5D-5L)], including index and thermometer scores. Capabilities, as a measure of well-being (ICEpop CAPability, ICECAP measure). Further therapeutic surgery or second-line treatment for endometriosis as a proxy for recurrence or ‘treatment failure’, defined as having undergone hysterectomy, surgery for endometriosis, laparoscopy or taking GnRHa treatment. Discontinuation rates of randomised treatment (time to first treatment change), with reasons for change. Serious adverse events. The economic evaluation estimated the cost per quality-adjusted life-years (QALYs) gained. Sample size To detect an 8-point difference on the EHP-30 pain domain with 90% power (p = 0.05) and assuming the standard deviation (SD) to be 22 points required 160 participants per group, 320 in total. To account for any loss to follow-up (estimated to be 20%), this target was inflated to 400. Eight points is equivalent to 0.36 SD, which can be considered halfway between a small (0.2 SD) to moderate (0.5 SD) effect size. This size of sample would also give us good power (80%) to detect a 10-point difference in the two stratified analyses of LNG-IUS versus COCP and DMPA versus COCP provided that these subgroups had a roughly even split. Results A total of 405 women were allocated to receive either LARC (N = 205) or COCP (N = 200) following laparoscopic surgery for endometriosis. The two randomised groups were comparable in terms of age [29.6 years (6.7 years) vs. 29.3 years (6.6 years)]; body mass index [27.0 kg/m2 (10.6 kg/m2) vs. 26.3 kg/m2 (5.5 kg/m2)]; early-stage endometriosis: stages I and II (79% vs. 79%); complete surgical excision (92% vs. 90%); white ethnicity (91% vs. 92%) and previous hormonal treatment (27% vs. 23%). Pain scores improved in both groups (24 and 23 points on average) compared with preoperative values but there was no statistically significant difference between LARC and COCP at 3 years [adjusted mean difference: −0.8; 95% confidence interval (CI) −5.7 to 4.2; p = 0.76]. The choice of LARC (LNG-IUS or DMPA) did not alter these findings. Most of the other domains of the EHP-30 were improved in both groups at all time points compared with preoperative scores, with no consistent evidence of any difference between groups when estimates of uncertainty were considered. Women in the LARC group had fewer surgical procedures or second-line treatments compared with those taking COCP (73 vs. 97 events, occurring in 50 vs. 61 women due to repeat interventions), translating to a 33% reduction in time to treatment failure [hazard ratio (HR) 0.67, 95% CI 0.44 to 1.00]. Participants in the LARC group had a slightly higher mean EQ-5D-5L score at 36 months compared with those in the COCP arm (0.693 and 0.686, respectively). The mean adjusted imputed QALY difference between the two arms was 0.043 (95% CI −0.069 to 0.152) in favour of COCP, where participants in LARC group had a lower QALY value than those randomised to COCP (1.937 and 1.976, respectively). Despite this, the COCP group was estimated to be more expensive than the LARC group by £533 (95% CI 52 to 983) per woman over 36 months of follow-up. Limitations In the absence of a no-treatment arm, we were unable to demonstrate the effect of surgery alone on preventing recurrence of pain symptoms. While we are able to comment on the effectiveness of a strategy of postoperative prescription of LARC versus COCP, the true impact of these interventions is difficult to gauge as the prolonged duration of follow-up meant that many women had discontinued their allocated treatments. The predominance of white women in the recruited sample limits our ability to be confident about how our results might apply to women from other ethnic backgrounds. Use of telephone follow-up to collect primary outcome data in those who failed to return full questionnaires resulted in missing data for some of the secondary outcomes. While all patients were recruited prior to the COVID-19 pandemic, the number of women who required further surgery may be underestimated, given the negative impact of COVID-19 on waiting lists for elective surgery throughout the UK. It is possible that this may have led to an increase in the use of GnRHa treatment by women who were unable to access surgery for their symptoms. Conclusions At 36 months, women allocated to LARCs or COCP had comparable levels of pain, with both groups showing around 40% improvement from presurgical levels. Although COCP is likely to be considered more cost-effective at a threshold of £20,000 per QALY, the difference between the two is marginal. LARCs may be preferred by some women as they are associated with lower rates of surgery, particular hysterectomy and operations for recurrence of endometriosis. Future work Future research needs to focus on evaluating newer hormonal preparations, a more holistic approach to symptom suppression and identification of biomarkers to diagnose endometriosis and its recurrence. Active measures need to be adopted to improve the inclusivity of clinical research and ensure that the ethnic mix within participants mirror that of the general population. Trial registration This trial is registered as ISRCTN97865475. https://doi.org/10.1186/ISRCTN97865475. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 11/114/01) and is published in full in Health Technology Assessment; Vol. 28, No. 55. See the NIHR Funding and Awards website for further award information.

Background We compared the relative benefits, harms and cost-effectiveness of hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery ± systemic chemotherapy versus cytoreductive surgery ± systemic chemotherapy or systemic chemotherapy alone in people with peritoneal metastases from colorectal, gastric or ovarian cancers by a systematic review, meta-analysis and model-based cost–utility analysis. Methods We searched MEDLINE, EMBASE, Cochrane Library and the Science Citation Index, ClinicalTrials.gov and WHO ICTRP trial registers until 14 April 2022. We included only randomised controlled trials addressing the research objectives. We used the Cochrane risk of bias tool version 2 to assess the risk of bias in randomised controlled trials. We used the random-effects model for data synthesis when applicable. For the cost-effectiveness analysis, we performed a model-based cost–utility analysis using methods recommended by The National Institute for Health and Care Excellence. Results The systematic review included a total of eight randomised controlled trials (seven randomised controlled trials, 955 participants included in the quantitative analysis). All comparisons other than those for stage III or greater epithelial ovarian cancer contained only one trial, indicating the paucity of randomised controlled trials that provided data. For colorectal cancer, hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy probably results in little to no difference in all-cause mortality (60.6% vs. 60.6%; hazard ratio 1.00, 95% confidence interval 0.63 to 1.58) and may increase the serious adverse event proportions compared to cytoreductive surgery ± systemic chemotherapy (25.6% vs. 15.2%; risk ratio 1.69, 95% confidence interval 1.03 to 2.77). Hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy probably decreases all-cause mortality compared to fluorouracil-based systemic chemotherapy alone (40.8% vs. 60.8%; hazard ratio 0.55, 95% confidence interval 0.32 to 0.95). For gastric cancer, there is high uncertainty about the effects of hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy versus cytoreductive surgery + systemic chemotherapy or systemic chemotherapy alone on all-cause mortality. For stage III or greater epithelial ovarian cancer undergoing interval cytoreductive surgery, hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy probably decreases all-cause mortality compared to cytoreductive surgery + systemic chemotherapy (46.3% vs. 57.4%; hazard ratio 0.73, 95% confidence interval 0.57 to 0.93). Hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy may not be cost-effective versus cytoreductive surgery + systemic chemotherapy for colorectal cancer but may be cost-effective for the remaining comparisons. Limitations We were unable to obtain individual participant data as planned. The limited number of randomised controlled trials for each comparison and the paucity of data on health-related quality of life mean that the recommendations may change as new evidence (from trials with a low risk of bias) emerges. Conclusions In people with peritoneal metastases from colorectal cancer with limited peritoneal metastases and who are likely to withstand major surgery, hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy should not be used in routine clinical practice (strong recommendation). There is considerable uncertainty as to whether hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy or cytoreductive surgery + systemic chemotherapy should be offered to patients with gastric cancer and peritoneal metastases (no recommendation). Hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy should be offered routinely to women with stage III or greater epithelial ovarian cancer and metastases confined to the abdomen requiring and likely to withstand interval cytoreductive surgery after chemotherapy (strong recommendation). Future work More randomised controlled trials are necessary. Study registration This study is registered as PROSPERO CRD42019130504. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/135/02) and is published in full in Health Technology Assessment; Vol. 28, No. 51. See the NIHR Funding and Awards website for further award information. Plain language summary What was the question? Cancers of the bowel, ovary or stomach can spread to the lining of the abdomen (‘peritoneal metastases’). Chemotherapy (the use of drugs that aim to kill cancer cells) given by injection or tablets (‘systemic chemotherapy’) is one of the main treatment options. There is uncertainty about whether adding cytoreductive surgery (cytoreductive surgery; an operation to remove the cancer) and ‘hyperthermic intraoperative peritoneal chemotherapy’ (warm chemotherapy delivered into the lining of the abdomen during cytoreductive surgery) are beneficial. What did we do? We reviewed all the information from medical literature published until 14 April 2022, to answer the above uncertainty. What did we find? We found the following from eight trials, including about 1000 participants. In people with peritoneal metastases from bowel cancer, hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy probably does not provide any benefits and increases harm compared to cytoreductive surgery + systemic chemotherapy, while cytoreductive surgery + systemic chemotherapy appears to increase survival compared to systemic chemotherapy alone. There is uncertainty about the best treatment for people with peritoneal metastases from stomach cancer. In women with peritoneal metastases from ovarian cancer who require systemic chemotherapy before cytoreductive surgery to shrink the cancer to allow surgery (‘advanced ovarian cancer’), hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy probably increases survival compared to cytoreductive surgery + systemic chemotherapy. What does this mean? In people who can withstand a major operation and in whom cancer can be removed, cytoreductive surgery + systemic chemotherapy should be offered to people with peritoneal metastases from bowel cancer, while hyperthermic intraoperative peritoneal chemotherapy + cytoreductive surgery + systemic chemotherapy should be offered to women with peritoneal metastases from ‘advanced ovarian cancer’. Uncertainty in treatment continues for gastric cancer. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/135/02) and is published in full in Health Technology Assessment; Vol. 28, No. 51. See the NIHR Funding and Awards website for further award information. Scientific summary Background There is uncertainty about whether hyperthermic intraoperative peritoneal chemotherapy (HIPEC) with cytoreductive surgery (CRS) improves survival and/or quality of life (QoL) compared to CRS or no treatment in addition to systemic chemotherapy in people with peritoneal metastases who can withstand major surgery. Objectives Primary objectives To compare the relative benefits and harms of HIPEC + CRS ± systemic chemotherapy versus CRS ± systemic chemotherapy or systemic chemotherapy alone in people with peritoneal metastases from colorectal, gastric or stage III or greater epithelial ovarian cancers eligible to undergo HIPEC + CRS by a systematic review and meta-analysis. Secondary objectives To compare the cost-effectiveness of HIPEC + CRS ± systemic chemotherapy versus CRS ± systemic chemotherapy or systemic chemotherapy alone from the NHS and personal social services (PSS) perspective using a model-based cost–utility analysis. Methods We performed a systematic review of literature by searching MEDLINE, EMBASE, Cochrane Library, Science Citation Index, Conference Proceedings Citation Index, as well as trial registers until 14 April 2022. We followed the standard guidance for performing a high-quality systematic review and meta-analysis. We included only randomised controlled trials (RCTs) and assessed the risk of bias using Risk of Bias version 2.0 (ROB 2.0). We were unable to perform an individual participant data (IPD) meta-analysis as planned because of unforeseen circumstances related to coronavirus disease 2019 (COVID-19). This led to trialists who were also clinical researchers being unable to engage for transfer of IPD. We did not foresee that study authors (surgeons) would be sufficiently engaged with providing IPD soon because of the backlog with surgeries and the fatigue induced by COVID-19. Therefore, we performed a meta-analysis based on aggregate data. We calculated the hazard ratio (HR), risk ratio (RR), rate ratio or mean difference (MD) with 95% confidence intervals (CIs) as appropriate. When applicable, we performed meta-analysis using the random-effects model using Review Manager 5.4. We used GRADE guidance to assess the certainty of evidence and determine the strength of recommendations. For the cost-effectiveness analysis, we performed a model-based cost–utility analysis using methods recommended by The National Institute for Health and Care Excellence (NICE). We estimated the costs and quality-adjusted life-years (QALYs) per patient using lifetime horizon. We calculated the incremental net monetary benefit (incremental NMB) for each comparison based on deterministic analysis and probabilistic sensitivity analysis (PSA). We also performed univariate sensitivity analysis and value of information analysis. Results The systematic review included a total of eight RCTs. A total of 955 participants in seven RCTs were included in the quantitative analysis. All comparisons other than those of ovarian cancer contained only one trial. For colorectal cancer, HIPEC + CRS + systemic chemotherapy probably results in little to no difference in all-cause mortality (60.6% in HIPEC + CRS + systemic chemotherapy vs. 60.6% in CRS + systemic chemotherapy; median follow-up 64 months; HR 1.00, 95% CI 0.63 to 1.58; one trial; 265 participants; moderate-certainty evidence) and may increase the number of people who developed serious adverse events compared to CRS +/– systemic chemotherapy (25.6% in HIPEC + CRS + systemic chemotherapy vs. 15.2% in CRS + systemic chemotherapy; RR 1.69, 95% CI 1.03 to 2.77; one trial; 265 participants; low-certainty evidence). HIPEC + CRS + systemic chemotherapy probably decreases all-cause mortality compared to fluorouracil-based systemic chemotherapy alone (40.8% in HIPEC + CRS + systemic chemotherapy vs. 60.8% in systemic chemotherapy alone; median follow-up 22 months; HR 0.55, 95% CI 0.32 to 0.95; one trial; 105 participants; moderate-certainty evidence). For gastric cancer, there is high uncertainty about the effect of HIPEC + CRS + systemic chemotherapy versus CRS + systemic chemotherapy on all-cause mortality (effect estimates not presented because of very low-certainty evidence). HIPEC + CRS + systemic chemotherapy probably decreases all-cause mortality compared to systemic chemotherapy (40.8% in HIPEC + CRS + systemic chemotherapy vs. 100% in systemic chemotherapy alone; minimum follow-up 24 months; HR 0.40, 95% CI 0.30 to 0.52; one trial; 17 participants; moderate-certainty evidence). For stage III or greater epithelial ovarian cancer requiring interval CRS, HIPEC + CRS + systemic chemotherapy probably decreases all-cause mortality compared to CRS + systemic chemotherapy (46.3% in HIPEC + CRS + systemic chemotherapy vs. 57.4% in CRS + systemic chemotherapy; median follow-up 32–70 months; HR 0.73, 95% CI 0.57 to 0.93; three trials; 500 participants; moderate-certainty evidence). It may result in little to no difference in health-related quality of life (HRQoL) (MD 4.85, 95% CI −7.74 to 17.44; one trial; 71 participants; moderate-certainty evidence) or number of people who developed serious adverse events compared to CRS + systemic chemotherapy (26.7% in HIPEC + CRS + systemic chemotherapy vs. 25.2% in CRS + systemic chemotherapy; RR 1.06, 95% CI 0.73 to 1.54; two trials; 316 participants; moderate-certainty evidence), although it probably increases the number of serious adverse events per participant compared to CRS + systemic chemotherapy (41.4 events per 100 participants in HIPEC + CRS + systemic chemotherapy vs. 32.6 events per 100 participants in CRS + systemic chemotherapy; rate ratio 1.27, 95% CI 1.09 to 1.49; one trial; 184 participants; moderate-certainty evidence). The cost-effectiveness analysis included the five comparisons described above: HIPEC + CRS + systemic chemotherapy versus CRS + systemic chemotherapy for each of the colorectal, gastric and ovarian cancers and HIPEC + CRS + systemic chemotherapy versus systemic chemotherapy alone for each of the colorectal and gastric cancers. In people with colorectal peritoneal metastases, the incremental NMBs at willingness to pay (WTP) of £20,000 and £30,000 were −£6162.83 and −£6164.19, respectively, indicating that HIPEC + CRS + systemic chemotherapy was not cost-effective compared to CRS + systemic chemotherapy in NHS. The likelihood of HIPEC + CRS + systemic chemotherapy being cost-effective compared to CRS + systemic chemotherapy was 46.5% and 47.6% at WTP of £20,000 and £30,000, respectively. In the same group of people, the incremental NMBs at WTP of £20,000 and £30,000 were £107,909.46 and £167,621.58, respectively, indicating that HIPEC + CRS + systemic chemotherapy may be cost-effective compared to systemic chemotherapy alone in NHS. The likelihood of HIPEC + CRS + systemic chemotherapy being cost-effective compared to systemic chemotherapy alone was 89.3% and 90.3% at WTP of £20,000 and £30,000, respectively. In people with gastric peritoneal metastases, the incremental NMBs at WTP of £20,000 and £30,000 were £14,174.73 and £22,955.89, respectively, for HIPEC + CRS + systemic chemotherapy versus CRS + systemic chemotherapy and £81,796.38 and £127,768.23, respectively, for HIPEC + CRS + systemic chemotherapy versus systemic chemotherapy, indicating that HIPEC + CRS + systemic chemotherapy may be cost-effective compared to CRS + systemic chemotherapy or systemic chemotherapy alone in NHS. The likelihood of HIPEC + CRS + systemic chemotherapy being cost-effective ranged between 60% and 70% for the different comparisons and different thresholds. In women with grade III or greater epithelial ovarian cancer requiring interval CRS, the incremental NMBs at WTP of £20,000 and £30,000 were £46,761.81 and £71,938.23, respectively, indicating that HIPEC + CRS + systemic chemotherapy was cost-effective compared to CRS +systemic chemotherapy in NHS. The likelihood of HIPEC + CRS + systemic chemotherapy being cost-effective compared to CRS + systemic chemotherapy was 71.9% and 72.4% at WTP of £20,000 and £30,000, respectively. The value of information analysis indicated that the expected value of perfect information (EVPI) ranged between £3 and £53 million for the different cancer types (when estimation was possible) for WTP of £20,000 and £30,000. Discussion and conclusions Limitations of the review We were unable to obtain IPD as planned. IPD would have allowed us to refine our effect estimates for subgroups of people with peritoneal metastases from colorectal, gastric or stage III or greater epithelial ovarian cancer. It is difficult to estimate whether our conclusions would have changed if we had IPD; however, our systematic review and meta-analysis support similar conclusions as the trial authors, suggesting that the impact of IPD may not be major enough to warrant an IPD once the health services have recovered from the impact of COVID-19. We estimated the HR for survival for gastric cancer trials from Kaplan–Meier curves. This might have introduced bias. However, because of the small number of participants and the estimations that we have performed to calculate the effect estimates, we have concluded that there is uncertainty in the benefit of HIPEC + CRS + systemic chemotherapy in gastric cancers. Because of the paucity of trials under each comparison, evidence from new RCTs of low risk of bias may change our recommendations. There are concerns regarding the clinical recommendations for people with colorectal peritoneal metastases based on the PRODIGE-7 trial. We have discussed in detail the different concerns raised and why these concerns should not be used as a justification for not basing clinical practice on PRODIGE-7 trial (in the full article). In summary, we based our clinical practice recommendations for colorectal peritoneal metastases on PRODIGE-7 trial because the trial was a low risk of bias trial for the comparison of HIPEC + CRS + systemic chemotherapy versus CRS + systemic chemotherapy. An appropriate analysis was used to analyse trial data, and there was no other trial of low of bias comparing HIPEC + CRS + systemic chemotherapy versus CRS + systemic chemotherapy. While the CRS + systemic chemotherapy was not directly compared with systemic chemotherapy alone, we recommended CRS + systemic chemotherapy in people with colorectal peritoneal metastases because of the lack of any ‘systemic chemotherapy alone’ treatments that provide equivalent median survival as that observed in the control arm (CRS + systemic chemotherapy) in the PRODIGE-7 trial. Because of the difficulties in estimating the Peritoneal Cancer Index (PCI) during surgery, we have not recommended HIPEC + CRS + systemic chemotherapy even for the subset of patients with PCI 11–15, but this exploratory subgroup analysis can guide future research. We have not based our recommendations on non-randomised studies, as we did not find any non-randomised study in which similar participants with colorectal peritoneal metastases underwent HIPEC + CRS + systemic chemotherapy versus CRS + systemic chemotherapy or systemic chemotherapy alone. CRS + systemic chemotherapy provides equivalent median survival of 41 months as HIPEC + CRS + systemic chemotherapy. When there is an existing, less invasive treatment that provides equivalent survival, it can hardly be considered life-threatening to warrant recommendations based on low-or very low-certainty evidence. Recommendations for clinical practice In people with peritoneal metastases from colorectal cancer, based on the results of PRODIGE-7 trial, HIPEC + CRS + systemic chemotherapy probably results in little to no difference in all-cause mortality or progression-free survival and results in increased complications compared to CRS + systemic chemotherapy. Therefore, HIPEC based on oxaliplatin regimen used in PRODIGE-7 trial + CRS + systemic chemotherapy should not be used in routine clinical practice (strong recommendation). Because of the lack of reliability of preoperative or perioperative PCI, the lack of pre-PRODIGE-7 trial standard classification of PCI into PCI 15 and pre-defined subgroup analysis based on the PCI classification, HIPEC based on oxaliplatin regimen used in PRODIGE-7 trial + CRS + systemic chemotherapy cannot be recommended for any subgroups. Because of the median survival observed in the CRS + systemic chemotherapy arm of PRODIGE-7 trial (41 months) and the poor survival observed in people with disseminated colorectal peritoneal metastases (

Background Strain and shear wave elastography which is commonly used with concurrent real-time imaging known as real-time ultrasound shear/strain wave elastography is a new diagnostic technique that has been reported to be useful in the diagnosis of nodules in several organs. There is conflicting evidence regarding its benefit over ultrasound-guided fine-needle aspiration cytology alone in thyroid nodules. Objectives To determine if ultrasound strain and shear wave elastography in conjunction with fine-needle aspiration cytology will reduce the number of patients who have a non-diagnostic first fine-needle aspiration cytology results as compared to conventional ultrasound-only guided fine-needle aspiration cytology. Design A pragmatic, unblinded, multicentre randomised controlled trial. Setting Eighteen centres with a radiology department across England. Participants Adults who had not undergone previous fine-needle aspiration cytology with single or multiple nodules undergoing investigation. Interventions Ultrasound shear/strain wave elastography-ultrasound guided fine-needle aspiration cytology (intervention arm) – strain or shear wave elastography-guided fine-needle aspiration cytology. Ultrasound-only guided fine-needle aspiration cytology (control arm) – routine ultrasound-only guided fine-needle aspiration cytology (the current standard recommended by the British Thyroid Association guidelines). Main outcome measure The proportion of patients who have a non-diagnostic cytology (Thy 1) result following the first fine-needle aspiration cytology. Randomisation Patients were randomised at a 1 : 1 ratio to the interventional or control arms. Results A total of 982 participants (80% female) were randomised: 493 were randomised to ultrasound shear/strain wave elastography-ultrasound guided fine-needle aspiration cytology and 489 were randomised to ultrasound-only guided fine-needle aspiration cytology. There was no evidence of a difference between ultrasound shear/strain wave elastography and ultrasound in non-diagnostic cytology (Thy 1) rate following the first fine-needle aspiration cytology (19% vs. 16% respectively; risk difference: 0.030; 95% confidence interval −0.007 to 0.066; p = 0.11), the number of fine-needle aspiration cytologies needed (odds ratio: 1.10; 95% confidence interval 0.82 to 1.49; p = 0.53) or in the time to reach a definitive diagnosis (hazard ratio: 0.94; 95% confidence interval 0.81 to 1.10; p = 0.45). There was a small, non-significant reduction in the number of thyroid operations undertaken when ultrasound shear/strain wave elastography was used (37% vs. 40% respectively; risk difference: −0.02; 95% confidence interval −0.06 to 0.009; p = 0.15), but no difference in the number of operations yielding benign histology – 23% versus 24% respectively, p = 0.70 (i.e. no increase in identification of malignant cases) – or in the number of serious adverse events (2% vs. 1%). There was no difference in anxiety and depression, pain or quality of life between the two arms. Limitations The study was not powered to detect differences in malignancy. Conclusions Ultrasound shear/strain wave elastography does not appear to have additional benefit over ultrasound-guided fine-needle aspiration cytology in the diagnosis of thyroid nodules. Future work The findings of the ElaTION trial suggest that further research into the use of shear wave elastography in the diagnostic setting of thyroid nodules is unlikely to be warranted unless there are improvements in the technology. The diagnostic difficulty in distinguishing between benign and malignant lesions still persists. Future studies might examine the role of genomic testing on fine-needle aspiration samples. There is growing use of targeted panels of molecular markers, particularly aimed at improving the diagnostic accuracy of indeterminate (i.e. Thy3) cytology results. The application of these tests is not uniform, and their cost effectiveness has not been assessed in large-scale trials. Study registration This study is registered as ISRCTN (ISRCTN18261857). Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 12/19/04) and is published in full in Health Technology Assessment; Vol. 28, No. 46. See the NIHR Funding and Awards website for further award information. Plain language summary About half the population will have lumps in their thyroid if examined by an ultrasound scan but may not know they have one. About one in twenty people will feel a thyroid lump in their neck at some time in their life, with about one in twenty of those being malignant. Currently, the recommended way of getting a diagnosis of thyroid nodules is by using ultrasound to guide a needle to get cells from the lump, called ultrasound-guided fine-needle aspiration cytology. These cells are examined to determine the cause of the lump. If there are enough cells, Doctors can then make a diagnosis of whether the lump is benign or malignant. If not, patients will undergo another ultrasound-guided fine-needle aspiration cytology. One in five ultrasound-guided fine-needle aspiration cytologies are non-diagnostic with an overall false-positive rate of approximately 24%. This means one in five patients, with benign disease, may undergo unnecessary diagnostic operations. Thyroid surgery carries risks of complications, which could be avoided if we had better ways to diagnose which patients actually need an operation. We conducted a randomised trial, ElaTION, to determine if a new technology called strain and shear wave elastography, commonly known as real-time elastography, would be better at helping the radiologist take a sufficient sample of cells and reduce the number of non-diagnostic results, reducing the number of fine-needle aspiration cytologies required to make a definitive diagnosis. Nine hundred eighty-two patients were recruited between 2015 and 2018 and followed up until the end of the trial. Patients were randomised into two groups: 489 patients received the standard ultrasound-guided fine-needle aspiration cytology alone, and 493 patients received ultrasound-guided fine-needle aspiration cytology + shear wave elastography. Ultrasound shear/strain wave elastography did not reduce non-diagnostic cytology at first fine-needle aspiration cytology or improve the likelihood of determining whether the lump is benign or malignant. The results of ElaTION do not support the use of shear wave elastography-fine-needle aspiration cytology in the diagnosis of thyroid nodules. Scientific summary Background Strain elastography (STE) which is commonly used with concurrent real-time imaging known as real-time elastography (RTE) and shear wave elastography (SWE) are new diagnostic techniques that have been reported to be useful in the diagnosis of nodules in several organs. There is conflicting evidence regarding its benefit over ultrasound-guided fine-needle aspiration cytology (US-FNAC) alone in thyroid nodules. Objectives The primary objective was to determine if strain/shear wave ultrasound elastography (USE) in conjunction with FNAC will reduce the number of patients with thyroid nodules who have a non-diagnostic first FNAC result as compared to conventional ultrasound (US)-only guided FNAC. Design ElaTION was a pragmatic, multicentre randomised controlled trial. Setting Eighteen centres with a radiology department across England. Participants Adults with thyroid nodules who had not undergone previous FNAC with single or multiple nodules undergoing investigation. Interventions Ultrasound shear/strain wave elastography-US guided FNAC (intervention arm) – strain or shear wave elastography (USE)-guided FNAC. US-only guided FNAC (control arm) – routine US-only guided FNAC (the current standard recommended by the British Thyroid Association guidelines). Main outcome measures Primary outcome The proportion of patients who have a non-diagnostic cytology (Thy 1) result following the first FNAC. Secondary outcomes The number of FNACs required to obtain a definitive diagnosis; time from first FNAC to a definitive diagnosis; the false-positive rate of nodules; the number of patients who have a non-diagnostic cytology result following any FNAC; the proportion of patients undergoing thyroidectomy; accuracy of US alone (overall); accuracy of first FNAC; accuracy of USE or US without FNAC compared to accuracy of USE or US with FNAC; and patient reported anxiety, pain and quality of life [by the Hospital Anxiety and Depression Scale (HADS), Visual Analogue Pain Scale (VAPS) and EuroQol (EQ)-5D questionnaire at baseline, 3, 6 and 12 months post randomisation]. Results A total of 982 participants were randomised: 493 were randomised to USE-US guided FNAC and 489 were randomised to US-only guided FNAC. There was no evidence of a difference between USE and US in non-diagnostic (Thy 1) rate following the first FNAC [19% vs. 16% respectively; risk difference (RD): 0.030; 95% confidence interval (CI), −0.007 to 0.066; p = 0.11], the number of FNACs needed [odds ratio (OR): 1.10; 95% CI, 0.82 to 1.49; p = 0.53] or in the time to reach a definitive diagnosis [hazard ratio (HR): 0.94; 95% CI, 0.81 to 1.10; p = 0.45]. There was a small, non-significant reduction in the number of thyroid operations undertaken when USE was used (37% vs. 40% respectively; RD: −0.02; 95% CI, −0.06 to 0.009; p = 0.15), but no difference in the number of operations yielding benign histology – 23% versus 24% respectively, p = 0.70 (i.e. no increase in identification of malignant cases) – or in the number of serious adverse events (2% vs. 1%). There was no difference in anxiety and depression, pain or quality of life between the two arms. Limitations The study was not powered to detect differences in malignancy. Conclusions Ultrasound shear/strain wave elastography does not appear to have additional benefit over US-FNAC in the diagnosis of thyroid nodules. Study registration This study is registered as ISRCTN (ISRCTN18261857). Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 12/19/04) and is published in full in Health Technology Assessment; Vol. 28, No. 46. See the NIHR Funding and Awards website for further award information.

Background Fetal growth restriction is associated with perinatal morbidity and mortality. Early identification of women having at-risk fetuses can reduce perinatal adverse outcomes. Objectives To assess the predictive performance of existing models predicting fetal growth restriction and birthweight, and if needed, to develop and validate new multivariable models using individual participant data. Design Individual participant data meta-analyses of cohorts in International Prediction of Pregnancy Complications network, decision curve analysis and health economics analysis. Participants Pregnant women at booking. External validation of existing models (9 cohorts, 441,415 pregnancies); International Prediction of Pregnancy Complications model development and validation (4 cohorts, 237,228 pregnancies). Predictors Maternal clinical characteristics, biochemical and ultrasound markers. Primary outcomes fetal growth restriction defined as birthweight 32 weeks). To assess if the performance of the prediction models is generalisable for various definitions of FGR, and assess the association between various birthweight centiles calculated using customised and population-based standards and perinatal morbidity and mortality. To estimate the net benefit (clinical utility) of the developed prediction models using decision curve analysis (DCA). To assess the costs and outcomes and the potential impact of resource use of the prediction models. Methods We followed existing recommendations for prediction model development and validation and reported in line with guidelines for prognostic research and IPD meta-analysis. Our meta-analysis utilised IPD within the IPPIC Network database. IPPIC is a living data repository of cleaned and harmonised data of pregnant women from large birth or population-based cohorts, study cohort data, registries or unpublished data from hospital records. The primary outcomes were (1) FGR defined as birthweight

Background Sustaining independence is important for older people, but there is insufficient guidance about which community health and care services to implement. Objectives To synthesise evidence of the effectiveness of community services to sustain independence for older people grouped according to their intervention components, and to examine if frailty moderates the effect. Review design Systematic review and network meta-analysis. Eligibility criteria Studies: Randomised controlled trials or cluster-randomised controlled trials. Participants: Older people (mean age 65+) living at home. Interventions: community-based complex interventions for sustaining independence. Comparators: usual care, placebo or another complex intervention. Main outcomes Living at home, instrumental activities of daily living, personal activities of daily living, care-home placement and service/economic outcomes at 1 year. Data sources We searched MEDLINE (1946–), Embase (1947–), CINAHL (1972–), PsycINFO (1806–), CENTRAL and trial registries from inception to August 2021, without restrictions, and scanned reference lists. Review methods Interventions were coded, summarised and grouped. Study populations were classified by frailty. A random-effects network meta-analysis was used. We assessed trial-result risk of bias (Cochrane RoB 2), network meta-analysis inconsistency and certainty of evidence (Grading of Recommendations Assessment, Development and Evaluation for network meta-analysis). Results We included 129 studies (74,946 participants). Nineteen intervention components, including ‘multifactorial-action’ (multidomain assessment and management/individualised care planning), were identified in 63 combinations. The following results were of low certainty unless otherwise stated. For living at home, compared to no intervention/placebo, evidence favoured: multifactorial-action and review with medication-review (odds ratio 1.22, 95% confidence interval 0.93 to 1.59; moderate certainty) multifactorial-action with medication-review (odds ratio 2.55, 95% confidence interval 0.61 to 10.60) cognitive training, medication-review, nutrition and exercise (odds ratio 1.93, 95% confidence interval 0.79 to 4.77) and activities of daily living training, nutrition and exercise (odds ratio 1.79, 95% confidence interval 0.67 to 4.76). Four intervention combinations may reduce living at home. For instrumental activities of daily living, evidence favoured multifactorial-action and review with medication-review (standardised mean difference 0.11, 95% confidence interval 0.00 to 0.21; moderate certainty). Two interventions may reduce instrumental activities of daily living. For personal activities of daily living, evidence favoured exercise, multifactorial-action and review with medication-review and self-management (standardised mean difference 0.16, 95% confidence interval −0.51 to 0.82). For homecare recipients, evidence favoured the addition of multifactorial-action and review with medication-review (standardised mean difference 0.60, 95% confidence interval 0.32 to 0.88). Care-home placement and service/economic findings were inconclusive. Limitations High risk of bias in most results and imprecise estimates meant that most evidence was low or very low certainty. Few studies contributed to each comparison, impeding evaluation of inconsistency and frailty. Studies were diverse; findings may not apply to all contexts. Conclusions Findings for the many intervention combinations evaluated were largely small and uncertain. However, the combinations most likely to sustain independence include multifactorial-action, medication-review and ongoing review of patients. Some combinations may reduce independence. Future work Further research is required to explore mechanisms of action and interaction with context. Different methods for evidence synthesis may illuminate further. Study registration This study is registered as PROSPERO CRD42019162195. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128862) and is published in full in Health Technology Assessment; Vol. 28, No. 48. See the NIHR Funding and Awards website for further award information. Plain language summary Which community services are best for helping older people to be independent? Key messages Due to a lack of robust evidence, the benefits and risks of most types of community services for older people are unclear. Individualised care planning, where medication is adjusted and there are regular follow-ups, probably helps people stay living at home. What are community services for older people? There are many kinds of community services for older people. For example, in some services, everyone is given exercise and dietary advice or an individualised care plan. These often aim to help older people age independently. What was the study about? Maintaining independence is important in later life. We wanted to find out which community services work best: to help people stay living at home, and to do day-to-day activities independently. We reviewed findings from previous studies that have tested different community services for older people. We combined these findings and compared different types of service with one another. We rated our confidence in the evidence. What did we find? We found 129 studies with 74,946 people. We found 63 different kinds of service have been studied. The studies were carried out in diverse populations around the world. Individualised care planning, where medication is adjusted and there are regular follow-ups, may help people age independently. It probably increases the chance of staying at home slightly. It may also help with doing day-to-day activities very slightly. Exercise and dietary advice may also help people stay living at home. However, there was some evidence that some services may reduce independence. We do not know what effect most services have. What are the limitations of the evidence? We generally had little confidence in the evidence because studies were small, and information was missing. How current is the evidence? The evidence is up to date to August 2021. Scientific summary Background The number and proportion of older people are growing in the UK and worldwide. Maintaining independence is a goal of community health and care services for older people. The concept of frailty can be used to distinguish between people who remain in robust health in later life and those who are at greater risk of losing independence and needing care. Previous research has suggested that community-based complex interventions are generally effective for supporting independence for older people, but only broad service models have been explored. There is insufficient guidance about which services to implement and the appropriateness of different services for different levels of frailty. We aimed to provide a rigorous, contemporary synthesis of trial evidence to identify how interventions might best be configured to improve outcomes for older people, and inform the commissioning and delivery of evidence-based services. Objectives (list of research questions) Do community-based complex interventions to sustain independence in older people increase living at home, independence and health-related quality of life? Do community-based complex interventions to sustain independence in older people reduce homecare usage, depression, loneliness, falls, hospitalisation, care-home placement, costs and mortality? How should interventions be grouped for network meta-analysis (NMA)? What is the optimal configuration of community-based complex interventions to sustain independence in older people? Do intervention effects differ by a population’s frailty level (robust; pre-frailty; frailty)? Methods Systematic review with NMA of trials evaluating community-based complex interventions to sustain independence in older people (mean age 65 years and over), compared with usual care or another complex intervention meeting our criteria, with follow-up for at least 24 weeks. We followed Cochrane methods, Grading of Recommendations Assessment, Development and Evaluation (GRADE) and Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) NMA guidance. Information sources We searched the following databases and trial registers from inception between 9 and 11 August 2021: Cochrane Central Register of Controlled Trials (CENTRAL) Wiley (1992–); MEDLINE Ovid (1946–); Embase and Embase Classic Ovid (1947–); CINAHL EBSCOhost (1972–); APA PsycINFO Ovid (1806–); US National Institutes of Health Ongoing Trials Register, ClinicalTrials.gov (www.clinicaltrials.gov); World Health Organization, International Clinical Trials Registry Platform (https://trialsearch.who.int). We scanned the reference lists of included studies. Study selection Eligibility criteria Randomised controlled trials (RCTs) or cluster-RCTs. Participants were older people living at home (mean age 65 years or older). Participants living in residential/nursing homes were excluded. With an intervention that: was both initiated and mainly provided in the community included two or more interacting components (intervention practices, structural elements and contextual factors) was targeted at the individual person, with provision of appropriate specialist care focused on sustaining (maintaining or improving) the person’s independence. Usual care, ‘placebo’ or attention control or a different complex intervention which met our criteria were eligible comparators. Outcome data were measured at a minimum 24 weeks (approximately 6 months) time point. Study selection process Two researchers independently evaluated eligibility of records (title and abstract) and reports (full text). Disagreements were resolved by consensus. Data collection process Two researchers independently collected data. Main outcomes Living at home. Activities of daily living (ADL): personal ADL (PADL)/instrumental ADL (IADL). Hospitalisation. Care-home placement. Homecare services (non-healthcare professional) usage. Costs. Cost-effectiveness. Additional outcomes Health status/health-related quality of life. Depression. Loneliness. Falls. Mortality. Data were extracted (including treatment effect estimates) and categorised into three time frames: short term (around 6 months): 24 weeks to 9 months medium term (around 12 months): > 9 months to 18 months long term (around 24 months): > 18 months with the medium term as our main time frame. Intervention grouping We grouped all eligible interventions (including comparators) in preparation for NMA in a three-stage process of coding and summarising based on the Template for Intervention Description and Replication framework, categorisation and grouping. Assessment of frailty Two reviewers with extensive clinical academic frailty expertise (AC and JG) independently categorised study level frailty (robust, pre-frailty, frailty) based on validated measures where available or participant characteristics and study inclusion criteria using the phenotype model as a framework. Risk-of-bias assessment Two reviewers independently assessed risk of bias (RoB) in each result of interest from each included study, using the Revised Cochrane risk-of-bias tool for randomised trials (RoB 2). Data synthesis Meta-analysis was conducted for living at home, PADL, IADL and care-home placement for each of the three time frames separately, and for hospitalisation, health status and depression in the medium term only. Other outcomes were narratively synthesised. Meta-analysis We meta-analysed the extracted effect estimates using modules within Stata. Random-effects meta-analyses were conducted. Initially, for each outcome and time frame, we performed a separate meta-analysis for each type of intervention versus control, to provide summary effectiveness results based only on direct evidence. An NMA was then conducted (for each outcome and time frame separately) using a multivariate random-effects meta-analysis framework via the network module in Stata using restricted maximum likelihood estimation. We produced summary (pooled) effect estimates for each pair of treatments in the network, with 95% confidence intervals (CIs). Based on the results, the ranking of intervention groups was calculated using resampling methods. The consistency assumption (that direct and indirect evidence are consistent with each other) was examined for each treatment comparison where possible and across the whole network. The effect of study-level frailty on each intervention group effect was examined where data allowed. Sensitivity analyses were conducted excluding results at the highest RoB, and funnel plots examined for small-study effects. Confidence in cumulative evidence We used the GRADE framework, adapted for NMA, to rate the certainty of the results of our NMA. Summary of economic evidence We followed the brief economic commentary framework to summarise, compare and contrast the principal findings from the included studies. Results We screened 40,112 records and assessed 973 reports for eligibility. We included 129 studies consisting of 496 reports. The studies assigned 74,946 participants (three studies missing data) to 266 eligible intervention arms. They were predominantly conducted in developed countries and most participants were described as white. Nonetheless, the overall population included a broad range of demographic characteristics. Study populations included all frailty levels. We identified 19 separate components of included interventions which were evaluated in 63 combinations including the absence of all of these components, which we termed available care (ac), and homecare (a common control group in populations where all participants were receiving homecare). Homecare involved frequent visits at home by professionals who typically supported domestic and self-care tasks. Five components were primarily about a process of ascertainment or assessment and planning with subsequent action: multifactorial-action from care planning (a process of individualised multidomain assessment and management) with or without routine review (scheduled, regular follow-ups), medication-review, monitoring and routine risk-screening. The 14 other components and their short labels (bold) were ADL training, providing aids and adaptations, alternative medicine, care voucher provision, cognitive training, health education, physical exercise, formal homecare, engagement in meaningful-activities, nutritional support, psychological (mood) therapy (psychology), social skills training, technology for communication and engagement (telecoms), welfare rights advice. Multifactorial-action was further delineated based on the presence or absence of an embedded medication-review and specific self-management strategies. We judged most results to be at high RoB, primarily due to missing outcome data. This led to serious concerns with RoB for many of the GRADE ratings of evidence. Findings Most networks were small and sparse, with few included studies contributing to most networks. We found little evidence of inconsistency but there was usually low power to detect this. All outcomes except mortality needed to be analysed in two separate NMAs as the networks were disconnected: one with ac as the reference comparator (‘available-care network’) and one with homecare as the reference comparator (‘homecare network’). Estimates are reported here only in comparison with the reference comparator. Comparisons with ac can be thought of as the effect of adding the intervention for a population who are not all receiving any particular care; comparisons with homecare are similarly an alternative intervention for a population already in receipt of homecare without associated reablement or multifactorial-action from care planning. Most estimates were low certainty or very low certainty due to RoB, imprecision or their combination, and we do not describe very low-certainty evidence below. Living at home For living at home in the medium term there were 21 studies (n = 16,937) with 14 intervention groups in the available-care network. There was moderate-certainty evidence that multifactorial-action and review with medication-review probably results in a slight increase in the chance of living at home [odds ratio (OR) 1.22, 95% CI 0.93 to 1.59; moderate certainty]. There was low-certainty evidence that multifactorial-action with medication-review [OR 2.55 (large), 95% CI 0.61 to 10.60]; cognitive training, medication-review, nutrition and exercise [OR 1.93 (large), 95% CI 0.79 to 4.77]; and ADL, nutrition and exercise [OR 1.79 (large), 95% CI 0.67 to 4.76] may result in an increase in the chance of living at home, and that risk-screening; education, multifactorial-action and review with medication-review; and education, multifactorial-action and review with medication-review and self-management may each result in some reduction in chance of living at home. Other comparisons with ac were of very low certainty. In the short- and long-term time frames, results were at best low certainty. For multifactorial-action and review with medication-review; and ADL, nutrition and exercise, estimates were similarly of small increases in the long term but of little to no difference in the short term. There were similar results in other time frames for education, multifactorial-action and review with medication-review and self-management; and risk-screening, but contrasting evidence of reduction followed by an increase in living at home for education, multifactorial-action and review with medication-review. The homecare network for living at home was smaller (five studies, n = 1978 in the medium term). In the short- and medium-term time frames, there was low-certainty evidence that homecare, ADL, multifactorial-action and review with self-management may result in a moderate or large reduction in the chance of living at home compared with homecare alone. Instrumental activities of daily living For the medium-term instrumental activities of daily living (IADL) available-care network there were 16 studies (n = 5309) with 14 intervention groups. Multifactorial-action and review with medication-review was associated with very slightly increased independence in IADL versus ac [standardised mean difference (SMD) 0.11, 95% CI 0.00 to 0.21; moderate-certainty evidence]. Two intervention groups may result in some reduction in IADL: ADL, aids and exercise; and ADL, aids, education, exercise, multifactorial-action and review with medication-review and self-management. There were contrasting findings for multifactorial-action and review with medication-review in the long term, with moderate-certainty evidence of a very slight reduction in IADL (SMD −0.08, 95% CI −0.21 to 0.05). For the homecare network, there was one low certainty finding in the short-term time frame of little to no difference for homecare, ADL, multifactorial-action and review with self-management with all other estimates being very low certainty. Personal activities of daily living For personal activities of daily living (PADL), 20 trials (n = 8583 participants) with 16 intervention groups contributed to the medium-term available-care network. One comparison was judged low certainty. Exercise, multifactorial-action and review with medication-review and self-management may result in a very slight increase in PADL (SMD 0.16, 95% CI −0.51 to 0.82). The homecare network included four trials (n = 632 participants) in the medium term. As for ac, only one comparison with homecare was low certainty: homecare, multifactorial-action and review with medication-review may result in an increase in PADL [SMD 0.60 (moderate), 95% CI 0.32 to 0.88]. Other outcomes For the service outcome of hospitalisation, there were low-certainty estimates of some reductions for education, exercise, multifactorial-action and review with medication-review and self-management; and education, multifactorial-action and review with medication-review; and of an increase for exercise, multifactorial-action and review with medication-review and self-management. For care-home placement, all estimates were rated very low certainty in the medium term. There was some evidence of both increases and decreases in use of homecare services with little pattern (not meta-analysed). For our additional outcomes, there was little evidence of any effect on self-reported health status, only low certainty beneficial findings regarding depression, very little evidence regarding loneliness and more complex interventions were associated with less falling than more falling (12 studies vs. 4 studies). For mortality, there was a large network of 65 studies (n = 38,351) and 41 intervention groups. There was low-certainty evidence of reductions for two, and increases for five, intervention groups. The summary of economic evidence included 39 studies. Based on the conclusions of 22 studies that performed a full economic evaluation, five intervention groups appeared promising compared with a standard intervention or ac from an economic perspective: ADL (medium-term time horizon); homecare, multifactorial-action and review with medication-review and self-management (short-term time horizon); meaningful-activities and education (short- and medium-term time horizon); multifactorial-action and review with medication-review (short- but not medium- or long-term time horizon); and exercise and multifactorial-action with medication-review (long-term time horizon). Summary across outcomes We found evidence that multifactorial-action and review with medication-review probably improves some important outcomes slightly (living at home, IADL), but there was also contradictory evidence for IADL in the long term. For some other intervention groups there was low-certainty evidence that they may improve or worsen particular outcomes but for most intervention groups evidence was either absent or very uncertain. Conclusions Available evidence suggests the community-based complex interventions most likely to sustain independence in older people involve multifactorial-action from multidomain assessment and individualised care planning, routine review and the incorporation of medication-review. There was also some positive evidence for the combination of exercise and nutritional support and multiple other intervention combinations. Decision-makers should be aware that there is plausible evidence that some community-based complex interventions may worsen outcomes such as living at home and ADL independence and that all of these findings are tentative. We recommend the uncertainty in these findings be addressed by: realist synthesis to explore the mechanisms and broader contextual factors relating to individual benefit or harm future robust, large-scale trials which compare alternative interventions with multifactorial-action and review with medication-review future Individual Participant Data meta-analysis (IPDMA) focusing on interventions with multifactorial action to explore factors relating to individual benefit or harm greater reporting of the organisational aspects of intervention implementation in complex intervention research. Study registration This study is registered as PROSPERO CRD42019162195. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128862) and is published in full in Health Technology Assessment; Vol. 28, No. 48. See the NIHR Funding and Awards website for further award information.

Background Renal cell carcinoma is the most common type of kidney cancer, comprising approximately 85% of all renal malignancies. Patients with advanced renal cell carcinoma are the focus of this National Institute for Health and Care Excellence multiple technology appraisal. A patient’s risk of disease progression depends on a number of prognostic risk factors; patients are categorised as having intermediate/poor risk or favourable risk of disease progression. Objectives The objectives of this multiple technology appraisal were to appraise the clinical effectiveness and cost-effectiveness of lenvatinib plus pembrolizumab versus relevant comparators listed in the final scope issued by the National Institute for Health and Care Excellence: sunitinib, pazopanib, tivozanib, cabozantinib and nivolumab plus ipilimumab. Methods The assessment group carried out clinical and economic systematic reviews and assessed the clinical and cost-effectiveness evidence submitted by Eisai, Hatfield, Hertfordshire, UK (the manufacturer of lenvatinib) and Merck Sharp & Dohme, Whitehouse Station, NJ, USA (the manufacturer of pembrolizumab). The assessment group carried out fixed-effects network meta-analyses using a Bayesian framework to generate evidence for clinical effectiveness. As convergence issues occurred due to sparse data, random-effects network meta-analysis results were unusable. The assessment group did not develop a de novo economic model, but instead modified the partitioned survival model provided by Merck Sharp & Dohme. Results The assessment group clinical systematic review identified one relevant randomised controlled trial (CLEAR trial). The CLEAR trial is a good-quality, phase III, multicentre, open-label trial that provided evidence for the efficacy and safety of lenvatinib plus pembrolizumab compared with sunitinib. The assessment group progression-free survival network meta-analysis results for all three risk groups should not be used to infer any statistically significant difference (or lack of statistically significant difference) for any of the treatment comparisons owing to within-trial proportional hazards violations or uncertainty regarding the validity of the proportional hazards assumption. The assessment group overall survival network meta-analysis results for the intermediate-/poor-risk subgroup suggested that there was a numerical, but not statistically significant, improvement in the overall survival for patients treated with lenvatinib plus pembrolizumab compared with patients treated with cabozantinib or nivolumab plus ipilimumab. Because of within-trial proportional hazards violations or uncertainty regarding the validity of the proportional hazards assumption, the assessment group overall survival network meta-analysis results for the favourable-risk subgroup and the all-risk population should not be used to infer any statistically significant difference (or lack of statistically significant difference) for any of the treatment comparisons. Only one cost-effectiveness study was included in the assessment group review of cost-effectiveness evidence. The study was limited to the all-risk population, undertaken from the perspective of the US healthcare system and included comparators that are not recommended by the National Institute for Health and Care Excellence for patients with untreated advanced renal cell carcinoma. Therefore, the extent to which resource use and results are generalisable to the NHS is unclear. The assessment group cost-effectiveness results from the modified partitioned survival model focused on the intermediate-/poor-risk and favourable-risk subgroups. The assessment group cost-effectiveness results, generated using list prices for all drugs, showed that, for all comparisons in the favourable-risk subgroup, treatment with lenvatinib plus pembrolizumab costs more and generated fewer benefits than all other treatments available to NHS patients. For the intermediate-/poor-risk subgroup, treatment with lenvatinib plus pembrolizumab costs more and generated more benefits than treatment with cabozantinib and nivolumab plus ipilimumab. Conclusions Good-quality clinical effectiveness evidence for the comparison of lenvatinib plus pembrolizumab with sunitinib is available from the CLEAR trial. For most of the assessment group Bayesian hazard ratio network meta-analysis comparisons, it is difficult to reach conclusions due to within-trial proportional hazards violations or uncertainty regarding the validity of the proportional hazards assumption. However, the data (clinical effectiveness and cost-effectiveness) used to populate the economic model are relevant to NHS clinical practice and can be used to inform National Institute for Health and Care Excellence decision-making. The assessment group cost-effectiveness results, generated using list prices for all drugs, show that lenvatinib plus pembrolizumab is less cost-effective than all other treatment options. Study registration This study is registered as PROSPERO CRD4202128587. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis Programme (NIHR award ref: NIHR134985) and is published in full in Health Technology Assessment; Vol. 28, No. 49. See the NIHR Funding and Awards website for further award information. Plain language summary What was the problem? Renal cell carcinoma is the most common type of kidney cancer. Several drug treatment options are available for NHS patients with advanced or metastatic disease, and the choice of treatment varies depending on a patient’s risk of disease progression. A new drug combination, lenvatinib plus pembrolizumab, may soon become available to treat NHS patients. This review explored whether treatment with lenvatinib plus pembrolizumab offered value for money to the NHS. What did we do? We reviewed the effectiveness of treatment with lenvatinib plus pembrolizumab versus other NHS treatment options. We also estimated the costs and benefits of treatment with lenvatinib plus pembrolizumab versus current NHS treatments for patients with higher and lower risks of disease progression. What did we find? Compared with current NHS treatments, treatment with lenvatinib plus pembrolizumab may increase the time that people with a higher risk of disease progression (i.e. worsening disease) were alive. However, for patients with a lower risk of disease progression, the available evidence is limited and only shows that treatment with lenvatinib plus pembrolizumab may prolong the time that patients have a stable level of disease. For all patients, compared to all current NHS treatments, treatment with lenvatinib plus pembrolizumab is very expensive. What does this mean? Compared with current NHS treatments for untreated renal cell carcinoma, using published prices (which do not include any discounts that are offered to the NHS), treatment with lenvatinib plus pembrolizumab may not provide good value for money to the NHS. Scientific summary Background Renal cell carcinoma (RCC) is the most common type of kidney cancer, comprising approximately 85% of all renal malignancies. Patients with advanced RCC (aRCC) have Stage 3 (locally advanced) or Stage 4 (metastatic) disease. A patient’s risk of disease progression depends on a number of prognostic risk factors. The International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) model is used in NHS clinical practice to categorise patients into one of two groups, namely favourable risk or intermediate/poor risk. This systematic review and cost-effectiveness analysis has been conducted to inform the following National Institute for Health and Care Excellence (NICE) multiple technology appraisal: lenvatinib with pembrolizumab for untreated aRCC (ID3760). In November 2021, the Medicines and Healthcare products Regulatory Agency approved the use of lenvatinib plus pembrolizumab as a treatment for all patients with untreated aRCC. Objectives The comparators listed in the final scope issued by NICE differ depending on the risk of disease progression. The objectives of this assessment were to appraise the clinical effectiveness and cost-effectiveness of lenvatinib plus pembrolizumab versus: cabozantinib and nivolumab plus ipilimumab in the intermediate-/poor-risk subgroup sunitinib, pazopanib and tivozanib in the favourable-risk subgroup sunitinib, pazopanib and tivozanib in the all-risk population. Clinical and economic systematic review methods The assessment group (AG) carried out a systematic review of clinical effectiveness evidence following the general principles outlined by the Centre for Reviews and Dissemination (CRD). The review was reported using the criteria recommended in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement. Searches were conducted between 11 October 2021 and 22 November 2021 in accordance with the general principles recommended by the European Network for Health Technology Assessment. The protocol is registered with PROSPERO (registration number: CRD42021285879). The AG reviewed only randomised controlled trials (RCTs) and full economic analyses identified by the searches. However, the group also considered evidence provided by the manufacturers of lenvatinib (Eisai Ltd) and pembrolizumab (Merck Sharp & Dohme, Whitehouse Station, NJ, USA) provided in submissions to NICE; company submission (CS) reference lists were searched for relevant RCTs. In line with the final scope issued by NICE, the outcomes considered by the AG were overall survival (OS), progression-free survival (PFS), objective tumour response rate, adverse events (AEs), health-related quality of life (HRQoL), incremental cost per life-year gained and incremental cost per quality-adjusted life-year (QALY) gained. Clinical effectiveness results Direct clinical effectiveness evidence (CLEAR trial) The AG systematic review included one RCT, the CLEAR trial. The CLEAR trial was a good-quality, phase III, multicentre, open-label RCT (with an ongoing extension phase) that provided evidence for the comparison of the efficacy of lenvatinib plus pembrolizumab versus sunitinib. Results for all outcomes were assessed at the third interim analysis (August 2020, median OS follow-up of 26.6 months), that is the final data cut-off for PFS. The companies also presented OS results from an updated OS analysis (March 2021, median OS follow-up of approximately 33 months). At the time of the third interim analysis, the CLEAR trial hazard ratio (HR) results showed statistically significant improvements in PFS and objective tumour response rate for patients treated with lenvatinib plus pembrolizumab versus patients treated with sunitinib for the intermediate-/poor-risk subgroup, the favourable-risk subgroup and the all-risk population. The HR results from the updated OS analysis showed a statistically significant improvement for patients treated with lenvatinib plus pembrolizumab versus patients treated with sunitinib for the intermediate-/poor-risk subgroup and the all-risk population; there were too few events in the favourable-risk subgroup for robust OS conclusions to be drawn. Eisai carried out a treatment-switching analysis to test whether adjusting for the effect of subsequent treatments affected OS results. Results were generated only for the all-risk population and were marked as academic-in-confidence. Nearly all the patients in the CLEAR trial lenvatinib plus pembrolizumab and sunitinib arms experienced at least one all-grade AE, with more Grade ≥ 3 AEs reported in the lenvatinib plus pembrolizumab arm than in the sunitinib arm. The proportion of patients who discontinued treatment of either lenvatinib or pembrolizumab due to AEs was approximately twice as high as patients who discontinued treatment of sunitinib; the proportion of patients who withdrew treatment of both lenvatinib and pembrolizumab due to AEs was approximately the same as the proportion of patients who withdrew treatment with sunitinib. Health-related quality of life was measured using three tools, including the EuroQol-5 Dimensions, three-level version questionnaire. When compared with treatment with sunitinib, treatment with lenvatinib plus pembrolizumab did not result in any clinically meaningful differences (as measured by predefined minimally important differences) in HRQoL measured using any of the three tools. Indirect clinical effectiveness evidence To compare the effectiveness of lenvatinib plus pembrolizumab versus relevant comparators other than sunitinib, the AG carried out Bayesian HR network meta-analyses. It was decided not to undertake a flexible modelling approach for network meta-analysis (NMA), which relaxes the proportional hazards (PH) assumption, such as fractional polynomial network meta-analyses because interpretation of the estimates provided by these complex modelling techniques can be difficult and results are often not intuitive. While deviance information criterion (DIC) statistics provide an approach to compare the fit of different models, they do not provide information about whether a model is a good fit to the data or whether the estimates generated by the model, including projections of results beyond the follow-up times of trials included in the NMA, are clinically plausible. Furthermore, flexible models, which appear similar according to model fit (i.e. according to DIC statistics), may generate very different long-term survival estimates. The AG assessed the feasibility of conducting Bayesian HR NMAs for the three population risk groups (intermediate-/poor-risk subgroup, favourable-risk subgroup and all-risk population) for all outcomes listed in the final scope issued by NICE. However, due to limited data availability, it was not possible to carry out NMAs for all outcomes for all three patient risk groups. Further, as networks were sparse, it was only possible to generate results using fixed-effect NMAs. The AG PFS NMA results for the intermediate-/poor-risk subgroup, the favourable-risk subgroup and the all-risk population should not be used to infer any statistically significant difference (or lack of statistically significant difference) for any of the treatment comparisons because of within-trial PH violations or uncertainty regarding the validity of the PHs assumption. The AG OS NMA results for the intermediate-/poor-risk subgroup suggested that there was a numerical, but not statistically significant, improvement in the OS for patients treated with lenvatinib plus pembrolizumab compared with patients treated with cabozantinib or nivolumab plus ipilimumab. Because of within-trial PH violations or uncertainty regarding the validity of the PH assumption, the AG OS NMA results for the favourable-risk subgroup and the all-risk population should not be used to infer any statistically significant difference (or lack of statistically significant difference) for any of the treatment comparisons. The AG objective tumour response rate NMA results for the intermediate-/poor-risk subgroup suggested that, although treatment with lenvatinib plus pembrolizumab led to a statistically significant improvement in objective tumour response rate compared to treatment nivolumab plus ipilimumab, it did not lead to a statistically significant improvement in objective tumour response rate for the comparison of lenvatinib plus pembrolizumab versus cabozantinib. It was not possible to generate results for the IMDC/MSKCC (Memorial Sloan-Kettering Cancer Center) favourable-risk subgroup due to data limitations. The AG objective tumour response rate NMA results for the all-risk population suggest that treatment with lenvatinib plus pembrolizumab led to a statistically significant improvement in objective tumour response rate versus treatment with sunitinib and versus treatment with pazopanib. The AG Grade ≥ 3 AE NMA results for the intermediate-/poor-risk subgroup suggested that treatment with lenvatinib plus pembrolizumab led to statistically significantly more Grade ≥ 3 AEs versus treatment with cabozantinib. It was not possible to generate results for the IMDC/MSKCC favourable-risk subgroup. The AG Grade ≥ 3 AE NMA results for the all-risk population suggested that treatment with lenvatinib led to statistically significantly more Grade ≥ 3 AEs versus treatment with sunitinib and versus treatment with pazopanib. Economic systematic review results The AG systematic review identified one relevant cost-effectiveness study. This study compared the cost-effectiveness of lenvatinib plus pembrolizumab versus sunitinib (and vs. other treatments). However, the study was undertaken from the perspective of the US healthcare system and generated results only for the all-risk population and included comparators that are not recommended by NICE as treatment options for patients with aRCC. Therefore, the extent to which these results were generalisable to the NHS was unclear. Cost-effectiveness analysis methods The Eisai and Merck Sharp & Dohme CSs to NICE included partitioned survival models built in Microsoft Excel. The AG considered that results from both models could be used to inform decision-making but that, in some instances, the companies could have made more appropriate assumptions and parameter choices. The AG did not develop a de novo economic model; instead, it modified the model provided by Merck Sharp & Dohme [referred to as the Merck Sharp & Dohme/Assessment Group (MSD/AG) model]. Neither of the companies produced cost-effectiveness results for the comparison of lenvatinib plus pembrolizumab versus nivolumab plus ipilimumab (intermediate-/poor-risk subgroup) despite both models having the functionality for this comparison. Furthermore, Eisai did not generate any cost-effectiveness results for the favourable-risk subgroup. The MSD/AG model was populated with OS, PFS and time to treatment discontinuation (TTD) data from the CLEAR trial (lenvatinib plus pembrolizumab versus sunitinib for favourable-risk subgroup and the all-risk population). The AG PFS and OS NMA results were used to estimate effectiveness for the comparison of lenvatinib plus pembrolizumab versus cabozantinib and versus nivolumab plus ipilimumab for the intermediate-/poor-risk population. NICE appraisal committees have concluded that sunitinib and pazopanib are of equivalent effectiveness and that, at best, tivozanib may have a similar effect to sunitinib or pazopanib. These conclusions were based on all-risk population data; the AG has assumed that this assumption holds for the favourable-risk population. The most important changes made by the AG to the Merck Sharp & Dohme model were different choices for estimating PFS, OS and TTD for the intervention and comparator treatments and for modelling two lines, rather than one line, of subsequent treatment. Cost-effectiveness analysis results The AG cost-effectiveness results presented in this report were estimated using list prices. Also, the AG cost-effectiveness results generated using confidential discounted prices were supplied to NICE in a confidential appendix, but cannot be presented here. For the intermediate-/poor-risk subgroup, the AG base-case cost-effectiveness results suggested that treatment with lenvatinib plus pembrolizumab generated more QALYs versus treatment with cabozantinib and versus nivolumab plus ipilimumab, but at a greater overall cost than either of these two treatments. Using list prices, the incremental cost-effectiveness ratios per QALY gained for the comparison of lenvatinib plus pembrolizumab versus cabozantinib and versus nivolumab plus ipilimumab exceed £100,000. For the favourable-risk subgroup, the AG base-case cost-effectiveness results suggested that treatment with sunitinib generated more QALYs than treatment with lenvatinib plus pembrolizumab at a lower overall cost, that is treatment with lenvatinib plus pembrolizumab was dominated by treatment with sunitinib (and, using the assumption of equivalent effectiveness, by pazopanib and tivozanib). The AG carried out extensive one-way sensitivity analyses, scenario analyses and probabilistic sensitivity analyses. Results from these analyses demonstrate that the AG base-case cost-effectiveness results are robust. Clinical and cost-effectiveness conclusions Good-quality clinical effectiveness evidence for the comparison of lenvatinib plus pembrolizumab versus sunitinib was available from the CLEAR trial. For most of the AG Bayesian HR NMA comparisons, it was difficult to reach conclusions due to within-trial PH violations or uncertainty regarding the validity of the PHs assumption. However, the data (clinical effectiveness and cost-effectiveness) used to populate the MSD/AG model are relevant to NHS clinical practice and can be used to inform NICE decision-making. The all-risk population comprises patients with intermediate-/poor-risk and patients with favourable-risk disease. The AG cost-effectiveness analyses have focused on the two subgroups, and the AG cost-effectiveness results, generated using list prices for all drugs, show that lenvatinib plus pembrolizumab is less cost-effective than all other treatment options. Study registration This study is registered as PROSPERO CRD4202128587. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis Programme (NIHR award ref: NIHR134985) and is published in full in Health Technology Assessment; Vol. 28, No. 49. See the NIHR Funding and Awards website for further award information.

Background Lung cancer is one of the most common types of cancer in the United Kingdom. It is often diagnosed late. The 5-year survival rate for lung cancer is below 10%. Early diagnosis may improve survival. Software that has an artificial intelligence-developed algorithm might be useful in assisting with the identification of suspected lung cancer. Objectives This review sought to identify evidence on adjunct artificial intelligence software for analysing chest X-rays for suspected lung cancer, and to develop a conceptual cost-effectiveness model to inform discussion of what would be required to develop a fully executable cost-effectiveness model for future economic evaluation. Data sources The data sources were MEDLINE All, EMBASE, Cochrane Database of Systematic Reviews, Cochrane CENTRAL, Epistemonikos, ACM Digital Library, World Health Organization International Clinical Trials Registry Platform, clinical experts, Tufts Cost-Effectiveness Analysis Registry, company submissions and clinical experts. Searches were conducted from 25 November 2022 to 18 January 2023. Methods Rapid evidence synthesis methods were employed. Data from companies were scrutinised. The eligibility criteria were (1) primary care populations referred for chest X-ray due to symptoms suggestive of lung cancer or reasons unrelated to lung cancer; (2) study designs that compared radiology specialist assessing chest X-ray with adjunct artificial intelligence software versus radiology specialists alone and (3) outcomes relating to test accuracy, practical implications of using artificial intelligence software and patient-related outcomes. A conceptual decision-analytic model was developed to inform a potential full cost-effectiveness evaluation of adjunct artificial intelligence software for analysing chest X-ray images to identify suspected lung cancer. Results None of the studies identified in the searches or submitted by the companies met the inclusion criteria of the review. Contextual information from six studies that did not meet the inclusion criteria provided some evidence that sensitivity for lung cancer detection (but not nodule detection) might be higher when chest X-rays are interpreted by radiology specialists in combination with artificial intelligence software than when they are interpreted by radiology specialists alone. No significant differences were observed for specificity, positive predictive value or number of cancers detected. None of the six studies provided evidence on the clinical effectiveness of adjunct artificial intelligence software. The conceptual model highlighted a paucity of input data along the course of the diagnostic pathway and identified key assumptions required for evidence linkage. Limitations This review employed rapid evidence synthesis methods. This included only one reviewer conducting all elements of the review, and targeted searches that were conducted in English only. No eligible studies were identified. Conclusions There is currently no evidence applicable to this review on the use of adjunct artificial intelligence software for the detection of suspected lung cancer on chest X-ray in either people referred from primary care with symptoms of lung cancer or people referred from primary care for other reasons. Future work Future research is required to understand the accuracy of adjunct artificial intelligence software to detect lung nodules and cancers, as well as its impact on clinical decision-making and patient outcomes. Research generating key input parameters for the conceptual model will enable refinement of the model structure, and conversion to a full working model, to analyse the cost-effectiveness of artificial intelligence software for this indication. Study registration This study is registered as PROSPERO CRD42023384164. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135755) and is published in full in Health Technology Assessment; Vol. 28, No. 50. See the NIHR Funding and Awards website for further award information. Plain language summary Lung cancer is one of the most common types of cancer in the United Kingdom. Early diagnosis may improve survival, as lung cancer is often diagnosed late. Chest X-rays can be used to identify features of lung cancer. There can be delays in getting X-rays, and sometimes features of lung cancer are not seen on them. Artificial intelligence software may help by finding features of cancer on chest X-rays and highlighting them. A radiologist will look at the X-rays and information from the software. There is a lack of information about how lung cancer diagnosis could change if artificial intelligence software is used and what the costs may be to the National Health Service. This project looked at the use of artificial intelligence software in the detection of lung cancer in people referred from primary care. Software companies were invited to provide evidence. There were no studies that looked at this topic among people from primary care. We summarised the closest evidence we could find instead. All of this had flaws, so we could not tell if the results were accurate or helpful to this review. It was not clear if artificial intelligence helped to find cancers or improve people’s health. We made a theoretical model to discuss the best way to assess if artificial intelligence software might be cost-effective in detecting lung cancer and what evidence would be needed to do this in a fully working model. Costs and alternative pricing models provided by five companies were used to calculate the cost of adding artificial intelligence software to review chest X-rays in people referred from their general practitioner, for the first 5 years, based on one National Health Service trust. Future studies are needed to identify the impact of adjunct artificial intelligence on test accuracy, clinical decision-making and patient outcomes (e.g. mortality and morbidity). Scientific summary Background Lung cancer occurs when abnormal cells multiply in an uncontrolled way to form a tumour in the lung. It is one of the most common types of cancer in the UK, and each year over 43,000 new cases are diagnosed. In the early stages of the disease, people usually do not have symptoms, which means that lung cancer is often diagnosed late. The 5-year survival rate for lung cancer is low, at below 10%. Early diagnosis may improve survival. The National Institute for Health and Care Excellence (NICE) has identified software that has an artificial intelligence (AI)-developed algorithm (referred to hereafter as AI software) as potentially useful in assisting with the identification of suspected lung cancer. AI combines computer science and data sets to enable problem-solving. Machine learning and deep learning are subfields of AI. They comprise AI algorithms that seek to create expert systems to make predictions or classifications based on data input. This assessment covers the use of AI software as an adjunct to an appropriate radiology specialist to assist in the identification of suspected lung cancer on chest X-rays (CXRs). AI technologies subject to this assessment are standalone software platforms developed with deep-learning algorithms to interpret CXRs. The algorithms are fixed but updated periodically. The AI software automatically interprets radiology images from the CXR to identify abnormalities or suspected abnormalities. The abnormalities detected and the methods of flagging the location and type of abnormalities differ between different AI technologies. For example, a CXR may be flagged as suspected lung cancer when a lung nodule, lung mass or hilar enlargement, or a combination of these, is identified. A technology may classify CXRs into those with and without a nodule, or it may identify several different abnormalities or lung diseases. Objectives The overall aim of this early value assessment (EVA) is to identify evidence on adjunct AI software for analysing CXRs for suspected lung cancer and identify evidence gaps to help direct data collection and further research. A conceptual modelling process was undertaken to inform discussion of what would be required to develop a fully executable cost-effectiveness model for future economic evaluation. The assessment is not intended to replace the need for a full assessment (Diagnostic Assessment Report) or to provide sufficient detail or synthesis to enable a recommendation to be made about whether AI software can be implemented in clinical practice at the present time. There are two populations of interest in this EVA: (1) people referred from primary care for a CXR because they have symptoms suggestive of lung cancer (symptomatic population) and (2) people referred from primary care for a CXR for reasons unrelated to lung cancer (incidental population). Based on the scope produced by NICE, we defined the following questions to inform future assessment on the benefits, harms and costs of adjunct AI for analysing on CXRs for suspected lung cancer compared with human reader alone in these populations: What is the test accuracy and test failure rate of adjunct AI software to detect lung cancer on CXRs? What are the practical implications of adjunct AI to detect lung cancer on CXRs? What is the clinical effectiveness of adjunct AI software applied to CXRs? What are the cost and resource use considerations relating to use of adjunct AI to detect lung cancer? What would a health economic model to estimate the cost-effectiveness of adjunct AI to detect lung cancer look like? Methods Data sources MEDLINE All (via Ovid), EMBASE (via Ovid), Cochrane Database of Systematic Reviews (via Wiley), Cochrane CENTRAL (via Wiley), Epistemonikos, ACM Digital Library, World Health Organization International Clinical Trials Registry Platform, clinical experts, and company submissions. Eligibility criteria Population: people referred for CXR from primary care because they have symptoms suggestive of lung cancer, and people referred for CXR from primary care for reasons unrelated to lung cancer. Intervention: radiology specialist with adjunct AI Comparator: radiology specialist without adjunct AI Outcomes: test accuracy, patient management, clinical effectiveness. Study selection, data extraction and assessment of risks of bias Titles and abstracts of all identified records were screened by one reviewer against the review eligibility criteria, with a random 20% screened by a second reviewer. Full texts of records considered potentially relevant by either reviewer were retrieved and assessed for inclusion by one reviewer. A random 20% sample were assessed independently by a second reviewer, with any disagreements resolved by consensus or discussion with a third reviewer. We planned to extract data into a piloted form, assess risk of bias and synthesise data using methods described in the research protocol; however, no studies met the inclusion criteria. Post hoc methods were determined following discussions with NICE to select and summarise the closest available evidence to the review inclusion criteria. Studies were selected that assessed eligible AI software in conjunction with radiologists compared with radiologists alone but in which the referral status and symptomatic status of the population was unclear. Data were extracted by one reviewer, with a random 20% checked by a second reviewer. Results were summarised narratively, and key biases were noted. Data synthesis A narrative data synthesis was performed. Modelling The conceptual modelling process explored both the structure and evidence requirements for parameter inputs, for future model development. An iterative approach was taken to facilitate the identification of cost outcomes, potential value drivers of AI software for this indication and evidence linkage requirements for longer-term outcomes, where time allowed. Costs associated with implementing AI software were also considered. Information to inform the conceptual model was obtained from a variety of sources including a literature review, current clinical guidelines, discussion with specialist clinical experts and the companies submitting evidence on AI software. Given the time available, the diagnostic component of the model was the primary focus of the health economics aspect of the report. Priority was given to the following considerations: input parameters to populate the model – including consideration of the type of evidence required, sources available and gaps in the evidence relevant outcome measures to compare the cost-effectiveness and clinical effectiveness of AI software in the detection of lung cancer identification of potential value drivers of the model – with recommendations of how these can be measured for inclusion in a cost-effectiveness model. Results Test accuracy, practical implications and clinical effectiveness No studies met the inclusion criteria of the review. Two ongoing studies with unclear eligibility were identified. In the absence of available evidence, we summarised data from six studies that had unclear populations but included a comparison of CXRs read by readers with and without the use of commercial AI software. Statistical comparisons were not undertaken in most of the studies, but there was some evidence that sensitivity might be higher among specialist radiologist with AI than among specialist radiologist without AI. This finding was not consistent between studies, however. No significant differences were observed for specificity, positive predictive value or number of cancers detected. None of the studies provided evidence on the clinical effectiveness of adjunct AI software. The summarised studies were small retrospective studies with important methodological limitations, and their generalisability to the UK population is unclear. Conclusions There is currently no evidence applicable to this review on the use of adjunct AI software for the detection of suspected lung cancer on CXRs in either people referred from primary care with symptoms of lung cancer or people referred from primary care for other reasons. Implications for service provision Lung cancer pathways are complex and contain many routes to diagnosis. Although national guidance and timelines for diagnosis exist, practice varies widely throughout radiology departments and lung cancer teams both within and across NHS trusts. With many ways to achieve these targets, changes in any area of the diagnostic pathway may have a significant impact elsewhere. There is some evidence for the impact of CXR results on the diagnostic pathway when performed without AI assistance, as is current practice. This is limited, and it is difficult to compare results because of the different study designs used and the different outcomes reported. There is no published evidence to link measures of progression through the diagnostic pathways with long-term outcomes such as stage at diagnosis and survival. There is currently no applicable evidence to show the impact of the addition of AI software to CXR review on the diagnosis of lung cancer. There may be multiple ways AI software could change measures along this pathway. These could include improved accuracy of lung cancer detection, directing patients along the quickest pathway to diagnosis, quicker report turnaround time to achieve earlier confirmatory testing, or prioritisation of cases for review including those without lung cancer who can be discharged more quickly and free up staff time and resources. AI software may also have a negative impact on pathways by increasing the number of benign lung nodules detected and increasing the number of patients who undergo a computed tomography scan that they might not have needed. This would be detrimental to the patients, with increased exposure to radiation and anxiety due to a positive CXR result, and has cost and resource use implications for the department. With a lack of evidence on AI software, the impact on service provision is unknown and may have significant implications in terms of progression through diagnostic pathways, resource use, costs and patient outcomes. Study registration This study is registered as PROSPERO CRD42023384164. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135755) and is published in full in Health Technology Assessment; Vol. 28, No. 50. See the NIHR Funding and Awards website for further award information.

Background Lynch syndrome is an inherited condition which leads to an increased risk of colorectal, endometrial and ovarian cancer. Risk-reducing surgery is generally recommended to manage the risk of gynaecological cancer once childbearing is completed. The value of gynaecological colonoscopic surveillance as an interim measure or instead of risk-reducing surgery is uncertain. We aimed to determine whether gynaecological surveillance was effective and cost-effective in Lynch syndrome. Methods We conducted systematic reviews of the effectiveness and cost-effectiveness of gynaecological cancer surveillance in Lynch syndrome, as well as a systematic review of health utility values relating to cancer and gynaecological risk reduction. Study identification included bibliographic database searching and citation chasing (searches updated 3 August 2021). Screening and assessment of eligibility for inclusion were conducted by independent researchers. Outcomes were prespecified and were informed by clinical experts and patient involvement. Data extraction and quality appraisal were conducted and results were synthesised narratively. We also developed a whole-disease economic model for Lynch syndrome using discrete event simulation methodology, including natural history components for colorectal, endometrial and ovarian cancer, and we used this model to conduct a cost–utility analysis of gynaecological risk management strategies, including surveillance, risk-reducing surgery and doing nothing. Results We found 30 studies in the review of clinical effectiveness, of which 20 were non-comparative (single-arm) studies. There were no high-quality studies providing precise outcome estimates at low risk of bias. There is some evidence that mortality rate is higher for surveillance than for risk-reducing surgery but mortality is also higher for no surveillance than for surveillance. Some asymptomatic cancers were detected through surveillance but some cancers were also missed. There was a wide range of pain experiences, including some individuals feeling no pain and some feeling severe pain. The use of pain relief (e.g. ibuprofen) was common, and some women underwent general anaesthetic for surveillance. Existing economic evaluations clearly found that risk-reducing surgery leads to the best lifetime health (measured using quality-adjusted life-years) and is cost-effective, while surveillance is not cost-effective in comparison. Our economic evaluation found that a strategy of surveillance alone or offering surveillance and risk-reducing surgery was cost-effective, except for path_PMS2 Lynch syndrome. Offering only risk-reducing surgery was less effective than offering surveillance with or without surgery. Limitations Firm conclusions about clinical effectiveness could not be reached because of the lack of high-quality research. We did not assume that women would immediately take up risk-reducing surgery if offered, and it is possible that risk-reducing surgery would be more effective and cost-effective if it was taken up when offered. Conclusions There is insufficient evidence to recommend for or against gynaecological cancer surveillance in Lynch syndrome on clinical grounds, but modelling suggests that surveillance could be cost-effective. Further research is needed but it must be rigorously designed and well reported to be of benefit. Study registration This study is registered as PROSPERO CRD42020171098. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR129713) and is published in full in Health Technology Assessment; Vol. 28, No. 41. See the NIHR Funding and Awards website for further award information. Plain language summary Lynch syndrome is an inherited condition which puts people at a higher risk of getting bowel cancer, womb cancer and ovarian cancer. Although people with Lynch syndrome are more likely to get these cancers, they are more likely to survive cancer if they get it. People diagnosed with Lynch syndrome get regular testing (surveillance) using a camera to check for bowel cancer or polyps. For womb and ovarian cancer, surveillance may also be an option, but it is less well studied in these cancers. This means that many women are not offered surveillance. Women with Lynch syndrome are recommended to have risk-reducing surgery when their risk starts rising, if they do not want any more children. We wanted to find out whether surveillance for womb and ovarian cancer would work and would be good value for money. Doctors and patients have said that these are important research questions. We searched for published research on this subject and found a lot of studies, but these studies were often small or not well designed, so they could only tell us a limited amount. Studies did not always measure the things that patients want to know. There was some evidence that people having surveillance might live longer than people not having surveillance, but there was also some evidence that risk-reducing surgery is better than surveillance. Surveillance has detected some cancers which had no symptoms, but there are also cancers diagnosed soon after a surveillance visit where nothing was found. People often find surveillance painful, but experiences vary. Our work shows that surveillance and surgery could be good value for money for many women with Lynch syndrome. We need better research to help patients and doctors decide whether surveillance is right for them. Scientific summary Background Lynch syndrome is an inherited cancer predisposition syndrome, which leads to an increased lifetime risk for colorectal, endometrial and ovarian cancers. These cancers are typically observed at younger ages in people with Lynch syndrome than in the general population. Cancer risks depend somewhat on which deoxyribonucleic acid (DNA) mismatch repair gene is affected in the patient, with path_MLH1 and path_MSH2 genotypes generally having the highest penetrance, path_MSH6 having high penetrance for endometrial cancer but lower penetrance for colorectal and ovarian cancer, and path_PMS2 having lower penetrance still. Survival from colorectal and ovarian cancer among patients with Lynch syndrome tends to be better than survival among unselected patients, and this may also be true for endometrial cancer, but survival from endometrial cancer is already more favourable. When Lynch syndrome is diagnosed, measures are put in place to manage cancer risks. These measures typically include biennial colonoscopic surveillance from 25 years of age and the offer of risk-reducing gynaecological surgery (hysterectomy with bilateral salpingo-oophorectomy) after completion of childbearing and before the individual faces a significant risk of gynaecological cancer (surgery is generally recommended from 35 years). In addition, some patients may use aspirin as chemoprophylaxis and may have surveillance of other organs besides the colon. Gynaecological cancer surveillance is contentious. It is perceived that there is a lack of evidence to support widespread adoption of colonoscopic surveillance but also that there is insufficient evidence that it is ineffective so should not be offered to patients. Many women with Lynch syndrome do want gynaecological surveillance, and some resort to private healthcare if it is not provided by their local NHS hospital. Some women may wish to receive colonoscopic surveillance for a time before opting for risk-reducing surgery when they are older; some women may not ever want to undergo risk-reducing surgery, and some may be unsuitable for surgery. Research aims We aimed to determine whether gynaecological surveillance was effective and/or cost-effective in Lynch syndrome. Our objectives were to conduct systematic reviews of clinical effectiveness and cost-effectiveness evidence, as well as a systematic review of health state utility values, and to develop a whole-disease economic model for Lynch syndrome and use it to conduct a cost-effectiveness analysis. Systematic review of clinical effectiveness evidence Methods We conducted a systematic review in line with a preregistered protocol (PROSPERO CRD42020171098). Our study identification methods included bibliographical database searches, citation chasing and hand screening of conference proceedings and clinical trials registries. Searches were updated to 3 August 2021. Study selection was conducted independently by two reviewers. A broad range of outcomes were determined a priori and a broad range of study designs were considered eligible for inclusion, including non-comparative observational studies (e.g. cross-sectional studies and case series). Risk of bias was assessed using one or more of three checklists, according to the study design. Narrative synthesis was performed, supported by cross-tabulation. Studies were too methodologically heterogeneous and insufficiently numerous to justify quantitative synthesis (i.e. meta-analysis). Results A total of 30 studies were included in the review, of which 20 were single-arm studies. Five studies compared colonoscopic surveillance with risk-reducing surgery, three compared time periods with different surveillance approaches and two compared surveillance with no intervention. There was a high likelihood of overlap between some studies. No studies provided high-quality evidence that is precise and at low risk of bias. The most significant contribution to the risk of bias in studies was failure to adequately consider and address confounding factors. Some studies adopted a diagnostic accuracy evaluation design and were generally reported too poorly to enable good assessment of the risk of bias or were found to be at high risk of bias. Mortality and survival Some evidence suggests that all-cause mortality rates are lower with surgery than with surveillance, and lower with surveillance than with no intervention. Some evidence suggests that endometrial cancer-specific mortality is lower with surveillance than with no intervention, but lower still with risk-reducing surgery. Endometrial cancer survival for cancers detected by surveillance was not significantly different from survival for occult cancers diagnosed upon risk-reducing surgery. A similar but even weaker pattern was observed for ovarian cancer. Stage at diagnosis Data were generally too sparse to be meaningful, but there was some evidence of ovarian cancers being diagnosed in earlier stages with surveillance than without surveillance. Fertility One study found that 5 of 41 participants in a surveillance programme gave birth over a 10-year period. Another study found that participants had concerns that hysteroscopy could lead to fertility issues (due to infection risk). Cancer detection Detection rates of endometrial cancer, ovarian cancer and premalignancies were low, with zero events in some studies. This and the lower number of comparative studies make it difficult to draw conclusions about the effect of surveillance on cancer detection rates. Symptomatic and asymptomatic cancers The proportion of cancers that were symptomatic detected during surveillance was extremely heterogeneous across the studies, with some studies reporting that all cancers were asymptomatic and others reporting that all were symptomatic. Only two studies had a mixture of symptomatic and asymptomatic cancers. It is clear that some cancers judged to be asymptomatic (at least by clinical researchers) can be detected by surveillance, albeit in small numbers. Interval and missed cancers Cancers detected due to symptoms soon after a negative surveillance visit (interval cancers) and occult cancers detected following risk-reducing surgery soon after a negative surveillance visit (missed cancers) were reported in a number of studies. Although numbers were generally low, it is clear that surveillance does not detect all cancers and that cancers can arise very soon after a negative surveillance visit. Test accuracy and test failures Five studies attempted to evaluate the accuracy of surveillance tests. False positive results were recorded for pelvic ultrasound and false negative results were recorded for hysteroscopy. Test failure rates were rarely reported, but did reach as high as 24% and 26% for endometrial biopsy and hysteroscopy in one study, while the failure rate was 4% for transvaginal ultrasound. Harms of surveillance One study found no uterine perforations among 69 hysteroscopies with endometrial sampling. Numerous studies measured pain and some also measured use of pain relief. On average, endometrial biopsy was reported as moderately painful, although some individuals experienced no pain and others experienced severe pain. One study found that transvaginal ultrasound was less painful than endometrial biopsy or hysteroscopy. Use of non-steroidal anti-inflammatories for pain relief was common. Around one in seven women in one survey had undergone general anaesthetic for surveillance (rising to around one in four when restricted to those receiving hysteroscopy). Factors that may affect adverse events Some evidence suggested that endometrial biopsy is more painful in postmenopausal women, but one study did not find this to be the case. Pain ratings for endometrial biopsy were higher for nulliparous participants compared with parous participants. Systematic review of cost-effectiveness evidence Methods Our study identification methods included bibliographical database searches and citation chasing. Study selection was conducted independently by two reviewers. Data extraction and quality appraisal were conducted and included the use of a set of bespoke quality appraisal questions. Narrative synthesis was conducted, supported by cross-tabulation. Results Three cost-effectiveness analyses were identified. All three studies were based on relatively simple decision analytical models. All studies included at least one surveillance arm and one risk-reducing surgery arm. Two studies included a ‘no intervention’ arm. Risk-reducing surgery was economically dominant (less costly and more effective than alternatives) in two studies and was highly cost-effective in the other. Surveillance was dominated by risk-reducing surgery in all analyses. If risk-reducing surgery strategies were removed, one study would find surveillance cost-effective versus no intervention, while the other study would find it not cost-effective (producing health benefits but at too great a cost). Systematic review of utility values Methods We sought utility values relating to endometrial cancer, ovarian cancer, gynaecological cancer surveillance and risk-reducing gynaecological surgery. We did not restrict the population to people with Lynch syndrome. In expectation that there would be insufficient data on risk-reducing gynaecological surgery, we also sought utility values relating to gynaecological surgery for benign gynaecological conditions. Our study identification methods included bibliographical database searches and citation chasing. Study selection was conducted independently by two reviewers. Data extraction and quality appraisal were conducted. Narrative synthesis was conducted, supported by cross-tabulation. Results Fifty-eight studies were identified, with more than half relating to ovarian cancer. Only four studies related to gynaecological surveillance and only two studies related to risk-reducing surgery for Lynch syndrome. The studies relating to surveillance and risk-reducing surgery asked participants to value hypothetical disease states, while most of the other studies asked patients to describe or value their own health. Utility values tended to be lower for more advanced endometrial or ovarian cancer. The studies reporting utility values for gynaecological surveillance were either methodologically flawed or reported minimal detail of their methods. The studies reporting utility values for risk-reducing surgery in Lynch syndrome were similarly methodologically flawed. For benign gynaecological conditions, utility generally drops sharply following surgery for a recovery period, and eventually reaches a level higher than preoperative utility. This finding is not expected to be replicated in risk-reducing surgery, but some studies may be a useful proxy for the utility of risk-reducing surgery, particularly if they include premenopausal bilateral oophorectomy. Model-based economic evaluation Methods We developed a whole-disease model using a discrete event simulation methodology. The model included natural history components for colorectal, endometrial and ovarian cancers that were calibrated to aggregate data from published studies, including the Prospective Lynch Syndrome Database. Clinical parameters (e.g. cancer survival) were estimated, where possible, from studies of Lynch syndrome populations. We used the model to conduct a cost–utility analysis of risk-reducing strategies for gynaecological cancer, including surveillance and risk-reducing surgery, and comparing these with a no intervention strategy. The economic evaluation was generally conducted in line with the National Institute for Health and Care Excellence (NICE) reference case and quality-adjusted life-years (QALYs) were the measure of health benefit. Results Risk reduction strategies are predicted to be cost-effective compared with no intervention, except for path_PMS2 Lynch syndrome. For other genotypes, surveillance (alone or with risk-reducing surgery also offered) is expected to lead to more QALYs (and greater costs) than only offering risk-reducing surgery and to be cost-effective. For path_PMS2, risk-reducing surgery and surveillance led to significant cost increases, since there were minimal changes to cancer outcomes to offset these costs. Value of information calculations suggest that further research to obtain more precise parameter estimates would be very valuable. Further value of information analyses may help to prioritise research. Conclusions Clinical effectiveness evidence for gynaecological cancer surveillance in Lynch syndrome is sparse and methodologically limited. There is some evidence that surveillance can prevent some deaths compared with no intervention, but there is also evidence that risk-reducing surgery prevents more deaths. Some asymptomatic cancers are detected by surveillance, but some cancers are also missed. Recipients of surveillance have a wide range of pain experiences. While existing publications have concluded that risk-reducing surgery is clearly cost-effective (generally leading to a substantial gain in QALYs while lowering or only slightly increasing costs) and that surveillance alone is not cost-effective if risk-reducing surgery is an option, we have found that surveillance can be a cost-effective way to manage the risk of gynaecological cancer. Further research is needed to reduce the uncertainty in model parameters, both to determine cost-effectiveness and to provide context to patients about the potential clinical value of risk-reducing strategies. Implications for health care People with Lynch syndrome should be informed that gynaecological cancer surveillance is not expected to reduce the risk of gynaecological cancer and cancer death to the same extent as risk-reducing surgery. There is some evidence that surveillance could be beneficial compared with no risk reduction (e.g. some asymptomatic cancers detected), but there is also evidence that some cancers are missed and that some individuals find surveillance severely painful. The prognosis from endometrial and ovarian cancer appears to be better for people with Lynch syndrome than for unselected patients. Gynaecological cancer surveillance is estimated to cost the NHS over £300 per year per patient, while risk-reducing surgery is estimated to cost over £6000. Recommendations for research Researchers should consult with biostatisticians or epidemiologists or other methodological experts before conducting trials and publishing further in this area – the quality of current research falls below the level needed to inform decision-making. More in-depth value of information analyses should be conducted to identify which parameters or groups of parameters are most critical to research further. Health utilities should be directly elicited from individuals with Lynch syndrome to identify the potential effects of surveillance and risk-reducing surgery on health-related quality of life and QALYs; relatedly, it may be beneficial to consider whether willingness to pay is a better indication of the value of undergoing or avoiding surveillance. Study registration This study is registered as PROSPERO CRD42020171098. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR129713) and is published in full in Health Technology Assessment; Vol. 28, No. 41. See the NIHR Funding and Awards website for further award information.

Background Transurethral resection of bladder tumour has been the mainstay of bladder cancer staging for > 60 years. Staging inaccuracies are commonplace, leading to delayed treatment of muscle-invasive bladder cancer. Multiparametric magnetic resonance imaging offers rapid, accurate and non-invasive staging of muscle-invasive bladder cancer, potentially reducing delays to radical treatment. Objectives To assess the feasibility and efficacy of the introducing multiparametric magnetic resonance imaging ahead of transurethral resection of bladder tumour in the staging of suspected muscle-invasive bladder cancer. Design Open-label, multistage randomised controlled study in three parts: feasibility, intermediate and final clinical stages. The COVID pandemic prevented completion of the final stage. Setting Fifteen UK hospitals. Participants Newly diagnosed bladder cancer patients of age ≥ 18 years. Interventions Participants were randomised to Pathway 1 or 2 following visual assessment of the suspicion of non-muscle-invasive bladder cancer or muscle-invasive bladder cancer at the time of outpatient cystoscopy, based upon a 5-point Likert scale: Likert 1–2 tumours considered probable non-muscle-invasive bladder cancer; Likert 3–5 possible muscle-invasive bladder cancer. In Pathway 1, all participants underwent transurethral resection of bladder tumour. In Pathway 2, probable non-muscle-invasive bladder cancer participants underwent transurethral resection of bladder tumour, and possible muscle-invasive bladder cancer participants underwent initial multiparametric magnetic resonance imaging. Subsequent therapy was determined by the treating team and could include transurethral resection of bladder tumour. Main outcome measures Feasibility stage: proportion with possible muscle-invasive bladder cancer randomised to Pathway 2 which correctly followed the protocol. Intermediate stage: time to correct treatment for muscle-invasive bladder cancer. Results Between 31 May 2018 and 31 December 2021, of 638 patients approached, 143 participants were randomised; 52.1% were deemed as possible muscle-invasive bladder cancer and 47.9% probable non-muscle-invasive bladder cancer. Feasibility stage: 36/39 [92% (95% confidence interval 79 to 98%)] muscle-invasive bladder cancer participants followed the correct treatment by pathway. Intermediate stage: median time to correct treatment was 98 (95% confidence interval 72 to 125) days for Pathway 1 versus 53 (95% confidence interval 20 to 89) days for Pathway 2 [hazard ratio 2.9 (95% confidence interval 1.0 to 8.1)], p = 0.040. Median time to correct treatment for all participants was 37 days for Pathway 1 and 25 days for Pathway 2 [hazard ratio 1.4 (95% confidence interval 0.9 to 2.0)]. Limitations For participants who underwent chemotherapy, radiotherapy or palliation for multiparametric magnetic resonance imaging-diagnosed stage T2 or higher disease, it was impossible to conclusively know whether these were correct treatments due to the absence of histopathologically confirmed muscle invasion, this being confirmed radiologically in these cases. All patients had histological confirmation of their cancers. Due to the COVID-19 pandemic, we were unable to realise the final stage. Conclusion The multiparametric magnetic resonance imaging-directed pathway led to a substantial 45-day reduction in time to correct treatment for muscle-invasive bladder cancer, without detriment to non-muscle-invasive bladder cancer participants. Consideration should be given to the incorporation of multiparametric magnetic resonance imaging ahead of transurethral resection of bladder tumour into the standard pathway for all patients with suspected muscle-invasive bladder cancer. The improved decision-making accelerated time to treatment, even though many patients subsequently needed transurethral resection of bladder tumour. A proportion of patients can avoid transurethral resection of bladder tumour completely, reducing costs and morbidity, given the much lower cost of magnetic resonance imaging and biopsy compared to transurethral resection of bladder tumour. Future work Further work to cross-correlate with the recently developed Vesical Imaging-Reporting and Data System will improve accuracy and aid dissemination. Longer follow-up to examine the effect of the pathway on outcomes is also required. Incorporation of liquid deoxyribonucleic acid-based biomarkers may further improve the quality of decision-making and should also be investigated further. Study registration This study is registered as ISRCTN 35296862. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 14/08/60) and is published in full in Health Technology Assessment; Vol. 28, No. 42. See the NIHR Funding and Awards website for further award information. Plain language summary The BladderPath trial explored how to accelerate diagnosis and avoid unnecessary surgery for patients with bladder cancer which had grown into the muscle wall of the bladder, referred to as muscle-invasive bladder cancer. Following initial outpatient diagnosis, bladder cancer patients currently undergo inpatient or day-case surgical tumour removal using a telescope (transurethral resection of bladder tumour). This surgery is fundamental to the treatment of early bladder cancer (non-muscle-invasive). However, for muscle-invasive disease, the main role of transurethral resection of bladder tumour is to confirm that the tumour has grown into the bladder muscle, and this is often inaccurate; the actual correct treatment for muscle-invasive bladder cancer patients should include chemotherapy, radiotherapy and/or bladder removal. For these patients, having transurethral resection of bladder tumour may delay this correct treatment and impact survival. Additionally, for patients determined to need palliative care due to advanced disease, the transurethral resection of bladder tumour may represent over-treatment. A magnetic resonance imaging scan with contrast agent (called multiparametric magnetic resonance imaging) gives a clearer picture of the bladder than normal scans, allowing distinction between invasive and non-invasive tumours. The BladderPath trial investigated adding multiparametric magnetic resonance imaging for patients with suspected muscle-invasive bladder cancer and the effect on treatment times. Subsequent therapy could include transurethral resection of bladder tumour if clinically determined as necessary by the treating team. Trial participants were randomly allocated either to the standard pathway (Pathway 1: all underwent transurethral resection of bladder tumour) or to a new pathway (Pathway 2). In Pathway 2, urologists conducting the initial outpatient diagnostic bladder inspections used a scale to assess whether tumours appeared to be either probably non-muscle-invasive or possibly muscle-invasive. Participants whose tumours appeared possibly muscle-invasive had initial multiparametric magnetic resonance imaging as their next investigation instead of transurethral resection of bladder tumour. We then compared the duration of time from initial diagnosis to receiving the correct treatment for participants in each pathway. Of the 143 participants, 75 (52.1%) were diagnosed as possibly muscle invasive. In Pathway 1, the duration for half of the participants in the group to have received their correct treatment for muscle-invasive bladder cancer was 98 days, which reduced to 53 days in Pathway 2. Furthermore, the duration for half of all the participants in the two groups to have received their correct treatment was 37 days for Pathway 1 and 31 days for Pathway 2. In summary, use of initial multiparametric magnetic resonance imaging in suspected muscle-invasive bladder cancer participants substantially reduced the time to correct treatment (surgery, radiotherapy, chemotherapy or instigation of palliative care) and avoided unnecessary surgery. There was no negative impact on participants with non-invasive disease. Adopting multiparametric magnetic resonance imaging into the pathway ahead of transurethral resection of bladder tumour for patients with suspected muscle-invasive bladder cancer is recommended. Scientific summary Background Bladder cancer (BC) is the fifth most common cancer in Western society. Standard management follows a pathway established > 60 years ago with the first description of transurethral resection of bladder tumour (TURBT), and prognosis has not improved for 30 years. Following visual diagnosis by outpatient flexible cystoscopy, TURBT is the subsequent diagnostic and staging tool for all patients. While TURBT is mostly well-tolerated and therapeutic for non-muscle-invasive BC (NMIBC), its role in muscle-invasive BC (MIBC) is predominantly diagnostic. Furthermore, for MIBC patients, initial TURBT often under-stages invasion (up to 30% of MIBCs are initially staged as high-grade NMIBC at first TURBT) and may contribute to extravesical tumour dissemination as a result of the piecemeal resection process. Subsequently, accurate staging by cross-sectional pelvic imaging post TURBT is impaired by post-surgical artefacts. Moreover, internationally, TURBT followed by histopathological review and multidisciplinary team (MDT) decision-making typically adds a number of weeks to the pathway, creating a delay in commencing correct radical treatment for MIBC patients and potentially worse outcomes. Thus, an ideal pathway would separate NMIBC patients from MIBC patients at the time of diagnosis by the faster and more accurate application of established technologies to expedite therapy, potentially improving outcomes. Imaging advances suggesting multiparametric (mp) magnetic resonance imaging (MRI) may allow the accurate discrimination of NMIBC from MIBC, theoretically offering a safer and faster route to radical treatment than TURBT. To test the hypothesis whether MIBC patients can be safely expedited to radical treatment by using initial mpMRI for local staging rather than TURBT, we undertook the BladderPath randomised controlled trial [NHS Research Ethics Committee (REC) approval 17/LO/1819, ISRCTN 35296862]. Objectives To assess the feasibility and efficacy of the substitution of TURBT with mpMRI in the staging of patients with suspected MIBC, hypothesising that image-directed (mpMRI) staging would shorten the time period to correct treatment for MIBC patients compared to the standard TURBT-based pathway. Methods BladderPath is a randomised trial comparing risk-stratified image-directed (mpMRI) care with TURBT for patients with newly diagnosed BC. Patients with symptoms suspicious of a new diagnosis of BC were identified via haematuria clinics, and they provided written informed consent for study participation. Ineligible patients were those unable or unwilling to undergo MRI, those with a previous BC diagnosis and those who had previously entered the study. Participants with possible MIBC (Likert 3–5 as visually assessed on a 5-point Likert scale at flexible cystoscopy) were randomised to standard TURBT assessment (Pathway 1) or mpMRI-based assessment (Pathway 2) with flexible cystoscopy tumour biopsy Pathway 2 probable NMIBC (Likert 1–2) participants underwent TURBT. Primary outcomes: Feasibility phase – proportion of Pathway 2 possible MIBC participants who correctly followed protocol (target: 80%); intermediate stage – time to first correct treatment (chemotherapy, radiotherapy, surgery, decision for palliative care) for participants with confirmed MIBC (target: 30-day improvement) and as time to TURBT or palliative care for NMIBC. Randomisation was achieved by using a computerised allocation program; stratification variables included participants’ sex, age and clinician’s initial visual assessment of muscle invasiveness of the tumour. Blinding of participants, caregivers and outcome assessors was not possible. Results Between 31 May 2018 and 31 December 2021, recruitment took place in 15 UK urology centres; 638 patients were screened as potentially eligible, of which 309 were registered and 143 were randomised (72 to Pathway 1, 71 to Pathway 2). The 166 registered patients not randomised were not found to have BC during initial cystoscopy. Three participants were subsequently found to be ineligible post randomisation (one in Pathway 1, two in Pathway 2). Seven participants withdrew from the study (three in Pathway 1, four in Pathway 2), including three participants who were confirmed as not having cancer. Nine protocol deviations were reported by nine participants (five in Pathway 1, four in Pathway 2). The primary outcome for the feasibility stage was the proportion of possible MIBC participants randomised to Pathway 2 who correctly followed the pathway protocol. In total, 36 of the 39 [92%; 95% confidence interval (CI) 79% to 98%] possible MIBC participants in Pathway 2 underwent mpMRI as per protocol. Three Pathway 2 possible MIBC participants did not undergo mpMRI post randomisation: one participant had metal in their eye, one patient withdrew (29 days post randomisation) and one underwent MRI prior to trial entry (scan was requested independently of the study). Of the 36 participants who underwent mpMRI, 17 were diagnosed as MIBC, 16 as NMIBC and 3 were inconclusive. The secondary outcome for the feasibility stage was the overall proportion of randomised participants who correctly followed the protocol in each pathway. For Pathway 1, this was defined as the number of probable NMIBC and possible MIBC participants randomly accrued who underwent TURBT at an appropriate stage, as a proportion of all participants randomised to that pathway. For Pathway 2, it was defined as the number of probable NMIBC participants who underwent TURBT plus the number of possible MIBC participants who underwent mpMRI, divided by all randomised to Pathway 2. The overall proportion of participants who correctly followed their respective protocol pathway was 96% (95% CI 88% to 99%) in each pathway. There was no statistical difference between the pathways. For the Intermediate stage, the primary outcome was time to correct treatment (TTCT) for participants who were initially classified as possible MIBC and were then confirmed to have MIBC (by TURBT or mpMRI). For the 25 participants who were initially classified as possible MIBC and were then confirmed as MIBC (14 in Pathway 1; 11 in Pathway 2), 24 participants received a correct treatment (the remaining patients died 81 days post randomisation, before a correct treatment; date last seen is used in the time-to-event analysis). Median TTCT for all participants who were initially classified as possible MIBC and were then confirmed to have MIBC (N = 25) was 77 days (95% CI 54 to 98). Median TTCT for Pathway 1 (N = 14) was 98 days (95% CI 72 to 125). Median TTCT for Pathway 2 (N = 11) was 53 days (95% CI 20 to 81). The p-value of 0.0201 suggests a statistical difference in TTCT between the pathways. A Cox model adjusting for the stratification factors of sex and age, with study centre included as a random effect, showed that the hazard ratio (HR) of an event for Pathway 2 versus Pathway 1 was 2.9 (95% CI 1.0 to 8.1, p = 0.04). An event in this model indicates a patient receiving a correct treatment; therefore, the HR of 2.9 indicates that participants in Pathway 2 received correct treatment 2.9 times quicker than those in Pathway 1. To assess the secondary outcome of TTCT for probable NMIBC participants confirmed as NMIBC, there were 58 participants initially classified as probable NMIBC and then confirmed as NMIBC (28 in Pathway 1 and 30 in Pathway 2), all of whom received correct treatment of TURBT. Median TTCT for probable NMIBC participants confirmed as NMIBC (N = 58) was 16 days (95% CI 11 to 23); median TTCT for Pathway 1 (N = 28) was 14 days (95% CI 10 to 29) and 17 days (95% CI 8 to 25) for Pathway 2 (N = 25), log-rank p = 0.6677. A Cox model adjusting for the stratification factors of sex and age showed that the HR for Pathway 2 versus Pathway 1 was 0.8 (95% CI 0.5 to 1.5). For the secondary outcome of TTCT for all randomised participants, 131 of 143 randomised participants had received a correct treatment (72 in Pathway 1 and 71 in Pathway 2); participants who had not received a correct treatment were censored at their date last seen and included in the time-to-treatment analysis. Median TTCT for all randomised participants (N = 143) was 31 days (95% CI 22 to 37); median TTCT for Pathway 1 (N = 72) was 37 days (95% CI 23 to 47) and 25 days (95% CI 18 to 35) for Pathway 2 (N = 71), log-rank p = 0.0295. A Cox model adjusting for the stratification factors of sex and age showed that the HR for Pathway 2 versus Pathway 1 was 1.4 (95% CI 0.9 to 2.0). To assess the secondary outcome of time to definitive treatment (TTDT) for all randomised participants, 137 randomised participants had received definitive treatment (6 participants did not receive definitive treatment and their date last seen was used in the time-to-event analysis). Median TTDT for all randomised participants (N = 143) was 23 days (95% CI 20 to 29); median TTDT for Pathway 1 (N = 72) was 23 days (95% CI 17 to 29) and for Pathway 2 (N = 71) was 22 days (95% CI 17 to 32), log-rank p-value of 0.9619. A Cox model adjusting for the stratification factors of sex and age showed that the HR for Pathway 2 versus Pathway 1 was 0.9 (95% CI 0.6 to 1.2). Clinical analysis Delays in administering the correct treatment for MIBC patients after initial urological consultation and disease diagnosis are internationally widespread [Russell B, Liedberg F, Khan MS, Nair R, Thurairaja R, Malde S, et al. A systematic review and meta-analysis of delay in radical cystectomy and the effect on survival in bladder cancer patients. Eur Urol Oncol 2020;3(2):239–49]. Prolonged delays contribute to poor prognosis, and so attempts to improve and refine the diagnostic and treatment pathways for BC patients are of international importance and a priority for patients and healthcare professionals alike [Russell et al. 2020; Bessa A, Maclennan S, Enting D, Bryan R, Josephs D, Hughes S, et al. Consensus in bladder cancer research priorities between patients and healthcare professionals using a four-stage modified Delphi method. Eur Urol 2019;76(2):258–9]. Although first described over 60 years ago, the piecemeal resection of bladder tumour(s), TURBT, remains the initial diagnostic and staging tool for all patients. The shortcomings of TURBT are well-reported [Bessa et al. 2019; Del Giudice F, Flammia RS, Pecoraro M, Moschini M, D’Andrea D, Messina E, et al. The accuracy of Vesical Imaging-Reporting and Data System (VI-RADS): an updated comprehensive multi-institutional, multi-readers systematic review and meta-analysis from diagnostic evidence into future clinical recommendations. World J Urol 2022;40(7):1617–28; Wallace DM, Bryan RT, Dunn JA, Begum G, Bathers S; West Midlands Urological Research Group. Delay and survival in bladder cancer. BJU Int 2002;89(9):868–78; Bryan RT, Collins SI, Daykin MC, Zeegers MP, Cheng KK, Wallace DMA, et al. Mechanisms of recurrence of Ta/T1 bladder cancer. Ann R Coll Surg Engl 2010;92(6):519–24], all of which may delay the correct radical treatment for MIBC patients or lead to incorrect therapy choices. Over the course of the last decade, data suggest that mpMRI may allow the accurate discrimination of NMIBC and MIBC, and so potentially offering a safer and faster route to radical treatment than TURBT (Panebianco et al. 2018; Del Giudice et al. 2022). We have shown that it is feasible to introduce mpMRI for initial staging into the pathway for those patients visually diagnosed with possible MIBC at outpatient diagnostic flexible cystoscopy. Moreover, we have demonstrated that by doing so, possible MIBC patients receive their correct therapy significantly quicker – 45 days quicker, even if some of these MIBC patients still require TURBT either to resolve diagnostic uncertainty or as part of their planned care (e.g. to debulk tumour prior to radiotherapy). Although the relationship between delay and survival in BC is complex (Wallace et al. 2002), it is reasonable to contemplate that administering correct treatment to MIBC patients more than 6 weeks earlier than the current standard of care can only be beneficial. Several studies report adverse outcomes associated with delays of over 3 months between bladder cancer diagnosis and radical cystectomy (Russell et al. 2020); the mpMRI-guided BladderPath pathway (Pathway 2) undercut this TTCT by a considerable margin (median TTCT 53 days), whereas the standard pathway did not (median TTCT 98 days). Unfortunately, with substantial interruptions to recruitment due to the COVID-19 pandemic, we have been unable to recruit sufficient patients to evaluate our a priori survival outcomes. Further limitations to the study are that, for patients who underwent systemic chemotherapy, radiotherapy or palliation for mpMRI-diagnosed MIBC, it is impossible to conclusively know whether these were ‘correct’ treatments in the sense that staging was radiological not pathological, that is by histological confirmation of muscle invasion. This is however the norm for staging of most cancers. All patients had histological confirmation of their cancers and all treatments were approved via the relevant MDT. An important component of this new pathway is the ability of urologists to accurately triage patients as probable NMIBC or possible MIBC at the time of outpatient diagnostic flexible cystoscopy based upon the macroscopic appearances of suspicious bladder lesions. Building upon previous evidence (Bryan et al. 2010), we have shown that 89% of visually diagnosed probable NMIBCs were pathologically confirmed as NMIBCs, demonstrating that urologists can reliably identify such tumours. Hence, the simple patient pathway change suggested by the BladderPath data described here is universally applicable and is easy to implement. Conclusions The mpMRI-directed pathway led to a substantial reduction in TTCT for MIBC participants without detriment to the TTCT for NMIBC participants. Consideration should be given to the incorporation of mpMRI ahead of TURBT in the standard pathway for all patients with suspected MIBC. A proportion of patients were able to avoid TURBT completely and the improved decision-making accelerated time to treatment, even though many patients subsequently needed TURBT as part of their treatment plan. Trial registration This trial is registered as ISRCTN 35296862. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 14/08/60) and is published in full in Health Technology Assessment; Vol. 28, No. 42. See the NIHR Funding and Awards website for further award information.

Background Low levels of testosterone cause male hypogonadism, which is associated with sexual dysfunction, tiredness and reduced muscle strength and quality of life. Testosterone replacement therapy is commonly used for ameliorating symptoms of male hypogonadism, but there is uncertainty about the magnitude of its effects and its cardiovascular and cerebrovascular safety. Aims of the research The primary aim was to evaluate the safety of testosterone replacement therapy. We also assessed the clinical and cost-effectiveness of testosterone replacement therapy for men with male hypogonadism, and the existing qualitative evidence on men’s experience and acceptability of testosterone replacement therapy. Design Evidence synthesis and individual participant data meta-analysis of effectiveness and safety, qualitative evidence synthesis and model-based cost-utility analysis. Data sources Major electronic databases were searched from 1992 to February 2021 and were restricted to English-language publications. Methods We conducted a systematic review with meta-analysis of individual participant data according to current methodological standards. Evidence was considered from placebo-controlled randomised controlled trials assessing the effects of any formulation of testosterone replacement therapy in men with male hypogonadism. Primary outcomes were mortality and cardiovascular and cerebrovascular events. Data were extracted by one reviewer and cross-checked by a second reviewer. The risk of bias was assessed using the Cochrane Risk of Bias tool. We performed one-stage meta-analyses using the acquired individual participant data and two-stage meta-analyses to integrate the individual participant data with data extracted from eligible studies that did not provide individual participant data. A decision-analytic Markov model was developed to evaluate the cost per quality-adjusted life-years of the use of testosterone replacement therapy in cohorts of patients of different starting ages. Results We identified 35 trials (5601 randomised participants). Of these, 17 trials (3431 participants) provided individual participant data. There were too few deaths to assess mortality. There was no difference between the testosterone replacement therapy group (120/1601, 7.5%) and placebo group (110/1519, 7.2%) in the incidence of cardiovascular and/or cerebrovascular events (13 studies, odds ratio 1.07, 95% confidence interval 0.81 to 1.42; p = 0.62). Testosterone replacement therapy improved quality of life and sexual function in almost all patient subgroups. In the testosterone replacement therapy group, serum testosterone was higher while serum cholesterol, triglycerides, haemoglobin and haematocrit were all lower. We identified several themes from five qualitative studies showing how symptoms of low testosterone affect men’s lives and their experience of treatment. The cost-effectiveness of testosterone replacement therapy was dependent on whether uncertain effects on all-cause mortality were included in the model, and on the approach used to estimate the health state utility increment associated with testosterone replacement therapy, which might have been driven by improvements in symptoms such as sexual dysfunction and low mood. Limitations A meaningful evaluation of mortality was hampered by the limited number of defined events. Definition and reporting of cardiovascular and cerebrovascular events and methods for testosterone measurement varied across trials. Conclusions Our findings do not support a relationship between testosterone replacement therapy and cardiovascular/cerebrovascular events in the short-to-medium term. Testosterone replacement therapy improves sexual function and quality of life without adverse effects on blood pressure, serum lipids or glycaemic markers. Future work Rigorous long-term evidence assessing the safety of testosterone replacement therapy and subgroups most benefiting from treatment is needed. Study registration The study is registered as PROSPERO CRD42018111005. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/68/01) and is published in full in Health Technology Assessment; Vol. 28, No. 43. See the NIHR Funding and Awards website for further award information. Plain language summary Testosterone is a hormone which is vital for sexual activity, bone growth and muscle development in men. Men with low testosterone levels may experience problems with erections and may suffer from brittle bones (osteoporosis), weakness, feeling down (low mood) and tiredness. The manifestations of low testosterone can be treated with testosterone replacement therapy. However, there is current uncertainty about the positive effects of testosterone replacement therapy and its safety. We brought together results from all available medical studies that looked at the use of testosterone replacement therapy in men with low testosterone and contacted the doctors who led these studies to gather further information on their participants. We found 35 studies (5601 participants) conducted in different countries, 17 of which provided additional information on their participants. We did not find any evidence to show that testosterone replacement therapy increases the risk of heart problems, or any evidence to show that some men who take testosterone replacement therapy benefit more than others. Men with low testosterone reported having low mood, poor concentration and lack of energy; however, medical studies often failed to prove that these manifestations improved with testosterone replacement therapy. Most medical studies were conducted among white men in North America using questionnaires designed specifically for them; therefore, the results may not reflect the experiences of men in other countries and from more diverse ethnic backgrounds. There is too much uncertainty about the benefits of testosterone replacement therapy to accurately estimate its value for money for the NHS. We think our findings offer some reassurance to doctors and patients that testosterone replacement therapy does not increase the risk of heart problems. New studies are needed to find out whether some groups of men (such as older or younger men) are more likely to benefit from testosterone replacement therapy more than others. It is also important to develop tools which better reflect the experience of men from a diverse range of social and ethnic backgrounds. To inform men with low testosterone about our findings, we are creating a website with dedicated YouTube video clips. Scientific summary Background and objectives Low levels of the testicular androgen hormone testosterone cause male hypogonadism (MH), which is associated with sexual dysfunction, hot flushes, reduced physical energy, mood disturbance and complications such as osteoporosis. Testosterone (androgen) replacement therapy (TRT) is the standard of treatment for MH. Prescribing rates of TRT in the UK have risen between 2001 and 2010 without an observable increase in the prevalence of hypogonadism, which may reflect increasing media awareness among men about MH. Randomised controlled trials (RCTs) of TRT in men with MH have produced varying and sometimes conflicting results regarding the safety and efficacy of TRT. In particular, there is considerable uncertainty regarding the impact of TRT on cardiovascular (CV) event risk in men with MH, which has led to mandatory safety labelling in some countries. Lack of clarity about the effects of TRT have led to variations in clinical practice among clinicians, which in turn exposes men with MH to inconsistent standards of clinical care and potentially life-threatening healthcare risks. The overarching aim of this project was to conduct a comprehensive and unbiased appraisal of quantitative, qualitative and economic evidence for the use of TRT in symptomatic men with MH. This evidence synthesis had three specific objectives: (1) to perform an individual patient data meta-analysis to estimate the clinical effectiveness and safety of TRT for men with testosterone deficiency syndrome; (2) to synthesise the existing qualitative evidence reporting men’s experience and acceptability of TRT; (3) to develop a decision model to estimate the cost-effectiveness of TRT for the treatment of symptomatic men with testosterone deficiency syndrome. Review methods Methods for the quantitative synthesis and individual participant data meta-analysis We conducted a quantitative synthesis including a meta-analysis of individual participant data (IPD) according to current methodological standards, to identify evidence from placebo-controlled RCTs evaluating the effects of TRT in men with low testosterone. The methods were pre-specified in a research protocol (PROSPERO database registration number: CRD42018111005; www.crd.york.ac.uk/prospero/display_record.php?RecordID=111005). Comprehensive searches of major electronic databases were conducted, including MEDLINE, MEDLINE In-process & Other Non-indexed Citations, MEDLINE Epub Ahead of Print, EMBASE, Science Citation Index and CENTRAL. Searches were restricted to publications from 1992 to date. Searches were carried out in August 2018 and updated in February 2021. The population considered was men with low testosterone. The intervention was TRT and the comparator was placebo. Primary outcomes were all-cause mortality and CV and cerebrovascular (CBV) events. Other outcomes of interest were sexual function, physical health parameters, functional activities, psychological symptoms, other comorbidities, prostate-related outcomes, physiological markers and quality of life (QoL). Outcomes were assessed at 12 months or at the closest time point to 12 months. Two reviewers independently screened all records identified by the searches and one reviewer screened all potentially relevant full texts, with a 10% sample independently checked by a second reviewer. Data were extracted by one reviewer with a random sample of 10% cross-checked by a second reviewer and all outcome data further cross-checked by the statistician. Risk of bias of the included RCTs was assessed by two independent reviewers using the Cochrane Risk of Bias tool (version 1). Data were analysed using both a one-stage meta-analysis for each outcome using the acquired IPD and a two-stage approach to enable the integration of the IPD along with the extracted published data from eligible studies without IPD. Methods for qualitative synthesis and patient-reported outcome measures analysis A comprehensive search was conducted to identify published papers reporting qualitative data and/or the development of patient-reported outcome measures (PROMs) for men with hypogonadism who received or were considered to receive TRT, and papers reporting the views of care providers. Electronic databases (Ovid MEDLINE, Embase, PsycInfo, EBSCO CINAHL and Proquest ASSIA) were searched for publication from 1992 to February 2020. The same eligibility criteria used for the IPD meta-analysis were applied. One review author (MA-M) independently screened all titles and abstracts with a randomly selected sample of 10% cross-checked by a second review author (KG). A third author (JC) was consulted when consensus could not be reached regarding eligibility. For the qualitative evidence synthesis, we appraised eligible studies for methodological rigour and theoretical relevance using the Critical Appraisal Skills Programme (CASP) tool. We conducted a thematic synthesis using both inductive and deductive approaches to analysis and we applied grading of recommendations assessment, development and evaluation-confidence in the evidence from reviews of qualitative research (GRADE-CERQual) to the findings of the thematic synthesis. For the PROMs analysis, we analysed the individual verbatim items from each PROM using a directed content analysis approach informed by the included papers and the relevant health domain was the World Health Organization International Classification of Functioning, Disability and Health (WHO-ICF). Domain mapping was conducted by two reviewers (MA-M and KG) independently, with any conflicts resolved through discussion. Descriptive statistics were used to describe general information and measure detail with narrative synthesis of the PROM and their inter-related domains. Methods for the systematic review of economic evaluations and the development of a new economic model The economic evidence on TRT was assessed through a systematic review (SR) of economic evaluations and a new model-based economic evaluation comparing TRT with standard of care (SoC; e.g. no treatment). Systematic review of economic evaluations Full economic evaluations (i.e. studies reporting cost and consequences – regardless of the way these were estimated – for at least two strategies including TRT) were included. Relevant electronic databases [i.e. MEDLINE, Embase, NHS Economic Evaluations Database (NEED), the HTA Database, Cost-Effectiveness Analysis Registry and Research Papers in Economics] were searched from 1992 until 4 February 2021. In addition, conference proceedings of key professional organisations (endocrinology, cardiology and men’s health) from 2018 to 2020 and International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Scientific Presentations Databases were searched. One reviewer screened titles and abstracts, selected studies for inclusion and extracted data following the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist. Economic model A cost–utility analysis was conducted using a cohort Markov model that was developed following best practice on decision modelling. The model care pathways were informed by existing clinical guidelines, the results of the IPD meta-analysis, and by discussion with clinical and methodological experts. Similar structures including five Markov states (no complications, post-Cardiac pathology, post-Peripheral Vascular System Pathology, post-Cerebrovascular System Pathology and Death) were used for TRT and SoC. The analysis of the TestES IPD provided key input data (i.e. all-cause mortality, CV and CBV complications and utility weights for TRT and SoC). NHS and personal and social services perspectives were adopted for costs. Ten-year and lifetime time horizons were considered. Cost and effects were discounted at 3.5% annual discount rate. We attached probability distributions to input mean parameter values and the model was run probabilistically using 10,000 Monte Carlo simulations. Three cohorts were defined according to the patients’ starting age (40-year-old cohort, 60-year-old cohort, 75-year-old cohort). Mean total cost, mean quality-adjusted life-years (QALYs) and incremental cost-effectiveness ratios (ICERs) were calculated. Probabilistic results were reported using cost-effectiveness acceptability curves (CEAC). Results Results of the quantitative synthesis and individual participant data meta-analysis Thirty-five trials with a total of 5601 randomised participants were included in the review of clinical effectiveness. Of these, 17 studies provided IPD for a total of 3431 participants. Overall risk of bias was assessed as being low for 12/17 studies providing IPD and unclear for the remaining five studies. Of the 18 studies not providing IPD, risk of bias was assessed as low for three studies, unclear for 13 studies and high for two studies. Mortality was lower in the TRT group (0.4%) than in the placebo group (0.8%) [odds ratio (OR) 0.46, 95% confidence interval (CI) 0.17 to 1.24; p = 0.13], but there were too few events for a reliable evaluation. There was no difference in the occurrence of CV and/or CBV events in the TRT group (7.5%) and the placebo group (7.2%; OR 1.07, 95% CI 0.81 to 1.42; p = 0.62). Effects of TRT on QoL were more variable, with some scales showing a difference in favour of TRT [International Index of Erectile Function-15 items (IIEF-15), ageing males’ symptoms (AMS), 3/10 subscales of the short form-36 items/short form-12 items (SF-36/SF-12)]. Serum testosterone was higher in the TRT group and serum cholesterol, triglycerides, haemoglobin and haematocrit were all lower in the TRT than the placebo group. Results of the qualitative synthesis and patient-reported outcome measures Five studies, reported in nine publications, were included in the qualitative evidence synthesis. Six broad themes (with several linked subthemes) were identified in relation to men’s, and their care providers’, experiences of low testosterone and receiving TRT as treatment. Five broad patient-facing themes were identified and ordered to reflect key timeline stages and decision points that a man with low testosterone may experience: symptoms of low testosterone and impact on daily life; diagnosis of low testosterone; access to treatment information; perceived effects of TRT; expectations, experience and preference of the type of TRT. A sixth theme on providers’ perception of diagnosis and treatment was also identified. We identified several interconnected subthemes highlighting the complexity of how low testosterone symptoms influence many aspects of men’s lives and their experiences of treatment. The PROMs search identified a total of nine eligible PROMs measuring experiences of men with low testosterone. The number of items varied across PROMs and ranged from 3 to 53 items (median = 7) with a cumulative total of 98 individual items across the nine PROMs. Our review identified 10 relevant health domains across the 98 items from the 9 PROMs. The 10 domains were defined according to the WHO-ICF classifications and were identified as: Cognition, Energy, General Well-being, Mood, Pain, Physical-General, Role, Sexual, Sleep, Social. The domain most frequently identified across PROMs was the sexual domain, with 29 (29.6% of total items) items measuring this concept across the PROMs. Results of the systematic review of economic evaluations and of the cost-effectiveness analyses Only one study met our inclusion criteria for the systematic review of economic evaluations. The study was a model-based cost–utility analysis conducted in Sweden. The model accounted for TRT benefits from reduced risk of fractures, depression, and type 2 diabetes and for a higher risk of CV and CBV events associated with diabetes. The authors concluded that the lifelong treatment with testosterone undecanoate (TU) depot injection was a cost-effective treatment option for men diagnosed with hypogonadism in Sweden. Our model results show that the cost-effectiveness of TRT is dependent on the relative risk (RR) of all-cause mortality and the methods used to derive health state utility scores [i.e. through the short form-6 dimensions (SF-6D) algorithm or a mapping exercise between Beck depression inventory (BDI) score and EuroQol-5 dimensions (EQ-5D) score] for the TRT versus SoC. When the RR of mortality favouring TRT and the BDI-based utility scores were used for the 10-year time horizon, ICERs remained below the £20,000 threshold, irrespective of the cohort starting age. ICERs also stayed below the £20,000 threshold for the 60- and 75-year-old cohorts when the RR of mortality favouring TRT and the SF-6D utility difference were applied for 10 years. However, ICERs increased above the £20,000 threshold when the difference in all-cause mortality between TRT and SoC was dropped, and the utility scores were defined using the SF-6D. Extending the model time horizon for lifetime for the later scenario further increased the ICER as the impact of complications became more pronounced, eroding the modest SF-6D-based QALY increment. Limitations Inclusion was restricted to studies published in English. Lack of long-term data on the effects and safety of TRT. A meaningful evaluation of the association between mortality and TRT was hampered by the limited number of defined events. Definition and reporting of CV and CBV events and methods for testosterone measurement varied across studies. Length of follow-up in many included trials precluded the accumulation of enough CV or CBV events. There were too few studies of oral testosterone to assess whether the incidence of CV or CBV events was affected by the mode of administration of testosterone therapy. There were only a limited number of available qualitative studies assessing the experience of men with low testosterone; these were predominantly conducted in North America among white men. There were limited data for generating preference-based health-related QoL weights to be used in the economic evaluation. Implications for health care Our collaborative IPD meta-analysis failed to show a relationship between TRT and CV or CBV events in the short-to-medium term (within 12 months). Our findings indicate that TRT improves sexual function and patients’ QoL without adverse effects on blood pressure, serum lipids or glycaemic markers. Overall, there is a paucity of qualitative evidence on the effects of low testosterone and consequences of TRT when compared with the number of existing clinical trials assessing the effectiveness of TRT. Our results indicate that the effects of low testosterone and subsequent treatment with TRT have multiple impacts and concerns for men. This study has shown the considerable variability that exists in disease-specific PROMs for men with low testosterone regarding development and domain coverage. It has also highlighted the lack of input from men in the development of these PROMs, bringing into questions their relevance and adequacy in capturing outcomes that matter to men with low testosterone. Results of our economic model suggest that the cost-effectiveness of TRT is dependent on its effects on all-cause mortality and health state utility resulting from improvements in symptoms associated with hypogonadism such as sexual dysfunction, low mood and reduced QoL. We also noticed that the choice of the instrument and approach to estimate QoL weights (BDI or SF-6D) was crucial to the cost-effectiveness of TRT. Further clarity on the CV safety of TRT in men with hypogonadism and more in-depth mapping of clinical outcomes to generic preference-based measures of health-related QoL will be crucial to inform more robust estimates of the cost-effectiveness of TRT for men with hypogonadism. Recommendations for future research Future research should assess the safety of long-term use of TRT including its impact on mortality, identify the threshold of baseline serum testosterone associated with improvement of sexual function, cognition, mood and QoL in men with hypogonadism and compare the effects of transdermal versus intramuscular TRT. Moreover, further research aiming at mapping QoL instruments commonly used in existing trials of MH to generic preference-based health-related QoL instruments would facilitate the integration of existing evidence into future economic evaluations and models. There is also the need to gather more data on the experience of men with hypogonadism from multiethnic populations and develop a holistic symptom score for TRT response. Study registration The study is registered as PROSPERO CRD42018111005. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/68/01) and is published in full in Health Technology Assessment; Vol. 28, No. 43. See the NIHR Funding and Awards website for further award information.

Background Glaucoma is a chronic disease of the optic nerve and a leading cause of severe visual loss in the UK. Once patients have been diagnosed, they need regular monitoring at hospital eye services. Recent advances in technology mean patients with glaucoma can now monitor their disease at home. This could be more convenient for patients and potentially reduce costs and increase capacity for the NHS. However, it is uncertain whether self-monitoring would be acceptable or possible for patients with glaucoma. Objectives The objectives were to: identify which patients are most appropriate for home monitoring; understand views of key stakeholders (patients, clinicians, researchers) on whether home glaucoma monitoring is feasible and acceptable; develop a conceptual framework for the economic evaluation of home glaucoma monitoring; and explore the need for and provide evidence on the design of a future study to evaluate the clinical and cost-effectiveness of digital technologies for home monitoring of glaucoma. Design In-home Tracking of glaucoma: Reliability, Acceptability, and Cost (I-TRAC) was a multiphase mixed-methods feasibility study with key components informed by theoretical and conceptual frameworks. Setting Expert glaucoma specialists in the UK recruited through professional glaucoma societies; study site staff and patient participants recruited through three UK hospital eye services (England, Scotland, Northern Ireland); and UK research teams recruited though existing networks. Intervention Home tonometer that measures intraocular pressure and a tablet computer with a visual function application. Patients were asked to use the technology weekly for 12 weeks. Results Forty-two patients were recruited. Retention and completion of follow-up procedures was successful, with 95% (n = 40) completing the 3-month follow-up clinic visits. Adherence to the interventions was generally high [adherence to both devices (i.e. ≥ 80% adherence) was 55%]. Overall, patients and healthcare professionals were cautiously optimistic about the acceptability of digital technologies for home monitoring of patients with glaucoma. While most clinicians were supportive of the potential advantages glaucoma home monitoring could offer, concerns about the technologies (e.g. reliability and potential to miss disease progression) and how they would fit into routine care need to be addressed. Additionally, clarity is required on defining the ideal population for this intervention. Plans for how to evaluate value for money in a future study were also identified. However, the study also highlighted several unknowns relating to core components of a future evaluative study that require addressing before progression to a definitive effectiveness trial. Limitations The main limitation relates to our sample and its generalisability, for example, the over-representation of educated persons of white ethnicity who were generally experienced with technology and research motivated. Conclusions The In-home Tracking of glaucoma: Reliability, Acceptability, and Cost study has demonstrated ‘cautious optimism’ when considering patients’ and healthcare professionals’ views on the acceptability of digital technologies for home monitoring of patients with glaucoma. However, the study also highlighted several unknowns relating to the research question and design of a future evaluative study that require addressing before progression to a randomised controlled trial. Future work Further research is required to determine the appropriate population (i.e. low vs. high risk of progression) and further refine the intervention components and delivery for planning of future evaluation studies. Study registration This study is registered as Research Registry #6213. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR129248) and is published in full in Health Technology Assessment; Vol. 28, No. 44. See the NIHR Funding and Awards website for further award information. Plain language summary What is this research about? The In-home Tracking of glaucoma: Reliability, Acceptability, and Cost study explored whether glaucoma patients who would normally be monitored in hospital could do some monitoring themselves at home, and whether self-monitoring at home would be acceptable or possible for them. How was the research done? We delivered In-home Tracking of glaucoma: Reliability, Acceptability, and Cost in four phases by: Surveying expert glaucoma specialists to understand which patients would benefit most from home monitoring. Providing glaucoma patients with an iPad tablet and a device which measures eye pressure to use once a week for 3 months. The patients who participated and the clinical staff delivering the study were interviewed about their experiences. Interviewing researchers with experience of running large studies testing digital technologies to monitor patients’ health at home to understand challenges. Reviewing other researchers’ work and comparing it with ours to help us understand whether home monitoring of glaucoma could be good value for money. What did the research find out? Overall, patients and healthcare professionals were cautiously optimistic about the digital technologies for home monitoring of glaucoma. Most patient participants were able to use the technologies, and half told us they preferred home monitoring. Most clinicians recognised the potential advantages of glaucoma home monitoring but had concerns about the technologies (specifically reliability and the risk of missing disease progression) and how they would fit into routine care. Plans for how to evaluate value for money in a future study were identified. The study did not aim to identify whether the digital technology was better than what happens currently; a different study design with many more patients would be required to answer that question. The study did identify several important questions to answer before designing a future larger study; for example, how to ensure diverse patient participation. These questions should be the focus of future research in this area. Scientific summary Background Glaucoma is a chronic neurodegenerative eye disease and the second commonest cause of severe visual loss in the UK. Diagnosed patients require regular, lifelong monitoring to detect progression and assess effectiveness of treatment, with monitoring typically delivered within the hospital eye services (HES). Ophthalmology is the busiest NHS outpatient specialty, accounting for 10% of all outpatient visits, and glaucoma represents a significant part of this workload. In England alone there are over 1 million clinic visits per year for patients with glaucoma. Providing regular surveillance and treatment is already a major challenge for the NHS but as the prevalence of glaucoma increases with age, demand for glaucoma care is increasing (and will continue to do so) due to the ageing population. The two main measurements used in the assessment of glaucoma are intraocular pressure (IOP) measurement and visual field (VF) testing. Recent advances in technology mean it is now possible for patients with glaucoma to measure their IOP and test their VFs in their own home. Home monitoring could theoretically replace or supplement standard care, perhaps allowing patients to require fewer outpatient visits, while increasing convenience and potentially reducing costs and increasing capacity for the NHS. It is also possible that frequent home testing may lead to better outcomes than achieved with standard infrequent visits to HES, potentially leading to earlier detection of progression or highlighting high IOP undetected through conventional outpatient clinic-based tests. Currently though, it is not known if home monitoring is acceptable to people with glaucoma, or if home monitoring in the general glaucoma population is feasible. The main aim of this study was to assess acceptability and feasibility of home monitoring, and to make recommendations about future research to test how the NHS could use home monitoring. Objectives The aim of In-home Tracking of glaucoma: Reliability, Acceptability, and Cost (I-TRAC) was to determine the feasibility and acceptability of digital technologies to monitor glaucoma at home and inform the possible need for and design of a definitive evaluative study. The specific research objectives were to: identify which patients with glaucoma are most appropriate for home monitoring (e.g. all patients, those with stable disease, or those with severe glaucoma?) understand the views of key stakeholders (patients, clinicians, researchers) on whether home monitoring is feasible and acceptable develop a conceptual framework for the economic evaluation of home monitoring for glaucoma explore the need for and provide evidence on the design of a future study to evaluate the clinical and cost-effectiveness of digital technologies for home monitoring of glaucoma. Methods Design The I-TRAC study was a multiphase mixed-methods feasibility study with key components informed by theoretical (i.e. the Theoretical Framework of Acceptability and Theoretical Domains Framework) and conceptual [A process for Decision-making after Pilot and feasibility Trials (ADePT)] frameworks, utilising various data sources from a survey, interviews, focus groups, an observational intervention study, and literature reviews. Recruitment Recruitment to each of the I-TRAC study phases was as follows: Expert glaucoma specialists from the UK were recruited through a professional glaucoma society and invited to participate in a survey. Those who responded were asked to indicate their willingness to participate in a follow-up focus group or interview. Patient participants for the intervention study were recruited through three secondary care ophthalmology glaucoma clinics. All site staff involved in the patient-facing delivery of I-TRAC across the three sites were invited to participate in a focus group or interview. External researchers, such as Chief Investigators and Trial/Study Managers, who had been involved in evaluating digital technologies for home monitoring of eye disease or other diseases (as either feasibility, pilot or full-size trials) and who were either known to the I-TRAC team or identified through published literature were invited to participate in an interview. Intervention Patient participants were asked to use two home monitoring technologies to measure IOP and visual function (through a contrast sensitivity assessment) on a weekly basis for a duration of 12 weeks. Participants were provided with an iCare HOME 2 handheld tonometer (iCare Oy, Vantaa, Finland) to measure IOP, and the OKKO Visual Health App (OKKO Health, Bristol, UK) was used to measure visual function on a tablet computer. Site staff and patient participants received training in how to use the devices in hospital in advance of the home monitoring period. Participants received a prompt each week to remind them to use their home monitoring equipment; they could opt between receiving e-mail or text message electronic reminders. Results The key findings from I-TRAC are presented below in relation to the study objectives. Objective 1: Identify which glaucoma patients are most appropriate for home monitoring The online survey, completed by 49 expert glaucoma clinicians, aimed to determine which glaucoma patients may be most suitable for home monitoring using digital technology. The survey findings demonstrated agreement among expert glaucoma clinicians that there is a place for home monitoring of glaucoma patients using digital technology. However, based on the scenarios used in this study, there is limited agreement among clinicians about which glaucoma patients are most suitable for home monitoring using digital technologies to measure IOP and visual function. Clinicians reported that they were generally not supportive of the home monitoring of high-risk patients as a replacement for standard care, due to the fear of missing disease progression or unreliable readings. However, they were generally supportive of home monitoring having a role within low-risk scenarios. Clinicians anticipated that the integration of home monitoring into the current healthcare system could act as an adjunct to increase hospital capacity for glaucoma patients who require face-to-face assessment. The survey highlighted a range of issues and challenges related to the home monitoring of glaucoma patients using digital technologies. A central theme was clinicians’ lack of trust in home monitoring technologies, related to concerns about the reliability, accuracy and clinical usefulness of these technologies. Clinicians expressed concerns about patient safety (if standard care were to be replaced with home tonometry), decreased rather than increased glaucoma progression detection, and concerns about how resource efficient (time and financial) this approach could be in comparison to current provision. Objective 2: Understand the views of key stakeholders on whether home monitoring is feasible and acceptable Several key stakeholder groups were included in various I-TRAC phases in order to achieve this objective. Firstly, three focus groups (n = 2, n = 4, n = 4) and five individual interviews with expert glaucoma clinicians suggested there is cautious optimism about the use of digital technologies to monitor patients’ glaucoma at home. Clinicians reported that they are interested and enthused by the potential of glaucoma home monitoring, but there are several areas of concern that need to be addressed before they would feel reassured to buy-in to this approach. They believe home monitoring could meet an existing clinical need, addressing the present difficulties of inadequate capacity to meet increasing demand. They can see potential patient and service benefits, but they require reassurances about the technologies and how such a service would be implemented into routine care. Contextually, the influence of the coronavirus disease discovered in 2019 pandemic was evident; clinicians’ experiences throughout the pandemic have both prompted a need to adapt and change the way they monitor patients with glaucoma and highlighted that monitoring outside the clinic is possible and, in many ways, can be done safely. The current care backlog post pandemic is driving enthusiasm for innovative solutions. The multiphase mixed-methods study exploring the intervention’s acceptability demonstrated that for both patient participants and site staff tasked with delivering I-TRAC there were many positives and the intervention was deemed broadly acceptable. Overall, the I-TRAC study recruited well, recruiting 95% of its proposed sample size (42 patient participants of 45) in the planned recruitment period (November 2021–August 2022, 10 months). Retention and completion of follow-up procedures was also successful, with 95% (n = 40) completing the 3-month follow-up clinic visits. Adherence to the interventions was generally high, especially considering that our predetermined adherence levels were above 80%, and satisfaction with the process and the training were also scored highly by patient participants. However, 48% (n = 20) of patient participants contacted site staff at least once when at home, resulting in additional input from site staff regarding study process or intervention delivery. The qualitative data from the interviews and focus groups with patient participants (n = 10) and site staff (n = 9) did corroborate some of these findings (e.g. stating recruitment worked well and study processes were easy to follow and low burden). Yet the qualitative data also highlighted important areas not identified through the quantitative pilot, such as: the need for a refinement of eligibility criteria and associated recognition of limited sample diversity in I-TRAC; issues relating to inadequate training (for both site staff and patients); a lack of confidence in the technology (and their ability) in relation to the purpose of home monitoring; familiarity with the device and physical dexterity issues; and some anxieties in relation to a lack of clinical oversight when monitoring at home. These findings highlight that several key factors need to be deliberated when considering the feasibility of future trials evaluating digital technology for home monitoring of glaucoma. Finally for this objective, we interviewed researchers from external teams (n = 8) who had been involved in the evaluation of digital technologies for home monitoring patients’ health. Although researchers reported multiple challenges encountered while carrying out digital health technology studies/trials, they also had many suggestions as to how these could be prevented or overcome. The common barriers were low stakeholder acceptability, lack of understanding of digital technologies, poor resource planning, insufficient peripheral infrastructure, problematic relationships with commercial partners, and the unsolved dilemma of digital exclusion. The findings illustrate that researchers in the UK carrying out digital health technology studies encounter a number of challenges impacting on the successful design, conduct, and delivery of digital health technology studies/trials and potentially leading to wasted research efforts. This broader exploration of feasibility issues surrounding digital health technology studies/trials highlights significant agreement about critical trial design and conduct issues that require consideration in studies/trials of this type. Objective 3: Develop a conceptual framework for the economic evaluation of home monitoring for glaucoma Several methods were used to develop a conceptual framework to consider the future economic evaluation of home monitoring for glaucoma. Data sources used included evidence from two systematic reviews conducted for the I-TRAC project and secondary analysis of the data collected through all previous I-TRAC phases to identify resource use and patient preference components. Overall, the key categories of intervention costs of glaucoma home monitoring identified in this study include equipment cost, patient training, ongoing patient support during home monitoring, potential spill-over costs (e.g. high readings trigger hospital visits) and costs of data integration (to the existing medical records) and evaluation by artificial intelligence, while key sources of patient utilities of glaucoma home monitoring are categorised as health-related quality of life (HRQoL)-related (e.g. more frequent disease monitoring and faster identification of disease progression) and non-HRQoL related (e.g. convenience). Given the complexity and scarcity of relevant evidence in the literature, it is recommended that further qualitative/quantitative research needs to be conducted to better understand the study population, care pathways of the compared interventions, cost categories and benefits of home monitoring, before a formal economic evaluation can be conducted. A step-by-step approach is then recommended to carefully explore what economic evaluation approach can be suitable in the context of glaucoma home monitoring. Objective 4: Explore the need for and provide evidence on the design of a future study to evaluate the clinical and cost-effectiveness of digital technologies for home monitoring of glaucoma The data from across all phases of the I-TRAC study were mapped to the ADePT framework in order to establish the feasibility of a future evaluative study of the clinical and cost-effectiveness of digital technologies for home monitoring of glaucoma. Many of the 14 items in ADePT were successfully achieved (e.g. 93% of sample size was recruited in the original recruitment window, and overall acceptability of the home monitoring technologies was good). The mixed-methods data collection and analysis allowed in-depth investigation of key areas and highlighted uncertainties that need to be addressed before moving to an evaluative study. Some of these focused on aspects related to the population; that is, clarity is still required on which glaucoma patients would be most suitable for home monitoring, and this is linked to whether home monitoring is considered as an additional service (i.e. in addition to routine monitoring through HES) or as a replacement service (i.e. patients would not attend HES and instead would be monitored at home). While the interventions were broadly deemed as acceptable to patients and clinicians, further refinement of the intervention is required (e.g. frequency and duration) and consideration of how it ‘fits’ within the healthcare system is required before evaluation. As well as the intervention, considerations about an appropriate comparator were not explored in I-TRAC but should be. Lastly, determination of appropriate outcomes (and their importance for a range of stakeholders) for the evaluation of digital technologies for glaucoma home monitoring also requires attention. Conclusions The I-TRAC study has demonstrated ‘cautious optimism’ when considering patients’ and healthcare professionals’ views on the acceptability of digital technologies for home monitoring patients with glaucoma. Much of the caution from clinicians related to concerns around the reliability of the technologies and the potential to miss progression of the disease if patients were not monitored in clinic. The study evidenced sufficient fidelity, good adherence to the interventions among the patient population, and feasibility of delivery of both the interventions and the study processes. However, I-TRAC also highlighted several unknowns relating to the research question and design of a future evaluative study that require addressing before progression to a randomised controlled trial. The I-TRAC study has also considered the wider ecosystem challenges of running digital health technology trials through evidencing the views of external research teams experienced in digital home monitoring technology delivery. Further research is required to determine the appropriate population (i.e. glaucoma patients with low vs. high risk of progression) and further refine intervention components and their delivery, to allow future evaluation studies to be planned. Study registration This study is registered as Research Registry #6213. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR129248) and is published in full in Health Technology Assessment; Vol. 28, No. 44. See the NIHR Funding and Awards website for further award information.

Background There is interest in using treatment breaks in oncology, to reduce toxicity without compromising efficacy. Trial design A Phase II/III multicentre, open-label, parallel-group, randomised controlled non-inferiority trial assessing treatment breaks in patients with renal cell carcinoma. Methods Participants Patients with locally advanced or metastatic renal cell carcinoma, starting tyrosine kinase inhibitor as first-line treatment at United Kingdom National Health Service hospitals. Interventions At trial entry, patients were randomised (1 : 1) to a drug-free interval strategy or a conventional continuation strategy. After 24 weeks of treatment with sunitinib/pazopanib, drug-free interval strategy patients took up a treatment break until disease progression with additional breaks dependent on disease response and patient choice. Conventional continuation strategy patients continued on treatment. Both trial strategies continued until treatment intolerance, disease progression on treatment, withdrawal or death. Objective To determine if a drug-free interval strategy is non-inferior to a conventional continuation strategy in terms of the co-primary outcomes of overall survival and quality-adjusted life-years. Co-primary outcomes For non-inferiority to be concluded, a margin of ≤ 7.5% in overall survival and ≤ 10% in quality-adjusted life-years was required in both intention-to-treat and per-protocol analyses. This equated to the 95% confidence interval of the estimates being above 0.812 and −0.156, respectively. Quality-adjusted life-years were calculated using the utility index of the EuroQol-5 Dimensions questionnaire. Results Nine hundred and twenty patients were randomised (461 conventional continuation strategy vs. 459 drug-free interval strategy) from 13 January 2012 to 12 September 2017. Trial treatment and follow-up stopped on 31 December 2020. Four hundred and eighty-eight (53.0%) patients [240 (52.1%) vs. 248 (54.0%)] continued on trial post week 24. The median treatment-break length was 87 days. Nine hundred and nineteen patients were included in the intention-to-treat analysis (461 vs. 458) and 871 patients in the per-protocol analysis (453 vs. 418). For overall survival, non-inferiority was concluded in the intention-to-treat analysis but not in the per-protocol analysis [hazard ratio (95% confidence interval) intention to treat 0.97 (0.83 to 1.12); per-protocol 0.94 (0.80 to 1.09) non-inferiority margin: 95% confidence interval ≥ 0.812, intention to treat: 0.83 > 0.812 non-inferior, per-protocol: 0.80 < 0.812 not non-inferior]. Therefore, a drug-free interval strategy was not concluded to be non-inferior to a conventional continuation strategy in terms of overall survival. For quality-adjusted life-years, non-inferiority was concluded in both the intention-to-treat and per-protocol analyses [marginal effect (95% confidence interval) intention to treat −0.05 (−0.15 to 0.05); per-protocol 0.04 (−0.14 to 0.21) non-inferiority margin: 95% confidence interval ≥ −0.156]. Therefore, a drug-free interval strategy was concluded to be non-inferior to a conventional continuation strategy in terms of quality-adjusted life-years. Limitations The main limitation of the study is the fewer than expected overall survival events, resulting in lower power for the non-inferiority comparison. Future work Future studies should investigate treatment breaks with more contemporary treatments for renal cell carcinoma. Conclusions Non-inferiority was shown for the quality-adjusted life-year end point but not for overall survival as pre-defined. Nevertheless, despite not meeting the primary end point of non-inferiority as per protocol, the study suggested that a treatment-break strategy may not meaningfully reduce life expectancy, does not reduce quality of life and has economic benefits. Although the treating clinicians’ perspectives were not formally collected, the fact that clinicians recruited a large number of patients over a long period suggests support for the study and provides clear evidence that a treatment-break strategy for patients with renal cell carcinoma receiving tyrosine kinase inhibitor therapy is feasible. Trial registration This trial is registered as ISRCTN06473203. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment Programme (NIHR award ref: 09/91/21) and is published in full in Health Technology Assessment; Vol. 28, No. 45. See the NIHR Funding and Awards website for further award information. Plain language summary Treatment breaks in cancer are of significant interest to patients and health professionals. Renal cell carcinoma is the most common type of kidney cancer. Sunitinib and pazopanib are both targeted treatments. They were commonly used to treat advanced kidney cancer but often cause side effects, sometimes requiring use of a reduced dose or even stopping treatment. The STAR trial was designed to see whether planned treatment breaks made patients with advanced kidney cancer being treated with sunitinib and pazopanib feel better, without substantially affecting how well the treatment worked. After 24 weeks of treatment, patients took sunitinib and pazopanib either as they normally would or in the alternative way with planned treatment breaks. Treating patients in this way was continued until drug-related side effects stopped treatment, patients’ disease worsened while taking treatment or the patient died. The trial compared how well the different treatment strategies worked in terms of how long patients lived and their quality of life over that time. This trial is the largest United Kingdom trial in advanced renal cell carcinoma. Patients took part from 60 United Kingdom centres between 2012 and 2017. It was funded by the National Institute for Health and Care Research Health Technology Assessment Programme and run by the Leeds Clinical Trials Research Unit. In total, 920 patients took part. Four hundred and sixty-one patients were allocated to continue treatment and 459 were allocated to start at least one treatment break. Treatment breaks lasted on average 87 days. The length of time patients lived in both arms of the trial appeared similar, but this cannot be concluded due to insufficient information. Being allocated to have treatment breaks rather than continuing treatment did not negatively impact a patient’s quality of life. Additionally, allocating patients to have treatment breaks was shown to have significant cost savings compared to just continuing treatment. Importantly planned treatment breaks were shown to be feasible. Scientific summary Background There is increasing interest in using treatment breaks in oncology, to reduce toxicity without compromising efficacy. STAR was designed to determine if a tyrosine kinase inhibitor (TKI) drug-free interval strategy (DFIS) was non-inferior to a conventional continuation strategy (CCS) in the first-line treatment of advanced renal cell carcinoma (RCC). Objectives The overall primary objective was to determine whether a sunitinib or pazopanib DFIS is non-inferior in terms of overall survival (OS) and quality-adjusted life-year (QALY) compared to a sunitinib or pazopanib CCS in patients with locally advanced and/or metastatic clear cell RCC. Secondary objectives included comparing a DFIS to a CCS in terms of summative progression-free interval, time to strategy failure, time to treatment failure, toxicity (common terminology criteria for adverse events v.4.0), quality of life (QoL) [Functional Assessment of Cancer Therapy – Kidney Symptom Index (FKSI-15), Functional Assessment of Cancer Therapy-G (FACT-G), EuroQol-5 Dimensions (EQ-5DTM) and EQ-Visual Analogue Scale (VASTM)], cost effectiveness, progression-free survival (PFS). Three ancillary studies were also included in the trial. The Patient Preference and Understanding Study was designed to understand the participants’ experiences of taking part in the trial. The dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) substudy was designed to investigate whether early DCE-MRI could predict progressive disease (PD) at 24 weeks. The computerised tomography (CT) substudy was designed to determine CT as a potential biomarker and predictor of PD at 24 weeks. Methods STAR was a UK Phase II/III multicentre, open-label, parallel-group, randomised controlled non-inferiority (NI) trial. The trial recruited adults, with histologically confirmed locally advanced or metastatic clear cell, uni-dimensionally measurable, RCC who required but had not received prior systemic therapy. Participants were required to have an Eastern Cooperative Oncology Group performance status of 0–1, meet pre-specified blood parameters prior to randomisation, provide written informed consent, be able and willing to comply with the terms of the protocol and follow approved pregnancy prevention guidelines. Exclusion criteria included pulmonary or mediastinal disease, life expectancy < 6 months, previous treatment with or known contraindications or hypersensitivity to TKIs, untreated brain metastases, concurrent or previous invasive cancers which could confuse diagnosis or end points, use of contraindicated concomitant medication or substances and poorly controlled hypertension. At trial entry, participants were randomised (1 : 1) centrally by Leeds CTRU to a CCS or a DFIS. Randomisation was stratified by Motzer prognostic group (favourable, intermediate, poor), trial site, gender, age (< 60, ≥ 60 years), disease status (metastatic, locally advanced), previous nephrectomy (yes, no), TKI (sunitinib, pazopanib). Both strategies received an initial 4 cycles of trial treatment [sunitinib: 1 cycle of treatment refers to 50 mg (starting dose) on days 1–28, repeated every 42 days; pazopanib: 1 cycle of treatment refers to 800 mg (starting dose) on days 1–42, repeated every 42 days] following which participants in the DFIS arm took up their first treatment break. Following disease progression on a treatment break, DFIS participants restarted their treatment. Additional breaks were permitted following a further 4 cycles of treatment at the treating clinician’s discretion. The trial strategy continued until death, progression while receiving treatment, unacceptable toxicity (clinical withdrawal) or participant withdrawal. Following cessation of the trial strategy, participants were followed up after 6 months and then annually thereafter until death, patient withdrawal or the end of the trial. All trial data were collected on trial-specific case report forms (CRFs). All CRFs, apart from the questionnaires for the patient-reported outcome measures (FSKI-15, FACT-G, EQ-5D and EQ-VAS) were completed by research staff at the site. Overall survival was defined as time from randomisation to death from any cause. For NI to be concluded, a margin of ≤ 7.5% in OS was pre-specified. The survival at 2 years in the CCS arm was assumed to be 48.5%. A 7.5% NI margin led to the survival at 2 years in the DFIS arm assumed to be at least 41%, leading to a hazard ratio (HR) of 0.812. This along with a one-sided 2.5% significance level, 5% dropout rate and 80% power required 920 participants. Where 80% power would be attained when 720 events were observed. Quality-adjusted life-years were defined as the time spent in each health state over trial and follow-up, calculated as the utility index of the EQ-5D questionnaire, which was collected frequently (6-weekly for 24 weeks, 2-weekly for 24 weeks, 6-weekly while on study, 6 months post end of trial strategy and annually thereafter). For NI to be concluded a margin of ≤ 10% in mean QALYs was pre-specified. The power for the QALY comparison was derived using simulations. Under the assumptions of a 0.9 HR in favour of CCS, a sample size of 920, 5.83 years of recruitment and 3.25 years of follow-up a CCS QALY estimate of 1.56 was derived along with a power of 69.94%. The CCS QALY estimate of 1.56 in conjunction with the 10% margin equates to a NI boundary of −0.156 for the difference in mean QALYs. It was pre-specified that NI was required in both intention-to-treat (ITT) and per-protocol (PP) analyses, in both end points, to be concluded overall. An in-depth qualitative inductive study of 11 patients on the DFIS arm was conducted to better understand their acceptability of extended treatment breaks. A thematic analysis using a comparative and contrastive approach was employed. An economic evaluation was also conducted which estimated the cost effectiveness of DFIS versus CCS at 2 years (within-trial analysis) and over the patients’ lifetime (decision modelling analysis). Results Participant flow Two thousand one hundred and ninety-seven patients were screened; of these, 920 were randomised (CCS: 461, DFIS: 459). In total, 878 (95.4%) participants ceased trial treatment prior to the end of follow-up. Radiological disease progression was the most frequent reason for treatment discontinuation at 43.7% (CCS: 48.4%, DFIS: 39.0%). Overall, 13,147 (78.6%, n = 16,726) (CCS: 5764/7401, 77.9%; DFIS: 7383/9327, 79.2%) of QoL questionnaire booklets were returned on the trial. The proportions of reasons for missing booklets were similar between the two arms, where the highest reason was ‘missed by site in error’ (overall: 27%, CCS: 28.4%, DFIS: 25.8%). However, the majority of reasons for missing were missing (overall: 52.1%, CCS: 50.6%, DFIS: 53.4%). Overall, 63 participants (6.8%, n = 920) withdrew from some aspect of the trial. This resulted in 21 (2.82%, n = 920) participants being formally lost to follow-up (CCS: 11, 2.4%; DFIS: 10, 2.2%). An additional DFIS participant was lost to follow-up due to the participant moving away. However, all participants were included in the analysis. Numbers analysed Nine hundred and nineteen participants (CCS: 461, DFIS: 458) were included in the ITT population as one DFIS participant did not have RCC. Similarly, 871 participants (CCS: 453, DFIS: 418) were included in the PP population, 916 participants (CCS: 485, DFIS: 431) were included in the safety population, 869 participants (CCS: 438, DFIS: 431) were included in the EuroQol-5 Dimensions, three-level version (EQ-5D-3L) population, 856 participants (CCS: 425, DFIS: 431) were included in the FACT-G population and 882 participants (CCS: 436, DFIS: 446) were included in the FKSI population. Baseline characteristics and treatment (on trial and in follow-up) Key demographic and disease-related characteristics were similar across both randomisation allocation and randomised under TKI for all analysis populations. STAR participants were predominantly white (ITT – 96.3%), male (ITT – 72.7%) and aged ≥ 60 at randomisation (ITT – 73.4%). On average, participants in both arms received a similar number of treatment cycles [ITT median (interquartile range, IQR): 5 (2, 10) vs. 4 (2, 9)]. Overall, 248 participants (56.3%, n = 459) in the DFIS arm continued on the trial after week 24. The majority of these (127, 51.2%) only started one treatment break. Thirty-eight (15.3%) of these had 2 treatment breaks and 68 (27.4%) took 3 or more treatment breaks. The maximum number of treatment breaks taken by 1 participant was 9, and 15 participants (6%) did not take up a treatment break despite continuing on study after week 24. Overall, 61.6% of participants received further systemic anticancer therapy treatment during follow-up, with a higher proportion in the CCS arm (68.5%, n = 461) compared to the DFIS arm (54.6%, n = 458). Co-primary end points Median (IQR) follow-up for OS in the PP population was 58 months (46, 72). In total 648 (74.4%, n = 871) of PP participants died prior to the end of follow-up (31 December 2020) [CCS: 330 (72.8%, n = 453), DFIS: 318 (76.1%, n = 418)]. Renal cancer was related to the cause of 607 (93.7%, n = 648) deaths [CCS: 314 (95.2%, n = 330), DFIS: 293 (92.1%, n = 318)]. Median OS [95% confidence interval (CI)] in the CCS arm was 28 months (24, 32) and 27 months (23, 31) in the DFIS arm. At 2 years post randomisation (95% CI), 54.2% (50.8%, 57.5%) of PP participants were alive. This was similar between the two arms [CCS: 55.2% (50.5%, 59.7%), DFIS: 53.1% (48.2%, 57.8%)]. On application of a Cox proportional hazards (PH) model adjusted for the stratification factors of the trial, the HR for randomisation allocation, CCS versus DFIS (95% CI) was 0.94 (0.80 to 1.09). Comparing the lower bound of the CI to the NI boundary of 0.812. At the 2.5% significance level, we do not reject the null hypothesis that DFIS is not non-inferior to CCS in terms of OS. There is insufficient evidence to conclude NI. However, the analyses conducted in the ITT population along with the sensitivity analysis conducted in the PP population using a piecewise hazards model and a Cox regression model where Motzer Score at randomisation was considered as two categories did conclude NI between the two arms. The mean QALY (95% CI) for PP participants in the CCS arm was 1.73 (1.60 to 1.86) and 1.80 (1.65 to 1.95) in the DFIS arm derived by combining the results across 52 imputed data sets. The marginal effect (95% CI) of randomisation allocation, DFIS versus CCS was 0.04 (−0.14 to 0.21). Comparing the lower bound of the CI to the NI boundary of 0.156. As the lower bound of the CI is above −0.156, we conclude that the DFIS arm is non-inferior to the CCS arm in terms of QALYs in the PP population. This conclusion remained consistent across all the sensitivity analyses; ITT population, derived from week 24, measured up to 12, 24 and 36 months post randomisation, complete case analysis, analysis conducted using a multivariate linear regression analysis and analysis conducted under missing not at random scenario 1 (worst-case scenario). Secondary end points All secondary time to event end-point analysis was conducted in the ITT population. With the exception of PFS, all analyses showed a significant difference in favour of the DFIS compared to the CCS when applied in a Cox PH model adjusted for the stratification factors of the trial. As PFS measures to the first progression and does therefore not account for successful rechallenge of TKI in the DFIS arm post progression this was expected. There was little difference in the summary statistics of the various QoL questionnaire scores and subscales (FKSI-15, FACT-G, EQ-5D-3L and EQ5D-VAS) when considered by randomisation allocation in each of their populations. A difference only appeared around week 294 when few participants remained on the CCS arm. The multilevel modelling conducted on the FKSI-15 and the FACT-G questionnaires concluded some significant differences between the DFIS and the CCS in favour of the DFIS. However, these differences were small and not consistently observed in the sensitivity analysis, which modelled the questionnaires up to regular intervals post randomisation. The within-trial economic evaluation at 2 years indicated that DFIS was highly likely to be cost-effective. DFIS was found to provide a small QALY benefit (0.049) and substantial cost savings (£3235) over CCS. The value-for-money metric was principally driven by savings in medicine costs. There were relatively minor differences in resource use between arms. The lifetime cost-effectiveness modelling led to the same conclusion yielding QALY gains of 0.08 and cost savings of £2420 (both probabilistic values) in favour of DFIS. The probabilistic sensitivity analyses indicated that DFIS had a 96% chance of being cost saving and a 68% chance of yielding QALY gains versus CCS. The conclusions of both the trial- and model-based analyses were relatively robust to a series of supplementary and sensitivity analyses. Safety All participants bar three in the safety population (CCS: 1, DFIS: 2) experienced an adverse event (AE) during the course of the trial. Overall, a higher proportion of participants in the DFIS arm experienced an AE grade 3 or above (CCS: 70.7%, n = 485, DFIS: 76.1%, n = 431). This difference between arms continues when considering serious adverse events (SAEs) by arm overall and pre and post week 24 (overall – CCS: 45.8%, n = 485, DFIS: 54.1%, n = 431, pre CCS: 32.8%, n = 485, DFIS: 35.5%, n = 431, post CCS: 32.8%, n = 265, DFIS: 48.4%, n = 223). However, this may be explained through the reporting requirements. Participants on DFIS will have reported AEs and SAEs for longer due to their increased time on trial strategy. On consideration of serious AEs deemed to be related to the participants TKI [serious adverse reactions (SARs)], which were collected for the duration of the trial, the difference, in favour of CCS, is more evident prior to week 24 when both strategies are the same (overall – CCS: 18.8%, n = 485, DFIS: 23.4%, n = 431, pre CCS: 13.2%, n = 485, DFIS: 19.3%, n = 431, post CCS: 11.7%, n = 265, DFIS: 9.4%, n = 223). The majority of SAEs and SARs resulted in recovery or recovery with sequelae (SAEs: 80.1%, n = 744, SARs: 87.6%, n = 226). There were 13 suspected unexpected serious adverse reactions (CCS: 5, DFIS: 8) on the trial. Substudies The qualitative substudy concluded that if a DFIS approach was implemented in practice some thought should be given to how patients should be supported during the extended break to cope with and alleviate worries related to ongoing monitoring and tumour growth. The magnetic resonance imaging (MRI) feasibility study demonstrated that DCE-MRI-derived biomarkers of tumour perfusion (perfused tumour volume, Ktrans and extracellular volume) as potential surrogate biomarkers to predict disease progression following TKI therapy in metastatic RCC and the technique of DCE-MRI assessment is reproducible. The CT substudy demonstrated that assessment of enhancement as well as size change alters the categorisation of response. Use of modified Choi (mChoi) criteria may allow for earlier detection of disease progression, and more representative separation of participants with partial response versus stable disease. While published literature has suggested an association of mChoi criteria with time to progression, in terms of prediction of early progression within 6 months of treatment initiation, no association was shown in this cohort. Conclusions Implications for health care The STAR trial is an exemplar trial, being one of the first to be powered on a QALY end point. It demonstrated no substantial detriment in terms of OS and QALY from a DFIS compared to a CCS. However, NI between the two arms cannot be concluded due to a lack of statistical power for the analysis of OS, as a result of fewer than expected OS events due to the changing landscape of effective treatments for relapsed advanced RCC during the trial. These results provide evidence-based reassurance for people who received National Institute for Health and Care Excellence (NICE)-approved treatment breaks in this setting during the COVID-19 pandemic. Through compliance with the protocol, treatment breaks within the trial were shown to be acceptable to patients and health professionals. Future research implications Further discussion is required with patient and public involvement (PPI) and NICE regarding the implications of these results after the NHS interim treatment options following the COVID-19 pandemic end. The STAR trial has provided significant learning regarding the development, design and delivery of intermittent therapy trials. Trial registration This trial is registered as ISRCTN06473203. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment Programme (NIHR award ref: 09/91/21) and is published in full in Health Technology Assessment; Vol. 28, No. 45. See the NIHR Funding and Awards website for further award information.

Background Second trimester miscarriage and preterm birth is a significant global problem. Surgical cervical cerclage is performed to prevent pregnancy loss and preterm birth. It utilises either a monofilament or braided suture. It is hypothesised that a braided material becomes colonised with pathogenic bacteria that causes vaginal dysbiosis, infection and cerclage failure. Objectives The primary objective of the study was to examine the effectiveness of using a monofilament suture material as opposed to a braided suture material on pregnancy loss in women requiring a vaginal cervical cerclage. Design Superiority open randomised controlled trial. Setting Seventy-five maternity sites across the UK. Participants Women experiencing a singleton pregnancy requiring a cervical cerclage. Interventions Monofilament suture or braided suture. Main outcome measures The primary outcome was pregnancy loss (miscarriage and perinatal mortality, including any stillbirth or neonatal death in the first week of life). Secondary outcomes included the core outcome set for preterm birth. Methods Women were randomised on a 1 : 1 basis to monofilament or braided cerclage utilising a bespoke randomisation service with minimisation dependent on the site, indication for cerclage, intention to use progesterone and planned surgical technique. The inclusion criteria were three or more previous mid-trimester losses or preterm births, insertion of a cerclage in a previous pregnancy, a history of a mid-trimester loss or preterm birth with a shortened cervical length in the current pregnancy or in women who clinicians deemed at risk of preterm birth. The exclusion criteria were an emergency or rescue cerclage, age of < 18 years, being unable to give informed consent or the cerclage having to be placed abdominally. The original sample size was calculated based on a relative risk reduction of 41% from a pregnancy loss rate of 19% in the braided group to 11% in the monofilament group with 90% power and alpha at p = 0.05. The independent data monitoring committee noted a lower-than-anticipated pooled event rate within the trial and recommended an increase in sample size to 2050. The outcome data were collected using clinical record forms from the maternal and neonatal medical records and reported to Birmingham Clinical Trials Unit. Results A total of 2049 women were randomised, after withdrawals and loss to follow-up, data on 1005 women in the monofilament group and 993 women in the braided group were included. The baseline demographics between the groups were similar. There was no evidence of a difference in pregnancy loss rates between the monofilament and braided groups (80/1003 vs. 75/993; adjusted risk ratio: 1.05, 95% confidence interval: 0.79 to 1.40; adjusted risk difference: 0.002, 95% confidence interval: −0.02 to 0.03). Limitations The trial did not collect long-term paediatric outcomes. There were no safety concerns. Conclusions There was no evidence of a difference in pregnancy loss between a monofilament suture and a braided suture. Future work Long-term follow-up of neonates born within the C-STICH (cerclage suture type for an insufficient cervix and its effects on health outcomes) trial. Trial registration This trial is registered as ISRCTN15373349. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 13/04/107) and is published in full in Health Technology Assessment; Vol. 28, No. 40. See the NIHR Funding and Awards website for further award information. Plain language summary Cervical cerclage is an operation performed in pregnancy to prevent miscarriage and preterm birth. A cervical cerclage is sometimes recommended in women who have had babies born prematurely before or who have had previous cervical surgery. A cerclage operation involves a stitch being inserted around the neck of the womb (cervix) to keep it closed during pregnancy and to prevent it opening prematurely. When performing the operation, the doctor can use different types of threads made of different materials. The threads used to perform the operation are called sutures. One suture type is a single strand or monofilament thread, and the other is a multifilament braided thread with lots of thin strands woven together. Some evidence has suggested that using a monofilament suture thread prevented pregnancy loss by preventing infection. Therefore, we performed a randomised controlled trial of the use of monofilament suture thread versus braided suture thread, aiming to reduce pregnancy loss in women who were having a cerclage as part of their routine care. The women consented to take part in the study and were randomly allocated to their cerclage performed with either a monofilament or braided suture thread; there was no other change to their planned pregnancy care. What happened in their pregnancy was recorded from their medical records and analysed. A total of 2049 women agreed to take part in the study and consented to the analysis of their pregnancy and neonatal outcomes. Cerclage suture type for an insufficient cervix and its effects on health outcomes showed that there was no difference in pregnancy loss between the two suture threads. There was decreased maternal sepsis and decreased chorioamnionitis (which is an infection inside the womb during labour) in the women who received a monofilament suture, which needs further investigation. Although more women who had a cerclage using the monofilament thread needed a small operation and an anaesthetic, often between 36 and 37 weeks, to remove the monofilament suture prior to a vaginal birth, there were no differences in the outcomes for their babies. Scientific summary Background Preterm birth and second trimester miscarriage is a significant cause of worldwide neonatal morbidity and mortality. The aetiology is complex and multifactorial with a number of causes, including cervical insufficiency. One treatment to prevent preterm birth and second trimester miscarriage caused by cervical insufficiency is the placement of a vaginal cervical cerclage (CC). A CC is the placement of a purse string suture thread around the cervix aiming to prevent pregnancy loss and preterm birth. A CC can be performed with either a monofilament suture thread or a braided suture thread. The choice of thread used during surgical operations is dependent on the properties of the thread, with braided threads being multifilament in nature with lots of single strands woven together, which predisposes it to potentially becoming colonised with pathogenic bacteria. A prior feasibility study suggested a difference in pregnancy loss outcomes between monofilament and braided suture threads and the acceptability of either thread for clinicians and women. Therefore, a randomised controlled trial was designed to explore this discrepancy, aiming to reduce pregnancy loss for women at high risk of preterm birth. Objectives The primary objective of the study was to examine the effectiveness of using a monofilament suture material compared to a braided suture material on minimising the risk of pregnancy loss in women requiring a vaginal CC. The secondary objectives of the study included exploring the impact of suture material on maternal and neonatal outcomes. Design A pragmatic open, parallel, multicentre, randomised superiority trial of monofilament versus braided suture type during CC to prevent pregnancy loss. Setting The study was conducted in hospital settings across the UK (75 sites) between 2015 and 2022. Participants Women were eligible for the trial if they required a vaginal CC as part of their routine care within their current pregnancy and they fulfilled the following eligibility criteria: Inclusion Singleton pregnancy. Indication for CC for either: a history of three or more previous mid-term losses or premature births (≤ 28 weeks), OR insertion of cervical sutures in previous pregnancies, OR a history of mid-trimester loss or premature birth with a shortened (≤ 25 mm) cervix, OR women whom clinicians deemed to be at risk of preterm birth either because of their history or because of the results of an ultrasound scan. Exclusion Had taken part in C-STICH previously. Aged < 18 years old at the time of presentation. Requiring a rescue cerclage. Unwilling or unable to give informed consent. Those in whom a cerclage was to be placed by any route other than vaginally (e.g. via an abdominal route). Immediate need for insertion of a suture. Membranes that had ruptured or were surfacing. Interventions Women were randomised at a 1 : 1 ratio via a secure internet facility to have their CC performed using either a monofilament or braided suture thread. Minimisation was employed to balance for the following: indication for cerclage, planned bladder dissection, intention to commence patient on progesterone and recruiting site. Outcome measures The primary outcome measure was pregnancy loss (miscarriage and perinatal mortality). Key secondary outcome: Time from conception to pregnancy end (any reason). Maternal outcomes: Miscarriage and previable neonatal death (defined as delivery < 24 weeks). Stillbirth (defined as intrauterine death ≥ 24 weeks). Gestation at delivery (in live births ≥ 24 weeks). Gestational age of < 28/< 32/< 37 weeks at delivery (in live births ≥ 24 weeks). Time from conception to onset of spontaneous vaginal delivery (in live births ≥ 24 weeks). Sepsis (at any time in pregnancy and until 7 days postnatal). Preterm prelabour rupture of membranes (PPROM). Mode of initiation of labour (spontaneous or induced). Mode of delivery (vaginal, operative vaginal or caesarean). Cerclage placement complications (cervical laceration/bleeding from cervix/ruptured membranes/bladder injury). Cerclage removal complications (cervical tears/need for anaesthetic/difficult to remove). Other maternal complications: vaginal bleeding/steroid use/chorioamnionitis/maternal pyrexia of 38°C (intrapartum/postnatal)/admission to high dependency unit (HDU) or intensive therapy unit (ITU) (pre/post delivery). Neonatal outcomes: Early neonatal death (defined as a death within 7 days after delivery) (in live births ≥ 24 weeks). Late neonatal death (defined as a death beyond 7 days and before 28 days after delivery) (in live births ≥ 24 weeks). Birth-weight centile adjusted for gestational age and sex (in live births ≥ 24 weeks). Small for gestational age and sex (< 10th centile, in live births ≥ 24 weeks). Resuscitation at birth (in live births ≥ 24 weeks). Additional care required [special care baby unit (SCBU)/neonatal intensive care unit (NICU)/HDU/transitional care] and length of stay in additional care (in live births ≥ 24 weeks). Antibiotics within 72 hours (in live births ≥ 24 weeks). Sepsis (clinically diagnosed/proven) (in live births ≥ 24 weeks). Early neurodevelopmental morbidity (severe abnormality on cranial ultrasound scan) (in live births ≥ 24 weeks). Respiratory support and days on respiratory support (in live births ≥ 24 weeks). Supplementary oxygen requirements at 36 weeks post menstrual age (in live births ≥ 24 weeks). Necrotising enterocolitis (Bell’s stage 2 or 3) (in live births ≥ 24 weeks). Retinopathy of prematurity requiring laser treatment (in live births ≥ 24 weeks). Disabilities (live births ≥ 24 weeks). Congenital anomalies (in live births ≥ 24 weeks). The core outcome set for preterm birth was fully collected within the trial. Sample size The original sample size was based on a meta-analysis of non-randomised studies, where the pooled pregnancy loss rate in the monofilament group was 7% compared to 19% in the braided group. To allow for the observational nature of these data, we powered the study to detect a more plausible relative reduction of 41% (19% with braided sutures to 11.2% with monofilament sutures) with 90% power (alpha = 0.05). We planned to recruit 900 women. As a result of some uncertainty around the sample size parameters, the data monitoring committee reviewed the pooled event rate throughout the study. The sample size was subsequently increased to 2050 women to allow for a lower-than-anticipated event rate in order to maintain sufficient power to detect the same relative risk reduction. The primary analysis was by intention to treat. Results The trial opened for recruitment in August 2015 and completed recruitment in January 2021. A total of 2049 women were randomised into the trial and data for analysis was available for 1003 women in the monofilament group and 993 women in the braided group. The baseline demographic characteristics of women in the monofilament and braided group were similar. There was no evidence of a difference in pregnancy loss rates between the monofilament and braided groups [80/1003 vs. 75/993; adjusted risk ratio (RR) 1.05, 95% confidence interval (CI) 0.79 to 1.40; adjusted risk difference 0.002, 95% CI −0.02 to 0.03]. There was no difference in conception to pregnancy end (median time to pregnancy end: 37.9 weeks vs. 38.0 weeks; adjusted hazard ratio 1.04, 95% CI 0.95 to 1.14). Regarding maternal outcomes there was a decrease in maternal sepsis in the monofilament group (4%) compared to the braided group (7%) (RR 0.58, 95% CI 0.40 to 0.82) and a decreased risk of chorioamnionitis in the monofilament group (3%) compared to the braided group (6%) (RR 0.45, 95% CI 0.29 to 0.71). There was no difference in any neonatal outcomes. CC removal complications showed an increase in the monofilament group (56% vs. 42%, RR 1.25, 95% CI 1.15 to 1.36), with increased difficulty of removal and an increased need for anaesthetic for removal being the most common complications. Conclusions There was no evidence of a difference between a monofilament suture thread and a braided suture for pregnancy outcomes. We can be relatively confident that using a monofilament suture is unlikely to have a substantial impact on pregnancy loss compared to a braided suture. The uncertainty around our comparative estimate for this outcome is at most 2% in favour of the monofilament in absolute terms. While this margin may not completely rule out missing a clinically important difference, we consider this scenario to be unlikely. Therefore, clinicians should consider the relative merits and disadvantages of the physical suture properties when selecting the material to perform a vaginal CC. The trial was robustly conducted with minimal limitations. An important strength of the study is the recognition early in the trial that the event rate was lower than anticipated and, therefore, the sample size was increased allowing the effectiveness of a monofilament versus braided suture thread to be fully evaluated. Trial registration This trial is registered as ISRCTN15373349. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 13/04/107) and is published in full in Health Technology Assessment; Vol. 28, No. 40. See the NIHR Funding and Awards website for further award project information.

Background Differences in the way autistic children experience the world can contribute to anxiety and stress. Carol Gray’s Social Stories™ are a highly personalised intervention to support children by providing social information about specific situations in an individual story. Objectives This randomised controlled trial aimed to establish whether Social Stories are clinically effective and cost-effective in improving social responsiveness and social and emotional health in children on the autism spectrum in schools. Design A multisite pragmatic cluster randomised controlled trial comparing Social Stories with care as usual. Setting Eighty-seven schools (clusters) across Yorkshire and the Humber. Participants Two hundred and forty-nine children were randomised via a bespoke system hosted at York Trials Unit (129 Social Stories and 120 care as usual). Recruitment was completed in May 2021. Participants were children aged 4–11 years with a diagnosis of autism, alongside teachers, interventionists and caregivers. Recruitment was via schools, NHS trusts, support groups and local publicity. Intervention The intervention included training for educational professionals and caregivers covering psychoeducation and implementation of Social Stories. Stories were written around contextualised goals around the child’s need for social information. Interventionists read the Social Story™ with the child at least six times over 4 weeks during school. Main outcome measure The primary outcome was the Social Responsiveness Scale-2 completed by teachers at 6 months (the primary end point), which measures social awareness, cognition, communication and behaviour. Data were collected from caregivers and educational professionals at 6 weeks and 6 months through questionnaires. Blinding of participants was not possible. Results At 6 months, the estimated difference in expected teacher-reported Social Responsiveness Scale-2 T-score (the primary end point) was −1.61 (95% confidence interval −4.18 to 0.96, p = 0.220), slightly favouring the intervention group. The estimated differences for the parent-reported secondary outcomes at 6 months were small and generally favoured the control group except the measure of children’s quality-adjusted life-year (+ 0.001, 95% confidence interval −0.032 to 0.035) and parental stress (−1.49, 95% confidence interval −5.43 to 2.46, p = 0.460), which favoured the intervention group. Children in the intervention group met their individual goals more frequently than children who received usual care alone (0.97 confidence interval 0.21 to 1.73, p = 0.012). The intervention is likely to save small costs (−£191 per child, 95% confidence interval −767.7 to 337.7) and maintain a similar quality of life compared to usual care. The probability of Social Stories being a preferred option is 75% if the society is willing to pay £20,000 per quality-adjusted life-year gained. Limitations include considerable disruptions during the coronavirus disease 2019 pandemic. Conclusion Social Stories are used in schools and represent a low-cost intervention. There is no clinically evident impact on social responsiveness, anxiety and/or depression, parental stress or general health. Benefits were observed for specific behavioural goals as assessed by the teacher, and Social Stories may serve as a useful tool for facilitating dialogue between children and school staff to address specific behavioural challenges. Usage should be at the school’s discretion. Future work Given the uncertainty of the results in light of coronavirus disease 2019, further work to establish the impact of Social Stories is merited. Trial registration This trial is registered as ISRCTN11634810. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 16/111/91) and is published in full in Health Technology Assessment; Vol. 28, No. 39. See the NIHR Funding and Awards website for further award information. Plain language summary Autism affects the way children experience the world, and some children find social situations stressful. We wanted to know whether Social Stories™, developed by Carol Gray, helped children with their social skills and behaviour in school and whether they offered value for money. A randomised controlled trial design was used, which gave schools an equal chance of being asked to deliver Social Stories or to continue providing care as usual. Two hundred and forty-nine children from 87 schools took part and we trained school staff and parents to write and deliver Social Stories. We agreed with teachers and parents, what each child needed help with and wrote stories with this in mind. Trained staff read the Social Story with the child at least six times over 4 weeks. Follow-up information was collected from parents and school staff at the start of the study, after 6 weeks and 6 months. After 6 months, teachers completed a questionnaire called the Social Responsiveness Scale-2 which measures the child’s social skills. Using these measures, the results suggest that Social Stories do not lead to any significant changes in social skills, mental health, parent stress, general health or quality of life but children in schools allocated to Social Stories met their goal more frequently and incurred less costs than children who did not. Parents and educational professionals found the Social Stories intervention and training beneficial. Based on our findings, Social Stories do not appear to improve general social skills in primary-aged autistic children. Benefits were observed for specific goals, and school-based costs were reduced. Scientific summary Sections of this summary have been adapted from the trial protocol by Wright et al. (Wright B, Teige C, Watson J, Hodkinson R, Marshall D, Varley D, et al. Autism Spectrum Social Stories In Schools Trial 2 (ASSSIST2): study protocol for a randomised controlled trial analysing clinical and cost effectiveness of Social Stories™ in primary schools. BMC Psychol 2020;8:60. https://doi.org/10.1186/s40359-020-00427-z). This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution and reproduction in any medium provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article unless otherwise stated. Background Children on the autism spectrum experience a higher prevalence of mental health problems than typically developing children, including anxiety and low mood. Many children on the autism spectrum struggle to manage social anxiety and feelings of frustration, which can lead to behaviours that challenge them. The International Society for Autism Research (INSAR) has highlighted that more research evaluating early interventions for children on the autism spectrum is needed to ensure practitioners and policy-makers have robust data on intervention effectiveness and implementation to secure optimal outcomes for children. One intervention that attempts to alleviate social difficulties while not being intrusive, time-consuming or requiring extensive involvement from experts is Carol Gray’s Social Stories™. Social Stories are a highly personalised intervention aiming to share accurate, meaningful information about a particular goal or topic that the child needs help with in a positive and reassuring way. Social Stories can be written and delivered by both parents and professionals in a range of settings and represent a less time-consuming and intrusive intervention than alternatives. Previous studies examining the use of Social Stories have yielded mostly positive results but have largely been single-case studies with a lack of evidence from randomised controlled trials (RCTs). Despite the lack of rigorous evidence, numerous schools and families are already accessing Social Stories training and delivering the intervention with children and therefore a fully powered trial is timely. Schools have limited resources and limited access to specialist practitioner interventions, and therefore it is important that interventions such as Social Stories undergo robust evaluation. If they are found to be clinically effective and cost-effective, they can be delivered within schools on a day-to-day basis. Objectives The aim was to assess whether Social Stories alongside care as usual is clinically effective and cost-effective in improving child social impairment, reducing anxiety and improving social and emotional health in children on the autism spectrum in primary and special educational needs (SEN) schools when compared with care as usual alone. Primary objective The primary objective of the study was to establish whether Social Stories can improve social responsiveness in children on the autism spectrum in primary schools across Yorkshire and the Humber, when compared to children who have received care as usual only. Social responsiveness can be broadly defined as social awareness, social cognition, social communication, social motivation and mannerisms. Secondary objectives The secondary objectives of this trial were: to investigate whether Social Stories can reduce behaviours that challenge children on the autism spectrum in primary schools to investigate whether Social Stories can improve social and emotional health in children on the autism spectrum in primary schools to assess the cost effectiveness of Social Stories to examine the effects of Social Stories delivered in the classroom on general measures of health-related quality of life to examine whether Social Stories improve classroom attendance to assess sustainability of Social Stories in an educational setting across a 6-month period to examine any changes in parental stress to examine any associations between treatment preference and outcomes to examine how elements of session delivery (e.g. session frequency, length and any associated problems/adverse events) are associated with outcomes. Methods This trial was a multisite pragmatic cluster RCT comparing Social Stories and care as usual with a control group receiving care as usual alone. Care as usual is defined as the existing support routinely provided for a child with autism spectrum condition (ASC) from educational and health services such as specialist autism teacher teams, mental health teams or other associated professionals. The trial included an internal pilot, economic evaluation and a nested process evaluation. Setting The Social Stories intervention was primarily delivered within educational settings by educational professionals. Educational settings included both mainstream primary schools and SEN schools. Parents/caregivers were invited to receive Social Stories training and had the option of also delivering the intervention within the home. Participants Inclusion criteria The child was aged 4–11 years at the time of recruitment. The child attended a participating primary or SEN school within Yorkshire and the Humber. The child has a clinical diagnosis of ASC and daily challenging behaviour. Parents/guardians of the child were able to self-complete the English language outcome measures (with assistance if necessary). Exclusion criteria The school had used Social Stories for any pupil in the current or preceding school term. The child or interventionist teacher had taken part in the previous Social Stories feasibility study (ASSSIST-1). Schools that have taken part were not excluded. Where families were recruited but the child’s school was unwilling or unable to participate, families were notified by a telephone call and/or an e-mail. Where children were confirmed eligible and consent had been taken from all relevant parties, baseline questionnaires were distributed to parents/carers and educational professionals. A school was considered ‘ready to randomise’ once all consent forms and baseline questionnaires had been received by the trial team. Interventions Children in the intervention arm of the trial received the Social Stories intervention in addition to their care as usual. The Social Stories intervention was delivered by a trained educational professional (the interventionist) who was employed within each school allocated to the intervention arm. The interventionist varied between the schools [e.g. a teacher, teaching assistant (TA) or Special Educational Needs Coordinator (SENCO)] but was most typically a TA. A core aspect of the intervention was to first agree a goal around which the story would be set. The goal was typically a behavioural challenge the child was struggling with, for example, sharing with their teacher how they were feeling. This goal was agreed during a collaborative ‘goal-setting meeting’ attended by the child’s teacher, a parent/caregiver, a member of the research team and sometimes the child’s TA. Occasionally, the children themselves were able to feed into the goal-setting process if deemed appropriate. The intervention is designed to provide social information to the child, and so the goal sought to reflect this. In this way, many goals sought to reduce the child’s anxiety or frustration by equipping them with information or providing reassurance around appropriate behaviours within a given circumstance. We assessed goal attainment in terms of how frequently a child was able to implement the desired behaviours, for example, if a goal for a particular child was to use calming strategies when they were upset, we asked teachers to rate how often a child was able to do this. Interventionists were trained by members of the research team, who had received training via a cascade model overseen by child psychiatrist, Professor Barry Wright. The training of interventionists included some psychoeducation around the differences in children on the autism spectrum to facilitate understanding around what factors may be driving the observed behaviours associated with the behavioural goal. Training also provided key information on the design and implementation of Social Stories, with materials based on those developed in the preceding feasibility study with the support of Carol Gray and a Social Stories manual produced by Professor Barry Wright and a Clinical Psychologist with expertise in autism. During the training session, interventionists constructed a Social Story. Parents/guardians were also invited to attend these sessions. Following training, all Social Stories were assessed against a fidelity checklist by a member of the research team to ensure they conformed to the 10 established criteria central to Carol Gray’s Social Stories. They were then delivered to children in the intervention arm by the interventionist at least six times over a 4-week period. Sample size The primary outcome was that the teacher completed the Social Responsiveness Scale, Second Edition (SRS-2) T-score at 6 months. Within the pilot data, outcomes were measured at 6 and 16 weeks. The correlation between baseline and 6 weeks [r = 0.67, 95% confidence interval (CI) 0.44 to 0.80] was lower than that at 16 weeks (r = 0.83, 95% CI 0.68 to 0.91) for the pilot data. To be conservative, the lower 95% confidence limit was chosen for the lowest correlation between baseline and follow-up that we observed within our pilot data (r = 0.44). Assuming a difference of 3 points, standard deviation (SD) = 7, 5% alpha, 90% power, average cluster size 1.35, intracluster correlation coefficient (ICC) = 0.34, correlation = 0.44 and 25% attrition, a total sample size of 278 was required. Support was gained from the trial steering committee and funder to stop recruitment with 249 children, which we modelled would still retain at least 80% power under a variety of different scenarios. Randomisation Randomisation was completed by unblinded members of the trial team via a bespoke trial management system. Stratified blocked randomisation was used to allocate school clusters, with randomly varying block sizes (4, 6 and 8), stratification by school type (SEN school or mainstream school) and the number of participating children within a school (≤ 5 or > 5 participating children). There were no stipulations regarding the minimum number of participants per school cluster (i.e. any school with ≥ 1 eligible/consented child with available baseline data was eligible for randomisation). Main outcome measure The primary outcome of the trial was the SRS-2 at 6 months post randomisation, as reported by the child’s teacher. This was also collected at 6 weeks post randomisation. The SRS-2 identifies the presence and severity of social difficulty within the autism spectrum condition and consists of 65 questions. For each question, the person completing the form picks a score from 1 to 4 (1 = not true, 2 = sometimes true, 3 = often true, 4 = almost always true) that best describes the child’s behaviour. A T-score is calculated based on the sex of the child and the person completing the form (teacher). Economic evaluation The primary analysis for the economic evaluation was a within-trial cost-utility analysis conducted from a societal perspective. Combining costs and quality-adjusted life-years (QALYs), an incremental cost-effectiveness ratio (ICER) of cost per QALY was calculated and evaluated against the willingness-to-pay threshold of £20,000–30,000 per QALY gained to assess the cost effectiveness of Social Stories compared to usual care. Regression models on an intention-to-treat basis were used to compare mean costs and QALYs. To take uncertainty into consideration, a non-parametric bootstrap resampling method was used to produce CIs around the cost and QALY differences and ICER. The following sensitivity analyses were conducted to test assumptions made in the primary analysis: a complete case analysis; a sensitivity analysis from the UK NHS and personal social services (PSS) perspective; and a sensitivity analysis from the joint perspectives of the NHS/PSS and education perspectives to representative a global public sector perspective. Process evaluation The process evaluation was cross-sectional and longitudinal, encompassing all aspects of the Social Stories intervention. The aim of the process evaluation was to assess the fidelity of the programme, consider the views of various stakeholders and identify barriers and facilitators to successful implementation. We aimed to achieve this through a combination of data collection techniques, including interviews, focus groups, questionnaires, surveys and diaries (session logs). Statistical analyses For the primary outcome, differences in expected SRS-2 scores were estimated using a linear mixed-effect covariance pattern model with both post-randomisation time points (6 weeks and 6 months) included as outcomes and fixed effects for treatment group, time point and their interaction. Further fixed effects were included for the following cluster and participant-level baseline covariates: school SEN status (binary, SEN/non-SEN), number of consented children attending school (binary, ≤ 5/> 5), baseline score (linear term), age at randomisation (linear term) and sex (binary, female/male). Dependence between participants within a cluster was modelled using school cluster-level random intercepts, and dependence between repeated measurements within participants was modelled using an unstructured correlation matrix for the residual errors. Similar models were used to analyse the secondary outcomes. Results The primary analysis included all 249 randomised children. After 6 months, a reduction of 1.61 points was found on the SRS-2 in children on the autism spectrum in the intervention group (95% CI −4.18 to 0.96; p = 0.220). No statistically significant differences were found in overall symptoms of anxiety and/or depression, parental stress or general health. Children in the intervention group met their individual goals more frequently than children who received usual care alone, and this difference was statistically significant (0.97, CI 0.21 to 1.73; p = 0.012). The primary analysis was conducted under the principles of intention to treat. Coronavirus disease 2019 (COVID-19) impacted upon delivery in schools and school routines, and only 62.8% of the sample is known to have received the intervention as per protocol. Sensitivity analyses suggested that there may be a dose effect with improved clinical effects in those receiving the intervention as per protocol compared to those with fewer sessions. The economic evaluation showed that, compared to usual care, Social Stories slightly decreased the service use costs over the 6-month period by £191 (95% CI −337.7 to 767.7) per child and maintained similar QALYs. If society is willing to pay £20,000 for extra QALY gained, then the probability of Social Stories being a dominant and preferred option is 75%. The results of both primary and sensitivity analyses, which considered costs derived from various perspectives, are consistent but limited to the given data within the short study time frame, number of missing data and disruption of COVID-19. The qualitative results suggest parents and educational professionals found the Social Stories training and intervention beneficial. Limitations include considerable disruptions during the COVID-19 pandemic. Conclusion We found no impact of Social Stories on autistic children’s overall social responsiveness skills. There was some evidence that Social Stories are well supported by parents and teachers, and it may be effective at addressing a wider set of individual goals and could reduce costs. Based on the evidence generated through this trial, we cannot recommend Social Stories for the purposes of improving social skills, anxiety and/or depression, parental stress, general health and quality of life in autistic children. The COVID-19 pandemic had a large impact on schools during this trial, and sensitivity analyses suggested that poor compliance may have reduced the effectiveness of the intervention and that effects were better in those receiving Social Stories as per protocol. We did not find any negative effects, and Social Stories are already frequently used in schools to support autistic children and represent a low-cost and potentially cost-saving intervention. Despite limited evident impact on global social skills, based on the data elicited through the process evaluation, it appears that Social Stories may serve as a useful tool for facilitating dialogue between children and school staff to better understand the needs of autistic children, and usage should be at the school’s discretion. Trial registration This trial was registered as ISRCTN11634810. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 16/111/91) and is published in full in Health Technology Assessment; Vol. 28, No. 39. See the NIHR Funding and Awards website for further award information.

Background Selected patients with advanced heart failure ineligible for heart transplantation could benefit from left ventricular assist device therapy as ‘destination therapy’. There is evidence of the efficacy of destination therapy; however, it is not currently commissioned within the United Kingdom National Health Service due to the lack of economic evidence. Objective What is the clinical and cost-effectiveness of a left ventricular assist device compared to medical management for patients with advanced heart failure ineligible for heart transplantation (destination therapy)? Methods A systematic review of evidence on the clinical and cost-effectiveness of left ventricular assist devices as destination therapy was undertaken including, where feasible, a network meta-analysis to provide an indirect estimate of the relative effectiveness of currently available left ventricular assist devices compared to medical management. For the systematic reviews, data sources searched (up to 11 January 2022) were Cochrane CENTRAL, MEDLINE and EMBASE via Ovid for primary studies, and Epistemonikos and Cochrane Database of Systematic Reviews for relevant systematic reviews. Trial registers were also searched, along with data and reports from intervention-specific registries. Economic studies were identified in EconLit, CEA registry and the NHS Economic Evaluation Database (NHS EED). The searches were supplemented by checking reference lists of included studies. An economic model (Markov) was developed to estimate the cost-effectiveness of left ventricular assist devices compared to medical management from the United Kingdom National Health Service/personal social service perspective. Deterministic and probabilistic sensitivity analyses were conducted to explore uncertainties. Where possible, all analyses focused on the only currently available left ventricular assist device (HeartMate 3TM, Abbott, Chicago, IL, USA) in the United Kingdom. Results The clinical effectiveness review included 134 studies (240 articles). There were no studies directly comparing HeartMate 3 and medical management (a randomised trial is ongoing). The currently available left ventricular assist device improves patient survival and reduces stroke rates and complications compared to earlier devices and relative to medical management. For example, survival at 24 months is 77% with the HeartMate 3 device compared to 59% with the HeartMate II (MOMENTUM 3 trial). An indirect comparison demonstrated a reduction in mortality compared to medical management [relative risk of death 0.25 (95% confidence interval 0.13 to 0.47); 24 months; this study]. The cost-effectiveness review included 5 cost analyses and 14 economic evaluations covering different generations of devices and with different perspectives. The reported incremental costs per quality-adjusted life-year gained compared to medical management were lower for later generations of devices [as low as £46,207 (2019 prices; United Kingdom perspective; time horizon at least 5 years)]. The economic evaluation used different approaches to obtain the relative effects of current left ventricular assist devices compared to medical management from the United Kingdom National Health Service/personal social service perspective. All gave similar incremental cost-effectiveness ratios of £53,496–58,244 per quality-adjusted life-year gained – lifetime horizon. Model outputs were sensitive to parameter estimates relating to medical management. The findings did not materially differ on exploratory subgroup analyses based on the severity of heart failure. Limitations There was no direct evidence comparing the clinical effectiveness of HeartMate 3 to medical management. Indirect comparisons made were based on limited data from heterogeneous studies regarding the severity of heart failure (Interagency Registry for Mechanically Assisted Circulatory Support score distribution) and possible for survival only. Furthermore, the cost of medical management of advanced heart failure in the United Kingdom is not clear. Conclusions Using cost-effectiveness criteria applied in the United Kingdom, left ventricular assist devices compared to medical management for patients with advanced heart failure ineligible for heart transplant may not be cost-effective. When available, data from the ongoing evaluation of HeartMate 3 compared to medical management can be used to update cost-effectiveness estimates. An audit of the costs of medical management in the United Kingdom is required to further decrease uncertainty in the economic evaluation. Study registration This study is registered as PROSPERO CRD42020158987. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128996) and is published in full in Health Technology Assessment; Vol. 28, No. 38. See the NIHR Funding and Awards website for further award information. Plain language summary The majority of patients with advanced heart failure would be unsuitable for heart transplantation due to their age and comorbidities but selected patients could benefit from a left ventricular assist device. Left ventricular assist device therapy for such patients is known as ‘destination therapy’. This is a long-term therapy that involves implanting a battery-powered pump to support the patient’s heart. The purpose of this project was to collect and assess the research evidence on the effectiveness of left ventricular assist devices when used for destination therapy, and to estimate value for money compared to medical management from the United Kingdom National Health Service/personal social service perspective. This research identified that the currently available left ventricular assist device improves patient survival as well as reducing stroke rates and complications compared to earlier devices and relative to medical management. However, there is uncertainty in the evidence due to the absence of studies directly comparing the current device to medical therapy alone. An ongoing clinical trial is currently assessing this. It also means there is uncertainty about whether left ventricular assist devices could provide value for money as determined currently for the United Kingdom National Health Service. Scientific summary Background Heart failure is a debilitating, progressive syndrome characterised by the inability of the heart to pump blood around the body. Pharmacological treatments are used as first-line treatment but may eventually become less effective and left ventricular assist devices (LVADs) or heart transplant (HT) are considered. LVADs are frequently used as bridge to transplant (BTT) or bridge to candidacy (BTC). However, some patients are ineligible for HT and either continue on medical management (MM) or could have a LVAD implanted as ‘destination therapy’ (DT). LVAD as DT is not currently commissioned within the United Kingdom National Health Service (UK NHS). The costs of LVADs are high, especially when compared to the alternative MM, but may also offer significant benefit in terms of survival. It is important to determine whether LVADs are both clinically and cost-effective as DT to inform decision-making from the UK NHS/personal social service (PSS) perspective on their potential as long-term treatment for advanced heart failure patients ineligible for HT. Aims and objectives What is the clinical and cost-effectiveness of a LVAD compared to MM for advanced heart failure (AHF) patients ineligible for HT (DT)? The specific objectives to address this aim were to undertake: – a systematic review of available evidence on the clinical effectiveness of a LVAD as DT, including a network meta-analysis (NMA) to provide an indirect estimate of the relative effectiveness of currently available LVADs compared to MM; – a systematic review of available economic evidence on the use of a LVAD as DT; and – the development of an economic model to estimate the cost-effectiveness of a LVAD compared to MM from the UK NHS/PSS perspective. Due to the withdrawal of the HeartWare ventricular assist device (HVAD) during the undertaking of this research, the analyses primarily focus on the HeartMate 3™ (Abbott, Chicago, IL, USA) device, the only LVAD available in the UK at this time. Methods Systematic review of clinical effectiveness A systematic review was undertaken of all LVADs as DT and reporting followed the general principles of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. The review was registered on PROSPERO (CRD42020158987). Eligibility criteria Studies of patients over 16 years of age with AHF who received any type of LVAD as DT were included. The review considered all devices, but the analysis focused on the HM3 device due to the recent withdrawal of the HeartWare HVAD (Medtronic, Dublin, Republic of Ireland). Eligible comparators (where relevant) were MM and other LVADs. Outcomes were survival, quality of life (QoL), hospitalisations, major events, complications and functional status. Study designs eligible were any clinical trial (whether randomised, non-randomised or single arm), observational studies (cohort, case-controls and case series) and reports from patient registries [e.g. INTERMACS, International Registry for Mechanically Assisted Circulation (IMACS)]. Studies were eligible if 50 or more DT patients were included. Systematic reviews were included and used to identify any additional potentially relevant primary studies. Searches and study selection Databases were searched from inception to 20 May 2020, with an updated search on 11 January 2022. Databases searched included Cochrane Library (CENTRAL), MEDLINE and EMBASE via Ovid, Epistemonikos, Cochrane Library of Systematic Reviews and World Health Organization (WHO) clinical trials portal (for ongoing studies). There were no restrictions by language or date of publication. Two reviewers independently undertook title and abstract screening and full-text selection via Covidence (Veritas Health Innovation, Melbourne, VIC, Australia). Disagreements were resolved by a third reviewer or consensus and reasons for exclusion were recorded. Risk of bias, data extraction and synthesis Quality assessment and data extraction were completed by one reviewer and checked by a second. Appropriate risk-of-bias tools dependent upon study design were applied. A hierarchical approach to synthesis was undertaken to avoid double-counting of studies with overlapping patient data and to manage the large volume of evidence. Randomised controlled trials (RCTs) and controlled non-randomised trials were considered in the first instance. Registry reports and uncontrolled observational studies were used to supplement findings for all outcomes. Data were tabulated and analysed in a narrative approach by device, and forest plots without summary estimates were presented (and where appropriate the feasibility of meta-analysis was considered). A network meta-analysis was considered for the main outcomes to produce an indirect comparison of the HM3 device (across LVAD generations) to MM, but only carried out for survival. Systematic review of cost-effectiveness A systematic review of the cost-effectiveness of LVADs was carried out utilising the same search strategy, and at the same time as the clinical effectiveness review, with the addition of three further specialist economics database searches in EconLit, Cost-Effectiveness Analysis (CEA) registry and the NHS Economic Evaluation Database (NHS EED). Appropriate risk-of-bias tools were applied and a narrative synthesis was undertaken. Economic evaluation The systematic reviews’ findings were used to inform the development of a cost–utility analysis (CUA), from the NHS/PSS perspective, using a Markov model with a lifetime horizon and 1-month cycles. Along with evidence from the reviews, the model was informed by guidance from clinical specialists, patients and commissioners. All costs used were in 2019 prices, and a discount rate of 3.5% was applied as per the national UK guidelines. To produce the base case, mortality risks for MM and LVAD arms required some assumptions and several methods for estimating the risks were identified. Two of these were primarily utilised: non-comparative net weight estimates and comparative estimates mapped to LVAD data from the recent relevant HM3 trial (MOMENTUM). The analysis was repeated incorporating a small probability of LVAD DT recipients transitioning to HT eligibility. The potential impacts of the severity of heart failure on cost-effectiveness were explored by considering subgroupings of profiles based on the INTERMACS classification. Uncertainty was explored via both deterministic and probabilistic sensitivity analyses, paying specific attention to the life expectancy and ongoing costs. Results Systematic review of clinical effectiveness There were 240 articles from 134 studies included in the clinical effectiveness review (5 randomised trials, 1 non-randomised trial, 86 observational studies, reports from 5 registries, 5 ongoing studies and 32 systematic reviews). Of the six trials that were included, only one of these assessed the HM3 and this was in comparison to the previous generation HeartMate II device (MOMENTUM RCT). The majority of HM3 data comes from this trial, with minimal additional data contributions from registry reports or observational studies of single cohorts. The MOMENTUM study was considered as having some concerns regarding risk of bias; however, this was primarily due to the per-protocol analysis for the DT participants and most other domains were considered low risk. There were 624 DT patients in the MOMENTUM 3 trial in total, with a mean age of 63 [standard deviation (SD) 12], 82.2% male and 52.1% INTERMACS level 3. At the longest follow-up point (24 months) survival was 76.7% in HM3 DT patients compared to 59% in HeartMate II patients. Clear and significant improvements in QoL from baseline were reported at 12 months and maintained at 24 months in the HM3 group; however, this was similar in the HeartMate II group. Major events and complications were present in both groups by the 24-month follow-up. There were eight stroke events per 100 patient-years in the HM3, as well as one pump thrombosis event and 70 bleeding events per 100 patient-years. These were all lower than that of the HeartMate II group. Rehospitalisations were also significantly lower in HM3 patients. While some reports included HM3 patients, there were no HM3 specific data reported in any patient registry reports. One observational study reported that HM3 patients (n = 15) had 0 pump thrombosis events in 24 months of follow-up. While it was not the focus due to withdrawal, survival levels were lower in the HeartWare HVAD trials when compared to the HM3 in the MOMENTUM trial and there were concerns with the stroke rates reported in the evidence. Risk of bias across the included trials varied with all but two studies reporting an overall high risk of bias or with some concerns for at least one outcome. The evidence contained within the remaining trials, observational studies and registry reports mostly relate to devices other than HM3. This evidence is summarised in the main part of this report. Indirect comparison of HeartMate 3 and medical management As there were no studies directly comparing HM3 and MM, indirect comparisons of the trial data were required utilising MM data from the older REMATCH trial (the first RCT comparing the first-generation HeartMate device to MM). Data were available to link through available studies for the survival outcome only. The network meta-analysis demonstrated a reduction in the risk of mortality, relative risk of death of 0.25 [95% confidence interval (CI) 0.13 to 0.47] 24 months, with the HM3 compared to MM. Systematic review of cost-effectiveness There were 19 studies reported in 20 articles included in the cost-effectiveness review: 5 cost analyses and 14 economic evaluations. Nine studies were US-based and four were UK-based. Most of the studies aimed to compare the health and cost outcomes of LVADs with MM. Most economic evaluations (n = 12) used a CUA approach and only two conducted a CEA. Markov-based modelling was applied in eight studies. The perspective, where stated, was the service provider in most studies. Healthcare resource use was usually estimated based on small numbers of patients from a single centre, which resulted in variability. In the studies comparing LVAD with MM for DT patients, the incremental cost per quality-adjusted life-year (QALY) gained estimates ranged between £46,207 and £238,401 in 2019 prices over a time horizon of 5 years or longer and from different perspectives. The overall quality of the studies was considered poor to moderate. Some limitations were limited consideration of uncertainty, insufficient time horizon and lack of consideration of some key complications and cost components. Only one study looked at the impact of disease severity on cost-effectiveness. More recent evaluations tended to have lower estimates of incremental cost-effectiveness, presumably reflecting better clinical outcomes of more recent devices. Two recent studies estimated the cost-effectiveness from a UK perspective, deriving incremental cost-effectiveness ratios of £47,361 and £46,207 per QALY gained for the HeartWare device (device withdrawn in 2021) and the HM3 device, respectively compared to MM. Economic evaluation The economic evaluation found similar results for each base case: Non-comparative net weight estimates approach: LVAD would produce an additional 2.86 QALYs per person, increase life expectancy by 3.73 years and the incremental cost to the NHS would be £152,735 per person. Incremental cost-effectiveness ratio (ICER): £53,496. Comparative estimates mapped to LVAD in MOMENTUM approach: LVAD would produce an additional 2.51 QALYs per person, increase life expectancy by 3.06 years and the incremental cost to the NHS would be £146,275 per person. ICER: £58,244. At a willingness to pay threshold of £50,000 per QALY gained, LVADs would not be considered cost-effective compared to MM for AHF patients ineligible for HT. The same applied when severity weighted ICER estimates based on QALY shortfall methods were used. The deterministic sensitivity analysis showed that inclusion of the probability of becoming eligible for a HT did not change these findings. Furthermore, the findings did not differ in subgroup analyses based on severity of heart failure. Model outputs were most sensitive to estimates related to outpatient costs for both LVAD and MM. Conclusions LVADs have significantly improved over time and the currently available HM3 LVAD is considered clinically effective in patients with end-stage heart failure ineligible for transplant, offering survival of over 75% at 2 years of follow-up with reduced complications and major events in comparison to older devices. However, the device compared to MM may not be considered cost-effective when using methods of defining this for end of life in the UK. Future research Currently, no RCT has been published that compares the HM3 device to MM; however, there is an ongoing trial (SweVAD) comparing the two, which is due to complete final study follow-up in December 2023. This randomised trial, undertaken in Sweden, should allow for relative effects to be determined between the two interventions. This will ultimately enable more robust data to be used to update the current model, rather than relying upon indirect comparisons with wide uncertainty. However, further issues around the true cost of MM are still present due to the lack of recent data on these costs in the UK. An audit of MM costs in DT patients in the UK would address this. Issues also persist in developing reliable subgroup analyses based on severity profiles to aid identification of whether a LVAD is (more) cost-effective for some groups of DT patients. Future trials and other studies should report results by patient severity profiles (e.g. INTERMACS classification), and if registry/observational studies then also by device implanted. Study registration This study is registered as PROSPERO CRD42020158987. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR128996) and is published in full in Health Technology Assessment; Vol. 28, No. 38. See the NIHR Funding and Awards website for further award information.

Background Insomnia is a prevalent and distressing sleep disorder. Multicomponent cognitive–behavioural therapy is the recommended first-line treatment, but access remains extremely limited, particularly in primary care where insomnia is managed. One principal component of cognitive–behavioural therapy is a behavioural treatment called sleep restriction therapy, which could potentially be delivered as a brief single-component intervention by generalists in primary care. Objectives The primary objective of the Health-professional Administered Brief Insomnia Therapy trial was to establish whether nurse-delivered sleep restriction therapy in primary care improves insomnia relative to sleep hygiene. Secondary objectives were to establish whether nurse-delivered sleep restriction therapy was cost-effective, and to undertake a process evaluation to understand intervention delivery, fidelity and acceptability. Design Pragmatic, multicentre, individually randomised, parallel-group, superiority trial with embedded process evaluation. Setting National Health Service general practice in three regions of England. Participants Adults aged ≥ 18 years with insomnia disorder were randomised using a validated web-based randomisation programme. Interventions Participants in the intervention group were offered a brief four-session nurse-delivered behavioural treatment involving two in-person sessions and two by phone. Participants were supported to follow a prescribed sleep schedule with the aim of restricting and standardising time in bed. Participants were also provided with a sleep hygiene leaflet. The control group received the same sleep hygiene leaflet by e-mail or post. There was no restriction on usual care. Main outcome measures Outcomes were assessed at 3, 6 and 12 months. Participants were included in the primary analysis if they contributed at least one post-randomisation outcome. The primary end point was self-reported insomnia severity with the Insomnia Severity Index at 6 months. Secondary outcomes were health-related and sleep-related quality of life, depressive symptoms, work productivity and activity impairment, self-reported and actigraphy-defined sleep, and hypnotic medication use. Cost-effectiveness was evaluated using the incremental cost per quality-adjusted life-year. For the process evaluation, semistructured interviews were carried out with participants, nurses and practice managers or general practitioners. Due to the nature of the intervention, both participants and nurses were aware of group allocation. Results We recruited 642 participants (n = 321 for sleep restriction therapy; n = 321 for sleep hygiene) between 29 August 2018 and 23 March 2020. Five hundred and eighty participants (90.3%) provided data at a minimum of one follow-up time point; 257 (80.1%) participants in the sleep restriction therapy arm and 291 (90.7%) participants in the sleep hygiene arm provided primary outcome data at 6 months. The estimated adjusted mean difference on the Insomnia Severity Index was −3.05 (95% confidence interval −3.83 to −2.28; p 65 years), sex, baseline insomnia severity [Insomnia Severity Index (ISI) score

Abstract Background Peripheral artery disease is a common condition caused by narrowing/blockage of the arteries, resulting in reduced blood supply. Peripheral artery disease is associated with an increased risk of vascular complications, but early treatment reduces mortality and morbidity. Leg ulcers are long-lasting wounds, usually treated by compression therapy. Compression therapy is not suitable for people with peripheral artery disease, as it can affect the arterial blood supply. In clinical practice, people with peripheral artery disease are identified by measurement of the ankle–brachial pressure index using a sphygmomanometer and manual Doppler device. However, this method can be uncomfortable for people with leg ulcers and automated devices have been proposed as a more acceptable alternative. The objective of this appraisal was to summarise the clinical and cost-effectiveness evidence on the use of automated devices to detect peripheral artery disease in people with leg ulcers. Methods Clinical effectiveness To identify reports of relevant studies, we searched major electronic databases and scrutinised the information supplied by the manufacturers of the automated devices under investigation. Due to the lack of evidence on people with leg ulcers, we considered evidence from studies of any design assessing automated devices versus an acceptable reference device in any population receiving ankle–brachial pressure index assessment. We summarised information on diagnostic accuracy of the automated devices and level of agreement with the reference device. For each device, when data permit, we pooled data across studies by conducting random-effects meta-analyses using a Hierarchical Summary Receiving Operating Characteristics model. Cost-effectiveness An economic model comprising a decision tree (24 weeks) and Markov models to capture lifetime costs and quality-adjusted life-years associated with venous, arterial and mixed aetiology disease in leg ulcer patients. Analyses were conducted from a United Kingdom National Health Service and Personal Social Services perspective. Costs and quality-adjusted life-years were discounted at 3.5% per year. Deterministic and several probabilistic analyses were used to capture uncertainty surrounding a range of optimistic and pessimistic assumptions about the impact of automated tests on health outcomes (ulcer healing and requirement for invasive management of arterial disease). Results Clinical effectiveness From the 116 records retrieved by the electronic searches, we included 24 studies evaluating five devices (BlueDop Vascular Expert, BOSO ABI-System 100, Dopplex Ability, MESI ankle–brachial pressure index MD and WatchBP Office ABI). Two studies assessing people with leg ulcers found that automated devices often gave higher ankle–brachial pressure index readings than manual Doppler (underestimation of arterial disease). In the 22 studies involving people without leg ulcers, automated devices generally demonstrated good specificity and moderate specificity. Meta-analysis of 12 studies showed a pooled sensitivity of 64% (95% confidence interval 57% to 71%) and a pooled specificity of 96% (95% confidence interval 92% to 98%) for detection of peripheral artery disease. Cost-effectiveness Automated devices cost less than manual Doppler to deliver. However, increased risks of invasive treatment requirements for inappropriately compressed arterial/mixed ulcers due to false-negative results, and increased healing times due to delayed compression of false-positive test results mean that in most scenarios manual Doppler was less costly and had slightly higher quality-adjusted life-years than automated devices. Results are highly uncertain, dependent on many assumptions and should be interpreted cautiously. Limitations and conclusions The limited evidence identified for each automated device, especially in people with leg ulcers, and its clinical heterogeneity precludes any firm conclusions on the diagnostic performance and cost-effectiveness of these devices in clinical practice. Study registration This study is registered as PROSPERO CRD42022327588. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135478) and is published in full in Health Technology Assessment; Vol. 28, No. 37. See the NIHR Funding and Awards website for further award information. Plain language summary Leg ulcers are long-lasting wounds mostly caused by problems in blood flow in the veins, which are treated by applying bandages or stockings to create a ‘compression’ effect. However, compression should not be used in people with a condition called peripheral artery disease. To identify people with peripheral artery disease who should not receive compression therapy, health professionals perform a test called ‘ankle–brachial pressure index’, which involves taking blood pressure of the arms and ankles using a device called ‘Doppler ultrasound’. The procedure is time-consuming and people with leg ulcers often find it uncomfortable. Automated devices have been proposed as a more acceptable option for assessing leg ulcers. However, we need to know whether these devices produce reliable results and represent good value for money for the National Health Service. We found 24 clinical studies that assessed 5 automated devices to measure ankle–brachial pressure index. The type of patients and clinical setting varied between studies. Two studies assessed people with leg ulcers and showed that the automated devices tended to give higher readings than standard Doppler and, therefore, may underestimate the presence of peripheral artery disease. Results of the 22 studies assessing people without leg ulcers showed that the automated devices could correctly identify people who did not have peripheral artery disease but were less precise in identifying people with peripheral artery disease. However, there was not enough evidence to confirm if these devices are reliable enough to be used in clinical practice. Compared to manual Doppler, the automated devices were less costly to deliver in clinical practice but had increased costs due to potentially inaccurate results. Our evaluation required many assumptions about how the devices would be used in practice, and there were no data on their impact on patient outcomes. Results are highly uncertain and should be interpreted cautiously. Given current evidence, it is unlikely that automated tests are a convenient option for the National Health Service. Scientific summary Background Peripheral artery disease (PAD) is a highly prevalent atherosclerotic condition characterised by the narrowing of the peripheral arteries resulting in restriction of blood supply to the affected limb. Although PAD is frequently asymptomatic, it can cause complications that can range from intermittent claudication (pain on walking which is relieved by rest) to critical limb ischaemia. Up to one-quarter of people with symptomatic PAD may require intervention, and amputation may be necessary if it is left untreated. Leg ulcers are defined as wounds that occur below the knee and either on or above the ankle (malleolus). Compression treatment (bandages or stockings) is recommended to treat venous leg ulcers, and there is a robust evidence base to support its effectiveness. However, compression therapy should be avoided in people with leg wounds and symptoms of arterial insufficiency, as compression may cause damage by impairing the arterial supply to the ulcerated leg. To improve PAD diagnosis and decide the most suitable treatment, people with leg ulcers are assessed using ankle–brachial pressure index (ABPI) measurements. ABPI is usually measured using a sphygmomanometer and manual Doppler device, which requires expertise from the relevant operator/healthcare professional. The procedure can be protracted and unpleasant for those with leg ulcers. Automated devices may be advantageous in reducing the length of time taken to assess ABPI and, thereby, any associated discomfort for the patient. In addition, automated devices may potentially be more accurate than manual processes in detecting PAD, thus conferring additional benefits such as reduced time to treatment and improved outcomes for people with leg ulcers. Objectives The specific objectives of this assessment were to: Determine the diagnostic performance and clinical utility of automated devices available in United Kingdom (UK) clinical practice [BlueDop Vascular Expert (BlueDop Medical), boso ABI-system 100 (BOSCH + SOHN), WatchBP Office ABI (Microlife), WatchBP Office Vascular (Microlife)], MESI ABPI MD (MESI), MESI mTABLET ABI (MESI), Dopplex Ability Automatic ABI System (Huntleigh Healthcare) for assessing the presence of PAD in people with leg ulcers. Develop an economic model to assess the cost-effectiveness of the automated devices available in UK clinical practice for assessing the presence of PAD in people with leg ulcers. Methods Clinical effectiveness Comprehensive electronic searches of databases including MEDLINE, EMBASE, Cochrane Library Web of Science and CINAHL were conducted to identify relevant reports of published studies. Evidence was considered from studies of any design assessing the relevant automated devices versus standard clinical assessment using a manual Doppler device. Initially, the population of interest was people with leg ulcers requiring measurement of ABPI, but, due to the dearth of available evidence, it was broadened to any population receiving ABPI measurement. Data on the diagnostic performance of the automated devices including data on the level of agreement between ABPI readings from automated devices and those from the reference device were extracted from the included studies. Information on the use of the devices in clinical practice was also recorded. Risk of bias was assessed using the Quality Assessment of Diagnostic Accuracy Studies – version 2 (QUADAS-2), QUADAS-C and the Review Body for Interventional Procedures (ReBIP) checklists, according to the type of study design. For each device, when sufficient data were available, we conducted random-effects meta-analyses using a Hierarchical Summary Receiving Operating Characteristic (HSROC) model. A two-stage, de novo decision analysis model was developed to assess cost-effectiveness. The first part was a decision tree model, which used a linked-evidence approach to capture the impact of test diagnostic accuracy on expected costs and quality-adjusted life-years (QALYs) for the first 24 weeks following test use. This included delayed venous ulcer healing due to false-positive (FP) test results (indicating PAD when the ulcer was venous) and increased risk of requiring invasive arterial treatment for inappropriately compressed arterial/mixed ulcers following a false-negative (FN) test result (indicating venous when underlying disease was arterial/mixed). It was assumed that any inaccurate tests would be identified within the 24-week time horizon of the decision tree. The surviving proportion of the cohort then entered arterial, mixed or venous ulcer Markov models depending on their true underlying disease classification. The venous disease model included five mutually exclusive health states, centred around ulcer healing (healed index ulcer, unhealed index ulcer, recurrence, healed post recurrence and death). The arterial and mixed disease models included four health states, focusing on the long-term outcomes of the arterial component of disease [critical limb ischaemia (CLI), healed post CLI, amputation and death]. The decision to structure the mixed Markov model similarly to the arterial-only model was based on discussion with clinical experts who explained that, in clinical practice, the arterial component of disease is likely to take priority in the patient’s care pathway. Costs were based on National Health Service and Personal Social Service perspective costs (2021 values) and included: micro-costing of the automated and manual Doppler devices costs of applying compression for the unhealed duration of a venous ulcer costs of referral to vascular services for test-positive patients, including the additional costs of unnecessary referral for patients with a FP test result costs of treating arterial disease, including endovascular and bypass procedures as well as follow-up nursing care long-term follow-up costs in the Markov model included the cost of managing recurrent venous ulcers, recurrent CLI and long-term health and social care costs of amputation. Health state utility values were obtained from the literature and were based on EuroQol-5 Dimensions data, valued using the UK value set where possible. Utilities were combined with mortality estimates for each health state to calculate QALYs. In the decision tree, utilities were dependent on the duration of ulcer healing time for venous ulcers, and whether patients had CLI for those with arterial/mixed disease. All utilities were adjusted for UK age- and sex-specific general population norms, allowing the cohort to experience reduced utility as they aged over subsequent model cycles. Expected costs and QALYs were accumulated over a lifetime horizon, in 6-monthly cycles and an annual discount rate of 3.5% per annum was applied to future costs and QALYs. Probabilistic analyses (Monte Carlo simulation with 1000 draws for each parameter) were conducted for a range of pessimistic and optimistic alternative base-case scenarios. A full range of deterministic scenarios explored the impact of alternative sources of model inputs and assumptions on cost-effectiveness results. Results Nature, description and quality of the available evidence The database searches identified 110 unique records, 79 records were supplied by the respective companies and 2 further studies were identified from reference lists. Twenty-four studies, published in 26 papers, were included in the systematic review of clinical effectiveness. Two studies enrolled specifically people with leg ulcers (167 participants in total) while the remaining studies (4258 participants in total) included people from primary care practices, cardiovascular risk services, vascular services and from epidemiological/general population-based studies. All studies used an ABPI threshold of 0.9. In healthy people, ABPI would be expected to be > 0.9. Most of the studies assessed the performance of a single automated device with only one study comparing two devices (WatchBP and MESI ABPI MD). Regarding the type of automated devices, two studies provided data on the BlueDop Vascular Expert device, four studies on the BOSO ABI-System 100, six studies on the Dopplex Ability, eight studies on the MESI ABPI MD and five studies on the WatchBP Office. No studies assessed the performance of the WatchBP Office Vascular and the MESI mTABLET ABI devices. Apart from one study conducted in New Zealand, all included studies were conducted in Europe (six in the UK). The risk of bias of included studies was assessed using the QUADAS-2 tool. Most studies were judged at low risk for the index test domain and at unclear risk for the patient selection, reference standard and flow and timing domains. The risk of applicability concerns was low in most studies. Summary of benefits and risks The two studies assessing people with leg ulcers did not provide sensitivity and specificity estimates but reported that automated devices gave generally higher readings than manual Doppler. The results of the 22 studies assessing people without leg ulcers varied. Seventeen studies reported sensitivity and specificity estimates for the detection of PAD and showed that the automated devices had good sensitivity but only moderate sensitivity indicating that a proportion of people with PAD would be missed. Sensitivity of BlueDop Vascular Expert ranged from 66% to 95% and specificity from 90% to 94% in two studies; sensitivity of the BOSO ABI-System 100 ranged from 61% to 77% and specificity from 94% to 98% in three studies; sensitivity of Dopplex Ability_ranged from 20% to 79% and specificity from 86% to 96% in four studies; sensitivity of the MESI ABPI MD ranged from 57% to 75% and specificity from 67% to 99% in five studies; sensitivity of the WatchBP Office ABI ranged from 44% to 83% and specificity from 97% to 100% in four studies; We were able to combine results across 12 studies (2004 participants in total) and 3 automated devices. The pooled sensitivity and specificity for PAD diagnosis using automated ABPI were 64% [95% confidence interval (CI) 57% to 71%] and 96% (95% CI 92% to 98%), respectively. Regarding the performance of individual devices, the pooled sensitivity for MESI ABPI MD was 67% (95% CI 59% to 74%) and the pooled specificity 94% (95% CI 83% to 98%); the pooled sensitivity for WatchBP Office ABI was 53% (95% CI 37% to 69%) and the pooled specificity 98% (95% CI 96% to 99%). For the remaining devices, we could not conduct meaningful meta-analyses due to the limited number of available studies. Summary of cost-effectiveness, including sensitivity analyses The uncertainties in the diagnostic accuracy evidence base and the unclear link between test results and patient management mean it is difficult to draw any firm conclusions on cost-effectiveness. A lack of evidence on the impact of the tests on important patient outcomes, the extent to which inaccurate test results would be identified in practice and the implications of acting on inaccurate test results contribute further uncertainty to the assessment of cost-effectiveness. Automated tests were less costly to deliver due to shorter testing times, but in most modelling scenarios, these cost savings were quickly offset by any additional risks and costs associated with withholding compression (FP) or inappropriately applying compression (FN). Given the current evidence base, it is unlikely that the automated tests would generate QALY gains or cost savings, unless a high proportion of FP and FN tests could be reliably identified in clinical practice through holistic patient assessment, and automated tests could deliver improvements in patient referral over manual Doppler testing. Discussion Strengths, limitations of the analyses and uncertainties The methods used to conduct this assessment were detailed, thorough and in line with current methodological standards. We identified only two studies assessing the performance of automated devices in determining ABPI in people with leg ulcers. Given the current lack of evidence in people with leg ulcers, we decided to widen our target population to include studies assessing the use of automated devices for measuring ABPI in different settings. We identified and summarised 22 studies focusing on people without leg ulcers. The main limitations of the clinical effectiveness assessment are summarised below. Lack of evidence on people with leg ulcers to draw any meaningful conclusion about this clinical population. Considerable clinical heterogeneity – in terms of characteristics of the patient population, setting and testing procedures – across studies that focused on people without leg ulcers. Suboptimal agreement between readings of the automated devices and those of the manual Doppler with a systematic tendency towards higher automated readings. Use of manual Doppler as the reference standard for detection of PAD. Variation in the prevalence of PAD across studies. Limited data on the performance of the automated devices in relevant subgroups of patients (e.g. diabetes patients). Uncertainty about the optimal threshold for automated ABPI measurement. Uncertainty about the potential role of automated devices in clinical practice (screening tool, alternative/adjuvant tool to current manual Doppler). Lack of data on the impact of the routine use of automated devices on health outcomes (e.g. the consequences of a delayed diagnosis because of FN results). No data on the WatchBP Office Vascular and MESI mTABLET ABI devices. With regard to the economic modelling, we identified the following areas of uncertainties that complement those identified for the review of clinical effectiveness evidence and raise doubt about the robustness of the cost-effectiveness results: A lack of data regarding the impact of different tests on patient-relevant outcomes such as ulcer healing. It is unclear whether automated tests could achieve tangible benefits in terms of a reduced time to compression therapy in patients with venous disease. Any benefits would rely on a lack of skills to complete manual Doppler assessment in the community, and it is unclear how widespread such a skill shortage might be. Uncertainty around whether inaccurate test results might be identified during clinical evaluation of patients during a testing appointment, and thus the extent to which inaccurate results would be acted upon in clinical practice [i.e. if tests would lead to inappropriate compression of arterial ulcers (FNs), or delayed time to compression (FPs)]. Limited data regarding the costs and outcomes specifically for mixed ulcer disease. Generalisability of the findings It is unclear how the results of studies assessing the accuracy of automated devices for measuring APBI in people without leg ulcers could be generalised to people with leg ulcers. Conclusions Future research is needed to evaluate the use of automated devices within specific populations (people with leg ulcers) and relevant settings. For the broader use of automated devices in clinical practice, more robust evidence is required to establish whether the use of automated devices is appropriate and cost-effective for the general screening of clinical populations with any vascular concerns. In addition, evidence is needed to support the use of automated devices as an alternative or adjunct to manual Doppler in people with symptoms of PAD. Study registration This study is registered as PROSPERO CRD42022327588. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135478) and is published in full in Health Technology Assessment; Vol. 28, No. 37. See the NIHR Funding and Awards website for further award information.

Background Estimation of glomerular filtration rate using equations based on creatinine is widely used to manage chronic kidney disease. In the UK, the Chronic Kidney Disease Epidemiology Collaboration creatinine equation is recommended. Other published equations using cystatin C, an alternative marker of kidney function, have not gained widespread clinical acceptance. Given higher cost of cystatin C, its clinical utility should be validated before widespread introduction into the NHS. Objectives Primary objectives were to: (1) compare accuracy of glomerular filtration rate equations at baseline and longitudinally in people with stage 3 chronic kidney disease, and test whether accuracy is affected by ethnicity, diabetes, albuminuria and other characteristics; (2) establish the reference change value for significant glomerular filtration rate changes; (3) model disease progression; and (4) explore comparative cost-effectiveness of kidney disease monitoring strategies. Design A longitudinal, prospective study was designed to: (1) assess accuracy of glomerular filtration rate equations at baseline (n = 1167) and their ability to detect change over 3 years (n = 875); (2) model disease progression predictors in 278 individuals who received additional measurements; (3) quantify glomerular filtration rate variability components (n = 20); and (4) develop a measurement model analysis to compare different monitoring strategy costs (n = 875). Setting Primary, secondary and tertiary care. Participants Adults (≥ 18 years) with stage 3 chronic kidney disease. Interventions Estimated glomerular filtration rate using the Chronic Kidney Disease Epidemiology Collaboration and Modification of Diet in Renal Disease equations. Main outcome measures Measured glomerular filtration rate was the reference against which estimating equations were compared with accuracy being expressed as P30 (percentage of values within 30% of reference) and progression (variously defined) studied as sensitivity/specificity. A regression model of disease progression was developed and differences for risk factors estimated. Biological variation components were measured and the reference change value calculated. Comparative costs of monitoring with different estimating equations modelled over 10 years were calculated. Results Accuracy (P30) of all equations was ≥ 89.5%: the combined creatinine–cystatin equation (94.9%) was superior (p 5%/year difference between mGFR and eGFR. Ability of equations to detect change was studied based on whether or not eGFR detected overall change, or decline only, in mGFR over 3 years against threshold changes variously defined as (1) > 10 ml/minute/1.73 m2; (2) > reference change value (RCV) (a > 21.5% increase or a > 17.7% decrease); (3) > 25% change; and (4) > 25% change and a change in disease stage. Sensitivity and specificity of eGFRs to identify progressive disease were evaluated. Estimated GFRs, in addition to urinary ACR, were also tested as predictors of progression and mortality. In the substudy of disease progression the change in mGFR, and the difference between mGFRs and eGFRs (bias), assessed every 12 months, were modelled over time using a longitudinal linear random coefficients regression model, to estimate average and variability in disease progression and bias. A model of disease progression based on mGFR was developed and differences in progression for risk factors estimated. In the biological variation substudy, analytical (CVA) and individual (CVI) components of variation were calculated and used to derive the RCV for significant changes in serial results for both mGFR and eGFR. Results from the main study informed a measurement model analysis. The trajectory of participants mGFR and eGFR over 10 years was used to estimate the proportion meeting the National Institute for Health and Care Excellence (NICE) definition of accelerated progression or of progression to CKD stage G4, assuming an annual testing schedule, and the number of participants expected to be incorrectly managed at each of the evaluated monitoring time points using different estimating equations. Based on the findings, the comparative costs of monitoring with GFR-estimating equations were calculated. Sample size Main study. Complete baseline data n = 1167. Three-year follow-up GFR data n = 875. Disease progression substudy. n = 278. Biological variation substudy. n = 20. Results All estimates of GFR relating to the primary study objectives were negatively biased compared to mGFR. There was no difference in median bias (ml/minute/1.73 m2) against mGFR between the MDRD (−3.7), CKD-EPIcreatinine (−2.8), CKD-EPIcystatin (−4.1) and CKD-EPIcreatinine-cystatin (−3.9) equations. Accuracy (P30) of the CKD-EPIcystatin equation (89.5%) did not differ from that of the MDRD (89.5%) and CKD-EPIcreatinine (90.2%) equations: accuracy of the CKD-EPIcreatinine-cystatin equation (94.9%) was superior (p 70% concordance. The CKD-EPIcreatinine-cystatin equation had better concordance than the other three primary study equations (p 83% in all cases), sensitivity for detecting change was < 63% in all cases. There was no clear difference in sensitivity or specificity between the four main study equations. For all equations and all thresholds, there was no clear evidence of improved performance of cystatin-containing equations compared to their matched creatinine-only equation. In the substudy of disease progression, modelling data showed a strong association between albuminuria status and rate of progression in mGFR and CKD-EPIcreatinine eGFR, with those with albuminuria having faster progression (steeper decline). Higher baseline mGFR values were associated with faster progression rate for mGFR. African-Caribbean ethnicity increased (slower decline) and South Asian ethnicity decreased (faster decline) the estimate of progression slope for mGFR and CKD-EPIcreatinine GFR. However, recruitment of ethnic minority participants in particular to the substudy fell short of target, limiting the strength of any conclusions. Within-subject biological variation of mGFR was 6.7%, with similar, although in some cases significantly lower, biological variation of eGFR (5.0, 5.3, 5.3 and 5.0% for MDRD, CKD-EPIcreatinine, CKD-EPIcystatin and CKD-EPIcreatinine-cystatin equations, respectively). Derived RCVs (%, positive/negative) were 21.5/−17.7 (mGFR), 15.1/−13.1 (MDRD), 15.9/−13.7 (CKD-EPIcreatinine), 15.9/−13.8 (CKD-EPIcystatin) and 15.1/−13.1 (CKD-EPIcreatinine-cystatin). We observed 62 deaths during the 3-year follow-up period. The study was not powered for hard end points. However, in agreement with earlier studies, regression models including each GFR-estimating equation separately demonstrated mortality was associated with lower eGFR, increasing age and male gender. An association with categorical albuminuria was not observed. There was no evidence of superiority of CKD-EPI equations, including the cystatin C-containing equations, as predictors of death compared to the MDRD equation. A measurement model analysis found no evidence to suggest that any of the estimating equations were superior for identifying CKD progression based on NICE-defined clinical end points. The average incremental per patient costs (compared to MDRD) of monitoring over a 10-year period using the cystatin C-based equations were estimated (CKD-EPIcystatin £42.20; CKD-EPIcreatinine-cystatin £43.32). Conclusions Most GFR equations achieved acceptable accuracy as judged by P30. There was little difference between the equations in accuracy, with evidence of superior accuracy for the CKD-EPIcreatinine-cystatin equation. Across several important characteristics (age, gender, diabetes, albuminuria, BMI, GFR level) we found no difference in accuracy of GFR-estimating equations. In relation to GFR estimation in African-Caribbean individuals, there was evidence to suggest caution should be exercised before advocating simple removal of the black race factor from the original CKD-EPI equations. In the longitudinal study, the CKD-EPIcreatinine-cystatin displayed slightly better concordance with mGFR than the other main study equations when tracking patients, but all study equations underestimated the mGFR decline. The sensitivity of GFR equations to detect clinically relevant threshold changes in mGFR, either overall or when considering decline in GFR only, was ≤ 63% for all equations. This is of concern given that such thresholds, including the NICE definition of accelerated progression and the change recognised as being true as determined by biological variation (RCV), were studied. Overall, data comparing the accuracy of different GFR-estimating equations demonstrated no notable benefit of using a cystatin C-containing equation in detecting GFR change. The measurement model underpinning the health economic analysis focused on the comparative accuracy of the estimating equations to detect accelerated progression. The analysis estimated accuracy over a longer trajectory than the main study and factored in measurement error, but found no clear benefit of using a cystatin C-based estimating equation. There was therefore no evidence to suggest that adding cystatin C measurement to current GFR monitoring protocols would be cost-effective. The disease progression modelling of the substudy data noted faster progression associated with higher baseline GFR and albuminuria; the latter consistent with other studies. Any conclusions relating to the influence of ethnicity were tempered by poor recruitment of ethnic minority individuals. The biological variability data have implications for monitoring of patients with CKD and clinical ability to understand CKD progression, both in clinical practice and research. The information presented provides an evidence base allowing clinicians to have meaningful discussions with their patients about the implications of changes in their GFR results. Inclusion of cystatin C in GFR-estimating equations was associated with marginal improvements in accuracy, but no clear advantages in terms of detecting GFR change over time. Problems of standardisation of cystatin C assays remain, despite the introduction of an international standard. The use of cystatin C increases the economic cost of CKD monitoring with little apparent gain. These data do not support the use of cystatin C for the routine monitoring of GFR in people with stage 3 CKD. Further research is warranted to investigate specific patient groups that may benefit from cystatin C use. Trial registration This trial is registered as ISRCTN42955626 www.controlled-trials.com/ISRCTN42955626 (accessed 26 July 2023). Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 11/103/01) and is published in full in Health Technology Assessment; Vol. 28, No. 35. See the NIHR Funding and Awards website for further award information.

Background There is no consensus on optimal management of pilonidal disease. Surgical practice is varied, and existing literature is mainly single-centre cohort studies of varied disease severity, interventions and outcome assessments. Objectives A prospective cohort study to determine: disease severity and intervention relationship most valued outcomes and treatment preference by patients recommendations for policy and future research. Design Observational cohort study with nested mixed-methods case study. Discrete choice experiment. Clinician survey. Three-stage Delphi survey for patients and clinicians. Inter-rater reliability of classification system. Setting Thirty-one National Health Service trusts. Participants Patients aged > 16 years referred for elective surgical treatment of pilonidal disease. Interventions Surgery. Main outcome measures Pain postoperative days 1 and 7, time to healing and return to normal activities, complications, recurrence. Outcomes compared between major and minor procedures using regression modelling, propensity score-based approaches and augmented inverse probability weighting to account for measured potential confounding features. Results Clinician survey: There was significant heterogeneity in surgeon practice preference. Limited training opportunities may impede efforts to improve practice. Cohort study: Over half of patients (60%; N = 667) had a major procedure. For these procedures, pain was greater on day 1 and day 7 (mean difference day 1 pain 1.58 points, 95% confidence interval 1.14 to 2.01 points, n = 536; mean difference day 7 pain 1.53 points, 95% confidence interval 1.12 to 1.95 points, n = 512). There were higher complication rates (adjusted risk difference 17.5%, 95% confidence interval 9.1 to 25.9%, n = 579), lower recurrence (adjusted risk difference −10.1%, 95% confidence interval −18.1 to −2.1%, n = 575), and longer time to healing (>34 days estimated difference) and time to return to normal activities (difference 25.9 days, 95% confidence interval 18.4 to 33.4 days). Mixed-methods analysis: Patient decision-making was influenced by prior experience of disease and anticipated recovery time. The burden involved in wound care and the gap between expected and actual time for recovery were the principal reasons given for decision regret. Discrete choice experiment: The strongest predictors of patient treatment choice were risk of infection/persistence (attribute importance 70%), and shorter recovery time (attribute importance 30%). Patients were willing to trade off these attributes. Those aged over 30 years had a higher risk tolerance (22.35–34.67%) for treatment failure if they could experience rapid recovery. There was no strong evidence that younger patients were willing to accept higher risk of treatment failure in exchange for a faster recovery. Patients were uniform in rejecting excision-and-leave-open because of the protracted nursing care it entailed. Wysocki classification analysis: There was acceptable inter-rater agreement (κ = 0.52, 95% confidence interval 0.42 to 0.61). Consensus exercise: Five research and practice priorities were identified. The top research priority was that a comparative trial should broadly group interventions. The top practice priority was that any interventions should be less disruptive than the disease itself. Limitations Incomplete recruitment and follow-up data were an issue, particularly given the multiple interventions. Assumptions were made regarding risk adjustment. Conclusions and future work Results suggest the burden of pilonidal surgery is greater than reported previously. This can be mitigated with better selection of intervention according to disease type and patient desired goals. Results indicate a framework for future higher-quality trials that stratify disease and utilise broad groupings of common interventions with development of a patient-centred core outcome set. Trial registration This trial is registered as ISRCTN95551898. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/17/02) and is published in full in Health Technology Assessment; Vol. 28, No. 33. See the NIHR Funding and Awards website for further award information. Plain language summary Background Pilonidal disease is caused by ingrowing hairs between the buttocks. It can cause pain and infection and may need surgery. We do not know which operation gives the best results, or who operations help. Objectives PITSTOP aimed to find out which operation is the best and what is important to patients when deciding on surgery, and to suggest ideas for better treatment and future research. Methods We looked at what operations were done and their outcomes. We interviewed patients about their experiences. Some completed a survey to help us understand what operations they might prefer based on risks and outcomes. Surgeons completed a survey about their experiences, and we explored whether a new tool could help us tell the difference between ‘mild’ and ‘bad’ disease. We used findings from these studies to help patients and surgeons give priorities for future practice and research. Results Six hundred and sixty-seven patients joined PITSTOP. People who had a major operation had more pain and took longer to return to normal activities. Some were still affected 6 months after surgery. However, disease recurrence was lower than after a minor procedure. Patients based decisions about treatment on the likelihood of success and the time to recover. The study and the surgeons’ survey both showed marked differences in practice. Surgeons tended to offer one or two operations learned during training. A classification tool put cases in similar groups, but this did not influence treatment choices. The consensus exercise identified five research priorities, the top one being to put types of surgery into two groups. Of the five practice priorities, the top one was that surgery should not make the patient worse than the disease. Conclusions There is variation in the treatment of pilonidal disease. Wound issues and impact on daily living should be avoided. The highlighted research questions should be addressed to improve care. Scientific summary Background Despite being a common condition, there is no clear consensus as to how pilonidal disease should be managed. Throughout the UK, surgical practice is varied, and existing literature largely consists of single-centre cohort studies using a range of disease classification systems, interventions and outcome assessments. There is a need to improve clinical management of this condition and define future research processes and priorities. Objectives PITSTOP aimed to investigate surgical options for the treatment of pilonidal sinus disease (PSD). The objectives were to: Follow patients with symptomatic pilonidal sinus referred to each collaborating site, prospectively recording details of their pit/track anatomy, surgical management, medical events and health-related quality of life (QoL) until 6 months after their operation. Describe the combination of interventions currently in use and quantify clinical and patient-reported outcomes (PROs) associated with each. Identify patient-specific disease features that might predict poor outcome in each treatment group by risk-modelling methods. Derive a case mix-adjusted estimate of the risks associated with common treatment options, using causal inference methods to provisionally rank the optimal management strategies among patients for whom more than one treatment is considered appropriate. Provide an overview of patient views and experiences. Collect the views of patients on which interventions they would rather avoid and which outcomes they most value. Reach a surgeon-based consensus on which subtypes of pilonidal disease may benefit from which treatment options. Reach a surgeon and patient-based consensus on research priorities. Design and setting PITSTOP was an observational cohort with nested mixed-methods and qualitative design which took place across 33 NHS Trust sites in the UK. The study had an additional clinician survey component and validation of a classification system and culminated in a three-stage Delphi exercise to identify research and practice priorities. Participants Eligible patients undergoing surgical management for PSD and interested in participating were consented to the study. Participants referred to a collaborating centre for definitive elective surgical treatment of PSD were required to meet the following criteria: Consenting patients over 16 years of age with PSD. Participants were ineligible if any of the following conditions were met: Asymptomatic Pregnant Unable to give consent Acute abscess Hypersensitivity to the sealants. Intervention This was a non-intervention study with the choice of treatment being a shared decision made between the surgeon and patient. Treatment of PSD typically comprises two essential components (with the exception of phenol injection, seton and fistuloscope/diathermy – which aim to induce fibrosis with or without obliteration of the tracks): Excision of the affected skin and fat (the amount of which varies among patients, and surgeons differ with respect to how the resultant wound is managed). Closure of the wound, which is either left open and heals slowly by secondary intention or closed using sutures, also known as primary wound closure. There are 18 excision–closure combinations that are theoretically possible. Therefore, the operative intervention was recorded by the following excision and closure techniques: Excision procedure Local excision. Curettage. Phenol injection. Pit picking. Seton. Laser-assisted excision. EPSiT (endoscopic pilonidal sinus treatment). Closure procedure Primary midline closure. Fibrin glue (FG). Marsupialisation. Lateral closure (Karydakis; Bascom’s cleft). Flap (Limberg; rhomboid). Follow-up Participants were followed up 1 day and 7 days after surgery (with day 0 the day of elective surgical treatment), at the routine clinic visit appointment, at 6 months, and at the end of study. Main outcome measures As PITSTOP is a cohort study examining current practice, there are no primary outcome measures. The following data were collected: pain (numeric rating scale) on day 1 and day 7 postoperatively and at each follow-up EuroQol five dimensions five levels questionnaire (EQ-5D-5L) QoL at each follow-up interactions with primary and secondary care length of time to healing return to normal activities complications recurrence infection. Statistical methods The study aimed to recruit 800 patients, with at least 100 within each of the front-running management strategies. Two treatment comparisons were undertaken: any major excisional procedure versus any minor procedure, and asymmetric closure versus any minor procedure (minimal excision). Regression modelling, propensity score-based approaches [inverse probability weighting (IPW) and nearest neighbour matching] and augmented IPW were used to account for measured potential confounding. Continuous outcomes (pain at day 1 and day 7) were modelled using linear regression; binary outcomes (recurrence, complication) were modelled using logistic regression, with marginal probabilities used to estimate absolute risk difference. Time to wound healing and time to return to normal activity were modelled using parametric accelerated survival time. The following features were considered as potential risk factors for outcome and treatment choice: sex, body mass index, depth of natal cleft, presence and type of gluteal hair, smoking status, pit density, presence of unilateral or bilateral disease, distance from furthest lateral opening to the nearest pit, presence of pus and Wysocki disease classification. Features affecting treatment choice were assessed using logistic regression with treatment choice as the outcome, and the same propensity score adjustments were used for each outcome. For each outcome, three regression models were fitted: one that adjusted for all listed features, one that adjusted for features associated with the outcome (based on Akaike’s information criteria, model c-statistic and clinical review) and one that adjusted for disease classification alone. The difference in predicted outcome and corresponding 95% confidence interval (CI) were estimated for each modelling strategy on each outcome. Mixed-methods case studies Mixed-methods case studies were undertaken to understand why people make and regret decisions on their treatment. Longitudinal semistructured interviews (20 cohort participants from 13 sites) were conducted at baseline and 6 months later with framework analysis mapped findings to Witt’s coping in deliberation framework and Sekhon’s acceptability framework. We triangulated findings with baseline scores CollaboRATE shared decision-making (SDM) survey scores and 6-month decision regret (DR) scores. Discrete choice experiment An online survey using the discrete choice experiment (DCE) method was used to ask patients to choose their favoured treatment when presented with competing hypothetical treatment profiles. Regression analyses were conducted on DCE responses. Clinician survey A survey was developed following the CHERRIES statement checklist of recommendations to identify the most frequently used interventions for specific clinical scenarios in current PSD practice. It was disseminated via the surgical trainee research collaboratives to practising consultants throughout the UK. Data were captured and stored in the REDCap software. Consensus exercise and validation of a classification system Two separate surveys were undertaken. The first comprised a Delphi consensus exercise in which surgeons and patients were asked to recommend best practice and further research. In the second, 15 surgeons were each asked to retrospectively assess photographs to quantify agreement in the Wysocki classification tool, with 90 patients each assessed by 6 surgeons. Results Cohort study Participants Thirty-one UK sites recruited participants over a 46-month period from May 2019 to March 2022. Seven hundred and twenty-nine participants consented to the study; after exclusions due to no procedure (n = 45), incorrect diagnosis (n = 7) and insufficient treatment information (n = 10), there were 667 participants included in the analysis cohort. Six-month follow-up data were available for 71% of participants; recurrence and complication data were available for 94% and 96% of participants, respectively. Main results Sixty per cent of patients (n = 397) received a major procedure; this comprised 272 (41%) asymmetric closure, 49 (7%) leave open and 76 (11%) midline closure. The remaining participants received minimal excision (n = 270, 40%), most commonly glue (n = 106, 16%) or pit picking (n = 60, 9%). Pain on day 1 and day 7 was higher for patients that received major procedures compared to minor procedures (augmented IPW-adjusted mean difference in day 1 pain 1.58 points, 95% CI 1.14 to 2.01 points, n = 536; mean difference in day 7 pain 1.53 points, 95% CI 1.12 to 1.95 points, n = 512). The difference was broadly consistent regardless of adjustment method. Complications were reported by 207/385 (54%) participants that had major procedures, and 94/258 (36%) participants that had minor procedures (augmented IPW-adjusted risk difference 17.5%, 95% CI 9.1 to 25.9%, n = 579). Recurrence was reported for 86/373 (23%) of major and 87/256 (34%) of minor procedures; the augmented IPW-adjusted risk difference was −10.1% (95% CI −18.1 to −2.1%, n = 575). The estimated difference between groups was smaller for treatment failure (a composite of recurrence, failure to heal, and failure to return to normal activities), where 45% of major procedure participants versus 47% of minor procedure participants experienced treatment failure (augmented IPW-adjusted risk difference −2.3%, 95% CI −10.9 to 6.2%). The estimates for treatment difference in recurrence and complication were consistent across the various adjustment methods. Treatment differences for time to healing and time to return to normal activities were less consistent between adjustment methods. Participants receiving major procedures took an estimated minimum of 20 days longer to return to normal activities (augmented IPW difference 25.9 days, 95% CI 18.4 to 33.4 days). Difference in time to healing was estimated to be at least 34 days more in the major group, but the estimates and corresponding CIs were inconsistent across models. At 6 months, around 25% of participants in both groups had wounds that had not healed; 12% of major procedure participants and 4% of minor procedure participants were yet to return to normal activity. Comparisons of asymmetric closure with minimal excision produced results similar to the major versus minor comparisons. Clinician survey The survey was completed by 109 participants (54.5% response) who routinely cared for patients with PSD. Respondents reported a median caseload of 15 patients per year [interquartile range (IQR) 10–20]. Of those estimating their recurrence rates, 28.8% predicted they were in the 16–30% range. A wide range of treatment strategies were employed, with 65.1% practising an excision-and-leave-open technique despite this being considered obsolete as a treatment option. Surgical training in this obsolete technique was experienced by 83.1%. Many recommended non-surgical treatments despite the lack of evidence. Mixed-methods case studies No choice of treatment was offered to 9/20 participants, although this was not always seen negatively on the CollaboRATE instrument. Decision-making was influenced by prior experience of pilonidal sinus and anticipated recovery time. Participants scoring highest on DR also had among the highest SDM scores. The burden involved in wound care and the gap between expected and actual time for recovery were the principal reasons given for DR. Discrete choice experiment The survey was completed by 111 participants. The strongest predictor of treatment choice was risk of infection/persistence (attribute importance 70%), followed by shorter recovery time (attribute importance 30%). Patients were willing to trade off recovery time against risk of infection/persistence. Patients aged over 30 years had a higher risk tolerance (22.35–34.67%) for treatment failure if they could experience rapid recovery. Younger patients were willing to accept smaller risks (maximum acceptable risk 1.51–2.15) in exchange for a faster recovery. Patients were uniform in rejecting excision-and-leave-open because of the protracted nursing care it entailed. Wysocki tool classification The Wysocki tool demonstrated acceptable agreement. Overall, the kappa statistic for agreement was 0.52 (95% CI 0.42 to 0.61), with five of the six surgeons reaching consensus in 53 (59%) patients. Agreement did not differ with regard to the surgeon’s experience. This, along with the prognostic value of the tool, suggests this could be used in practice to classify PSD. Consensus exercise The top five practice statements included: any treatment should not be worse than the disease itself; minimally invasive procedures should be preferred when feasible; surgeons should have opportunities to learn new techniques; a classification tool will help inform treatment options; and delayed return to activities is an important outcome. The top five research statements included: future trials should compare broad groups of operations – minimally invasive procedures versus major excisional techniques, with stratification by severity of disease; a core outcome set and PROs should be used; there should be an attempt to develop an algorithm or decision tree to aid surgeon decision-making. Conclusions The burden of surgery for PSD is significantly greater than that reported in the literature. Many surgeons perceive this but continue to practise outdated procedures. While minimally invasive procedures may reduce this burden in many patients, they are not always offered when they could be. This practice may be driven by the desire to achieve cure at the expense of protracted recovery, but this is not always what patients want. Many would trade reduced chance of cure for more rapid recovery. Future practice priorities should follow the ethos of not making the surgery worse than the disease itself and appreciate that patients need to be properly informed about the burden associated with wound care and the likelihood of recurrence associated with different procedures, to minimise DR. Future trials should compare broad groups of interventions (minimally invasive vs. major excisions) stratified by disease severity and utilising a reliable and validated Wysocki classification system. Such trials should incorporate a core set of PROs. Trial registration This trial is registered as ISRCTN95551898. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/17/02) and is published in full in Health Technology Assessment; Vol. 28, No. 33. See the NIHR Funding and Awards website for further award information. Study protocol The original protocol for the PITSTOP study can be found here. https://figshare.shef.ac.uk/articles/journal_contribution/The_PITSTOP_Study_PIlonidal_sinus_Treatment_STudying_the_OPtions_-_Protocol/7578242. A list of the protocol amendments since the initial Research Ethics Committee (REC) approval can be found below. Version number Changes made Date of REC/HRA approval Amendment number 1.2 Errors in the flow diagramErrors that were overlooked such as age where it stated 18 years instead of 16 yearsTypological errors in the textChanges to the call at day one post surgery – this is now a phone call by a research nurse 03 May 2019/03 April 2019 SA01 1.2 Changes of collaborator Christine Moffatt University affiliation – previously University of Nottingham, now Nottingham Trent University (p7) 1.2 Addition of optional consent item described which is the taking of photos pre-surgery to assist with the development of the classification system for PSD (p19) 1.2 A follow-up call on day 1 post surgery as opposed to a text message (p19, p34, table 1) 1.2 Removal for the 4–6 week time frame from routine clinic visit follow-up as this varies from clinic to clinic (p34) 1.2 Addition of infection as outcome measure (p33) 1.3 Clarification: Patients must have a minimum of 24 hours between receiving the Patient Information Sheet and consenting to the study (p16) N/A/19 November 2019 MA07 1.3 Clarification: The postoperative routine clinic visit may also be conducted by the telephone if considered routine at site (p19) 1.3 Clarification: The recruitment end date is July 2020 (p29) 1.4 An additional telephone follow-up can be made if the participant does not attend the pre-arranged face-to-face clinic visit (p19) N/A/03 December 2019 MA08 1.4 Clarification: Data can be collected by trained research personnel or a delegated member of the research team (p19) 1.4 Addition of the REC reference on the front page (p1) 1.5 Clarification: The photographs will be used to aid the validation of the classification system for PSD (p19) 02 March 2020/07 April 2020 SA03 1.5 Clarification: Preoperative photographs of the surgical site will be uploaded on the REDCap data capture system (p20) 1.5 Clarification: A copy of the consent form will be uploaded on to the REDCap data capture system for monitoring purposes (p20) 2.0 Clarification: Sponsor and CI e-mail addresses have been updated (p6) 03 August 2020/05 August 2020 SA04 2.0 Clarification: Study coordinating staff list has been updated (p8) 2.0 Response to COVID-19: Addition of postal consent if preoperative consultations are conducted remotely OR social distancing measures inhibit research personnel to gain consent in person (p15) 2.0 Surgeon survey is also accessible via an online link hosted on REDCap (p18) 2.0 Research personnel may text a participant to arrange a follow-up call (p18) 2.0 Response to COVID-19: Baseline measures can be taken over the phone (p19) 2.0 Response to COVID-19: Follow-up data can be collected by study coordinating team if research personnel at NHS trust do not have capacity (p19) 2.0 Response to COVID-19: missed 6-month follow-up data can be collected at the end of study (p19) 2.0 The DCE questionnaire can be completed by non-PITSTOP patients. The questionnaire can be advertised using a leaflet and/or social media (p24) 2.1 Research personnel may e-mail a participant to arrange a follow-up call (p18) N/A/27 October 2020 MA11 2.2 Update to study coordinating team (p7) N/A/29 September 2022 NSA13 2.2 Clarification of study procedures for the PSD seton surgical procedure (p19) 2.2 Clarification of follow-up procedures (p20) 2.2 Update to recruitment end date (p30) 3.0 Amendment to consensus technique (p25) 18 November 2021 SA05

Background Vaccination of infants with pneumococcal conjugate vaccines is recommended by the World Health Organization. Evidence is mixed regarding the differences in immunogenicity and efficacy of the different pneumococcal vaccines. Objectives The primary objective was to compare the immunogenicity of pneumococcal conjugate vaccine-10 versus pneumococcal conjugate vaccine-13. The main secondary objective was to compare the seroefficacy of pneumococcal conjugate vaccine-10 versus pneumococcal conjugate vaccine-13. Methods We searched the Cochrane Library, EMBASE, Global Health, MEDLINE, ClinicalTrials.gov and trialsearch.who.int up to July 2022. Studies were eligible if they directly compared either pneumococcal conjugate vaccine-7, pneumococcal conjugate vaccine-10 or pneumococcal conjugate vaccine-13 in randomised trials of children under 2 years of age, and provided immunogenicity data for at least one time point. Individual participant data were requested and aggregate data used otherwise. Outcomes included the geometric mean ratio of serotype-specific immunoglobulin G and the relative risk of seroinfection. Seroinfection was defined for each individual as a rise in antibody between the post-primary vaccination series time point and the booster dose, evidence of presumed subclinical infection. Each trial was analysed to obtain the log of the ratio of geometric means and its standard error. The relative risk of seroinfection (‘seroefficacy’) was estimated by comparing the proportion of participants with seroinfection between vaccine groups. The log-geometric mean ratios, log-relative risks and their standard errors constituted the input data for evidence synthesis. For serotypes contained in all three vaccines, evidence could be synthesised using a network meta-analysis. For other serotypes, meta-analysis was used. Results from seroefficacy analyses were incorporated into a mathematical model of pneumococcal transmission dynamics to compare the differential impact of pneumococcal conjugate vaccine-10 and pneumococcal conjugate vaccine-13 introduction on invasive pneumococcal disease cases. The model estimated the impact of vaccine introduction over a 25-year time period and an economic evaluation was conducted. Results In total, 47 studies were eligible from 38 countries. Twenty-eight and 12 studies with data available were included in immunogenicity and seroefficacy analyses, respectively. Geometric mean ratios comparing pneumococcal conjugate vaccine-13 versus pneumococcal conjugate vaccine-10 favoured pneumococcal conjugate vaccine-13 for serotypes 4, 9V and 23F at 1 month after primary vaccination series, with 1.14- to 1.54-fold significantly higher immunoglobulin G responses with pneumococcal conjugate vaccine-13. Risk of seroinfection prior to the time of booster dose was lower for pneumococcal conjugate vaccine-13 for serotype 4, 6B, 9V, 18C and 23F than for pneumococcal conjugate vaccine-10. Significant heterogeneity and inconsistency were present for most serotypes and for both outcomes. Twofold higher antibody after primary vaccination was associated with a 54% decrease in risk of seroinfection (relative risk 0.46, 95% confidence interval 0.23 to 0.96). In modelled scenarios, pneumococcal conjugate vaccine-13 or pneumococcal conjugate vaccine-10 introduction in 2006 resulted in a reduction in cases that was less rapid for pneumococcal conjugate vaccine-10 than for pneumococcal conjugate vaccine-13. The pneumococcal conjugate vaccine-13 programme was predicted to avoid an additional 2808 (95% confidence interval 2690 to 2925) cases of invasive pneumococcal disease compared with pneumococcal conjugate vaccine-10 introduction between 2006 and 2030. Limitations Analyses used data from infant vaccine studies with blood samples taken prior to a booster dose. The impact of extrapolating pre-booster efficacy to post-booster time points is unknown. Network meta-analysis models contained significant heterogeneity which may lead to bias. Conclusions Serotype-specific differences were found in immunogenicity and seroefficacy between pneumococcal conjugate vaccine-13 and pneumococcal conjugate vaccine-10. Higher antibody response after vaccination was associated with a lower risk of subsequent infection. These methods can be used to compare the pneumococcal conjugate vaccines and optimise vaccination strategies. For future work, seroefficacy estimates can be determined for other pneumococcal vaccines, which could contribute to licensing or policy decisions for new pneumococcal vaccines. Study registration This study is registered as PROSPERO CRD42019124580. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/148/03) and is published in full in Health Technology Assessment; Vol. 28, No. 34. See the NIHR Funding and Awards website for further award information. Plain language summary Pneumococcal disease is a serious illness caused by a bacterial infection that can result in death. Children in the United Kingdom receive a vaccine to prevent this disease that protects against 13 different types of pneumococcal diseases. It is very effective, but other vaccines are also available, such as one that contains 10 types of pneumococcal diseases. Vaccines in the United Kingdom are bought by the government and the choice of which vaccine to provide is based on the cost of the vaccine as well as the benefits to our health. However, there is very little information comparing different vaccines and it is often assumed they are the same. We did a large analysis combining all studies of the two main licensed pneumococcal vaccines to determine which vaccine provides better protection against infection and how this affects costs. We used information from studies published in medical journals, and also data from studies done by the companies that own the vaccines. Our results showed that pneumococcal conjugate vaccine-13 vaccine provided better protection than pneumococcal conjugate vaccine-10 for 5 of the 10 serotypes that are contained in both vaccines. When we used these results to model what might have happened had either of these vaccines been introduced into the United Kingdom vaccination programme in 2006, we found that both vaccines caused a rapid decrease in the amount of disease, but that the decrease in disease was faster with pneumococcal conjugate vaccine-13 than pneumococcal conjugate vaccine-10. This resulted in 2808 cases of diseases prevented over a 25-year time frame with pneumococcal conjugate vaccine-13 compared with pneumococcal conjugate vaccine-10. Our methods can be used to compare other vaccines and we recommend this type of study be done in future when making decisions on vaccine product choice. Scientific summary Streptococcus pneumoniae (pneumococcus) causes severe diseases, including bacterial pneumonia, meningitis and sepsis, leading to substantial morbidity and mortality worldwide, with the highest disease burden being in young children and older adults. Three pneumococcal conjugate vaccines (PCVs) have been widely deployed worldwide in the past two decades: PCV7 (Prevnar; Pfizer, headquartered in New York City, New York, USA), PCV10 (Synflorix; GlaxoSmithKline, headquartered in Brentford, London, UK) and PCV13 (Prevenar 13; Pfizer, headquartered in New York City, New York, USA), resulting in substantial reduction in disease. Between 2009 and 2011, PCV7 was gradually replaced by PCV13 and PCV10 and is no longer available. The World Health Organization (WHO) does not preferentially endorse one PCV over another. Both PCV13 and PCV10 have been shown to provide both direct and indirect protection against pneumococcal pneumonia, invasive pneumococcal disease and nasopharyngeal carriage. Although there are 10 common serotypes in these 2 vaccines, the components of the vaccines differ, with different carrier proteins used in the conjugation process, as well as different amounts of polysaccharide, and these differences may contribute to differences in protection. Large randomised controlled trials directly comparing different PCVs with invasive pneumococcal disease as the primary outcome are not feasible. We previously used ‘seroinfection’ as an outcome for analysis of PCVs, where seroinfection is defined as an increase in antibody levels between the primary vaccination series (typically complete at 5–7 months of age) and the booster dose (typically administered at 9–18 months of age). Seroinfection can be regarded as evidence of exposure to the pathogen and a resultant subclinical infection, given antibody responses wane rapidly during this period otherwise. Seroinfection rates for different vaccines can be compared by calculating the relative risk (RR) of seroinfection, referred to herein as ‘seroefficacy’. We meta-analysed data from studies of PCVs to compare the immunogenicity and seroefficacy of PCV10 with PCV13 for each serotype. We aimed to determine if serotype-specific immune responses were higher for either vaccine and whether this resulted in greater protection again seroinfection. In addition, we explored the overall relationship between the higher immune response and protection against seroinfection in infants. Following this, we show how serotype-specific estimates of seroefficacy can be incorporated in vaccine cost-effectiveness models. Objectives The primary objective of the systematic review was to compare the immunogenicity of PCV10 versus PCV13 for each serotype contained in the vaccines. The secondary objectives were: to compare the seroefficacy of PCV10 versus PCV13 for each serotype contained in the vaccines for PCV10 and PCV13 separately, to estimate immunogenicity and seroefficacy in comparison with the older PCV7 vaccine to determine how the comparisons of immunogenicity and efficacy of PCV10 to PCV13 are affected by the co-administration of different routine vaccines. Methods Systematic review We conducted a systematic review identifying studies that compared the immunogenicity of licensed PCVs in trials which randomised children to one of two different PCVs. The PCVs included in the review were PCV7 (Prevnar; Pfizer), PCV10 (Synflorix; GlaxoSmithKline) and PCV13 (Prevenar 13; Pfizer); PCV7 was included even though no longer available, so that we could compare PCV13 and PCV10 indirectly through them each being compared with PCV7 for the same serotypes. Data sources The databases searched were Cochrane Database of Systematic Reviews and Cochrane Central Register of Controlled Trials, EMBASE, Global Health and MEDLINE. The trial registers searched were ClinicalTrials.gov (https://clinicaltrials.gov/) and WHO International Clinical Trials Registry Platform (https://trialsearch.who.int/). The search comprised title/abstract keywords and subject headings for pneumococcal vaccines and children. A methodological search filter for randomised controlled trials taken from the Cochrane Handbook was used to limit to randomised controlled trials. Pharmaceutical company websites (GlaxoSmithKline and Pfizer) were also hand-searched for relevant studies. No date or language limits were applied. Study selection Randomised controlled trials were included if they provided direct comparisons of either PCV7, PCV10 or PCV13 among infants and children ˂ 2 years of age, and if they provided estimates of antibody responses [serotype-specific anti-pneumococcal immunoglobulin G (IgG) to PCVs for at least one time point of 1] between 4 and 6 weeks after the primary vaccination series and/or 1 month after a booster vaccination. Individual participant-level data were retrieved if available. Aggregate data from publications were extracted if individual participant data were not available. Risk of bias in results of the included studies was assessed independently by two reviewers using the Cochrane Risk of Bias Tool. Data synthesis Each trial with individual participant-level data available was analysed to obtain the log of the ratio of geometric means (log-GMR) and its standard error (SE) for each serotype and time point of interest. The RR of seroinfection was estimated by comparing the proportion of participants with seroinfection between vaccine groups. When no seroinfection occurred in any group (numerator of absolute risk was 0), a small non-zero value (0.5) was added to both sero-infected and sero-non-infected groups to allow estimation of the RR. The log-GMRs, log-RRs and their SEs constituted the input data for evidence synthesis. Only trials supplying individual participant data were included in seroefficacy analyses. For serotypes contained in all three vaccines, evidence could be synthesised using a network meta-analysis (NMA) of all comparisons. For other serotypes, meta-analysis was used for evidence synthesis. To estimate the overall association between antibody geometric mean ratio (GMR) and RR across all serotypes, we fitted a mixed-effect model regressing study-level RRs of seroinfection on GMRs across serotypes, weighted by the sample size of each study. Fixed effects included GMR, serotype and interactions between GMR and serotype (allowing serotype-specific association), while study was included as a random effect. Mathematical modelling and retrospective economic evaluation To illustrate the use of serotype-specific estimates of seroefficacy in modelling vaccine impact and cost-effectiveness, we developed a serotype-specific mathematical model of pneumococcal transmission dynamics to compare the differential impact of PCV10 and PCV13 introduction on invasive pneumococcal disease cases with vaccine serotypes in England and Wales. The model estimated the impact over a 25-year time period from 2006 to 2030. We subsequently assessed the cost-effectiveness of introducing infant vaccination with PCV13 compared with introducing PCV10 from a healthcare payer perspective in England and Wales. More specifically, we retrospectively estimated the additional threshold price per dose below which PCV13 would be more cost-effective than PCV10 had they both been available at the time of introduction of the PCV vaccine programme in England and Wales in 2006. Results Database registry and hand searches identified 4699 publication records of which 47 studies (78 publication reports) satisfied our eligibility criteria. Nineteen studies (24 publication reports) were excluded from the analysis: 6 studies did not provide individual patient or aggregate data and 13 studies (18 publication reports) were studies with the vaccines of interest, but it was not possible to form a loop within the NMA to provide indirect evidence. The remaining 28 studies (54 publication records) from 2009 to 2023 were included in the NMAs. Twenty-two studies provided individual participant data with a further five studies reporting aggregate data. Immunogenicity Geometric mean ratios for comparisons between PCV13 versus PCV10 for any primary series schedule were higher for PCV13 for serotypes 4, 7F, 9V and 23F at 1 month after primary vaccination series, with 1.14- to 1.54-fold higher IgG responses with PCV13. Additional serotypes contained only in the PCV13 vaccine (3, 6A and 19A) also favoured PCV13 as expected. GMRs were similar for the remaining serotypes (1, 5, 6B, 14, 18C and 19F). GMRs favoured PCV7 over either PCV13 or PCV10 for serotypes 4, 6B, 9V, 14 and 23F. There was no difference in GMRs for serotypes 18C and 19F across three vaccines. At the pre-booster time point, data were available from 18 cohorts. IgG responses were lower with PCV13 compared with PCV10 for all PCV7 serotypes except for serotype 14, with the point estimates of GMRs comparing PCV13 versus PCV10 ranging from 0.44 to 0.78. IgG responses were higher for PCV13 for serotypes 1, 5 and 7F. GMRs comparing PCV13 versus PCV7 showed higher IgG with PCV7 for serotypes 4, 6B, 9V, 14 and 23F and higher IgG with PCV13 for serotype 19F. At 28 days post booster, data were available from 26 cohorts. GMRs favoured PCV13 over PCV10 for serotype 6B, 9V, 14 and 23F and favoured PCV10 over PCV13 for serotype 18C. For serotype 1, 5 and 7F, antibody responses were higher in PCV13 compared with PCV10. PCV7 recipients had higher geometric mean concentrations (GMCs) compared with PCV13 for all PCV7 serotypes except 6B for which there was no difference, and 19F, which favoured PCV13. For PCV13-only serotypes (3, 6A and 19A), GMRs favour PCV13 at all three time points. Substantial heterogeneity and network inconsistency were present for most serotypes at all three time points. To explore potential reasons for the observed heterogeneity, we summarised cohort-level GMRs and RRs for each vaccine comparison. These descriptive analyses revealed a lack of consistency in the direction of study-level estimates within each vaccine comparison, resulting in the significant heterogenicity. There was also no observable pattern in any trial-level variable (region, co-administered vaccines, vaccine schedule), from which one might propose a mechanism that would adequately explain this variation in GMRs. Seroefficacy There were 12 studies (15 cohorts) with available individual participant antibody data at both post-primary and prior to the booster dose, allowing serotype-specific estimation of seroefficacy from a total of 5152 participants. Of these 15 cohorts, 6 compared PCV10 versus PCV7, 3 compared PCV13 versus PCV7 and 6 compared PCV13 versus PCV10. Among PCV7 serotypes, the risk of seroinfection was lower with PCV13 than PCV10 for serotypes 4, 6B, 9V, 18C and 23F, while no difference was seen for serotype 14 and 19F. The RRs of seroinfection (PCV13 vs. PCV10) for PCV7 serotypes ranged from 0.32 (95% CI 0.19 to 0.52) for serotype 4 to 1.28 (95% CI 0.95 to 1.74) for serotype 14. For serotypes 1, 5 and 7F, evidence was summarised from six studies directly comparing PCV13 with PCV10. Comparisons between PCV13 and PCV7 favoured neither vaccine over the other, whereas comparisons between PCV7 and PCV10 favoured PCV7 for serotypes 5, 6B, 9V, 18C and 23F. The I2 and p-values indicated some heterogeneity for all PCV7 serotypes except for serotype 4 and 19F. In the mixed-effects model of all serotypes combined, vaccines that produced the same amount of antibody (GMR = 1) had very similar protection (adjusted RR 0.80, 95% CI 0.41 to 1.58). The model estimate indicates that for each twofold increase in antibody response, the risk of seroinfection was halved (GMR of 2.0; RR 0.46, 95% CI 0.23 to 0.96). Mathematical model and economic evaluation Mathematical model results showed that in the absence of any vaccine programme, an increase in invasive pneumococcal disease cases caused by all five serotypes would be seen over the 25-year time frame. With the introduction of either PCV13 or PCV10 vaccine programmes in 2006, case counts would have decreased, achieving near eradication of all serotypes within the time frame modelled. The decrease in cases was most rapid for serotype 6B and least rapid for serotype 4. The decrease in cases was less rapid for PCV10 than for PCV13 due to the lower seroefficacy. The introduction of an infant PCV13 programme was predicted to avoid an additional 2808 (95% CI 2690 to 2925) cases of invasive pneumococcal disease compared with PCV10 introduction between 2006 and 2030. This includes an estimated 326 cases of meningitis, 578 cases of sepsis, 1770 cases of invasive pneumonia and 30,680 cases of non-invasive pneumonia. Under base-case assumptions, this resulted in discounted healthcare savings of £13 million (95% CI £12 to £14 million). Including non-invasive pneumonia increased the savings to £27 million (95% CI £25 to £29 million). Conclusions In our study, we used a novel methodology to define seroinfection from immunogenicity data to compare the relative efficacy of PCVs in preventing infection. Our results using individual-level data from a global meta-analysis provide the first estimates of the comparative protection afforded by different pneumococcal vaccines and show that for many serotypes, carriage events are less common after PCV13 than PCV10, likely due to a higher antibody response. In addition, we quantify the relationship between the immune response to vaccination and protection against infection, measured serologically, and show that higher antibody responses in infants are associated with greater protection from infection. Licensure of new vaccines is based on non-inferiority comparisons with current vaccines and the proportion of antibody responses above the agreed threshold as a minimum requirement. Once a vaccine meets this ‘at-least-as-good-as’ immunogenicity criteria, it has previously not been clear whether exceeding it is of benefit, and the WHO position paper on pneumococcal vaccines states ‘It is unknown whether a lower serotype-specific GMC of antibody indicates less efficacy’. Our results show that lower protection against subclinical infection does indeed follow from lower antibody production and that two vaccines that produce a similar level of antibody will provide similar levels of protection. The implications of these findings are of greatest importance when a new vaccine roll-out is being considered. Lower antibody production or lower seroefficacy for one vaccine product does not necessarily imply limited effectiveness against invasive pneumococcal diseases when considering vaccines such as PCV10 and PCV13 which are highly effective vaccines in many settings. Instead, lower antibody responses lead to less rapidly observed indirect protection after implementation into a national programme as a smaller proportion of transmission events are blocked by the vaccine. This is evident in the mathematical modelling which showed less rapid decreases in the number of cases of invasive disease when introducing PCV10 compared with PCV13. Implications for practice This evidence of differences in serotype-specific protection can be incorporated into cost-effectiveness models used to compare vaccine products. Cost-effectiveness studies have highlighted the lack of evidence of comparative efficacy for different PCVs, resulting in previous cost-effectiveness models that ignore serotype-specific differences and assume equivalent efficacy for all serotypes covered by different PCVs. Our study fills this evidence gap and allows researchers and policy-makers to use more accurate vaccine-specific models in decision-making. Our cost-effectiveness analysis of a hypothetical scenario showed that introducing infant PCV13 was predicted to avert a higher burden of pneumococcal disease compared with PCV10. This would have realised a small saving of £13 million discounted over 24 years. When considering the introduction of new pneumococcal vaccines into the routine immunisation schedule, we recommend that differences in antibody responses for different vaccines be considered in modelling scenarios as higher antibody responses result in reduced transmission and greater impact on invasive diseases. Vaccine-specific threshold prices can then be determined for cost-effective vaccines. Our analysis showed that due to its higher efficacy against some serotypes, a higher threshold price per dose could be paid for PCV13 while remaining cost-effective. Study registration This study is registered as PROSPERO CRD42019124580. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/148/03) and is published in full in Health Technology Assessment; Vol. 28, No. 34. See the NIHR Funding and Awards website for further award information.

Background Atopic eczema is a common childhood skin problem linked with asthma, food allergy and allergic rhinitis that impairs quality of life. Objectives To determine whether advising parents to apply daily emollients in the first year can prevent eczema and/or other atopic diseases in high-risk children. Design A United Kingdom, multicentre, pragmatic, two-arm, parallel-group randomised controlled prevention trial with follow-up to 5 years. Setting Twelve secondary and four primary care centres. Participants Healthy infants (at least 37 weeks’ gestation) at high risk of developing eczema, screened and consented during the third trimester or post delivery. Interventions Infants were randomised (1 : 1) within 21 days of birth to apply emollient (Doublebase Gel®; Dermal Laboratories Ltd, Hitchin, UK or Diprobase Cream®) daily to the whole body (excluding scalp) for the first year, plus standard skin-care advice (emollient group) or standard skin-care advice only (control group). Families were not blinded to allocation. Main outcome measures Primary outcome was eczema diagnosis in the last year at age 2 years, as defined by the UK Working Party refinement of the Hanifin and Rajka diagnostic criteria, assessed by research nurses blinded to allocation. Secondary outcomes up to age 2 years included other eczema definitions, time to onset and severity of eczema, allergic rhinitis, wheezing, allergic sensitisation, food allergy, safety (skin infections and slippages) and cost-effectiveness. Results One thousand three hundred and ninety-four newborns were randomised between November 2014 and November 2016; 693 emollient and 701 control. Adherence in the emollient group was 88% (466/532), 82% (427/519) and 74% (375/506) at 3, 6 and 12 months. At 2 years, eczema was present in 139/598 (23%) in the emollient group and 150/612 (25%) in controls (adjusted relative risk 0.95, 95% confidence interval 0.78 to 1.16; p = 0.61 and adjusted risk difference −1.2%, 95% confidence interval −5.9% to 3.6%). Other eczema definitions supported the primary analysis. Food allergy (milk, egg, peanut) was present in 41/547 (7.5%) in the emollient group versus 29/568 (5.1%) in controls (adjusted relative risk 1.47, 95% confidence interval 0.93 to 2.33). Mean number of skin infections per child in the first year was 0.23 (standard deviation 0.68) in the emollient group versus 0.15 (standard deviation 0.46) in controls; adjusted incidence rate ratio 1.55, 95% confidence interval 1.15 to 2.09. The adjusted incremental cost per percentage decrease in risk of eczema at 2 years was £5337 (£7281 unadjusted). No difference between the groups in eczema or other atopic diseases was observed during follow-up to age 5 years via parental questionnaires. Limitations Two emollient types were used which could have had different effects. The median time for starting emollients was 11 days after birth. Some contamination occurred in the control group (< 20%). Participating families were unblinded and reported on some outcomes. Conclusions We found no evidence that daily emollient during the first year of life prevents eczema in high-risk children. Emollient use was associated with a higher risk of skin infections and a possible increase in food allergy. Emollient use is unlikely to be considered cost-effective in this context. Future research To pool similar studies in an individual patient data meta-analysis. Trial registration This trial is registered as ISRCTN21528841. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 12/67/12) and is published in full in Health Technology Assessment; Vol. 28, No. 29. See the NIHR Funding and Awards website for further award information. Plain language summary Eczema is a troublesome itchy skin condition affecting 1 in 5 children and 1 in 10 UK adults. There is no cure and affected children are more likely to develop food allergies. We wanted to see if we could prevent eczema by protecting the skin of babies at higher risk of developing eczema (with an immediate relative with eczema, asthma or hay fever) with moisturisers used to treat dry skin. Previous research suggested that protecting the skin barrier might also prevent food allergy. One thousand three hundred and ninety-four families took part in a study; half of them were asked to apply moisturiser every day to their newborn baby for the first year and half to look after their baby’s skin in the normal way. At the age of 2 years, we did not see any difference in how common eczema was between the two groups: 23% had eczema in the moisturiser group and 25% in the normal care group. It did not matter how we defined eczema – whether examined by a researcher or parent report. We did not find any differences in related conditions like asthma or hay fever either. We found that children using moisturisers had seen their doctor slightly more often for mild skin infections. There was a hint that food allergy might have been increased in the moisturiser group, but there was not enough data to be sure. We followed up the children to age 5 years, but we still did not find any benefits from using moisturisers in early life. Since this study, other similar research has been done using newer types of moisturisers, but their results are the same. This study shows that using daily moisturisers on healthy babies with a high risk of eczema does not prevent eczema. It is one less thing for busy families to worry about. Scientific summary Background Eczema, also known as atopic eczema (AE) or atopic dermatitis, is a common inflammatory skin disease that typically starts in early life. Eczema affects around 20% of children and 10% of adults in developed countries. Although most cases are mild, moderate to severe disease can have a major impact on the quality of life of an affected child and their family. The main symptom of this systemic inflammatory disease is itching which leads to scratching, bleeding, secondary infection as well as sleep loss and the social stigma of a visible skin disease. Genetic factors such as mutations in the gene encoding filaggrin (FLG) – a protein that is important for maintaining skin barrier function – have been shown to be important in increasing eczema risk, severity and persistence. Epidemiological studies showing that eczema is more common in smaller families, migrant populations and in higher socioeconomic status also suggest that the environment is critically important in determining disease expression. Eczema is closely related to other ‘atopic’ conditions including asthma, hay fever and food allergy. Some evidence suggests that the impaired skin barrier caused by eczema in early life is an important route for the development of food allergy as a result of sensitisation through the skin. Treatment of eczema depends on severity: mild eczema can be treated with emollients (moisturisers) and weak topical corticosteroids. Moderate eczema usually needs potent topical corticosteroids and other anti-inflammatory treatments such as topical tacrolimus, whereas severe eczema is treated by ultraviolet light or systemic treatments including ciclosporin, methotrexate, dupilumab and an emerging pipeline of new biologics. While good progress has been made with treatment, prevention of eczema remains challenging. Strategies include avoidance of allergens (foods and house dust mite) during pregnancy and early life, exclusive breastfeeding, or different timing of introducing solids, but these have not shown convincing preventative effects. Probiotics and prebiotics have shown some benefit, but the exact strain or combination and timing of intervention remain unclear. One previously unexplored eczema prevention strategy is enhancement of the skin barrier in early life. The hypothesis is that enhancing the skin barrier could interrupt an early cascade of inflammatory events that can lead to chronic auto-immune eczema and potentially prevent skin sensitisation to common allergens that can lead to food allergy. Two small pilot studies suggested that protecting the skin of babies who had a first-degree relative with eczema, asthma or hay fever with emollients could prevent the development of eczema in the first year of life. We wanted to see whether daily emollient application for the first year of life in such high-risk babies can prevent eczema and other allergic diseases in a sustained and convincing way by means of a definitive pragmatic randomised controlled trial. Objectives The primary objective was to determine whether advising parents to apply emollient daily for the first year of life could prevent eczema at 2 years of age when compared with standard skin-care advice alone in children at high risk of developing eczema. Secondary objectives included evaluating whether prophylactic use of emollients would delay the onset of eczema or reduce eczema severity when compared to standard skin-care advice, whether any preventive effect at age 2 years was sustained up to age 5 years and to determine the safety and cost-effectiveness of such a strategy. We also sought to determine whether emollients could prevent the development of other associated allergic conditions including food allergy, asthma and hay fever up to the age of 5 years. Methods We conducted a multicentre, pragmatic, two-arm parallel group randomised controlled prevention trial which recruited participants from 12 hospitals and 4 general practices in the UK. Participants were infants of at least 37 weeks’ gestation at high risk of developing eczema defined by having at least one first-degree relative with a parental report of doctor-diagnosed eczema, allergic rhinitis or asthma. Mothers had to be 16 years or older, and the consenting adult had to understand English. We excluded babies with a severe widespread skin condition that would make eczema assessment difficult; a serious health issue that would make it difficult for the family to take part in the trial; or a condition that would make the use of emollient inadvisable. Informed consent was obtained from mothers during pregnancy, or mother, father or guardian post delivery. The intervention group was advised to use one of two study emollients (Doublebase Gel® or Diprobase Cream®) at least once daily to the whole body (excluding the scalp) until the child reached 1 year, plus standard skin-care advice (designated the ‘emollient group’). The control group was advised to use standard skin-care advice only. Standard skin-care guidance, in booklet and video format at the time of randomisation, provided advice to use mild cleansers and shampoos specifically formulated for infants, and to avoid soap, bubble bath and baby wipes, and was given to both groups. Intervention adherence was measured at 3, 6 and 12 months and defined as satisfactory if emollients were applied at least three to four times per week to most of the child’s body. A similar definition was used to define contamination in the control group. The primary outcome was presence of eczema in the last year when the baby was aged 2 years, defined as meeting the United Kingdom Working Party (UKWP) Diagnostic Criteria for AE for children under 4 years. Secondary outcomes encompassed other ways of defining eczema including any parental report of a clinical diagnosis of eczema up to 2 years, parent completion of UKWP criteria at 1 and 2 years, and presence of visible eczema at 2 years recorded by a nurse masked to treatment allocation. Other eczema secondary outcomes included time to onset of eczema (based on first parent report of clinician diagnosis and first topical corticosteroid or immunosuppressant prescription); severity of eczema measured by trained nurses using the Eczema Area and Severity Index at 2 years and parent-reported Patient-Oriented Eczema Measure at 1 and 2 years. Other secondary outcomes included parent-reported wheezing and allergic rhinitis between 1 and 2 years; allergic sensitisation (masked skin prick tests) to milk, egg, peanut, cat dander, grass pollen or dust mite at 2 years; parent-reported food allergy; parental report of clinical diagnosis of food allergy at 1 and 2 years; and allergy to milk, egg or peanut at 2 years confirmed either by oral food challenge or for cases in which no oral food challenge was done, an expert panel of experienced paediatric allergists masked to treatment allocation. The panel decisions were made using an algorithm, validated using data from a previous trial, which incorporates all available data including skin prick test results, previous reaction history, frequency of food ingestion and allergy tests done outside the trial. The main economic outcome measure was incremental cost per percentage decrease in risk of eczema at 2 years in a cost-effectiveness analysis (CEA). Cost–utility analyses using quality-adjusted life-years (QALYs) derived from the proxy Child Health Utility instrument-9 domains (CHU-9D) for the child was also conducted as a secondary analysis. Safety outcomes were parent-reported skin infections (parents were asked what the doctor called the infection) and emollient-related infant slippages during the intervention period (year 1). Tertiary outcomes at 3, 4 and 5 years were to evaluate if emollients in early life could also prevent later-onset allergic diseases including asthma, hay fever and food allergy, and to see whether any early possible benefits of eczema prevention were sustained. Outcomes included parental-reported presence of eczema; severity of eczema; wheezing; allergic rhinitis; food allergy symptoms and clinical diagnoses of asthma, allergic rhinitis and food allergy. Randomisation (1 : 1) within 21 days of birth was performed using computer-generated pseudo-random code with permuted blocks of randomly varying size and concealed from the trial investigators via a web-based randomisation system. Randomisation was stratified by recruiting centre and number of first-degree relatives with eczema, asthma or hay fever (1, 2 or > 2). Research nurses conducting outcome assessments were masked to participant treatment allocation. Families were not masked to the allocated interventions and were reminded not to reveal allocation to research nurses. The trial was powered to detect a relative reduction of 30% in the primary outcome at the 5% significance level (two-sided) with 90% power assuming that 30% of children in the control group would have eczema and 20% attrition, resulting in a sample size of 1282. The target sample size was reached faster than anticipated at which point the Trial Steering Committee permitted that all pregnant mothers who had already consented to the trial could be randomised on the birth of the baby giving a maximum possible sample size of 1400. We analysed participants as randomised regardless of adherence with allocation using available data (i.e. without imputation for missing data) with sensitivity analyses for missing data. The adjusted relative risk (RR) and risk difference for the primary outcome were estimated using generalised estimating equations with the binomial family and log/identity link respectively, with an exchangeable correlation matrix to account for randomisation stratification by centre and number of immediate family members with atopic disease (one, two, or more than two) included as a covariate. A number of planned subgroup analyses for the primary outcome were conducted including according to FLG genotype, to test the hypothesis that FLG null genotype affects response to intensive emollient use from birth. Results One thousand three hundred and ninety-four newborns were randomised between 19 November 2014 and 18 November 2016; 693 to the emollient group and 701 to the control group. Primary outcome data at 2 years were collected for 1210 infants (87%). Unblinding of research nurses prior to skin examination occurred for 12 infants in the intervention group and 6 in the control group. Adherence in the emollient group was 88% (466/532) at 3 months, 82% (427/519) at 6 months and 74% (375/506) at 12 months. In the control group, contamination due to self-directed use of emollients was reported for 18% (82/457), 17% (62/372) and 15% (49/324) at 3, 6 and 12 months, respectively, for infants who did not have a parental report of a doctor diagnosis of eczema. At age 2 years, eczema was present in 139 (23%) of 598 infants in the emollient group and in 150 (25%) of 612 infants in the control group [adjusted RR 0.95, 95% confidence interval (CI) 0.78 to 1.16; p = 0.61; adjusted risk difference −1.2%, 95% CI −5.9% to 3.6%]. All sensitivity analyses conducted, including using multiple imputation for missing primary outcome data, were consistent with the primary analysis. There was no evidence of an interaction effect with allocated group for the primary eczema outcome in any of the subgroup analyses (including FLG genotype). Other eczema definitions supported the results of the primary analysis. Eczema severity and time to onset of eczema were also similar in the two groups. Mean number of skin infections per child in year 1 was 0.23 [standard deviation (SD) 0.68] in the emollient group versus 0.15 (SD 0.46) in the control group; adjusted incidence rate ratio 1.55 (95% CI 1.15 to 2.09). Infant slippage incidents were reported for 15/584 (2.6%) in the emollient group and 11/584 (1.9%) in the control group. Food allergies to milk, egg or peanut at 2 years were confirmed in 41/547 (7.5%) infants in the emollient group and 29/568 (5.1%) in the control group (adjusted RR 1.47, 95% CI 0.93 to 2.33). The largest difference was in the proportion of infants with confirmed food allergy to egg, with an adjusted RR of 1.56 (95% CI 0.92 to 2.65). Results of other measures of food allergy and food sensitisation were similar. The proportion of infants with allergic rhinitis, wheezing and allergic sensitisation to cat dander, grass pollen and dust mite was similar between groups at 2 years. The differences in quality of life (using CHU-9D for the child and EuroQol-5 Dimensions, five-level version for the main carer) between the two groups were very small. Although the emollient intervention period was the first year of life, parents in the emollient group continued to report more frequent moisturiser application through to 5 years than in the control group. By 5 years, 188/608 (31%) parents in the emollient group had reported a clinical diagnosis of eczema in their child since 12 months, compared with 178/631 (28%) in the control group (adjusted RR 1.10, 95% CI 0.93 to 1.30). A diagnosis of food allergy by 5 years was reported for 92/609 (15%) allocated to emollients and 87/632 (14%) allocated to control (adjusted RR 1.11, 95% CI 0.84 to 1.45). Similarly, the percentage of parents reporting that their child had a clinical diagnosis of asthma or allergic rhinitis by 5 years were similar in the two groups. In the complete-case CEA mean cost was £398.23 (SD 1408.39) per child in the emollient group (n = 598) and £312.16 (SD 1105.04) in the control group (n = 610). When intervention use was combined with other health resource use, the adjusted incremental cost was £87.45 (95% CI −54.31 to 229.27). The adjusted difference in the proportion of children without eczema at 2 years was 0.0164 (95% CI −0.0329 to 0.0656) higher in the emollient group compared to the control group. The adjusted incremental cost per percentage decrease in risk of eczema was £5337 at 2 years. Adjusted QALYs for children were very slightly improved (i.e. higher) in the emollient group at 2 years. Conclusions We found no evidence of a useful preventive effect of emollients for eczema at our primary outcome time of 2 years. The failure to show a reduction in eczema was consistent regardless of how eczema was assessed. Some evidence of an increase in skin infections during the intervention period and a possible increase in food allergy at age 2 years was observed. No benefit was observed for time to onset of eczema or eczema severity, and no benefits were observed for eczema, asthma, hay fever or food allergy in longer-term follow-up to 5 years. Emollient use is unlikely to be considered cost effective in this context. Inclusion of individual patient data from all similar eczema prevention studies in further meta-analysis may provide a clearer assessment of whether emollients can prevent eczema and related diseases and provide more certainty about potential harms. Implications for health care The study does not support the use of emollients to prevent eczema and has found a small signal of possible harms, so this intervention cannot be recommended for health care or public health use. As the study relates to prevention of eczema, emollients should continue to be used as part of standard treatment for eczema. Trial registration This trial is registered as ISRCTN21528841. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 12/67/12) and is published in full in Health Technology Assessment; Vol. 28, No. 29. See the NIHR Funding and Awards website for further award information.

Background Parkinson’s disease is a brain condition causing a progressive loss of co ordination and movement problems. Around 145,500 people have Parkinson’s disease in the United Kingdom. Levodopa is the most prescribed treatment for managing motor symptoms in the early stages. Patients should be monitored by a specialist every 6–12 months for disease progression and treatment of adverse effects. Wearable devices may provide a novel approach to management by directly monitoring patients for bradykinesia, dyskinesia, tremor and other symptoms. They are intended to be used alongside clinical judgement. Objectives To determine the clinical and cost-effectiveness of five devices for monitoring Parkinson’s disease: Personal KinetiGraph, Kinesia 360, KinesiaU, PDMonitor and STAT-ON. Methods We performed systematic reviews of all evidence on the five devices, outcomes included: diagnostic accuracy, impact on decision-making, clinical outcomes, patient and clinician opinions and economic outcomes. We searched MEDLINE and 12 other databases/trial registries to February 2022. Risk of bias was assessed. Narrative synthesis was used to summarise all identified evidence, as the evidence was insufficient for meta-analysis. One included trial provided individual-level data, which was re-analysed. A de novo decision-analytic model was developed to estimate the cost-effectiveness of Personal KinetiGraph and Kinesia 360 compared to standard of care in the UK NHS over a 5-year time horizon. The base-case analysis considered two alternative monitoring strategies: one-time use and routine use of the device. Results Fifty-seven studies of Personal KinetiGraph, 15 of STAT-ON, 3 of Kinesia 360, 1 of KinesiaU and 1 of PDMonitor were included. There was some evidence to suggest that Personal KinetiGraph can accurately measure bradykinesia and dyskinesia, leading to treatment modification in some patients, and a possible improvement in clinical outcomes when measured using the Unified Parkinson’s Disease Rating Scale. The evidence for STAT-ON suggested it may be of value for diagnosing symptoms, but there is currently no evidence on its clinical impact. The evidence for Kinesia 360, KinesiaU and PDMonitor is insufficient to draw any conclusions on their value in clinical practice. The base-case results for Personal KinetiGraph compared to standard of care for one-time and routine use resulted in incremental cost-effectiveness ratios of £67,856 and £57,877 per quality-adjusted life-year gained, respectively, with a beneficial impact of the Personal KinetiGraph on Unified Parkinson’s Disease Rating Scale domains III and IV. The incremental cost-effectiveness ratio results for Kinesia 360 compared to standard of care for one-time and routine use were £38,828 and £67,203 per quality-adjusted life-year gained, respectively. Limitations The evidence was limited in extent and often low quality. For all devices, except Personal KinetiGraph, there was little to no evidence on the clinical impact of the technology. Conclusions Personal KinetiGraph could reasonably be used in practice to monitor patient symptoms and modify treatment where required. There is too little evidence on STAT-ON, Kinesia 360, KinesiaU or PDMonitor to be confident that they are clinically useful. The cost-effectiveness of remote monitoring appears to be largely unfavourable with incremental cost-effectiveness ratios in excess of £30,000 per quality-adjusted life-year across a range of alternative assumptions. The main driver of cost-effectiveness was the durability of improvements in patient symptoms. Study registration This study is registered as PROSPERO CRD42022308597. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135437) and is published in full in Health Technology Assessment; Vol. 28, No. 30. See the NIHR Funding and Awards website for further award information. Plain language summary Parkinson’s disease is a brain condition causing loss of co-ordination and movement problems. Levodopa is the most prescribed treatment for early disease. Patients should be seen by a specialist every 6–12 months to assess their treatment needs. Wearable devices (like smart watches) may aid management by directly monitoring patients for disease symptoms including tremor and slowness of movement (bradykinesia), or side effects of treatment like involuntary movement (dyskinesia). This assessment considered the clinical and economic value of five wearable devices: Personal KinetiGraph, STAT-ON, Kinesia 360, KinesiaU and PDMonitor. We searched medical databases to find all studies of the five devices. We assessed the quality of these studies and reviewed their results. We found 77 studies of the devices. There was some evidence to suggest that Personal KinetiGraph can accurately measure bradykinesia and dyskinesia, leading to treatment modification in some patients, and a possible improvement in symptoms. The evidence for STAT-ON suggested it may be of value for diagnosing symptoms, but there is currently no evidence on its clinical value. There was insufficient evidence for Kinesia 360, KinesiaU and PDMonitor to draw any conclusions. An economic analysis was conducted to investigate whether using any of these technologies is economically viable. The economic analysis found that the quality-of-life benefits generated by remote monitoring devices were small relative to the additional costs of implementing them in the NHS. As such, none of the remote monitoring devices were good value for money when compared with the current standard of care. Scientific summary Background Parkinson’s disease (PD) is a condition that affects the brain, resulting in a progressive loss of co-ordination as well as movement problems. In the early stages of PD, the three main motor symptoms are shaking (tremor), slowness of movement (bradykinesia) and muscle stiffness (rigidity). There are around 145,500 people living with PD in the UK. The risk of developing the disease increases sharply with age. Levodopa is the most prescribed treatment for managing the motor symptoms of PD in the early stages. However, it may be associated with significant motor complications, including response fluctuations and dyskinesias (involuntary movements). Response fluctuations are characterised by large variations in motor performance, with normal function during the ‘on’ period, and weakness and restricted mobility during the ‘off’ period. Deep brain stimulation and levodopa–carbidopa intestinal gel can be considered in people with advanced PD whose symptoms do not respond adequately to best medical therapy. The National Institute for Health and Care Excellence (NICE) recommends that people with Parkinson’s disease (PwP) should be seen by a specialist every 6–12 months initially, then more often with increasing disease complexity, although this is often difficult because of the increasingly ageing population and demands on PD services. Remote monitoring devices are intended to be used alongside clinical judgement to assess disease severity and help manage PD symptoms and adverse effects of treatment. Results of the monitoring are analysed remotely and a summary provided to the specialist physician and/or to the patient. The data should be used to determine whether any changes in the treatment regimen are desirable, in consultation with the patient. This assessment considers only wearable, remote monitoring devices that produce results with no input, or limited input, from the user. Five relevant devices were identified for consideration: Personal KinetiGraph (PKG) Movement Recording System (Global Kinetics); Kinesia 360 motor assessment system (Great Lakes NeuroTechnologies); KinesiaU motor assessment system (Great Lakes NeuroTechnologies); PDMonitor (PD Neurotechnology); and STAT-ON (Sense4Care). Objectives To determine the clinical and cost-effectiveness of the five included remote monitoring devices in PwP. Methods Systematic review Systematic reviews were conducted following the general principles recommended in the Centre for Reviews and Dissemination (CRD) guidance and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. Comprehensive searches of the literature were conducted to identify all studies relating to the use of the five remote continuous monitoring devices. MEDLINE, EMBASE, CENTRAL, ClinicalTrials.gov and other databases and registries were searched on 1 February 2022. Two reviewers independently screened all titles and abstracts. All clinical studies of any of the five included devices, where used in people with PD (of any severity or stage), were eligible for inclusion. The key comparator was clinical judgement of disease symptoms without the use of remote monitoring devices; however, included studies did not have to have a comparator group. Outcomes of interest included: association and diagnostic accuracy between outputs of remote monitoring (such as bradykinesia score, dyskinesia score) and clinical measures [such as Unified Parkinson’s Disease Rating Scale (UPDRS) score or clinical judgement of symptoms]; all impacts on clinical decision-making, such as changes in therapy and dose modification; all clinical outcomes, such as UPDRS or Hoehn and Yahr scores, morbidities and mortality; and all patient, carer or clinician opinions on the technologies. Data reported in publications were extracted by one reviewer and independently checked by a second reviewer. Study quality was assessed using suitable tools, such as Quality Assessment of Diagnostic Accuracy Studies-2 for diagnostic accuracy studies and the Cochrane Risk-of-Bias tool for clinical trials. Evidence was synthesised using a narrative synthesis approach. The results of data extraction were presented in structured tables and as a narrative summary. A broad thematic synthesis was used to identify key issues arising from the extracted evidence. Due to the diversity of reporting across studies, meta-analysis was not feasible for any outcomes. One clinical trial provided its individual participant data; this was re-analysed for this report. Economic analysis Two cost-effectiveness reviews were conducted: (1) a review of remote monitoring devices for people with PD, and (2) a review of existing decision models evaluating treatments for people with PD. The titles and abstracts of all reports identified by the bibliographic searches were screened independently by two researchers. Key findings were summarised narratively. A de novo decision-analytic model was developed to assess the potential health gains and costs associated with implementing remote monitoring in the NHS. The base-case analysis considers only the cost-effectiveness of PKG and Kinesia 360, which are compared on a pairwise basis with current standard of care (SoC). The cost-effectiveness of other remote monitoring technologies was explored in scenario analysis. Based on company information, real-world applications of PKG and expert clinical advice, the External Assessment Group (EAG) assessed two alternative monitoring strategies: (1) one-time use: remote monitoring implemented at model baseline and as a one-time aid to clinical assessment, and (2) routine use: remote monitoring used at every follow-up assessment (i.e. over the review period at regular intervals) to routinely assist clinical judgement. The EAG model was based on a Markov model structure, which sought to capture changes in the Movement Disorder Society-Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) domain scale scores, as an indicator of the level of symptom control associated with the use of remote continuous monitoring devices relative to SoC. These changes in MDS-UPDRS were informed by the clinical literature and were linked to health-related quality of life to assess quality-adjusted life-year (QALY) changes associated with remote monitoring. The economic analysis also captured cost differentials between SoC and alternative remote monitoring strategies considering: (1) costs associated with using each remote monitoring device, (2) changes in levodopa-equivalent medication use, (3) implementation costs and (4) follow-up consultations. The costs applied were independent of MDS-UPDRS scores modelled. Changes in levodopa-equivalent medication were, however, informed by the relevant clinical effectiveness literature so as to align with the applied treatment effects. Clinical effectiveness results Seventy-seven studies of clinical effectiveness were included in the systematic review. There were 57 studies of PKG. The diagnostic accuracy studies suggested that PKG has good accuracy for assessing bradykinesia, dyskinesia and tremor, but lower accuracy to detect sleep disturbance. Studies reporting changes in management found that PKG provided additional information leading to a change in the clinical management plan in 31.8–79% of patients (depending on study), most commonly an increase in treatment dose. One non-randomised comparative trial of 162 patients provided individual participant data to the EAG. The results show that the use of PKG appears to improve UPDRS scores, particularly UPDRS III (by around 3.1 points) and UPDRS IV (by around 1.2 points). This is likely to be because PKG use is reducing time with bradykinesia (by 2.1 percentage points) dyskinesia (by 1.5 percentage points) and tremor (by 0.6 percentage points), although none of these reductions achieved statistical significance. The trial data suggested that PKG use predominately improves symptoms (particularly bradykinesia and UPDRS scores) in people who were not ‘in target’ and whose condition was not adequately controlled. Other trials reporting clinical outcomes were not comparative, but generally supported the evidence that PKG use improves UPDRS scores. Patient opinion was broadly supportive of PKG use, particularly as a reminder to take medication. Patients mostly felt that PKG provided additional useful information on their symptoms (59–79% felt this), but clinicians were more equivocal; only 33–47% felt PKG provided additional information. There were 15 included studies of the STAT-ON device. STAT-ON had high diagnostic accuracy to detect treatment ‘on–off’ times and bradykinesia, and high sensitivity but lower specificity to detect freezing of gait. There were no studies that presented evidence on the intermediate or clinical impact of STAT-ON. There were three included studies for the Kinesia 360 motor assessment system. It had moderate-to-good diagnostic accuracy to detect bradykinesia, dyskinesia and tremor. Two small randomised controlled trials (RCTs, 64 patients) found some inconclusive evidence that Kinesia 360 improved UPDRS III and Parkinson’s Disease Quality of Life 39 Questions scores when compared to standard management. One small cohort study (16 patients) of KinesiaU motor assessment system was included. The EAG consider this as too small to draw any meaningful conclusions. For PDMonitor only one conference abstract and one small case study were included. The EAG consider this as insufficient to draw any meaningful conclusions. Cost-effectiveness results Estimated changes in UPDRS associated with PKG show that PKG is associated with small unfavourable changes in UPDRS domains I and IV. In consideration of these highly uncertain results, the base-case analysis considered two alternative efficacy configurations for PKG: (1) an unrestricted analysis (considering all UPDRS domains), and (2) a restricted analysis (considering only UPDRS domains III and IV). The deterministic base-case incremental cost-effectiveness ratio (ICER) for PKG using a one-time use strategy was £67,856 and £202,363 per QALY for the restricted and unrestricted analysis, respectively. Considering a routine strategy, the deterministic base-case ICER was £57,877 per QALY in the restricted analysis and £172,602 per QALY in the unrestricted analysis. The deterministic base-case ICERs for Kinesia 360 using one-time use and routine remote monitoring strategies were £38,828 and £67,203 per QALY, respectively. Probabilistic results for PKG and Kinesia 360 aligned with the deterministic values. Sensitivity analysis demonstrated that cost-effectiveness results were sensitive to assumptions regarding the durability of modelled treatment effects. Scenarios with low or zero waning of the treatment effect improved cost-effectiveness markedly. Results were otherwise broadly robust to a range of alternative assumptions and parameter inputs. The EAG was not able to evaluate the cost-effectiveness of STAT-ON, KinesiaU or PDMonitor due to a lack of comparative clinical effectiveness evidence. In a cost comparison (assuming a 5-year time horizon), modelled device costs were lowest for PKG provided three devices or less were ordered per annum, followed by KinesiaU, STAT-ON, Kinesia 360 and PDMonitor. Discussion This assessment includes a comprehensive investigation of the diagnostic accuracy and clinical efficacy of remote monitoring devices for PD. The review used extensive database searches to identify all published evidence on the included technologies and followed rigorous recommended review methods to identify relevant publications, assess their risk of bias and undertake a narrative synthesis of the results. The review identified a substantial literature on the diagnostic accuracy of PKG, and a smaller literature on clinical efficacy. Evidence for other remote monitoring devices was generally limited. PKG appears to accurately measure several symptoms of PD, including dyskinesia, bradykinesia, tremor and treatment-related outcomes. PKG also appears to generate clinical benefits compared with clinical management alone, with improvements in UPDRS III and IV scores. However, the available evidence was generally low quality, particularly for diagnostic accuracy. This casts some doubt on the validity of the results reported in the identified studies. The cost-effectiveness of remote monitoring appears to be largely unfavourable with ICERs in excess of £30,000 per QALY. Cost-effectiveness results were largely robust to alternative assumptions and parameter inputs. The key drivers identified were: (1) the direction and magnitude of changes on UPDRS associated with remote monitoring strategies, (2) the persistence in changes to UPDRS (treatment waning) and (3) the number of devices requested (PKG). Insurmountable limitations in the evidence base meant that the EAG were unable to assess the cost-effectiveness of STAT-ON, KinesiaU or PDMonitor. Comparative evidence for Kinesia 360 was also extremely limited and unlikely to be comparable with that used for PKG, thereby making comparisons problematic. This essentially limits comparisons across alternative monitoring devices to a cost-minimisation exercise, which necessarily implies strong assumptions about relative efficacy. Conclusions The EAG considers that the evidence for PKG shows that it could be of use in clinical practice, provided it can be made cost-effective. It provides useful information on key symptoms of PD, including bradykinesia, dyskinesia and tremor. The use of PKG leads to changes in treatment management for at least some patients, and possible improvement in symptoms. Although there is some promising evidence for STAT-ON and Kinesia 360, the EAG considers that the evidence is currently not sufficient to be confident that these technologies will produce clinical benefits for patients. The EAG considers that there is too little evidence for KinesiaU or PDMonitor to draw any conclusions regarding their clinical value. Almost all current evidence relates to patients receiving pharmacological therapy, mainly levodopa. The EAG notes that, at present, it is unclear whether PKG or other remote monitoring technologies offer any clinical benefit in other patients, such as those receiving advanced therapies. Concerns about potential bias, together with the other limitations in the available evidence, means that cost-effectiveness estimates are highly uncertain. Key uncertainties relate to the magnitude and durability of treatment effects. The results of the economic analysis are largely unfavourable with ICERs in excess of thresholds typically adopted by NICE. Suggested research priorities The primary research priority should be to conduct further studies into the clinical impact of remote monitoring devices. This should focus on expanding the evidence base for PKG and Kinesia 360, where there is currently limited evidence on clinical effects, as well as conducting studies of STAT-ON, KinesiaU and PDMonitor, where there is currently no evidence of clinical effects. Any future studies of comparative effectiveness should address the methodological limitations of the current evidence, as identified by this report. These would preferably be RCTs with pre-specified outcome measures. Studies should be carefully designed to consider the most applicable remote monitoring schedules and settings, as there is significant potential for variation in how remote monitoring devices could be used in practice. Specific consideration should be given to longer-term routine use of remote monitoring devices; currently all evidence pertains to short-term applications. Future studies of remote monitoring devices for PD may also consider patients with early and advanced disease. There is currently no evidence in these populations for any device. Implementing remote monitoring may have a range of resource consequences that are currently not fully understood and may impact significantly on cost-effectiveness. This may include impacts on healthcare professionals’ time and administration of the devices, as well as risks such as loss, damage or theft of devices. Where possible, future studies should seek to address these uncertainties by collecting appropriate data on resource implications. Collecting further diagnostic accuracy evidence is considered a lower priority, but could be useful for Kinesia 360, KinesiaU and PDMonitor, where evidence is lacking. Diagnostic accuracy studies should evaluate the accuracy of these technologies for measuring bradykinesia and dyskinesia. Care should be taken to ensure the reference standard is robust and at a low risk of bias. It may be helpful for such studies to compare the technologies to PKG. Study registration This study is registered as PROSPERO CRD42022308597. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135437) and is published in full in Health Technology Assessment; Vol. 28, No. 30. See the NIHR Funding and Awards website for further award information.

Background The CaRi-Heart® device estimates risk of 8-year cardiac death, using a prognostic model, which includes perivascular fat attenuation index, atherosclerotic plaque burden and clinical risk factors. Objectives To provide an Early Value Assessment of the potential of CaRi-Heart Risk to be an effective and cost-effective adjunctive investigation for assessment of cardiac risk, in people with stable chest pain/suspected coronary artery disease, undergoing computed tomography coronary angiography. This assessment includes conceptual modelling which explores the structure and evidence about parameters required for model development, but not development of a full executable cost-effectiveness model. Data sources Twenty-four databases, including MEDLINE, MEDLINE In-Process and EMBASE, were searched from inception to October 2022. Methods Review methods followed published guidelines. Study quality was assessed using Prediction model Risk Of Bias ASsessment Tool. Results were summarised by research question: prognostic performance; prevalence of risk categories; clinical effects; costs of CaRi-Heart. Exploratory searches were conducted to inform conceptual cost-effectiveness modelling. Results The only included study indicated that CaRi-Heart Risk may be predictive of 8 years cardiac death. The hazard ratio, per unit increase in CaRi-Heart Risk, adjusted for smoking, hypercholesterolaemia, hypertension, diabetes mellitus, Duke index, presence of high-risk plaque features and epicardial adipose tissue volume, was 1.04 (95% confidence interval 1.03 to 1.06) in the model validation cohort. Based on Prediction model Risk Of Bias ASsessment Tool, this study was rated as having high risk of bias and high concerns regarding its applicability to the decision problem specified for this Early Value Assessment. We did not identify any studies that reported information about the clinical effects or costs of using CaRi-Heart to assess cardiac risk. Exploratory searches, conducted to inform the conceptual cost-effectiveness modelling, indicated that there is a deficiency with respect to evidence about the effects of changing existing treatments or introducing new treatments, based on assessment of cardiac risk (by any method), or on measures of vascular inflammation (e.g. fat attenuation index). A de novo conceptual decision-analytic model that could be used to inform an early assessment of the cost effectiveness of CaRi-Heart is described. A combination of a short-term diagnostic model component and a long-term model component that evaluates the downstream consequences is anticipated to capture the diagnosis and the progression of coronary artery disease. Limitations The rapid review methods and pragmatic additional searches used to inform this Early Value Assessment mean that, although areas of potential uncertainty have been described, we cannot definitively state where there are evidence gaps. Conclusions The evidence about the clinical utility of CaRi-Heart Risk is underdeveloped and has considerable limitations, both in terms of risk of bias and applicability to United Kingdom clinical practice. There is some evidence that CaRi-Heart Risk may be predictive of 8-year risk of cardiac death, for patients undergoing computed tomography coronary angiography for suspected coronary artery disease. However, whether and to what extent CaRi-Heart represents an improvement relative to current standard of care remains uncertain. The evaluation of the CaRi-Heart device is ongoing and currently available data are insufficient to fully inform the cost-effectiveness modelling. Future work A large (n = 15,000) ongoing study, NCT05169333, the Oxford risk factors and non-invasive imaging study, with an estimated completion date of February 2030, may address some of the uncertainties identified in this Early Value Assessment. Study registration This study is registered as PROSPERO CRD42022366496. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135672) and is published in full in Health Technology Assessment; Vol. 28, No. 31. See the NIHR Funding and Awards website for further award information. Plain language summary Coronary artery disease affects around 2.3 million people in the United Kingdom. It is caused by a build-up of fatty plaques on the walls of the blood vessels that supply the heart muscle. This can reduce the flow of blood to the heart and result in people experiencing chest pain (angina), especially when they exercise. Over time, the fatty plaques can grow and block more or all of the artery and blood clots can also form, causing blockage. A heart attack happens when the supply of blood to the heart muscle is blocked. People who have episodes of chest pain, whose doctors think that they may have coronary artery disease, can have a type of imaging (computed tomography coronary angiography) which shows whether there is any narrowing of their coronary arteries. When offering treatment, specialist heart doctors are likely to consider a person’s symptoms and other risk factors (such as family history of heart disease, diabetes and smoking history), as well as how much narrowing of the arteries has happened. CaRi-Heart® is a computer programme that uses information about inflammation in a person’s coronary arteries, together with recognised risk factors, such as age, sex, smoking, high cholesterol levels, high blood pressure and diabetes, to estimate an individual’s risk of dying from a heart attack in the next 8 years. There is evidence that CaRi-Heart® is better at estimating this risk than using information recognised risk factors alone. However, there is a lack of information about how treatment could change as a result of using CaRi-Heart® and whether any changes would improve outcomes for patients. There is also a lack of information about how much CaRi-Heart® would cost the National Health Service. Scientific summary Background Coronary artery disease (CAD) and acute myocardial infarction (AMI) are a significant health burden in the UK, with ischaemic heart disease being the leading cause of death in males. Guidelines from the National Institute for Health and Care Excellence (NICE) and the European Society of Cardiology (ESC) recommend computed tomography coronary angiography (CTCA) for the investigation of CAD in people with stable chest pain. CTCA provides a visualisation of the coronary arteries, which is used to identify plaques, to quantify the extent of any stenosis of the coronary arteries and the length and location of the affected area, and to quantify the extent of coronary artery calcification. Information provided by CTCA is structural rather than functional. Acute coronary events can arise from unstable, but anatomically non-significant, atherosclerotic plaques. The vascular inflammatory response is a modulator of atherogenesis and can be a factor in plaque rupture, leading to acute coronary events. CaRi-Heart® is a cloud-based CE-marked medical device (Caristo Diagnostics Ltd, Oxford, UK) that analyses images from CTCA scans to provide information about inflammation in the coronary arteries. The CaRi-Heart device uses this information to generate a perivascular fat attenuation index (FAI) score. It then estimates individual patient risk of 8-year cardiac death with a prognostic model, which includes the perivascular FAI score, as well as atherosclerotic plaque burden and clinical risk factors. This Early Value Assessment (EVA) considers whether CaRi-Heart Risk has potential to provide an effective, safe and cost-effective adjunctive investigation for assessment of cardiac risk, in people with stable chest pain/suspected CAD, who are undergoing CTCA. This assessment does not include the development of an executable cost-effectiveness model but does include conceptual modelling which explores the structure and evidence about parameters required for model development. Objectives A series of research questions were defined that could inform both a full assessment of the clinical and cost effectiveness of using CaRi-Heart, as an adjunctive investigation for assessment of cardiac risk, in people with stable chest pain/suspected CAD, who are undergoing CTCA and consideration of the potential of this technology to be cost-effective: What is the prognostic performance of CaRi-Heart, in people with stable chest pain, who are undergoing CTCA, where: the dependent variable is cardiac death? the dependent variable is a major adverse cardiovascular event (MACE)? What is the prevalence of ‘low’, ‘medium’ and ‘high’ CaRi-Heart Risk in people with no evidence of CAD, people with evidence of non-obstructive CAD and people with evidence of obstructive CAD, based on currently available CTCA imaging? What are the clinical effects of using CaRi-Heart to assess cardiac risk? How does CaRi-Heart Risk affect treatment decisions and patient adherence in people with no evidence of CAD, people with evidence of non-obstructive CAD and people with evidence of obstructive CAD, based on currently available CTCA imaging? What are the clinical effects of any changes to treatment, based on CaRi-Heart Risk, in people with no evidence of CAD, people with evidence of non-obstructive CAD and people with evidence of obstructive CAD, based on currently available CTCA imaging? What are the costs, from a UK NHS and Personal Social Services (PSS) perspective, using CaRi-Heart, as an adjunctive investigation for assessment of cardiac risk, in people with stable chest pain, who are undergoing CTCA? How might a conceptual model be specified in terms of structure and evidence required for parameterisation in order to estimate the cost-effectiveness of CaRi-Heart in people with stable chest pain, who are undergoing CTCA? Methods Questions 1–4 were addressed using a rapid review process. Twenty-four databases were searched from inception to October 2022, using a variety of databases including MEDLINE, EMBASE, Cochrane, Database of Abstracts of Reviews of Effects (DARE), Health Technology Assessment (HTA), Cumulative Index to Nursing and Allied Health Literature (CINAHL), Kleijnen Systematic Reviews Ltd (KSR) Evidence and Epistemonikos. One reviewer screened titles and abstracts of all reports identified by the searches, and a minimum of 20% were independently screened by a second reviewer. Full copies of all studies deemed potentially relevant, by either reviewer, were obtained and both reviewers independently assessed these for inclusion; any disagreements were resolved by consensus or discussion with a third reviewer. Data were extracted by one reviewer and checked by a second reviewer; any disagreements were resolved by consensus or discussion with a third reviewer. Study quality was assessed using appropriate risk-of-bias tools. Results were summarised by research question: prognostic performance; prevalence of risk categories; clinical effects; costs of CaRi-Heart. In addition to the rapid review, evidence that might be required to inform parameterisation of a future cost-effectiveness model was explored, as part of the conceptual modelling process, using a pragmatic, iterative searching approach; model parameterisation questions, other than costs, were not included in the rapid review. Results Rapid review The rapid review identified one relevant model development and validation study, which included a total of 3912 patients who were undergoing clinically indicated CTCA for the evaluation of stable coronary disease. The training/development (USA) cohort comprised 2040 patients, with a median (range) follow-up duration of 53.8 (4–105) months; a total of 85 deaths were reported during follow-up, of which 48 were cardiac. The validation (Germany) cohort comprised 1872 patients, with a median (range) follow-up duration of 72 (51–109) months; there were a total of 114 deaths during follow-up, of which 26 were confirmed cardiac deaths and 16 were deaths of unknown cause. Based on Prediction model Risk Of Bias ASsessment Tool (PROBAST), this study was rated as having high risk of bias and high concerns regarding its applicability to the decision problem specified for this EVA. Importantly, there has been no external validation of the CaRi-Heart Risk model, as the reported validation data set was used in a previous study to develop methods and thresholds for the main imaging predictors (FAI scores). With respect to applicability, the CaRi-Heart study evaluated CaRi-Heart Risk for the prediction of 8-year cardiac death; it did not consider prediction of cardiac risk, as specified in the scope for this EVA (i.e. including risk of non-fatal adverse cardiovascular events). In addition, it is unclear whether the clinical comparator model can be considered representative of standard of care in the UK NHS. The included study provided information relevant to research question 1: ‘What is the prognostic performance of CaRi-Heart, in people with stable chest pain, who are undergoing CTCA where: (1) the dependent variable is cardiac death? (2) the dependent variable is MACE?’ The hazard ratio (HR) for 8-year cardiac death, per unit increase in CaRi-Heart Risk, adjusted for ‘traditional risk factors’ (smoking, hypercholesterolaemia, hypertension, diabetes mellitus, Duke index, presence of high-risk plaque features and epicardial adipose tissue volume), was 1.05 [95% confidence interval (CI) 1.03 to 1.06] in the training/development cohort and 1.04 (95% CI 1.03 to 1.06) in the validation cohort. With respect to the subgroups of clinical interest, the predictive value of the CaRi-Heart Risk model was consistent across patients with and without obstructive CAD. In addition, the results of the included study indicated that the CaRi-Heart Risk model showed improved risk discrimination, when compared to a baseline clinical risk model, which included age, sex, hypertension, hypercholesterolaemia, diabetes mellitus and smoking (Δ c-statistic 0.149, p 10%) CaRi-Heart Risk scores, estimated from this study, was 3060/3912 (78.2%), 423/3912 (10.8%) and 429/3912 (11.0%), respectively. No studies were identified which addressed research question 3 (‘What are the clinical effects of using CaRi-Heart to assess cardiac risk?) or research question 4 (‘What are the costs, from a UK NHS and PSS perspective, using CaRi-Heart®, as an adjunctive investigation for assessment of cardiac risk, in people with stable chest pain, who are undergoing CTCA?’). Exploratory searches to inform model parameterisation Additional exploratory searches were conducted to inform the conceptual cost-effectiveness modelling, focusing on changes to treatment that might be made based on CaRi-Heart Risk and potential alternative technologies to CaRi-Heart. The results of these searches indicated that there is a deficiency with respect to evidence about the effects of changing existing treatments or introducing new treatments, based on assessment of cardiac risk (by any method), or on measures of vascular inflammation, such as perivascular FAI. However, the evidence is broadly supportive of a positive relationship between FAI and risk of adverse coronary events and hence of the future inclusion of FAI as an alternative technology (in evaluations of the CaRi-Heart device) should a method of measurement become commercially available in the UK NHS. The evidence also supports the efficacy of colchicine for secondary prevention of adverse cardiac events in unselected patients with CAD but does not provide unequivocal evidence about the mechanism by which this effect is mediated. Importantly, for the aims of this EVA, the evidence identified does not provide any indication of the efficacy of targeting colchicine treatment using CaRi-Heart Risk or separate measures of coronary inflammation, such as FAI. It should also be noted that colchicine is not currently recommended by NICE, or licensed in the UK, for this indication. Finally, the evidence suggests some uncertainty about whether and to what extent the efficacy of statins, for the secondary prevention of MACE in people with CAD, may vary with baseline risk assessed using currently available methods. In addition, there is currently no information about the effects of introducing statin treatment or changing the dose of existing statin treatment, based on CaRi-Heart Risk or on any assessment of coronary artery inflammation. Conceptual modelling A de novo conceptual decision-analytic model that could be used to inform an early assessment of the cost effectiveness of CaRi-Heart has been described. A combination of a short-term diagnostic model component and a long-term model component that evaluated the downstream consequences is anticipated to capture the diagnosis and the progression of CAD, respectively. It is expected that for the CaRi-Heart strategy, the initial diagnostic groups based on CTCA only would be, in turn, further split by the CaRi-Heart information into groups of low, medium or high CaRi-Heart Risk. If other competing alternatives are identified, those could be added to the model, if there is sufficient available evidence. Conclusions The rapid review methods and pragmatic approach to additional exploratory searches used to inform this EVA mean that, although areas of potential uncertainty have been described, our findings cannot be used to definitively state where there are evidence gaps. The evidence about the clinical utility of CaRi-Heart Risk is, as yet, sparse and is subject to considerable limitations, both in terms of risk of bias and applicability to UK clinical practice. There is some evidence to indicate that CaRi-Heart Risk may be predictive of an individual patient’s 8-year risk of cardiac death, for patients undergoing CTCA for suspected CAD. However, whether and to what extent CaRi-Heart represents an improvement relative to current standard of care remains unclear. Currently available data are insufficient to fully inform cost-effectiveness modelling. A large (n = 15,000) ongoing study, NCT05169333, the Oxford risk factors and non-invasive imaging (ORFAN) study, with an estimated completion date of February 2030, may address some of the uncertainties identified in this EVA. Suggested research priorities External validation of the CaRi-Heart Risk model should be considered a high priority. The external validation process could also be used to address some of the applicability concerns, for example, the ability of CaRi-Heart Risk to predict non-fatal adverse cardiovascular events (in addition to cardiac death) could be considered. It remains unclear whether and to what extent CaRi-Heart represents an improvement relative to current standard of care; this is largely because the definition of standard of care and hence the applicability of the comparator used in the CaRi-Heart study are uncertain. If a consensus could be reached, among clinical experts, as to what should constitute standard of care, then the prognostic performance of CaRi-Heart Risk could be compared to this standard. There is currently a lack of information about the costs, from a UK NHS and PSS perspective, using CaRi-Heart, as an adjunctive investigation for assessment of cardiac risk, in people with stable chest pain, who are undergoing CTCA. The company have indicated that the ORFAN study will collect data on costs to the NHS of adding CaRi-Heart to CTCA, including cardiologists’ time in training to interpret CaRi-Heart analyses, and the implementation costs per CTCA (if any) added to the price of a CaRi-Heart analysis to estimate the total cost per CTCA of introducing CaRi-Heart into the NHS. There is also a lack of information about the clinical effects of any changes to treatment/management that may be made as a result of adding assessment using CaRi-Heart Risk to current standard of care, in patients undergoing CTCA for suspected CAD. The optimum study design would be a randomised controlled trial (RCT) or cluster RCT, where patients or study centres are randomised to receive CTCA with or without the addition of CaRi-Heart Risk assessment, and information about changes to treatment/management and long-term clinical effects is collected. Observational study designs, including ‘before and after’ implementation studies or using matching techniques to provide a control, could provide an alternative approach. Acknowledging that the collection of data about the long-term clinical effects of using CaRi-Heart Risk will take a number of years, alternative, pragmatic approaches to populating this component of a full cost-effectiveness model may be considered useful. Such approaches could include estimation of the potential effects of treatment changes based on risk-stratified effects of treatment (e.g. statins), where risk stratification has been based on methods other than CaRi-Heart Risk and/or estimation of the potential effects of introducing new treatments (e.g. colchicine) where the ‘target condition’ (coronary inflammation) has been assessed by methods other than CaRi-Heart Risk (e.g. FAI alone). Study registration This study is registered as PROSPERO CRD42022366496. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135672) and is published in full in Health Technology Assessment; Vol. 28, No. 31. See the NIHR Funding and Awards website for further award information.

Background Most neovascular age-related macular degeneration treatments involve long-term follow-up of disease activity. Home monitoring would reduce the burden on patients and those they depend on for transport, and release clinic appointments for other patients. The study aimed to evaluate three home-monitoring tests for patients to use to detect active neovascular age-related macular degeneration compared with diagnosing active neovascular age-related macular degeneration by hospital follow-up. Objectives There were five objectives: Estimate the accuracy of three home-monitoring tests to detect active neovascular age-related macular degeneration. Determine the acceptability of home monitoring to patients and carers and adherence to home monitoring. Explore whether inequalities exist in recruitment, participants’ ability to self-test and their adherence to weekly testing during follow-up. Provide pilot data about the accuracy of home monitoring to detect conversion to neovascular age-related macular degeneration in fellow eyes of patients with unilateral neovascular age-related macular degeneration. Describe challenges experienced when implementing home-monitoring tests. Design Diagnostic test accuracy cohort study, stratified by time since starting treatment. Setting Six United Kingdom Hospital Eye Service macular clinics (Belfast, Liverpool, Moorfields, James Paget, Southampton, Gloucester). Participants Patients with at least one study eye being monitored by hospital follow-up. Reference standard Detection of active neovascular age-related macular degeneration by an ophthalmologist at hospital follow-up. Index tests KeepSight Journal: paper-based near-vision tests presented as word puzzles. MyVisionTrack®: electronic test, viewed on a tablet device. MultiBit: electronic test, viewed on a tablet device. Participants provided test scores weekly. Raw scores between hospital follow-ups were summarised as averages. Results Two hundred and ninety-seven patients (mean age 74.9 years) took part. At least one hospital follow-up was available for 317 study eyes, including 9 second eyes that became eligible during follow-up, in 261 participants (1549 complete visits). Median testing frequency was three times/month. Estimated areas under receiver operating curves were < 0.6 for all index tests, and only KeepSight Journal summary score was significantly associated with the lesion activity (odds ratio = 3.48, 95% confidence interval 1.09 to 11.13, p = 0.036). Older age and worse deprivation for home address were associated with lower participation (χ2 = 50.5 and 24.3, respectively, p 6 and < 42 months earlier. We tried to recruit equal numbers by time since starting treatment in the first-treated study eye (6–17 months; 18–29 months; 30–41 months) to ensure test performance was estimated across this range of duration of nAMD. Patients were followed for at least 6 months. Reference standard The reference standard was the reviewing ophthalmologist’s decision about the activity status of a study eye at a HES monitoring visit, recorded as active, inactive or uncertain. There were no additional hospital visits for the study. Such decisions are usually based on clinical examination and the results of hospital-based retinal imaging investigations, for example, colour fundus photographs and ocular coherence tomograms (OCTs). The reference standard grouped uncertain with inactive judgements for analyses. Index tests Three home-monitoring (‘index’) tests were evaluated, spanning low to moderate cost and complexity. These were: KeepSight Journal (KSJ): a paper-based booklet of near-vision tests presented as word puzzles developed in the United States and adapted by the study team for use in the study. MyVisionTrack® (mVT®): electronic vision test, intended to be viewed on a tablet device. MultiBit (MBT): electronic vision test, intended to be viewed on a tablet device. Specific thresholds indicating a significant clinical change were not provided for any of index tests in advance by their developers. Outcomes The primary outcome was classification of a study eye at a monitoring visit as having active or inactive disease (active, inactive, uncertain), that is, the reviewing ophthalmologist’s decision. A secondary outcome (new reference standard) for Objective A was a change from inactive to active status from one management visit to the next. This was considered better to represent how home monitoring might be implemented. For Objective C, outcomes investigated were willingness in principle to participate; ability to carry out index tests; adherence to weekly testing. For Objective D, the outcome was conversion of a fellow eye to active nAMD as judged by an ophthalmologist, that is, same reference standard. For Objective E, the following outcomes were described as measures of the technical and logistical challenges identified during the study: Frequency and reason for incoming calls made to the helpline and outgoing calls made to participants. Frequency and duration of events leading to the digital tests being unavailable for testing. Other technical and logistical challenges. Objective B study recruitment and data collection Recruitment to the qualitative component began 3 months after the monitoring for neovascular age-related macular degeneration reactivation at home (MONARCH) study began recruiting. During the consent process for participation in the study, patients could consent to further contact from the qualitative research team to discuss participation in the qualitative study. Maximum variation sampling ensured a range of perspectives were captured in relation to: age (young-old 50–69 years and older-old > 70 years), gender, one or both eyes with nAMD, time since first treatment (defined as above) and adherence to home monitoring (test data from the two electronic tests were used to categorise participants into ‘regular’ testers and ‘irregular’ testers). Patients who declined to participate in MONARCH but provided consent to be contacted about the qualitative study, informal ‘carers’, supporters or significant others in the lives of patients and healthcare professionals who interacted with participants at study sites visits were also approached to gather their perspectives about the acceptability of home monitoring. Statistical analysis Objective A: The test accuracy of index tests was estimated by fitting a logistic regression model to predict the reference standard from summary test scores for the interval between monitoring visits, adjusting for participants’ baseline data. Accuracy was estimated for the primary outcome using all index test data, data only for the 4 weeks preceding the monitoring visit, the reference standard based on reading centre decisions made from OCT images and for the secondary outcome. Test scores were summarised as: means (MBT and mVT); median (KSJ reported near visual acuity (VA), ordinal six-point scale); proportions (KSJ reported VA, Amsler grid and household object appearance reported worse than baseline vs. same or better). All four scores were fitted in the KSJ model and a single area under the receiving operator curve (AUROC) was estimated. Separate models were fitted for each test for the primary outcome, the two sensitivity analyses and the secondary outcome. Model performance was quantified by the odds ratio (OR) for the index test summary score(s) and the estimate of the AUROC and their respective confidence intervals (CIs). AUROCs were based on predicted probabilities calculated using only the fixed effects in the models. Sensitivity, specificity, positive and negative predictive values and 95% CIs were calculated using cut-off thresholds corresponding to Youden’s index for each model, which minimises overall misclassifications. Average test scores above and below the thresholds were also calculated. Analyses took account of the structure within the data, that is, the nesting of visits and eyes within patients. Objective B: All interviews were audio-recorded and transcribed. A directed content analysis approach based on deductive and inductive coding was used. NVivo version 12 was used to manage data and facilitate the analysis process, which, in summary, included the following stages: (1) independent transcription, (2) data familiarisation, (3) independent coding, (4) development of an analytical framework, (5) indexing, (6) charting and (7) interpreting data. Objective C: Willingness in principle to participate was defined as an approached eligible patient agreeing to attend a research visit for training. Ability to perform an index test was defined as the proportion of monitoring visits for which some valid index test data were available. Adherence was defined as the proportion of weeks between monitoring visits for which some valid data for an index test were available. The ability and adherence models were performed for each test separately at the level of the patient. Regression models estimated associations of age, sex, Index of Multiple Deprivation (IMD), stratum of time since first diagnosis and baseline visual acuity at diagnosis on the outcomes of willingness to participate, ability to perform tests and adherence to weekly testing; models for the latter two outcomes were fitted for each index test. Associations were reported with 95% CIs. Analyses of adherence and ability took account of nesting of visits within participants. Objective D: The test accuracy of the index tests for the reference standard of an ophthalmologist’s classification of a fellow eye as having active disease at a monitoring visit, that is, conversion to active nAMD, was estimated by the same methods as described for Objective A. Two sensitivity analyses were carried out: (1) the same reference standard but using test data only for the 4 weeks preceding the management visit and (2) the alternative reference standard of classification of a fellow eye having active disease based on reading centre grading of OCTs carried out during the monitoring visits. Objective E: This objective used descriptive summary descriptive statistics only. Results The study recruited 297 patients (consented participants) between 21 August 2018 and 31 March 2020. Half of recruited participants were first treated for nAMD 6–17 months before consenting, 28% 18–29 months before consenting and 22% 30–41 months before consenting. At the end of the study, data for at least one monitoring visit after starting to use the index monitoring tests were available for 357 study eyes in 297 patients. Data were available for at least one complete monitoring visit after starting to use the index monitoring tests for 317 study eyes of 261 patients. More participants were women (58.6%). Participants’ mean age was 74.9 (6.6) years [standard deviation (SD)]. The mean visual acuity of study eyes (better seeing eyes if participants had two study eyes) was 0.2 (0.2) Logarithm of the minimum angle of resolution (LogMAR) (SD). Objective A: Median testing frequency was three times per month. In the primary analysis, estimated AUROCs were < 0.6 for all index tests, and only KSJ summary score was significantly associated with the lesion activity (OR = 3.48, 95% CI 1.09 to 11.13; p = 0.036). Estimated AUROCs were < 0.6 for all tests in both sensitivity analyses and for the secondary outcome of change from inactive to active status between adjacent management visits. Objective B: Two overarching meta-themes emerged from the qualitative studies related to acceptance or non-acceptance of home monitoring. Meta-theme 1 encompassed four main themes: (1) the role of home monitoring; (2) suitability of procedures and instruments; (3) experience of home monitoring; and (4) feasibility of home monitoring in usual practice. Meta-theme 2 consisted of one main theme covering key inhibitors to acceptability. The main factors influencing acceptability included a participant’s understanding about the purpose of home monitoring and their experience of using it. While home monitoring was generally seen as a relatively straightforward exercise to undertake and non-burdensome, training and ongoing support were regarded as essential to its success. Home monitoring was acceptable to patients and its potential to reduce clinic visits during non-active treatment phases was recognised. Objective C: A minority of patients who were approached were willing in principle to participate. Increasing age and worse deprivation index for home address were associated with being unwilling in principle to participate (χ2 = 50.5 and 24.3, respectively, both p ≤ 0.001). Recruiting site was also associated with willingness in principle to participate, believed to be due to sites adopting different strategies for approaching and recruiting patients. Increasing age and worse deprivation were not consistently associated with either being able to self-monitor with the index tests or adherence to weekly testing (χ2 for all tests 0.08 for ability and adherence, except for worse IMD being associated with better adherence for the KSJ, χ2 = 12.15, p = 0.016). Recruiting site was also associated with being able to test and adhering to weekly testing. Objective D: There were 132 fellow eyes with data from 544 monitoring visits, 17 of which (12.9%) had nAMD recorded at one or more management visits over about 100 participant-years. This rate of conversion was higher than expected based on epidemiological studies of conversion rates in unaffected fellow eyes, potentially due to study eyes having had nAMD longer ago. Some predictors could not be fitted in models and estimates of associations were imprecise. The no-test model predicted conversion better than for Objective A (AUROC = 0.73) and electronic tests did not increase this (AUROCs = 0.73 and 0.76 for MBT and mVT, respectively). The estimated AUROC for the KSJ was 0.85, due to a strong positive association of the household object summary score with conversion (OR 15.3, p = 0.036). Objective E: Despite two-thirds of the population having previously used a smartphone, there were still a variety of challenges experienced with the electronic devices while testing at home that contributed to both reduced adherence and ultimately withdrawals from the study. Strengths and limitations The study had several strengths. Estimates of the diagnostic test accuracy of index tests were at low risk of bias: the study population was appropriate for the intended use of the tests, and summary test scores were not available to ophthalmologists providing the reference standard, which was judged after the index test data were collected. Limitations A smaller-than-planned sample size (less than half the target number of monitoring visits); nonetheless, 95% CIs for AUROCs were narrow (± 0.04) and estimates were able to rule out tests providing adequate accuracy for diagnosing active nAMD to enable patients to be monitored without hospital review. Tests were sometimes not available for technical reasons that were beyond the control of the research team. The study had no control over monitoring visits and participants are likely to have reported their subjective visual experience to their consultants, which might have influenced the reference standard. We could not define test thresholds a priori, and instead estimated AUROCs. We did not compare AUROCs for tests due to their poor accuracy. The ways in which patients were approached and screened varied across sites, generating a site effect in analyses of potential inequalities; variations may have reflected the pre-conceptions of research staff regarding the capabilities of patients to use the electronic tests. Conclusions Based on the detection of lesion activity assessed by clinicians in the clinic, we have shown that none of the index tests provides acceptable test accuracy for home monitoring in this context. Associations of increasing age and deprivation index for home address with unwillingness in principle to participate despite provision of hardware highlight the potential for inequality with interventions of the kind evaluated. While a proportion of nAMD patients are willing and interested in the potential for home monitoring, substantial practical and technological issues are encountered in the implementation of such, requiring a significant support infrastructure, including a study helpline. Future work Future research should focus on the methodological challenge of efficiently evaluating mobile health technologies which deal with constantly emerging new technology. The clear evidence of inequalities in participation and retention should prompt future research on ways to encourage participant and adoption of mobile health technologies by underserved populations. Focus should also be placed on methods to improve adherence and retention in longitudinal studies involving electronic testing, particularly around the nature of feedback to participants. Trial registration This trial is registered as ISRCTN79058224. Funding This award was funded by the National Institute of Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 15/97/02) and is published in full in Health Technology Assessment; Vol. 28, No. 32. See the NIHR Funding and Awards website for further award information.