Search

Aim: To describe the protocol for a multi-centre randomised controlled trial to determine whether treatment protocols monitoring daily CRP (C-reactive protein) or PCT (procalcitonin) safely allow a reduction in duration of antibiotic therapy in hospitalised adult patients with sepsis. Design: Multicentre three-arm randomised controlled trial. Setting: UK NHS hospitals. Target population: Hospitalised critically ill adults who have been commenced on intravenous antibiotics for sepsis. Health technology: Three protocols for guiding antibiotic discontinuation will be compared: (a) standard care; (b) standard care + daily CRP monitoring; (c) standard care + daily PCT monitoring. Standard care will be based on routine sepsis management and antibiotic stewardship. Measurement of outcomes and costs. Outcomes will be assessed to 28 days. The primary outcomes are total duration of antibiotics and safety outcome of all-cause mortality. Secondary outcomes include: escalation of care/re-admission; infection re-lapse/recurrence; antibiotic dose; length and level of critical care stay and length of hospital stay. Ninety-day all-cause mortality rates will also be collected. An assessment of cost effectiveness will be performed. Conclusion: In the setting of routine NHS care, if this trial finds that a treatment protocol based on monitoring CRP or PCT safely allows a reduction in duration of antibiotic therapy, and is cost effective, then this has the potential to change clinical practice for critically ill patients with sepsis. Moreover, if a biomarker-guided protocol is not found to be effective, then it will be important to avoid its use in sepsis and prevent ineffective technology becoming widely adopted in clinical practice.

Purpose:\ud To compare the effectiveness of the intravenous (IV) and intraosseous (IO) routes for drug administration in adults with a cardiac arrest enrolled in the Pre-Hospital Assessment of the Role of Adrenaline: Measuring the Effectiveness of Drug Administration in Cardiac Arrest (PARAMEDIC2) randomised, controlled trial.\ud \ud Methods:\ud Patients were recruited from five National Health Service Ambulance Services in England and Wales from December 2014 through October 2017. Patients with an out-of-hospital cardiac arrest who were unresponsive to initial resuscitation attempts were randomly assigned to 1 mg adrenaline or matching placebo. Intravascular access was established as soon as possible, and IO access was considered if IV access was not possible after two attempts.\ud \ud Results:\ud Among patients with out-of-hospital cardiac arrest, 3631 received adrenaline and 3686 received placebo. Amongst these, 1116 (30.1%) and 1121 (30.4%) received the study drug via the IO route. The odds ratios were similar in the IV and IO groups for return of spontaneous circulation (ROSC) at hospital handover [adjusted odds ratio (aOR) 4.07 (95% CI 3.42–4.85) and (aOR 3.98 (95% CI 2.86–5.53), P value for interaction 0.90]; survival to 30 days [aOR 1.67 (1.18–2.35) versus 0.9 (0.4–2.05), P = 0.18]; and favourable neurological outcome [aOR 1.39 (0.93–2.06) versus 0.62 (0.23–1.67), P = 0.14].\ud \ud Conclusion:\ud There was no significant difference in treatment effect (adrenaline versus placebo) on ROSC at hospital handover between drugs administered by the intraosseous route or by the intravenous route. We could not detect any difference in the treatment effect between the IV and IO routes on the longer term outcomes of 30-day survival or favourable neurological outcome at discharge (ISRCTN73485024).

Background There is a growing interest in the use of Bayesian adaptive designs in late-phase clinical trials. This includes the use of stopping rules based on Bayesian analyses in which the frequentist type I error rate is controlled as in frequentist group-sequential designs. Methods This paper presents a practical comparison of Bayesian and frequentist group-sequential tests. Focussing on the setting in which data can be summarised by normally distributed test statistics, we evaluate and compare boundary values and operating characteristics. Results Although Bayesian and frequentist group-sequential approaches are based on fundamentally different paradigms, in a single arm trial or two-arm comparative trial with a prior distribution specified for the treatment difference, Bayesian and frequentist group-sequential tests can have identical stopping rules if particular critical values with which the posterior probability is compared or particular spending function values are chosen. If the Bayesian critical values at different looks are restricted to be equal, O’Brien and Fleming’s design corresponds to a Bayesian design with an exceptionally informative negative prior, Pocock’s design to a Bayesian design with a non-informative prior and frequentist designs with a linear alpha spending function are very similar to Bayesian designs with slightly informative priors.This contrasts with the setting of a comparative trial with independent prior distributions specified for treatment effects in different groups. In this case Bayesian and frequentist group-sequential tests cannot have the same stopping rule as the Bayesian stopping rule depends on the observed means in the two groups and not just on their difference. In this setting the Bayesian test can only be guaranteed to control the type I error for a specified range of values of the control group treatment effect. Conclusions Comparison of frequentist and Bayesian designs can encourage careful thought about design parameters and help to ensure appropriate design choices are made.

Background: Dysregulated inflammation is associated with poor outcomes in COVID-19. We aimed to assess the efficacy of namilumab (a granulocyte-macrophage colony stimulating factor inhibitor) and infliximab (a tumour necrosis factor inhibitor) in hospitalised patients with COVID-19, to prioritise agents for phase 3 trials. Methods: In this randomised, multicentre, multi-arm, multistage, parallel-group, open-label, adaptive, phase 2, proof-of-concept trial (CATALYST), we recruited patients (aged ≥16 years) admitted to hospital with COVID-19 pneumonia and C-reactive protein (CRP) concentrations of 40 mg/L or greater, at nine hospitals in the UK. Participants were randomly assigned with equal probability to usual care or usual care plus a single intravenous dose of namilumab (150 mg) or infliximab (5 mg/kg). Randomisation was stratified by care location within the hospital (ward vs intensive care unit [ICU]). Patients and investigators were not masked to treatment allocation. The primary endpoint was improvement in inflammation, measured by CRP concentration over time, analysed using Bayesian multilevel models. This trial is now complete and is registered with ISRCTN, 40580903. Findings: Between June 15, 2020, and Feb 18, 2021, we screened 299 patients and 146 were enrolled and randomly assigned to usual care (n=54), namilumab (n=57), or infliximab (n=35). For the primary outcome, 45 patients in the usual care group were compared with 52 in the namilumab group, and 29 in the usual care group were compared with 28 in the infliximab group. The probabilities that the interventions were superior to usual care alone in reducing CRP concentration over time were 97% for namilumab and 15% for infliximab; the point estimates for treatment–time interactions were –0·09 (95% CI –0·19 to 0·00) for namilumab and 0·06 (–0·05 to 0·17) for infliximab. 134 adverse events occurred in 30 (55%) of 55 patients in the namilumab group compared with 145 in 29 (54%) of 54 in the usual care group. 102 adverse events occurred in 20 (69%) of 29 patients in the infliximab group compared with 112 in 17 (50%) of 34 in the usual care group. Death occurred in six (11%) patients in the namilumab group compared with ten (19%) in the usual care group, and in four (14%) in the infliximab group compared with five (15%) in the usual care group. Interpretation: Namilumab, but not infliximab, showed proof-of-concept evidence for reduction in inflammation—as measured by CRP concentration—in hospitalised patients with COVID-19 pneumonia. Namilumab should be prioritised for further investigation in COVID-19.

Introduction Severe SARS-CoV-2 infection is associated with a dysregulated immune response. Inflammatory monocytes and macrophages are crucial, promoting injurious, proinflammatory sequelae. Immunomodulation is, therefore, an attractive therapeutic strategy and we sought to test licensed and novel candidate drugs. Methods and analysis The CATALYST trial is a multiarm, open-label, multicentre, phase II platform trial designed to identify candidate novel treatments to improve outcomes of patients hospitalised with COVID-19 compared with usual care. Treatments with evidence of biomarker improvements will be put forward for larger-scale testing by current national phase III platform trials. Hospitalised patients >16 years with a clinical picture strongly suggestive of SARS-CoV-2 pneumonia (confirmed by chest X-ray or CT scan, with or without a positive reverse transcription PCR assay) and a C reactive protein (CRP) ≥40 mg/L are eligible. The primary outcome measure is CRP, measured serially from admission to day 14, hospital discharge or death. Secondary outcomes include the WHO Clinical Progression Improvement Scale as a principal efficacy assessment. Ethics and dissemination The protocol was approved by the East Midlands-Nottingham 2 Research Ethics Committee (20/EM/0115) and given urgent public health status; initial approval was received on 5 May 2020, current protocol version (V.6.0) approval on 12 October 2020. The MHRA also approved all protocol versions. The results of this trial will be disseminated through national and international presentations and peer-reviewed publications. Trial registration numbers EudraCT2020-001684-89,NCT40580903.

SummaryBackgroundDysregulated inflammation is associated with poor outcomes in Coronavirus disease 2019 (COVID-19). We assessed the efficacy of namilumab, a granulocyte-macrophage colony-stimulating factor inhibitor and infliximab, a tumour necrosis factor inhibitor in hospitalised patients with COVID-19 in order to prioritise agents for phase 3 trials.MethodsIn this randomised, multi-arm, parallel group, open label, adaptive phase 2 proof-of-concept trial (CATALYST) we recruited hospitalised patients ≥ 16 years with COVID-19 pneumonia and C-reactive protein (CRP) ≥ 40mg/L in nine UK hospitals. Participants were randomly allocated with equal probability to usual care, or usual care plus a single 150mg intravenous dose of namilumab (150mg) or infliximab (5mg/kg). Randomisation was stratified for ward versus ICU. The primary endpoint was improvement in inflammation in intervention arms compared to control as measured by CRP over time, analysed using Bayesian multi-level models. ISRCTN registry number 40580903.FindingsBetween 15thJune 2020 and 18thFebruary 2021 we randomised 146 participants: 54 to usual care, 57 to namilumab and 35 to infliximab. The probabilities that namilumab and infliximab were superior to usual care in reducing CRP over time were 97% and 15% respectively. Consistent effects were seen in ward and ICU patients and aligned with clinical outcomes, such that the probability of discharge (WHO levels 1-3) at day 28 was 47% and 64% for ICU and ward patients on usual care, versus 66% and 77% for patients treated with namilumab. 134 adverse events occurred in 30/55 (54.5%) namilumab patients compared to 145 in 29/54 (53.7%) usual care patients. 102 events occurred in 20/29 (69.0%) infliximab patients versus 112 events in 17/34 (50.0%) usual care patients.InterpretationNamilumab, but not infliximab, demonstrated proof-of-concept evidence for reduction in inflammation in hospitalised patients with COVID-19 pneumonia which was consistent with secondary clinical outcomes. Namilumab should be prioritised for further investigation in COVID-19.FundingMedical Research Council.

Background Adrenaline has been used as a treatment for cardiac arrest for many years, despite uncertainty about its effects on long-term outcomes and concerns that it may cause worse neurological outcomes. Objectives The objectives were to evaluate the effects of adrenaline on survival and neurological outcomes, and to assess the cost-effectiveness of adrenaline use. Design This was a pragmatic, randomised, allocation-concealed, placebo-controlled, parallel-group superiority trial and economic evaluation. Costs are expressed in Great British pounds and reported in 2016/17 prices. Setting This trial was set in five NHS ambulance services in England and Wales. Participants Adults treated for an out-of-hospital cardiac arrest were included. Patients were ineligible if they were pregnant, if they were aged Interventions Participants were randomised to either adrenaline (1 mg) or placebo in a 1 : 1 allocation ratio by the opening of allocation-concealed treatment packs. Main outcome measures The primary outcome was survival to 30 days. The secondary outcomes were survival to hospital admission, survival to hospital discharge, survival at 3, 6 and 12 months, neurological outcomes and health-related quality of life through to 6 months. The economic evaluation assessed the incremental cost per quality-adjusted life-year gained from the perspective of the NHS and Personal Social Services. Participants, clinical teams and those assessing patient outcomes were masked to the treatment allocation. Results From December 2014 to October 2017, 8014 participants were assigned to the adrenaline (n = 4015) or to the placebo (n = 3999) arm. At 30 days, 130 out of 4012 participants (3.2%) in the adrenaline arm and 94 out of 3995 (2.4%) in the placebo arm were alive (adjusted odds ratio for survival 1.47, 95% confidence interval 1.09 to 1.97). For secondary outcomes, survival to hospital admission was higher for those receiving adrenaline than for those receiving placebo (23.6% vs. 8.0%; adjusted odds ratio 3.83, 95% confidence interval 3.30 to 4.43). The rate of favourable neurological outcome at hospital discharge was not significantly different between the arms (2.2% vs. 1.9%; adjusted odds ratio 1.19, 95% confidence interval 0.85 to 1.68). The pattern of improved survival but no significant improvement in neurological outcomes continued through to 6 months. By 12 months, survival in the adrenaline arm was 2.7%, compared with 2.0% in the placebo arm (adjusted odds ratio 1.38, 95% confidence interval 1.00 to 1.92). An adjusted subgroup analysis did not identify significant interactions. The incremental cost-effectiveness ratio for adrenaline was estimated at £1,693,003 per quality-adjusted life-year gained over the first 6 months after the cardiac arrest event and £81,070 per quality-adjusted life-year gained over the lifetime of survivors. Additional economic analyses estimated incremental cost-effectiveness ratios for adrenaline at £982,880 per percentage point increase in overall survival and £377,232 per percentage point increase in neurological outcomes over the first 6 months after the cardiac arrest. Limitations The estimate for survival with a favourable neurological outcome is imprecise because of the small numbers of patients surviving with a good outcome. Conclusions Adrenaline improved long-term survival, but there was no evidence that it significantly improved neurological outcomes. The incremental cost-effectiveness ratio per quality-adjusted life-year exceeds the threshold of £20,000–30,000 per quality-adjusted life-year usually supported by the NHS. Future work Further research is required to better understand patients’ preferences in relation to survival and neurological outcomes after out-of-hospital cardiac arrest and to aid interpretation of the trial findings from a patient and public perspective. Trial registration Current Controlled Trials ISRCTN73485024 and EudraCT 2014-000792-11. Funding This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 25, No. 25. See the NIHR Journals Library website for further project information.

IntroductionSevere SARS-CoV-2 infection is associated with a dysregulated immune response. Inflammatory monocytes and macrophages are crucial, promoting injurious, pro-inflammatory sequelae. Immunomodulation is, therefore, an attractive therapeutic strategy and we sought to test licensed and novel candidate drugs.Methods and analysisThe CATALYST trial is a multi-arm, open-label, multi-centre, phase II platform trial designed to identify candidate novel treatments to improve outcomes of patients hospitalised with COVID-19 compared with usual care. Treatments with evidence of biomarker improvements will be put forward for larger-scale testing by current national phase III platform trials. Hospitalised patients >16 years with a clinical picture strongly suggestive of SARS-CoV-2 pneumonia (confirmed by chest X-ray or CT scan, with or without a positive reverse transcription polymerase chain reaction (RT-PCR) assay) and a C-Reactive Protein (CRP) ≥40 mg/L are eligible. The primary outcome measure is CRP, measured serially from admission to day 14, hospital discharge or death. Secondary outcomes include the WHO Clinical Progression Improvement Scale as a principal efficacy assessment.Ethics and disseminationThe protocol was approved by the East Midlands - Nottingham 2 Research Ethics Committee (20/EM/0115) and given Urgent Public Health status; initial approval was received on 05-May-2020, current protocol version (v6.0) approval on 12-Oct-2020. The MHRA also approved all protocol versions. The results of this trial will be disseminated through national and international presentations and peer-reviewed publications.Trial registration numberEudraCT Number: 2020-001684-89ISRCTN Number: 40580903Strengths and limitations of this trialCATALYST will provide a rapid readout on the safety and proof-of-concept of candidate novel treatmentsCATALYST will enable phase III trial resources to be focussed and allocated for agents with a high likelihood of successCATALYST uses Bayesian multi-level models to allow for nesting of repeated measures data, with factors for each individual patient and treatment arm, and allowing for non-linear responsesCATALYST is not designed to provide a definitive signal on clinical outcomes

Aim\ud To assess and evaluate patient safety incidents and in particular, medication errors, during a large multi-centre pre-hospital trial of emergency therapy (PARAMEDIC2), in order to inform and improve future pre-hospital medicines trials.\ud Methods\ud The PARAMEDIC2 trial was undertaken across five NHS Ambulance Services in England and Wales with randomisation between December 2014 and October 2017. Patients with an out -of-hospital cardiac arrest unresponsive to initial resuscitation were randomly assigned to 1 mg intravenous adrenaline or matching placebo. Records were reviewed to identify trial medication errors involving documentation and/or clinical protocol errors occurring in trial participants. Causes of medication errors, including root cause analysis where available, were reviewed to identify patterns and themes contributing to these errors.\ud Results\ud 8,016 patients were enrolled, of whom 4902 received trial medication. 'A total of 331 patient safety incidents was reported, involving 295 patients, representing an overall rate of 3.6% Of these, 166 (50.2%) were documentation errors while 165 (49.8%) were clinical protocol/medication errors. An overall rate of 0-4.5% was reported across all five ambulance services, with a mean of 2.0%. These errors had no impact on patient care or the trial and were all resolved\ud Conclusion\ud The overall medication error rate of 1.8 % primarily consisted of administration of open-label adrenaline and confusion with trial medication packs. A similar number of patients had documentation errors. This study is the first to provide data on patient safety incidents relating to medication errors encountered during a pre-hospital trial of emergency medication administration and will provide supporting data for planning future trials in this area.

Background The ‘Prehospital Assessment of the Role of Adrenaline: Measuring the Effectiveness of Drug Administration In Cardiac Arrest’ (PARAMEDIC2) trial showed that adrenaline improves overall survival, but not neurological outcomes. We sought to determine the within-trial and lifetime health and social care costs and benefits associated with adrenaline, including secondary benefits from organ donation. Methods We estimated the costs, benefits (quality-adjusted life years (QALYs)) and incremental cost-effectiveness ratios (ICERs) associated with adrenaline during the 6-month trial follow-up. Model-based analyses explored how results altered when the time horizon was extended beyond 6 months and the scope extended to include recipients of donated organs. Results The within-trial (6 months) and lifetime horizon economic evaluations focussed on the trial population produced ICERs of £1,693,003 (€1,946,953) and £81,070 (€93,231) per QALY gained in 2017 prices, respectively, reflecting significantly higher mean costs and only marginally higher mean QALYs in the adrenaline group. The probability that adrenaline is cost-effective was less than 1% across a range of cost-effectiveness thresholds. Combined direct economic effects over the lifetimes of survivors and indirect economic effects in organ recipients produced an ICER of £16,086 (€18,499) per QALY gained for adrenaline with the probability that adrenaline is cost-effective increasing to 90% at a £30,000 (€34,500) per QALY cost-effectiveness threshold. Conclusions Adrenaline was not cost-effective when only directly related costs and consequences are considered. However, incorporating the indirect economic effects associated with transplanted organs substantially alters cost-effectiveness, suggesting decision-makers should consider the complexity of direct and indirect economic impacts of adrenaline. Trial registration ISRCTN73485024. Registered on 13 March 2014.

Reproductive medicine is imbued with debates over the results of key trials. This has resulted in heterogeneity in clinical practice and a disconnect between researchers and the patient group they aim to treat. The criticisms of trials originate from the nature of reproductive health conditions and limitations imposed in designing trials to assess effect in a patient group with heterogenous pathologies leading to the same condition. This leads to challenges in balancing the difficulties of recruiting an enriched patient cohort versus the dilutionary effect and need for subgroup analysis from wider recruitment. These challenges manifest as a failure to achieve traditional statistical significance. One potential solution to overcoming these inherent challenges is that of a Bayesian statistical approach. Using examples from the literature we demonstrate the benefits of a Bayesian approach. Taking published data and using a flat prior (no background information used), a Bayesian re-analysis of the PRISM and EAGeR trials is presented. This demonstrated a 94.7% chance of progesterone and a 95.3% probability of aspirin preventing miscarriage, in contrast to the original trial conclusions. These highlight the role a Bayesian approach can play in overcoming the challenges of trials within reproductive health.

Aims We recently reported early outcomes in patients enrolled in a randomised trial of adrenaline in out-of-hospital cardiac arrest: the PARAMEDIC2 (Prehospital Assessment of the Role of Adrenaline: Measuring the Effectiveness of Drug Administration in Cardiac Arrest) trial. The purpose of the present paper is to report long-term survival, quality of life, functional and cognitive outcomes at 3, 6 and 12-months. Methods PARAMEDIC2 was a pragmatic, individually randomised, double blind, controlled trial with an economic evaluation. Patients were randomised to either adrenaline or placebo. This paper reports results on the modified Rankin Scale scores at 6-months, survival at 6 and 12-months, as well as other cognitive, functional and quality of life outcomes collected at 3 and 6 months (Two Simple Questions, the Mini Mental State Examination, the Informant Questionnaire on Cognitive Decline Evaluation for Cardiac Arrest, Hospital Anxiety and Depression Scale, the Post Traumatic Stress Disorder Checklist - Civilian Version, Short-Form 12-item Health Survey and the EuroQoL EQ-5D-5L). Results 8014 patients were randomised with confirmed trial drug administration. At 6-months, 78 (2.0%) of the patients in the adrenaline group and 58 (1.5%) of patients in the placebo group had a favourable neurological outcome (adjusted odds ratio 1.35 [95% confidence interval: 0.93, 1.97]). 117 (2.9%) patients were alive at 6-months in the adrenaline group compared with 86 (2.2%) in the placebo group (1.43 [1.05, 1.96], reducing to 107 (2.7%) and 80 (2.0%) respectively at 12-months (1.38 [1.00, 1.92]). Measures of 3 and 6-month cognitive, functional and quality of life outcomes were reduced, but there was no strong evidence of differences between groups. Conclusion Adrenaline improved survival through to 12-months follow-up. The study did not find evidence of improvements in favourable neurological outcomes. (ISCRTN 73485024)

Background Bayesian adaptive methods are increasingly being used to design clinical trials and offer several advantages over traditional approaches. Decisions at analysis points are usually based on the posterior distribution of the treatment effect. However, there is some confusion as to whether control of type I error is required for Bayesian designs as this is a frequentist concept. Methods We discuss the arguments for and against adjusting for multiplicities in Bayesian trials with interim analyses. With two case studies we illustrate the effect of including interim analyses on type I/II error rates in Bayesian clinical trials where no adjustments for multiplicities are made. We propose several approaches to control type I error, and also alternative methods for decision-making in Bayesian clinical trials. Results In both case studies we demonstrated that the type I error was inflated in the Bayesian adaptive designs through incorporation of interim analyses that allowed early stopping for efficacy and without adjustments to account for multiplicity. Incorporation of early stopping for efficacy also increased the power in some instances. An increase in the number of interim analyses that only allowed early stopping for futility decreased the type I error, but also decreased power. An increase in the number of interim analyses that allowed for either early stopping for efficacy or futility generally increased type I error and decreased power. Conclusions Currently, regulators require demonstration of control of type I error for both frequentist and Bayesian adaptive designs, particularly for late-phase trials. To demonstrate control of type I error in Bayesian adaptive designs, adjustments to the stopping boundaries are usually required for designs that allow for early stopping for efficacy as the number of analyses increase. If the designs only allow for early stopping for futility then adjustments to the stopping boundaries are not needed to control type I error. If one instead uses a strict Bayesian approach, which is currently more accepted in the design and analysis of exploratory trials, then type I errors could be ignored and the designs could instead focus on the posterior probabilities of treatment effects of clinically-relevant values.

Background Optimising techniques to wean patients from invasive mechanical ventilation (IMV) remains a key goal of intensive care practice. The use of non-invasive ventilation (NIV) as a weaning strategy (transitioning patients who are difficult to wean to early NIV) may reduce mortality, ventilator-associated pneumonia and intensive care unit (ICU) length of stay. Objectives Our objectives were to determine the cost effectiveness of protocolised weaning, including early extubation onto NIV, compared with weaning without NIV in a UK National Health Service setting. Methods We conducted an economic evaluation alongside a multicentre randomised controlled trial. Patients were randomised to either protocol-directed weaning from mechanical ventilation or ongoing IMV with daily spontaneous breathing trials. The primary efficacy outcome was time to liberation from ventilation. Bivariate regression of costs and quality-adjusted life-years (QALYs) provided estimates of the incremental cost per QALY and incremental net monetary benefit (INMB) overall and for subgroups [presence/absence of chronic obstructive pulmonary disease (COPD) and operative status]. Long-term cost effectiveness was determined through extrapolation of survival curves using flexible parametric modelling. Results NIV was associated with a mean INMB of £620 ($US885) (cost-effectiveness threshold of £20,000 per QALY) with a corresponding probability of 58% that NIV is cost effective. The probability that NIV is cost effective was higher for those with COPD (84%). NIV was cost effective over 5 years, with an estimated incremental cost-effectiveness ratio of £4618 ($US6594 per QALY gained). Conclusions The probability of NIV being cost effective relative to weaning without NIV ranged between 57 and 59% overall and between 82 and 87% for the COPD subgroup. Electronic supplementary material The online version of this article (10.1007/s41669-020-00210-1) contains supplementary material, which is available to authorized users.

Purpose To examine the time to drug administration in patients with a witnessed cardiac arrest enrolled in the Pre-Hospital Assessment of the Role of Adrenaline: Measuring the Effectiveness of Drug Administration in Cardiac Arrest (PARAMEDIC2) randomised controlled trial. Methods The PARAMEDIC2 trial was undertaken across 5 NHS ambulance services in England and Wales with randomisation between December 2014 and October 2017. Patients with an out-of-hospital cardiac arrest who were unresponsive to initial resuscitation attempts were randomly assigned to 1 mg intravenous adrenaline or matching placebo according to treatment packs that were identical apart from treatment number. Participants and study staff were masked to treatment allocation. Results 8016 patients were enrolled, 4902 sustained a witnessed cardiac arrest of whom 2437 received placebo and 2465 received adrenaline. The odds of return of spontaneous circulation decreased in both groups over time but at a greater rate in the placebo arm odds ratio (OR) 0.93 (95% CI 0.92–0.95) compared with the adrenaline arm OR 0.96 (95% CI 0.95–0.97); interaction OR: 1.03, 95% CI 1.01–1.05, p = 0.005. By contrast, although the rate of survival and favourable neurological outcome decreased as time to treatment increased, the rates did not differ between the adrenaline and placebo groups. Conclusion The rate of return of spontaneous circulation, survival and favourable neurological outcomes decrease over time. As time to drug treatment increases, adrenaline increases the chances of return of spontaneous circulation. Longer term outcomes were not affected by the time to adrenaline administration. (ISRCTN73485024). Electronic supplementary material The online version of this article (10.1007/s00134-019-05836-2) contains supplementary material, which is available to authorized users.

Background Bayesian adaptive designs can be more efficient than traditional methods for multi-arm randomised controlled trials. The aim of this work was to demonstrate how Bayesian adaptive designs can be constructed for multi-arm phase III clinical trials and assess potential benefits that these designs offer. Methods We constructed several alternative Bayesian adaptive designs for the Collaborative Ankle Support Trial (CAST), which was a randomised controlled trial that compared four treatments for severe ankle sprain. These designs incorporated response adaptive randomisation (RAR), arm dropping, and early stopping for efficacy or futility. We studied the operating characteristics of the Bayesian designs via simulation. We then virtually re-executed the trial by implementing the Bayesian adaptive designs using patient data sampled from the CAST study to demonstrate the practical applicability of the designs. Results We constructed five Bayesian adaptive designs, each of which had high power and recruited fewer patients on average than the original designs target sample size. The virtual executions showed that most of the Bayesian designs would have led to trials that declared superiority of one of the interventions over the control. Bayesian adaptive designs with RAR or arm dropping were more likely to allocate patients to better performing arms at each interim analysis. Similar estimates and conclusions were obtained from the Bayesian adaptive designs as from the original trial. Conclusions Using CAST as an example, this case study shows how Bayesian adaptive designs can be constructed for phase III multi-arm trials using clinically relevant decision criteria. These designs demonstrated that they can potentially generate earlier results and allocate more patients to better performing arms. We recommend the wider use of Bayesian adaptive approaches in phase III clinical trials. Trial registration CAST study registration ISRCTN, ISRCTN37807450. Retrospectively registered on 25 April 2003.

Background Phase III trials often require large sample sizes, leading to high costs and delays in clinical decision-making. Group sequential designs can improve trial efficiency by allowing for early stopping for efficacy and/or futility and thus may decrease the sample size, trial duration and associated costs. Bayesian approaches may offer additional benefits by incorporating previous information into the analyses and using decision criteria that are more practically relevant than those used in frequentist approaches. Frequentist group sequential designs have often been used for phase III studies, but the use of Bayesian group sequential designs is less common. The aim of this work was to explore how Bayesian group sequential designs could be constructed for phase III trials conducted in emergency medicine. Methods The PARAMEDIC2 trial was a phase III randomised controlled trial that compared the use of adrenaline to placebo in out-of-hospital cardiac arrest patients on 30-day survival rates. It used a frequentist group sequential design to allow early stopping for efficacy or harm. We constructed several alternative Bayesian group sequential designs and studied their operating characteristics via simulation. We then virtually re-executed the trial by applying the Bayesian designs to the PARAMEDIC2 data to demonstrate what might have happened if these designs had been used in practice. Results We produced three alternative Bayesian group sequential designs, each of which had greater than 90% power to detect the target treatment effect. A Bayesian design which performed interim analyses every 500 patients recruited produced the lowest average sample size. Using the alternative designs, the PARAMEDIC2 trial could have declared adrenaline superior for 30-day survival with approximately 1500 fewer patients. Conclusions Using the PARAMEDIC2 trial as a case study, we demonstrated how Bayesian group sequential designs can be constructed for phase III emergency medicine trials. The Bayesian framework enabled us to obtain efficient designs using decision criteria based on the probability of benefit or harm. It also enabled us to incorporate information from previous studies on the treatment effect via the prior distributions. We recommend the wider use of Bayesian approaches in phase III clinical trials. Trial registration PARAMEDIC2 Trial registration ISRCTN, ISRCTN73485024. Registered 13 March 2014, http://www.isrctn.com/ISRCTN73485024

AIM: \ud \ud To assess the cost-effectiveness of LUCAS-2, a mechanical device for cardiopulmonary resuscitation (CPR) as compared to manual chest compressions in adults with non-traumatic, out-of-hospital cardiac arrest.\ud \ud METHODS: \ud \ud We analysed patient-level data from a large, pragmatic, multi-centre trial linked to administrative secondary care data from the Hospital Episode Statistics (HES) to measure healthcare resource use, costs and outcomes in both arms. A within-trial analysis using quality adjusted life years derived from the EQ-5D-3L was conducted at 12-month follow-up and results were extrapolated to the lifetime horizon using a decision-analytic model.\ud \ud RESULTS: \ud \ud 4471 patients were enrolled in the trial (1652 assigned to the LUCAS-2 group, 2819 assigned to the control group). At 12 months, 89 (5%) patients survived in the LUCAS-2 group and 175 (6%) survived in the manual CPR group. In the vast majority of analyses conducted, both within-trial and by extrapolation of the results over a lifetime horizon, manual CPR dominates LUCAS-2. In other words, patients in the LUCAS-2 group had poorer health outcomes (i.e. lower QALYs) and incurred higher health and social care costs.\ud \ud CONCLUSION: \ud \ud Our study demonstrates that the use of the mechanical chest compression device LUCAS-2 represents poor value for money when compared to standard manual chest compression in out-of-hospital cardiac arrest.

To assess and evaluate patient safety incidents and in particular, medication errors, during a large multi-center pre-hospital trial of emergency therapy (PARAMEDIC2), in order to inform and improve future pre-hospital medicines trials.The PARAMEDIC2 trial was undertaken across five NHS Ambulance Services in England and Wales with randomisation between December 2014 and October 2017. Patients with an out -of-hospital cardiac arrest unresponsive to initial resuscitation were randomly assigned to 1 mg intravenous adrenaline or matching placebo. Records were reviewed to identify trial medication errors involving documentation and/or clinical protocol errors occurring in trial participants. Causes of medication errors, including root cause analysis where available, were reviewed to identify patterns and themes contributing to these errors.Eight thousand sixteen patients were enrolled, of whom 4902 received trial medication. A total of 331 patient safety incidents was reported, involving 295 patients, representing an overall rate of 3.6% of these, 166 (50.2%) were documentation errors while 165 (49.8%) were clinical protocol/medication errors. An overall rate of 0-4.5% was reported across all five ambulance services, with a mean of 2.0%. These errors had no impact on patient care or the trial and were all resolved CONCLUSION: The overall medication error rate of 1.8% primarily consisted of administration of open-label adrenaline and confusion with trial medication packs. A similar number of patients had documentation errors. This study is the first to provide data on patient safety incidents relating to medication errors encountered during a pre-hospital trial of emergency medication administration and will provide supporting data for planning future trials in this area.

We report a case of hypersensitivity pneumonitis (HP) in a young person secondary to vaping. He presented with a putative diagnosis of asthma and required extracorporeal membrane oxygenationbecause of intractable respiratory failure. He developed a critical illness and steroid myopathy and required prolonged rehabilitation. Our patient fulfils diagnostic criteria for HP secondary to e-cigarettes with a positive exposure history, deterioration after skin prick testing, specific serum IgM antibodies against the implicated liquid raising the possibility that the relevant antigen was present in that liquid and radiological and histopathological features compatible with acute HP. There are two learning points. The first is always to consider a reaction to e-cigarettes in someone presenting with an atypical respiratory illness. The second is that we consider e-cigarettes as ‘much safer than tobacco’ at our peril.

Background Invasive mechanical ventilation (IMV) is a life-saving intervention. Following resolution of the condition that necessitated IMV, a spontaneous breathing trial (SBT) is used to determine patient readiness for IMV discontinuation. In patients who fail one or more SBTs, there is uncertainty as to the optimum management strategy. Objective To evaluate the clinical effectiveness and cost-effectiveness of using non-invasive ventilation (NIV) as an intermediate step in the protocolised weaning of patients from IMV. Design Pragmatic, open-label, parallel-group randomised controlled trial, with cost-effectiveness analysis. Setting A total of 51 critical care units across the UK. Participants Adult intensive care patients who had received IMV for at least 48 hours, who were categorised as ready to wean from ventilation, and who failed a SBT. Interventions Control group (invasive weaning): patients continued to receive IMV with daily SBTs. A weaning protocol was used to wean pressure support based on the patient’s condition. Intervention group (non-invasive weaning): patients were extubated to NIV. A weaning protocol was used to wean inspiratory positive airway pressure, based on the patient’s condition. Main outcome measures The primary outcome measure was time to liberation from ventilation. Secondary outcome measures included mortality, duration of IMV, proportion of patients receiving antibiotics for a presumed respiratory infection and health-related quality of life. Results A total of 364 patients (invasive weaning, n = 182; non-invasive weaning, n = 182) were randomised. Groups were well matched at baseline. There was no difference between the invasive weaning and non-invasive weaning groups in median time to liberation from ventilation {invasive weaning 108 hours [interquartile range (IQR) 57–351 hours] vs. non-invasive weaning 104.3 hours [IQR 34.5–297 hours]; hazard ratio 1.1, 95% confidence interval [CI] 0.89 to 1.39; p = 0.352}. There was also no difference in mortality between groups at any time point. Patients in the non-invasive weaning group had fewer IMV days [invasive weaning 4 days (IQR 2–11 days) vs. non-invasive weaning 1 day (IQR 0–7 days); adjusted mean difference –3.1 days, 95% CI –5.75 to –0.51 days]. In addition, fewer non-invasive weaning patients required antibiotics for a respiratory infection [odds ratio (OR) 0.60, 95% CI 0.41 to 1.00; p = 0.048]. A higher proportion of non-invasive weaning patients required reintubation than those in the invasive weaning group (OR 2.00, 95% CI 1.27 to 3.24). The within-trial economic evaluation showed that NIV was associated with a lower net cost and a higher net effect, and was dominant in health economic terms. The probability that NIV was cost-effective was estimated at 0.58 at a cost-effectiveness threshold of £20,000 per quality-adjusted life-year. Conclusions A protocolised non-invasive weaning strategy did not reduce time to liberation from ventilation. However, patients who underwent non-invasive weaning had fewer days requiring IMV and required fewer antibiotics for respiratory infections. Future work In patients who fail a SBT, which factors predict an adverse outcome (reintubation, tracheostomy, death) if extubated and weaned using NIV? Trial registration Current Controlled Trials ISRCTN15635197. Funding This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 23, No. 48. See the NIHR Journals Library website for further project information.

Introduction\ud \ud Previous research suggests there may be differences in the effects of adrenaline related to the initial cardiac arrest rhythm. The aim of this study was to assess the effect of adrenaline compared with placebo according to whether the initial cardiac arrest rhythm was shockable or non-shockable.\ud \ud Methods\ud \ud Return of spontaneous circulation (ROSC), survival and neurological outcomes according to the initial arrest rhythm were compared amongst patients enrolled in the PARAMEDIC-2 randomised, placebo controlled trial. The results of the PARAMEDIC-2 and PACA out of hospital cardiac arrest trials were combined and meta-analysed.\ud \ud Results\ud \ud The initial rhythm was known for 3,929 (98.2%) in the placebo arm and 3,919 (97.6%) in the adrenaline arm. The effect on the rate of ROSC of adrenaline relative to placebo was greater in patients with non-shockable cardiac rhythms (1002/3003 (33.4%) versus 222/3005 (7.4%), adjusted OR: 6.5, (95% CI 5.6-7.6)) compared with shockable rhythms 349/716 (48.7%) versus (208/702 (29.6%), adjusted OR: 2.3, 95%CI: 1.9-2.9)). The adjusted odds ratio for survival at discharge for non-shockable rhythms was 2.5 (1.3, 4.8) and 1.3 (0.9, 1.8) for shockable rhythms (P value for interaction 0.065) and 1.8(0.8-4.1) and 1.1 (0.8-1.6) respectively for neurological outcome at discharge (P value for interaction 0.295). Meta-analysis found similar results.\ud \ud Conclusion\ud \ud Relative to placebo, the effects of adrenaline ROSC are greater for patients with an initially non-shockable rhythm than those with a shockable rhythms. Similar patterns are observed for longer term survival outcomes and favourable neurological outcomes, although the differences in effects are less pronounced.

ObjectiveThe traditional approach of null hypothesis testing dominates the design and analysis of randomised controlled trials. This study aimed to demonstrate how a simple Bayesian analysis could have been used to analyse the Optimisation of Perioperative Cardiovascular Management to Improve Surgical Outcome (OPTIMISE) trial to obtain more clinically interpretable results.Design, setting, participants and interventionsThe OPTIMISE trial was a pragmatic, multicentre, observer-blinded, randomised controlled trial of 734 high-risk patients undergoing major gastrointestinal surgery in 17 acute care hospitals in the UK. Patients were randomly allocated to a cardiac output-guided haemodynamic therapy algorithm for intravenous fluid and inotropic drug administration during and in the 6 hours following surgery (n=368) or to standard care (n=366). The primary outcome was a binary outcome consisting of a composite of predefined 30-day moderate or major complications and mortality.MethodsWe repeated the primary outcome analysis of the OPTIMISE trial using Bayesian statistical methods to calculate the probability that the intervention was superior, and the probability that a clinically relevant difference existed. We explored the impact of a flat prior and an evidence-based prior on our analyses.ResultsAlthough OPTIMISE was not powered to detect a statistically significant difference between the treatment arms for the observed effect size (relative risk=0.84, 95% CI 0.70 to 1.01; p=0.07), by using Bayesian analyses we were able to demonstrate that there was a 96.9% (flat prior) to 99.5% (evidence-based prior) probability that the intervention was superior to the control.ConclusionsThe use of a Bayesian analytical approach provided a different interpretation of the findings of the OPTIMISE trial (compared with the original frequentist analysis), and suggested patient benefit from the intervention. Incorporation of information from previous studies provided further evidence of a benefit from the intervention. Bayesian analyses can produce results that are more easily interpretable and relevant to clinicians and policy-makers.Trial registration numberISRCTN04386758; Post-results.

Adrenaline and vasopressin are widely used to treat people with cardiac arrest, but there is uncertainty about the safety, effectiveness and the optimal dose.To determine whether adrenaline or vasopressin, or both, administered during cardiac arrest, afford any survival benefit.We searched the Cochrane Central Register of Controlled Trials, MEDLINE, Embase and DARE from their inception to 8 May 2018, and the International Liaison Committee on Resuscitation 2015 Advanced Life Support Consensus on Science and Treatment Recommendations. We also searched four trial registers on 5 September 2018 and checked the reference lists of the included studies and review papers to identify potential papers for review.Any randomised controlled trial comparing: standard-dose adrenaline versus placebo; standard-dose adrenaline versus high-dose adrenaline; and adrenaline versus vasopressin, in any setting, due to any cause of cardiac arrest, in adults and children. There were no language restrictions.Two review authors independently identified trials for review, assessed risks of bias and extracted data, resolving disagreements through re-examination of the trial reports and by discussion. We used risk ratios (RRs) with 95% confidence intervals (CIs) to compare dichotomous outcomes for clinical events. There were no continuous outcomes reported. We examined groups of trials for heterogeneity. We report the quality of evidence for each outcome, using the GRADE approach.We included 26 studies (21,704 participants).Moderate-quality evidence found that adrenaline increased survival to hospital discharge compared to placebo (RR 1.44, 95% CI 1.11 to 1.86; 2 studies, 8538 participants; an increase from 23 to 32 per 1000, 95% CI 25 to 42). We are uncertain about survival to hospital discharge for high-dose compared to standard-dose adrenaline (RR 1.10, 95% CI 0.75 to 1.62; participants = 6274; studies = 10); an increase from 33 to 36 per 1000, 95% CI 24 to 53); standard-dose adrenaline versus vasopressin (RR 1.25, 95% CI 0.84 to 1.85; 6 studies; 2511 participants; an increase from 72 to 90 per 1000, 95% CI 60 to 133); and standard-dose adrenaline versus vasopressin plus adrenaline (RR 0.76, 95% CI 0.47 to 1.22; 3 studies; 3242 participants; a possible decrease from 24 to 18 per 1000, 95% CI 11 to 29), due to very low-quality evidence.Moderate-quality evidence found that adrenaline compared with placebo increased survival to hospital admission (RR 2.51, 95% CI 1.67 to 3.76; 2 studies, 8489 participants; an increase from 83 to 209 per 1000, 95% CI 139 to 313). We are uncertain about survival to hospital admission when comparing standard-dose with high-dose adrenaline, due to very low-quality evidence. Vasopressin may improve survival to hospital admission when compared with standard-dose adrenaline (RR 1.27, 95% CI 1.04 to 1.54; 3 studies, 1953 participants; low-quality evidence; an increase from 260 to 330 per 1000, 95% CI 270 to 400), and may make little or no difference when compared to standard-dose adrenaline plus vasopressin (RR 0.95, 95% CI 0.83 to 1.08; 3 studies; 3249 participants; low-quality evidence; a decrease from 218 to 207 per 1000 (95% CI 181 to 236).There was no evidence that adrenaline (any dose) or vasopressin improved neurological outcomes.The rate of return of spontaneous circulation (ROSC) was higher for standard-dose adrenaline versus placebo (RR 2.86, 95% CI 2.21 to 3.71; participants = 8663; studies = 3); moderate-quality evidence; an increase from 115 to 329 per 1000, 95% CI 254 to 427). We are uncertain about the effect on ROSC for the comparison of standard-dose versus high-dose adrenaline and standard-does adrenaline compared to vasopressin, due to very low-quality evidence. Standard-dose adrenaline may make little or no difference to ROSC when compared to standard-dose adrenaline plus vasopressin (RR 0.97, 95% CI 0.87 to 1.08; 3 studies, 3249 participants; low-quality evidence; a possible decrease from 299 to 290 per 1000, 95% CI 260 to 323).The source of funding was not stated in 11 of the 26 studies. The study drugs were provided by the manufacturer in four of the 26 studies, but neither drug represents a profitable commercial option. The other 11 studies were funded by organisations such as research foundations and government funding bodies.This review provides moderate-quality evidence that standard-dose adrenaline compared to placebo improves return of spontaneous circulation, survival to hospital admission and survival to hospital discharge, but low-quality evidence that it did not affect survival with a favourable neurological outcome. Very low -quality evidence found that high-dose adrenaline compared to standard-dose adrenaline improved return of spontaneous circulation and survival to admission. Vasopressin compared to standard dose adrenaline improved survival to admission but not return of spontaneous circulation, whilst the combination of adrenaline and vasopressin compared with adrenaline alone had no effect on these outcomes. Neither standard dose adrenaline, high-dose adrenaline,vasopressin nor a combination of adrenaline and vasopressin improved survival with a favourable neurological outcome. Many of these studies were conducted more than 20 years ago. Treatment has changed in recent years, so the findings from older studies may not reflect current practice.

IntroductionIn 2013, a single-centre study reported the safe use of esmolol in patients with septic shock and tachycardia who required vasopressor therapy for more than 24 hours. Although not powered to detect a change in mortality, marked improvements were seen in survival (adjusted HR, 0.39; 95% CI, 0.26 to 0.59; pMethods and analysisSTRESS-L is a randomised, open-label, non-blinded clinical trial which is enrolling a total of 340 patients with septic shock as defined by Sepsis-3 consensus definition and a tachycardia (heart rate ≥95 beats per minute (bpm)) after vasopressor treatment of at least 24 hours. Standard randomisation (1:1 ratio) allocates patients to receive usual care (according to international standards) versus usual care and a continuous landiolol infusion to reduce the heart rate between 80 and 94 bpm. The primary endpoint is the mean Sequential Organ Failure Assessment score over 14 days from entry into the trial and while in intensive care unit. Results will inform current clinical practice guidelines.Ethics and disseminationThis trial has clinical trial authorisation from the UK competent authority, the Medicines and Healthcare products Regulatory Agency, and has been approved by the East of England-Essex Research Ethics Committee (reference: 17/EE/0368).The results of the trial will be reported first to trial collaborators. The main report will be drafted by the trial coordinating team, and the final version will be agreed by the Trial Steering Committee before submission for publication, on behalf of the collaboration.RegistrationThe trial is funded by the National Institute for Health Research Efficacy and Mechanism Evaluation (EME) (Project Number: EME-14/150/85) and registeredISRCTN12600919and EudraCT: 2017-001785-14.

IntroductionShoulder pain due to irreparable rotator cuff tears can cause substantial disability, but treatment options are limited. A balloon spacer is a relatively simple addition to a standard arthroscopic debridement procedure, but it is costly and there is no current randomised trial evidence to support its use. This trial will evaluate the clinical and cost-effectiveness of a subacromial balloon spacer for individuals undergoing arthroscopic debridement for irreparable rotator cuff tears.New surgical procedures can provide substantial benefit to patients. Good quality randomised controlled trials (RCTs) are needed, but trials in surgery are typically long and expensive, exposing patients to risk and the healthcare system to substantial costs. One way to improve the efficiency of trials is with an adaptive sample size. Such methods are well established in drug trials but have rarely, if ever, been used in surgical trials.Methods and analysisSubacromial spacer for Tears Affecting Rotator cuff Tendons: a Randomised, Efficient, Adaptive Clinical Trial in Surgery (START:REACTS) is a participant and assessor blinded, adaptive, multicentre RCT comparing arthroscopic debridement with the InSpace balloon (Stryker, USA) to arthroscopic debridement alone for people with a symptomatic irreparable rotator cuff tear. It uses a group sequential adaptive design where interim analyses are performed using all of the 3, 6 and 12-month data that are available at each time point. A maximum of 221 participants will be randomised (1:1 ratio), this will provide 90% power (at the 5% level) for a 6 point difference in the primary outcome; the Oxford Shoulder Score at 12 months. A substudy will use deltoid-active MRI scans in 56 participants to assess the function of the balloon. Analysis will be on an intention-to-treat basis and reported according to principles established in the Consolidated Standards of Reporting Trials statement.Ethics and disseminationNRES number 18/WM/0025. The results will be disseminated via peer-reviewed publications, presentations at conferences, lay summaries and social media.Trial registration numberISRCTN17825590

Purpose A systematic review and meta-analysis was conducted to answer the question ‘In adults with respiratory failure requiring invasive ventilation for more than 24 h, does a weaning strategy with early extubation to non-invasive ventilation (NIV) compared to invasive ventilation weaning reduce all-cause hospital mortality?’ Methods We included randomised and quasi-randomised controlled trials that evaluated the use of non-invasive ventilation, compared to invasive ventilation, as a weaning strategy in adults mechanically ventilated for at least 24 h. The EMBASE, MEDLINE and Cochrane Central Register of Controlled Trials (CENTRAL) bibliographic databases were searched from inception to February 2018. Bayesian hierarchical models were used to perform the meta-analysis. The primary outcome was mortality at hospital discharge. Secondary outcomes included mortality (30, 60, 90 and 180 days), quality of life, duration of invasive ventilation, weaning failure, length of stay [intensive care unit (ICU) and hospital] and adverse events. Results Twenty-five relevant studies involving 1609 patients were included in the quantitative analysis. Studies had moderate to high risk of bias due to risk of performance and detection bias. Mortality at hospital discharge was lower in the NIV weaning group compared to the invasive weaning group [pooled odds ratio (OR) 0.58, 95% highest density interval (HDI) 0.29–0.89]. Subgroup analyses showed lower pooled mortality at hospital discharge rates in NIV weaning than those in the control group in chronic obstructive pulmonary disease (COPD) patients (pooled OR 0.43, 95% HDI 0.13–0.81) and the effect is less certain in the mixed ICU population (pooled OR 0.88, 95% HDI 0.25–1.48). NIV weaning reduced the duration of invasive ventilation in patients [standardised mean difference (SMD) − 1.34, 95% HDI − 1.92 to − 0.77] and ICU length of stay (SMD − 0.70, 95% HDI − 0.94 to − 0.46). Reported rates of ventilator associated pneumonia (VAP) were lower in the NIV group. NIV weaning did not reduce overall hospital length of stay or long-term mortality. There were insufficient data to compare other adverse events and health-related quality of life. Conclusions The use of NIV in weaning from mechanical ventilation decreases hospital mortality, the incidence of VAP and ICU length of stay. NIV as a weaning strategy appears to be most beneficial in patients with COPD. Electronic supplementary material The online version of this article (10.1007/s00134-018-5434-z) contains supplementary material, which is available to authorized users.