Search

Abstract Background Early evidence that patients with (multiple) pre-existing diseases are at highest risk for severe COVID-19 has been instrumental in the pandemic to allocate critical care resources and later vaccination schemes. However, systematic studies exploring the breadth of medical diagnoses are scarce but may help to understand severe COVID-19 among patients at supposedly low risk. Methods We systematically harmonized >12 million primary care and hospitalisation health records from ~500,000 UK Biobank participants into 1448 collated disease terms to systematically identify diseases predisposing to severe COVID-19 (requiring hospitalisation or death) and its post-acute sequalae, Long COVID. Results Here we identify 679 diseases associated with an increased risk for severe COVID-19 (n = 672) and/or Long COVID (n = 72) that span almost all clinical specialties and are strongly enriched in clusters of cardio-respiratory and endocrine-renal diseases. For 57 diseases, we establish consistent evidence to predispose to severe COVID-19 based on survival and genetic susceptibility analyses. This includes a possible role of symptoms of malaise and fatigue as a so far largely overlooked risk factor for severe COVID-19. We finally observe partially opposing risk estimates at known risk loci for severe COVID-19 for etiologically related diseases, such as post-inflammatory pulmonary fibrosis or rheumatoid arthritis, possibly indicating a segregation of disease mechanisms. Conclusions Our results provide a unique reference that demonstrates how 1) complex co-occurrence of multiple – including non-fatal – conditions predispose to increased COVID-19 severity and 2) how incorporating the whole breadth of medical diagnosis can guide the interpretation of genetic risk loci.

Abstract Background Our study examined whether prevalent and incident comorbidities are increased in idiopathic pulmonary fibrosis (IPF) patients when compared to matched chronic obstructive pulmonary disease (COPD) patients and control subjects without IPF or COPD. Methods IPF and age, gender and smoking matched COPD patients, diagnosed between 01/01/1997 and 01/01/2019 were identified from the Clinical Practice Research Datalink GOLD database multiple registrations cohort at the first date an ICD-10 or read code mentioned IPF/COPD. A control cohort comprised age, gender and pack-year smoking matched subjects without IPF or COPD. Prevalent (prior to IPF/COPD diagnosis) and incident (after IPF/COPD diagnosis) comorbidities were examined. Group differences were estimated using a t-test. Mortality relationships were examined using multivariable Cox proportional hazards adjusted for patient age, gender and smoking status. Results Across 3055 IPF patients, 38% had 3 or more prevalent comorbidities versus 32% of COPD patients and 21% of matched control subjects. Survival time reduced as the number of comorbidities in an individual increased (p

Objective The HDRUK Phenotype Library (phenotypes.healthdatagateway.org) shares definitions used to measure concepts of interest (such as diagnoses or treatments) within health datasets. It already holds more than 1000 phenotypes, with researchers able to contribute their work via an API. We aimed to create a more user-friendly method of contributing to the Library. Approach We designed a phenotype creation workflow enabling users to create and submit new content via web interface. Goals included clarity and ease of use for a broad range of users. Results A home page shows researchers’ own content. Researchers can create new phenotypes using a web form, entering metadata such as name, authors, description, and publications. Code lists are defined via one or more rules, including search terms or referring to another phenotype, or by CSV upload. Users can make their phenotypes accessible to a research group or to all authenticated users, as well as publish content on the web. Publication requests are reviewed to ensure content is complete and appropriate. Editing, with full history and version control, is also supported. Conclusions We implemented and released the new features. We are currently engaging researchers to get feedback and invite content submission. Implications The benefit of tools to support research transparency and repeatability is only realized when they are adopted. We hope that a GUI to support phenotype creation will broaden the Library’s user base help it serve as an enabler of higher-quality, more efficient research across the worldwide health data research community.

Background Despite the increasing availability of electronic healthcare record (EHR) data and wide availability of plug-and-play machine learning (ML) Application Programming Interfaces, the adoption of data-driven decision-making within routine hospital workflows thus far, has remained limited. Through the lens of deriving clusters of diagnoses by age, this study investigated the type of ML analysis that can be performed using EHR data and how results could be communicated to lay stakeholders.Methods Observational EHR data from a tertiary paediatric hospital, containing 61 522 unique patients and 3315 unique ICD-10 diagnosis codes was used, after preprocessing. K-means clustering was applied to identify age distributions of patient diagnoses. The final model was selected using quantitative metrics and expert assessment of the clinical validity of the clusters. Additionally, uncertainty over preprocessing decisions was analysed.Findings Four age clusters of diseases were identified, broadly aligning to ages between: 0 and 1; 1 and 5; 5 and 13; 13 and 18. Diagnoses, within the clusters, aligned to existing knowledge regarding the propensity of presentation at different ages, and sequential clusters presented known disease progressions. The results validated similar methodologies within the literature. The impact of uncertainty induced by preprocessing decisions was large at the individual diagnoses but not at a population level. Strategies for mitigating, or communicating, this uncertainty were successfully demonstrated.Conclusion Unsupervised ML applied to EHR data identifies clinically relevant age distributions of diagnoses which can augment existing decision making. However, biases within healthcare datasets dramatically impact results if not appropriately mitigated or communicated.

Background Electronic health records (EHRs) have the potential to be used to produce detailed disease burden estimates. In this study we created disease estimates using national EHR for three high burden conditions, compared estimates between linked and unlinked datasets and produced stratified estimates by age, sex, ethnicity, socio-economic deprivation and geographical region. Methods EHRs containing primary care (Clinical Practice Research Datalink), secondary care (Hospital Episode Statistics) and mortality records (Office for National Statistics) were used. We used existing disease phenotyping algorithms to identify cases of cancer (breast, lung, colorectal and prostate), type 1 and 2 diabetes, and lower back pain. We calculated age-standardised incidence of first cancer, point prevalence for diabetes, and primary care consultation prevalence for low back pain. Results 7.2 million people contributing 45.3 million person-years of active follow-up between 2000–2014 were included. CPRD-HES combined and CPRD-HES-ONS combined lung and bowel cancer incidence estimates by sex were similar to cancer registry estimates. Linked CPRD-HES estimates for combined Type 1 and Type 2 diabetes were consistently higher than those of CPRD alone, with the difference steadily increasing over time from 0.26% (2.99% for CPRD-HES vs. 2.73 for CPRD) in 2002 to 0.58% (6.17% vs. 5.59) in 2013. Low back pain prevalence was highest in the most deprived quintile and when compared to the least deprived quintile the difference in prevalence increased over time between 2000 and 2013, with the largest difference of 27% (558.70 per 10,000 people vs 438.20) in 2013. Conclusions We use national EHRs to produce estimates of burden of disease to produce detailed estimates by deprivation, ethnicity and geographical region. National EHRs have the potential to improve disease burden estimates at a local and global level and may serve as more automated, timely and precise inputs for policy making and global burden of disease estimation.

Abstract Raynaud’s phenomenon (RP) is a common vasospastic disorder that causes severe pain and ulcers, but despite its high reported heritability, no causal genes have been robustly identified. We conducted a genome-wide association study including 5,147 RP cases and 439,294 controls, based on diagnoses from electronic health records, and identified three unreported genomic regions associated with the risk of RP (p

BackgroundThe occurrence of a range of health outcomes following myocardial infarction (MI) is unknown. Therefore, this study aimed to determine the long-term risk of major health outcomes following MI and generate sociodemographic stratified risk charts in order to inform care recommendations in the post-MI period and underpin shared decision making.Methods and findingsThis nationwide cohort study includes all individuals aged ≥18 years admitted to one of 229 National Health Service (NHS) Trusts in England between 1 January 2008 and 31 January 2017 (final follow-up 27 March 2017). We analysed 11 non-fatal health outcomes (subsequent MI and first hospitalisation for heart failure, atrial fibrillation, cerebrovascular disease, peripheral arterial disease, severe bleeding, renal failure, diabetes mellitus, dementia, depression, and cancer) and all-cause mortality. Of the 55,619,430 population of England, 34,116,257 individuals contributing to 145,912,852 hospitalisations were included (mean age 41.7 years (standard deviation [SD 26.1]); n = 14,747,198 (44.2%) male). There were 433,361 individuals with MI (mean age 67.4 years [SD 14.4)]; n = 283,742 (65.5%) male). Following MI, all-cause mortality was the most frequent event (adjusted cumulative incidence at 9 years 37.8% (95% confidence interval [CI] [37.6,37.9]), followed by heart failure (29.6%; 95% CI [29.4,29.7]), renal failure (27.2%; 95% CI [27.0,27.4]), atrial fibrillation (22.3%; 95% CI [22.2,22.5]), severe bleeding (19.0%; 95% CI [18.8,19.1]), diabetes (17.0%; 95% CI [16.9,17.1]), cancer (13.5%; 95% CI [13.3,13.6]), cerebrovascular disease (12.5%; 95% CI [12.4,12.7]), depression (8.9%; 95% CI [8.7,9.0]), dementia (7.8%; 95% CI [7.7,7.9]), subsequent MI (7.1%; 95% CI [7.0,7.2]), and peripheral arterial disease (6.5%; 95% CI [6.4,6.6]). Compared with a risk-set matched population of 2,001,310 individuals, first hospitalisation of all non-fatal health outcomes were increased after MI, except for dementia (adjusted hazard ratio [aHR] 1.01; 95% CI [0.99,1.02];p = 0.468) and cancer (aHR 0.56; 95% CI [0.56,0.57];p < 0.001). The study includes data from secondary care only-as such diagnoses made outside of secondary care may have been missed leading to the potential underestimation of the total burden of disease following MI.ConclusionsIn this study, up to a third of patients with MI developed heart failure or renal failure, 7% had another MI, and 38% died within 9 years (compared with 35% deaths among matched individuals). The incidence of all health outcomes, except dementia and cancer, was higher than expected during the normal life course without MI following adjustment for age, sex, year, and socioeconomic deprivation. Efforts targeted to prevent or limit the accrual of chronic, multisystem disease states following MI are needed and should be guided by the demographic-specific risk charts derived in this study.

Abstract Background Low-frequency variants play an important role in breast cancer (BC) susceptibility. Gene-based methods can increase power by combining multiple variants in the same gene and help identify target genes. Methods We evaluated the potential of gene-based aggregation in the Breast Cancer Association Consortium cohorts including 83,471 cases and 59,199 controls. Low-frequency variants were aggregated for individual genes’ coding and regulatory regions. Association results in European ancestry samples were compared to single-marker association results in the same cohort. Gene-based associations were also combined in meta-analysis across individuals with European, Asian, African, and Latin American and Hispanic ancestry. Results In European ancestry samples, 14 genes were significantly associated (q

Most genetic studies of disease have been done in European ancestry cohorts, and the relevance to other populations is not guaranteed. Here, the authors use data from 22,000 British South Asian individuals and find that the transferability of polygenic scores was high for lipids and blood pressure, and lower for BMI and coronary artery disease.

Abstract Aim The optimal strategy for diabetes control in patients with heart failure (HF) following myocardial infarction (MI) remains unknown. Metformin, a guideline‐recommended therapy for patients with chronic HF and type 2 diabetes mellitus (T2DM), is associated with reduced mortality and HF hospitalizations. However, worse outcomes have been reported when used at the time of MI. We compared outcomes of patients with T2DM and HF of ischaemic aetiology according to antidiabetic treatment. Methods and results This study used linked data from primary care, hospital admissions, and death registries for 4.7 million inhabitants in England, as part of the CALIBER resource. The primary endpoint was a composite of cardiovascular mortality and HF hospitalization. The secondary endpoints were the individual components of the primary endpoint and all‐cause mortality. To evaluate the effect of temporal changes in diabetes treatment, antidiabetic medication was included as time‐dependent covariates in survival analyses. The study included 1172 patients with T2DM and prior MI and incident HF between 3 January 1998 and 26 February 2010. Five hundred and ninety‐six patients had the primary outcome over median follow‐up of 2.53 (IQR: 0.98–4.92) years. Adjusted analyses showed a reduced hazard of the composite endpoint for exposure to all antidiabetic medication with hazard ratios (HRs) of 0.50 [95% confidence interval (CI): 0.42–0.59], 0.66 (95% CI: 0.55–0.80), and 0.53 (95% CI: 0.43–0.65), respectively. A similar effect was seen for all‐cause mortality [HRs of 0.43 (95% CI: 0.35–0.52), 0.57 (95% CI: 0.46–0.70), and 0.34 (95% CI: 0.27–0.43), respectively]. Conclusions When considering changes in antidiabetic treatment over time, all drug classes were associated with reduced risk of cardiovascular mortality and HF hospitalization.

Abstract Background Assessing the spectrum of disease risk associated with hypertriglyceridemia is needed to inform potential benefits from emerging triglyceride lowering treatments. We sought to examine the associations between a full range of plasma triglyceride concentration with five clinical outcomes. Methods We used linked data from primary and secondary care for 15 M people, to explore the association between triglyceride concentration and risk of acute pancreatitis, chronic pancreatitis, new onset diabetes, myocardial infarction and all-cause mortality, over a median of 6–7 years follow up. Results Triglyceride concentration was available for 1,530,411 individuals (mean age 56·6 ± 15·6 years, 51·4% female), with a median of 1·3 mmol/L (IQR: 0.9.to 1.9). Severe hypertriglyceridemia, defined as > 10 mmol/L, was identified in 3289 (0·21%) individuals including 620 with > 20 mmol/L. In multivariable analyses, a triglyceride concentration > 20 mmol/L was associated with very high risk for acute pancreatitis (Hazard ratio (HR) 13·55 (95% CI 9·15–20·06)); chronic pancreatitis (HR 25·19 (14·91–42·55)); and high risk for diabetes (HR 5·28 (4·51–6·18)) and all-cause mortality (HR 3·62 (2·82–4·65)) when compared to the reference category of ≤ 1·7 mmol/L. An association with myocardial infarction, however, was only observed for more moderate hypertriglyceridaemia between 1.7 and 10 mmol/L. We found a risk interaction with age, with higher risks for all outcomes including mortality among those ≤ 40 years compared to > 40 years. Conclusions We highlight an exponential association between severe hypertriglyceridaemia and risk of incident acute and chronic pancreatitis, new diabetes, and mortality, especially at younger ages, but not for myocardial infarction for which only moderate hypertriglyceridemia conferred risk.

Abstract Background Several common conditions have been widely recognised as risk factors for COVID-19 related death, but risks borne by people with rare diseases are largely unknown. Therefore, we aim to estimate the difference of risk for people with rare diseases comparing to the unaffected. Method To estimate the correlation between rare diseases and COVID-19 related death, we performed a retrospective cohort study in Genomics England 100k Genomes participants, who tested positive for Sars-Cov-2 during the first wave (16-03-2020 until 31-July-2020) of COVID-19 pandemic in the UK (n = 283). COVID-19 related mortality rates were calculated in two groups: rare disease patients (n = 158) and unaffected relatives (n = 125). Fisher’s exact test and logistic regression was used for univariable and multivariable analysis, respectively. Results People with rare diseases had increased risk of COVID19-related deaths compared to the unaffected relatives (OR [95% CI] = 3.47 [1.21– 12.2]). Although, the effect was insignificant after adjusting for age and number of comorbidities (OR [95% CI] = 1.94 [0.65–5.80]). Neurology and neurodevelopmental diseases was significantly associated with COVID19-related death in both univariable (OR [95% CI] = 4.07 [1.61–10.38]) and multivariable analysis (OR [95% CI] = 4.22 [1.60–11.08]). Conclusions Our results showed that rare disease patients, especially ones affected by neurology and neurodevelopmental disorders, in the Genomics England cohort had increased risk of COVID-19 related death during the first wave of the pandemic in UK. The high risk is likely associated with rare diseases themselves, while we cannot rule out possible mediators due to the small sample size. We would like to raise the awareness that rare disease patients may face increased risk for COVID-19 related death. Proper considerations for rare disease patients should be taken when relevant policies (e.g., returning to workplace) are made.

Summary: Background: Although chronic kidney disease (CKD) is associated with high multimorbidity, polypharmacy, morbidity and mortality, existing classification systems (mild to severe, usually based on estimated glomerular filtration rate, proteinuria or urine albumin-creatinine ratio) and risk prediction models largely ignore the complexity of CKD, its risk factors and its outcomes. Improved subtype definition could improve prediction of outcomes and inform effective interventions. Methods: We analysed individuals ≥18 years with incident and prevalent CKD (n = 350,067 and 195,422 respectively) from a population-based electronic health record resource (2006–2020; Clinical Practice Research Datalink, CPRD). We included factors (n = 264 with 2670 derived variables), e.g. demography, history, examination, blood laboratory values and medications. Using a published framework, we identified subtypes through seven unsupervised machine learning (ML) methods (K-means, Diana, HC, Fanny, PAM, Clara, Model-based) with 66 (of 2670) variables in each dataset. We evaluated subtypes for: (i) internal validity (within dataset, across methods); (ii) prognostic validity (predictive accuracy for 5-year all-cause mortality and admissions); and (iii) medications (new and existing by British National Formulary chapter). Findings: After identifying five clusters across seven approaches, we labelled CKD subtypes: 1. Early-onset, 2. Late-onset, 3. Cancer, 4. Metabolic, and 5. Cardiometabolic. Internal validity: We trained a high performing model (using XGBoost) that could predict disease subtypes with 95% accuracy for incident and prevalent CKD (Sensitivity: 0.81–0.98, F1 score:0.84–0.97). Prognostic validity: 5-year all-cause mortality, hospital admissions, and incidence of new chronic diseases differed across CKD subtypes. The 5-year risk of mortality and admissions in the overall incident CKD population were highest in cardiometabolic subtype: 43.3% (42.3–42.8%) and 29.5% (29.1–30.0%), respectively, and lowest in the early-onset subtype: 5.7% (5.5–5.9%) and 18.7% (18.4–19.1%). Medications: Across CKD subtypes, the distribution of prescription medication classes at baseline varied, with highest medication burden in cardiometabolic and metabolic subtypes, and higher burden in prevalent than incident CKD. Interpretation: In the largest CKD study using ML, to-date, we identified five distinct subtypes in individuals with incident and prevalent CKD. These subtypes have relevance to study of aetiology, therapeutics and risk prediction. Funding: AstraZeneca UK Ltd, Health Data Research UK.

Abstract Background Cardiovascular and renal diseases (CVRD) are major causes of mortality in individuals with type 2 diabetes (T2D). Studies of lifetime risk have neither considered all CVRD together nor the relative contribution of major risk factors to combined disease burden. Methods In a population-based cohort study using national electronic health records, we studied 473,399 individuals with T2D in England 2007–2018. Lifetime risk of individual and combined major adverse renal cardiovascular events, MARCE (including CV death and CVRD: heart failure; chronic kidney disease; myocardial infarction; stroke or peripheral artery disease), were estimated, accounting for baseline CVRD status and competing risk of death. We calculated population attributable risk for individual CVRD components. Ideal cardiovascular health was defined by blood pressure, cholesterol, glucose, smoking, physical activity, diet, and body mass index (i.e. modifiable risk factors). Results In individuals with T2D, lifetime risk of MARCE was 80% in those free from CVRD and was 97%, 93%, 98%, 89% and 91% in individuals with heart failure, chronic kidney disease, myocardial infarction, stroke and peripheral arterial disease, respectively at baseline. Among CVRD-free individuals, lifetime risk of chronic kidney disease was highest (54%), followed by CV death (41%), heart failure (29%), stroke (20%), myocardial infarction (19%) and peripheral arterial disease (9%). In those with HF only, 75% of MARCE after index T2D can be attributed to HF after adjusting for age, gender, and comorbidities. Compared with those with > 1,

Abstract Aims The HERMES (HEart failure Molecular Epidemiology for Therapeutic targetS) consortium aims to identify the genomic and molecular basis of heart failure. Methods and results The consortium currently includes 51 studies from 11 countries, including 68 157 heart failure cases and 949 888 controls, with data on heart failure events and prognosis. All studies collected biological samples and performed genome‐wide genotyping of common genetic variants. The enrolment of subjects into participating studies ranged from 1948 to the present day, and the median follow‐up following heart failure diagnosis ranged from 2 to 116 months. Forty‐nine of 51 individual studies enrolled participants of both sexes; in these studies, participants with heart failure were predominantly male (34–90%). The mean age at diagnosis or ascertainment across all studies ranged from 54 to 84 years. Based on the aggregate sample, we estimated 80% power to genetic variant associations with risk of heart failure with an odds ratio of ≥1.10 for common variants (allele frequency ≥ 0.05) and ≥1.20 for low‐frequency variants (allele frequency 0.01–0.05) at P

Abstract Background An Informatics Consult has been proposed in which clinicians request novel evidence from large scale health data resources, tailored to the treatment of a specific patient. However, the availability of such consultations is lacking. We seek to provide an Informatics Consult for a situation where a treatment indication and contraindication coexist in the same patient, i.e., anti-coagulation use for stroke prevention in a patient with both atrial fibrillation (AF) and liver cirrhosis. Methods We examined four sources of evidence for the effect of warfarin on stroke risk or all-cause mortality from: (1) randomised controlled trials (RCTs), (2) meta-analysis of prior observational studies, (3) trial emulation (using population electronic health records (N = 3,854,710) and (4) genetic evidence (Mendelian randomisation). We developed prototype forms to request an Informatics Consult and return of results in electronic health record systems. Results We found 0 RCT reports and 0 trials recruiting for patients with AF and cirrhosis. We found broad concordance across the three new sources of evidence we generated. Meta-analysis of prior observational studies showed that warfarin use was associated with lower stroke risk (hazard ratio [HR] = 0.71, CI 0.39–1.29). In a target trial emulation, warfarin was associated with lower all-cause mortality (HR = 0.61, CI 0.49–0.76) and ischaemic stroke (HR = 0.27, CI 0.08–0.91). Mendelian randomisation served as a drug target validation where we found that lower levels of vitamin K1 (warfarin is a vitamin K1 antagonist) are associated with lower stroke risk. A pilot survey with an independent sample of 34 clinicians revealed that 85% of clinicians found information on prognosis useful and that 79% thought that they should have access to the Informatics Consult as a service within their healthcare systems. We identified candidate steps for automation to scale evidence generation and to accelerate the return of results. Conclusion We performed a proof-of-concept Informatics Consult for evidence generation, which may inform treatment decisions in situations where there is dearth of randomised trials. Patients are surprised to know that their clinicians are currently not able to learn in clinic from data on ‘patients like me’. We identify the key challenges in offering such an Informatics Consult as a service.

Abstract Background Machine learning (ML) is increasingly used in research for subtype definition and risk prediction, particularly in cardiovascular diseases. No existing ML models are routinely used for cardiovascular disease management, and their phase of clinical utility is unknown, partly due to a lack of clear criteria. We evaluated ML for subtype definition and risk prediction in heart failure (HF), acute coronary syndromes (ACS) and atrial fibrillation (AF). Methods For ML studies of subtype definition and risk prediction, we conducted a systematic review in HF, ACS and AF, using PubMed, MEDLINE and Web of Science from January 2000 until December 2019. By adapting published criteria for diagnostic and prognostic studies, we developed a seven-domain, ML-specific checklist. Results Of 5918 studies identified, 97 were included. Across studies for subtype definition (n = 40) and risk prediction (n = 57), there was variation in data source, population size (median 606 and median 6769), clinical setting (outpatient, inpatient, different departments), number of covariates (median 19 and median 48) and ML methods. All studies were single disease, most were North American (n = 61/97) and only 14 studies combined definition and risk prediction. Subtype definition and risk prediction studies respectively had limitations in development (e.g. 15.0% and 78.9% of studies related to patient benefit; 15.0% and 15.8% had low patient selection bias), validation (12.5% and 5.3% externally validated) and impact (32.5% and 91.2% improved outcome prediction; no effectiveness or cost-effectiveness evaluations). Conclusions Studies of ML in HF, ACS and AF are limited by number and type of included covariates, ML methods, population size, country, clinical setting and focus on single diseases, not overlap or multimorbidity. Clinical utility and implementation rely on improvements in development, validation and impact, facilitated by simple checklists. We provide clear steps prior to safe implementation of machine learning in clinical practice for cardiovascular diseases and other disease areas.

Background Most adults presenting with chest pain will not receive a diagnosis and be recorded with unattributed chest pain. The objective was to assess if they have increased risk of cardiovascular disease compared with those with noncoronary chest pain and determine whether investigations and interventions are targeted at those at highest risk. Methods and Results We used records from general practices in England linked to hospitalization and mortality information. The study population included patients aged 18 years or over with a new record of chest pain with a noncoronary cause or unattributed between 2002 and 2018, and no cardiovascular disease recorded up to 6 months (diagnostic window) afterward. We compared risk of a future cardiovascular event by type of chest pain, adjusting for cardiovascular risk factors and alternative explanations for chest pain. We determined prevalence of cardiac diagnostic investigations and preventative medication during the diagnostic window in patients with estimated cardiovascular risk ≥10%. There were 375 240 patients with unattributed chest pain (245 329 noncoronary chest pain). There was an increased risk of cardiovascular events for patients with unattributed chest pain, highest in the first year (hazard ratio, 1.25 [95% CI, 1.21–1.29]), persistent up to 10 years. Patients with unattributed chest pain had consistently increased risk of myocardial infarction over time but no increased risk of stroke. Thirty percent of patients at higher risk were prescribed lipid‐lowering medication. Conclusions Patients presenting to primary care with unattributed chest pain are at increased risk of cardiovascular events. Primary prevention to reduce cardiovascular events appears suboptimal in those at higher risk.

Abstract Reducing the burden of late-life morbidity requires an understanding of the mechanisms of ageing-related diseases (ARDs), defined as diseases that accumulate with increasing age. This has been hampered by the lack of formal criteria to identify ARDs. Here, we present a framework to identify ARDs using two complementary methods consisting of unsupervised machine learning and actuarial techniques, which we applied to electronic health records (EHRs) from 3,009,048 individuals in England using primary care data from the Clinical Practice Research Datalink (CPRD) linked to the Hospital Episode Statistics admitted patient care dataset between 1 April 2010 and 31 March 2015 (mean age 49.7 years (s.d. 18.6), 51% female, 70% white ethnicity). We grouped 278 high-burden diseases into nine main clusters according to their patterns of disease onset, using a hierarchical agglomerative clustering algorithm. Four of these clusters, encompassing 207 diseases spanning diverse organ systems and clinical specialties, had rates of disease onset that clearly increased with chronological age. However, the ages of onset for these four clusters were strikingly different, with median age of onset 82 years (IQR 82–83) for Cluster 1, 77 years (IQR 75–77) for Cluster 2, 69 years (IQR 66–71) for Cluster 3 and 57 years (IQR 54–59) for Cluster 4. Fitting to ageing-related actuarial models confirmed that the vast majority of these 207 diseases had a high probability of being ageing-related. Cardiovascular diseases and cancers were highly represented, while benign neoplastic, skin and psychiatric conditions were largely absent from the four ageing-related clusters. Our framework identifies and clusters ARDs and can form the basis for fundamental and translational research into ageing pathways.

Background A minority of acute coronary syndrome (ACS) cases are associated with ventricular arrhythmias (VA) and/or cardiac arrest (CA). We investigated the effect of VA/CA at the time of ACS on long‐term outcomes. Methods and Results We analyzed routine clinical data from 5 National Health Service trusts in the United Kingdom, collected between 2010 and 2017 by the National Institute for Health Research Health Informatics Collaborative. A total of 13 444 patients with ACS, 376 (2.8%) of whom had concurrent VA, survived to hospital discharge and were followed up for a median of 3.42 years. Patients with VA or CA at index presentation had significantly increased risks of subsequent VA during follow‐up (VA group: adjusted hazard ratio [HR], 4.15 [95% CI, 2.42–7.09]; CA group: adjusted HR, 2.60 [95% CI, 1.23–5.48]). Patients who suffered a CA in the context of ACS and survived to discharge also had a 36% increase in long‐term mortality (adjusted HR, 1.36 [95% CI, 1.04–1.78]), although the concurrent diagnosis of VA alone during ACS did not affect all‐cause mortality (adjusted HR, 1.03 [95% CI, 0.80–1.33]). Conclusions Patients who develop VA or CA during ACS who survive to discharge have increased risks of subsequent VA, whereas those who have CA during ACS also have an increase in long‐term mortality. These individuals may represent a subgroup at greater risk of subsequent arrhythmic events as a result of intrinsically lower thresholds for developing VA.

ImportanceA lack of internationally agreed standards for combining available data sources at scale risks inconsistent disease phenotyping limiting research reproducibility.ObjectiveTo develop and then evaluate if a rules-based algorithm can identify coronary artery disease (CAD) sub-phenotypes using electronic health records (EHR) and questionnaire data from UK Biobank (UKB).DesignCase-control and cohort study.SettingProspective cohort study of 502K individuals aged 40-69 years recruited between 2006-2010 into the UK Biobank with linked hospitalization and mortality data and genotyping.ParticipantsWe included all individuals for phenotyping into 6 predefined CAD phenotypes using hospital admission and procedure codes, mortality records and baseline survey data. Of these, 408,470 unrelated individuals of European descent had a polygenic risk score (PRS) for CAD estimated.ExposureCAD Phenotypes.Main outcomes and measuresAssociation with baseline risk factors, mortality (n = 14,419 over 7.8 years median f/u), and a PRS for CAD.ResultsThe algorithm classified individuals with CAD into prevalent MI (n = 4,900); incident MI (n = 4,621), prevalent CAD without MI (n = 10,910), incident CAD without MI (n = 8,668), prevalent self-reported MI (n = 2,754); prevalent self-reported CAD without MI (n = 5,623), yielding 37,476 individuals with any type of CAD. Risk factors were similar across the six CAD phenotypes, except for fewer men in the self-reported CAD without MI group (46.7% v 70.1% for the overall group). In age- and sex- adjusted survival analyses, mortality was highest following incident MI (HR 6.66, 95% CI 6.07-7.31) and lowest for prevalent self-reported CAD without MI at baseline (HR 1.31, 95% CI 1.15-1.50) compared to disease-free controls. There were similar graded associations across the six phenotypes per SD increase in PRS, with the strongest association for prevalent MI (OR 1.50, 95% CI 1.46-1.55) and the weakest for prevalent self-reported CAD without MI (OR 1.08, 95% CI 1.05-1.12). The algorithm is available in the open phenotype HDR UK phenotype library (https://portal.caliberresearch.org/).ConclusionsAn algorithmic, EHR-based approach distinguished six phenotypes of CAD with distinct survival and PRS associations, supporting adoption of open approaches to help standardize CAD phenotyping and its wider potential value for reproducible research in other conditions.

Heart failure is a complex syndrome that is associated with many different underlying risk factors. Here, to increase power, the authors jointly analyse cases of heart failure of different aetiologies in a genome-wide association study and identify 11 loci of which ten had not been previously reported.

Abstract Background The use of metformin after acute myocardial infarction (AMI) has been associated with reduced mortality in people with type 2 diabetes mellitus (T2DM). However, it is not known if it is acutely cardioprotective in patients taking metformin at the time of AMI. We compared patient outcomes according to metformin status at the time of admission for fatal and non-fatal AMI in a large cohort of patients in England. Methods This study used linked data from primary care, hospital admissions and death registry from 4.7 million inhabitants in England, as part of the CALIBER resource. The primary endpoint was a composite of acute myocardial infarction requiring hospitalisation, stroke and cardiovascular death. The secondary endpoints were heart failure (HF) hospitalisation and all-cause mortality. Results 4,030 patients with T2DM and incident AMI recorded between January 1998 and October 2010 were included. At AMI admission, 63.9% of patients were receiving metformin and 36.1% another oral hypoglycaemic drug. Median follow-up was 343 (IQR: 1–1436) days. Adjusted analyses showed an increased hazard of the composite endpoint in metformin users compared to non-users (HR 1.09 [1.01–1.19]), but not of the secondary endpoints. The higher risk of the composite endpoint in metformin users was only observed in people taking metformin at AMI admission, whereas metformin use post-AMI was associated with a reduction in risk of all-cause mortality (0.76 [0.62–0.93], P = 0.009). Conclusions Our study suggests that metformin use at the time of first AMI is associated with increased risk of cardiovascular disease and death in patients with T2DM, while its use post-AMI might be beneficial. Further investigation in well-designed randomised controlled trials is indicated, especially in view of emerging evidence of cardioprotection from sodium-glucose co-transporter-2 (SGLT2) inhibitors.

Abstract Background Clinical guidelines and public health authorities lack recommendations on scalable approaches to defining and monitoring the occurrence and severity of bleeding in populations prescribed antithrombotic therapy. Methods We examined linked primary care, hospital admission and death registry electronic health records (CALIBER 1998–2010, England) of patients with newly diagnosed atrial fibrillation, acute myocardial infarction, unstable angina or stable angina with the aim to develop algorithms for bleeding events. Using the developed bleeding phenotypes, Kaplan-Meier plots were used to estimate the incidence of bleeding events and we used Cox regression models to assess the prognosis for all-cause mortality, atherothrombotic events and further bleeding. Results We present electronic health record phenotyping algorithms for bleeding based on bleeding diagnosis in primary or hospital care, symptoms, transfusion, surgical procedures and haemoglobin values. In validation of the phenotype, we estimated a positive predictive value of 0.88 (95% CI 0.64, 0.99) for hospitalised bleeding. Amongst 128,815 patients, 27,259 (21.2%) had at least 1 bleeding event, with 5-year risks of bleeding of 29.1%, 21.9%, 25.3% and 23.4% following diagnoses of atrial fibrillation, acute myocardial infarction, unstable angina and stable angina, respectively. Rates of hospitalised bleeding per 1000 patients more than doubled from 1.02 (95% CI 0.83, 1.22) in January 1998 to 2.68 (95% CI 2.49, 2.88) in December 2009 coinciding with the increased rates of antiplatelet and vitamin K antagonist prescribing. Patients with hospitalised bleeding and primary care bleeding, with or without markers of severity, were at increased risk of all-cause mortality and atherothrombotic events compared to those with no bleeding. For example, the hazard ratio for all-cause mortality was 1.98 (95% CI 1.86, 2.11) for primary care bleeding with markers of severity and 1.99 (95% CI 1.92, 2.05) for hospitalised bleeding without markers of severity, compared to patients with no bleeding. Conclusions Electronic health record bleeding phenotyping algorithms offer a scalable approach to monitoring bleeding in the population. Incidence of bleeding has doubled in incidence since 1998, affects one in four cardiovascular disease patients, and is associated with poor prognosis. Efforts are required to tackle this iatrogenic epidemic.

Abstract Background Free text in electronic health records (EHR) may contain additional phenotypic information beyond structured (coded) information. For major health events – heart attack and death – there is a lack of studies evaluating the extent to which free text in the primary care record might add information. Our objectives were to describe the contribution of free text in primary care to the recording of information about myocardial infarction (MI), including subtype, left ventricular function, laboratory results and symptoms; and recording of cause of death. We used the CALIBER EHR research platform which contains primary care data from the Clinical Practice Research Datalink (CPRD) linked to hospital admission data, the MINAP registry of acute coronary syndromes and the death registry. In CALIBER we randomly selected 2000 patients with MI and 1800 deaths. We implemented a rule-based natural language engine, the Freetext Matching Algorithm, on site at CPRD to analyse free text in the primary care record without raw data being released to researchers. We analysed text recorded within 90 days before or 90 days after the MI, and on or after the date of death. Results We extracted 10,927 diagnoses, 3658 test results, 3313 statements of negation, and 850 suspected diagnoses from the myocardial infarction patients. Inclusion of free text increased the recorded proportion of patients with chest pain in the week prior to MI from 19 to 27%, and differentiated between MI subtypes in a quarter more patients than structured data alone. Cause of death was incompletely recorded in primary care; in 36% the cause was in coded data and in 21% it was in free text. Only 47% of patients had exactly the same cause of death in primary care and the death registry, but this did not differ between coded and free text causes of death. Conclusions Among patients who suffer MI or die, unstructured free text in primary care records contains much information that is potentially useful for research such as symptoms, investigation results and specific diagnoses. Access to large scale unstructured data in electronic health records (millions of patients) might yield important insights.

Abstract Background We characterised the phenotypic consequence of genetic variation at the PCSK9 locus and compared findings with recent trials of pharmacological inhibitors of PCSK9. Methods Published and individual participant level data (300,000+ participants) were combined to construct a weighted PCSK9 gene-centric score (GS). Seventeen randomized placebo controlled PCSK9 inhibitor trials were included, providing data on 79,578 participants. Results were scaled to a one mmol/L lower LDL-C concentration. Results The PCSK9 GS (comprising 4 SNPs) associations with plasma lipid and apolipoprotein levels were consistent in direction with treatment effects. The GS odds ratio (OR) for myocardial infarction (MI) was 0.53 (95% CI 0.42; 0.68), compared to a PCSK9 inhibitor effect of 0.90 (95% CI 0.86; 0.93). For ischemic stroke ORs were 0.84 (95% CI 0.57; 1.22) for the GS, compared to 0.85 (95% CI 0.78; 0.93) in the drug trials. ORs with type 2 diabetes mellitus (T2DM) were 1.29 (95% CI 1.11; 1.50) for the GS, as compared to 1.00 (95% CI 0.96; 1.04) for incident T2DM in PCSK9 inhibitor trials. No genetic associations were observed for cancer, heart failure, atrial fibrillation, chronic obstructive pulmonary disease, or Alzheimer’s disease – outcomes for which large-scale trial data were unavailable. Conclusions Genetic variation at the PCSK9 locus recapitulates the effects of therapeutic inhibition of PCSK9 on major blood lipid fractions and MI. While indicating an increased risk of T2DM, no other possible safety concerns were shown; although precision was moderate.

Abstract Background COPD is a highly heterogeneous disease composed of different phenotypes with different aetiological and prognostic profiles and current classification systems do not fully capture this heterogeneity. In this study we sought to discover, describe and validate COPD subtypes using cluster analysis on data derived from electronic health records. Methods We applied two unsupervised learning algorithms (k-means and hierarchical clustering) in 30,961 current and former smokers diagnosed with COPD, using linked national structured electronic health records in England available through the CALIBER resource. We used 15 clinical features, including risk factors and comorbidities and performed dimensionality reduction using multiple correspondence analysis. We compared the association between cluster membership and COPD exacerbations and respiratory and cardiovascular death with 10,736 deaths recorded over 146,466 person-years of follow-up. We also implemented and tested a process to assign unseen patients into clusters using a decision tree classifier. Results We identified and characterized five COPD patient clusters with distinct patient characteristics with respect to demographics, comorbidities, risk of death and exacerbations. The four subgroups were associated with 1) anxiety/depression; 2) severe airflow obstruction and frailty; 3) cardiovascular disease and diabetes and 4) obesity/atopy. A fifth cluster was associated with low prevalence of most comorbid conditions. Conclusions COPD patients can be sub-classified into groups with differing risk factors, comorbidities, and prognosis, based on data included in their primary care records. The identified clusters confirm findings of previous clustering studies and draw attention to anxiety and depression as important drivers of the disease in young, female patients.

Integration of omics data remains a challenge. Here, the authors introduce iCell, a framework to integrate tissue-specific protein–protein interaction, co-expression and genetic interaction data, enabling identification of the most rewired genes in cancer, unidentifiable in individual data layers.

Objectives To estimate the impact of the COVID-19 pandemic on cancer care services and overall (direct and indirect) excess deaths in people with cancer.Methods We employed near real-time weekly data on cancer care to determine the adverse effect of the pandemic on cancer services. We also used these data, together with national death registrations until June 2020 to model deaths, in excess of background (pre-COVID-19) mortality, in people with cancer. Background mortality risks for 24 cancers with and without COVID-19-relevant comorbidities were obtained from population-based primary care cohort (Clinical Practice Research Datalink) on 3 862 012 adults in England.Results Declines in urgent referrals (median=−70.4%) and chemotherapy attendances (median=−41.5%) to a nadir (lowest point) in the pandemic were observed. By 31 May, these declines have only partially recovered; urgent referrals (median=−44.5%) and chemotherapy attendances (median=−31.2%). There were short-term excess death registrations for cancer (without COVID-19), with peak relative risk (RR) of 1.17 at week ending on 3 April. The peak RR for all-cause deaths was 2.1 from week ending on 17 April. Based on these findings and recent literature, we modelled 40% and 80% of cancer patients being affected by the pandemic in the long-term. At 40% affected, we estimated 1-year total (direct and indirect) excess deaths in people with cancer as between 7165 and 17 910, using RRs of 1.2 and 1.5, respectively, where 78% of excess deaths occured in patients with ≥1 comorbidity.Conclusions Dramatic reductions were detected in the demand for, and supply of, cancer services which have not fully recovered with lockdown easing. These may contribute, over a 1-year time horizon, to substantial excess mortality among people with cancer and multimorbidity. It is urgent to understand how the recovery of general practitioner, oncology and other hospital services might best mitigate these long-term excess mortality risks.

Objective To evaluate the extent to which the inter-institutional, inter-disciplinary mobilisation of data and skills in the Farr Institute contributed to establishing the emerging field of data science for health in the UK. Design and Outcome measures We evaluated evidence of six domains characterising a new field of science: • defining central scientific challenges, • demonstrating how the central challenges might be solved, • creating novel interactions among groups of scientists, • training new types of experts, • re-organising universities, • demonstrating impacts in society. We carried out citation, network and time trend analyses of publications, and a narrative review of infrastructure, methods and tools. Setting Four UK centres in London, North England, Scotland and Wales (23 university partners), 2013-2018. Results 1. The Farr Institute helped define a central scientific challenge publishing a research corpus, demonstrating insights from electronic health record (EHR) and administrative data at each stage of the translational cycle in 593 papers with at least one Farr Institute author affiliation on PubMed. 2. The Farr Institute offered some demonstrations of how these scientific challenges might be solved: it established the first four ISO27001 certified trusted research environments in the UK, and approved more than 1000 research users, published on 102 unique EHR and administrative data sources, although there was no clear evidence of an increase in novel, sustained record linkages. The Farr Institute established open platforms for the EHR phenotyping algorithms and validations (>70 diseases, CALIBER). Sample sizes showed some evidence of increase but remained less than 10% of the UK population in primary care-hospital care linked studies. 3.The Farr Institute created novel interactions among researchers: the co-author publication network expanded from 944 unique co-authors (based on 67 publications in the first 30 months) to 3839 unique co-authors (545 papers in the final 30 months). 4. Training expanded substantially with 3 new masters courses, training >400 people at masters, short-course and leadership level and 48 PhD students. 5. Universities reorganised with 4/5 Centres established 27 new faculty (tenured) positions, 3 new university institutes. 6. Emerging evidence of impacts included: > 3200 citations for the 10 most cited papers and Farr research informed eight practice-changing clinical guidelines and policies relevant to the health of millions of UK citizens. Conclusion The Farr Institute played a major role in establishing and growing the field of data science for health in the UK, with some initial evidence of benefits for health and healthcare. The Farr Institute has now expanded into Health Data Research (HDR) UK but key challenges remain including, how to network such activities internationally.

Background Patients presenting with atrial fibrillation (AF) often undergo a blood test to measure troponin, but interpretation of the result is impeded by uncertainty about its clinical importance. We investigated the relationship between troponin level, coronary angiography, and all‐cause mortality in real‐world patients presenting with AF. Methods and Results We used National Institute of Health Research Health Informatics Collaborative data to identify patients admitted between 2010 and 2017 at 5 tertiary centers in the United Kingdom with a primary diagnosis of AF. Peak troponin results were scaled as multiples of the upper limit of normal. A total of 3121 patients were included in the analysis. Over a median follow‐up of 1462 (interquartile range, 929–1975) days, there were 586 deaths (18.8%). The adjusted hazard ratio for mortality associated with a positive troponin (value above upper limit of normal) was 1.20 (95% CI, 1.01–1.43; P

ObjectivesWe aimed to model the impact of coronavirus (COVID-19) on the clinical academic response in England, and to provide recommendations for COVID-related research.DesignA stochastic model to determine clinical academic capacity in England, incorporating the following key factors which affect the ability to conduct research in the COVID-19 climate: (i) infection growth rate and population infection rate (from UK COVID-19 statistics and WHO); (ii) strain on the healthcare system (from published model); and (iii) availability of clinical academic staff with appropriate skillsets affected by frontline clinical activity and sickness (from UK statistics).SettingClinical academics in primary and secondary care in England.ParticipantsEquivalent of 3200 full-time clinical academics in England.InterventionsFour policy approaches to COVID-19 with differing population infection rates: "Italy model" (6%), "mitigation" (10%), "relaxed mitigation" (40%) and "do-nothing" (80%) scenarios. Low and high strain on the health system (no clinical academics able to do research at 10% and 5% infection rate, respectively.Main outcome measuresNumber of full-time clinical academics available to conduct clinical research during the pandemic in England.ResultsIn the "Italy model", "mitigation", "relaxed mitigation" and "do-nothing" scenarios, from 5 March 2020 the duration (days) and peak infection rates (%) are 95(2.4%), 115(2.5%), 240(5.3%) and 240(16.7%) respectively. Near complete attrition of academia (87% reduction, ConclusionsPandemic COVID-19 crushes the science needed at system level. National policies mitigate, but the academic community needs to adapt. We highlight six key strategies: radical prioritisation (eg 3-4 research ideas per institution), deep resourcing, non-standard leadership (repurposing of key non-frontline teams), rationalisation (profoundly simple approaches), careful site selection (eg protected sites with large academic backup) and complete suspension of academic competition with collaborative approaches.

Abstract Background The ability of external investigators to reproduce published scientific findings is critical for the evaluation and validation of biomedical research by the wider community. However, a substantial proportion of health research using electronic health records (EHR), data collected and generated during clinical care, is potentially not reproducible mainly due to the fact that the implementation details of most data preprocessing, cleaning, phenotyping and analysis approaches are not systematically made available or shared. With the complexity, volume and variety of electronic health record data sources made available for research steadily increasing, it is critical to ensure that scientific findings from EHR data are reproducible and replicable by researchers. Reporting guidelines, such as RECORD and STROBE, have set a solid foundation by recommending a series of items for researchers to include in their research outputs. Researchers however often lack the technical tools and methodological approaches to actuate such recommendations in an efficient and sustainable manner. Results In this paper, we review and propose a series of methods and tools utilized in adjunct scientific disciplines that can be used to enhance the reproducibility of research using electronic health records and enable researchers to report analytical approaches in a transparent manner. Specifically, we discuss the adoption of scientific software engineering principles and best-practices such as test-driven development, source code revision control systems, literate programming and the standardization and re-use of common data management and analytical approaches. Conclusion The adoption of such approaches will enable scientists to systematically document and share EHR analytical workflows and increase the reproducibility of biomedical research using such complex data sources.

Summary: Background: To effectively prevent, detect, and treat health conditions that affect people during their lifecourse, health-care professionals and researchers need to know which sections of the population are susceptible to which health conditions and at which ages. Hence, we aimed to map the course of human health by identifying the 50 most common health conditions in each decade of life and estimating the median age at first diagnosis. Methods: We developed phenotyping algorithms and codelists for physical and mental health conditions that involve intensive use of health-care resources. Individuals older than 1 year were included in the study if their primary-care and hospital-admission records met research standards set by the Clinical Practice Research Datalink and they had been registered in a general practice in England contributing up-to-standard data for at least 1 year during the study period. We used linked records of individuals from the CALIBER platform to calculate the sex-standardised cumulative incidence for these conditions by 10-year age groups between April 1, 2010, and March 31, 2015. We also derived the median age at diagnosis and prevalence estimates stratified by age, sex, and ethnicity (black, white, south Asian) over the study period from the primary-care and secondary-care records of patients. Findings: We developed case definitions for 308 disease phenotypes. We used records of 2 784 138 patients for the calculation of cumulative incidence and of 3 872 451 patients for the calculation of period prevalence and median age at diagnosis of these conditions. Conditions that first gained prominence at key stages of life were: atopic conditions and infections that led to hospital admission in children (

Background: In 2017, 15.6% of the people living in England were born abroad, yet we have a limited understanding of their use of health services and subsequent health conditions. This linked population-based cohort study aims to describe the hospital-based healthcare and mortality outcomes of 1.5 million non-European Union (EU) migrants and refugees in England. Methods and analysis: We will link four data sources: first, non-EU migrant tuberculosis pre-entry screening data; second, refugee pre-entry health assessment data; third, national hospital episode statistics; and fourth, Office of National Statistics death records. Using this linked dataset, we will then generate a population-based cohort to examine hospital-based events and mortality outcomes in England between Jan 1, 2006, and Dec 31, 2017. We will compare outcomes across three groups in our analyses: 1) non-EU international migrants, 2) refugees, and 3) general population of England. Ethics and dissemination: We will obtain approval to use unconsented patient identifiable data from the Secretary of State for Health through the Confidentiality Advisory Group and the National Health Service Research Ethics Committee. After data linkage, we will destroy identifying data and undertake all analyses using the pseudonymised dataset. The results will provide policy makers and civil society with detailed information about the health needs of non-EU international migrants and refugees in England.

Summary: Background: We aimed to estimate socioeconomic disparities in the incidence of hospitalisation for first-ever stroke, quality of care, and post-stroke survival for the adult population of England. Methods: In this cohort study, we obtained data collected by a nationwide register on patients aged 18 years or older hospitalised for first-ever acute ischaemic stroke or primary intracerebral haemorrhage in England from July 1, 2013, to March 31, 2016. We classified socioeconomic status at the level of Lower Super Output Areas using the Index of Multiple Deprivation, a neighbourhood measure of deprivation. Multivariable models were fitted to estimate the incidence of hospitalisation for first stroke (negative binomial), quality of care using 12 quality metrics (multilevel logistic), and all-cause 1 year case fatality (Cox proportional hazards). Findings: Of the 43·8 million adults in England, 145 324 were admitted to hospital with their first-ever stroke: 126 640 (87%) with ischaemic stroke, 17 233 (12%) with intracerebral haemorrhage, and 1451 (1%) with undetermined stroke type. We observed a socioeconomic gradient in the incidence of hospitalisation for ischaemic stroke (adjusted incidence rate ratio 2·0, 95% CI 1·7–2·3 for the most vs least deprived deciles) and intracerebral haemorrhage (1·6, 1·3–1·9). Patients from the lowest socioeconomic groups had first stroke a median of 7 years earlier than those from the highest (p

BackgroundThere is limited knowledge of the scale and impact of multimorbidity for patients who have had an acute myocardial infarction (AMI). Therefore, this study aimed to determine the extent to which multimorbidity is associated with long-term survival following AMI.Methods and findingsThis national observational study included 693,388 patients (median age 70.7 years, 452,896 [65.5%] male) from the Myocardial Ischaemia National Audit Project (England and Wales) who were admitted with AMI between 1 January 2003 and 30 June 2013. There were 412,809 (59.5%) patients with multimorbidity at the time of admission with AMI, i.e., having at least 1 of the following long-term health conditions: diabetes, chronic obstructive pulmonary disease or asthma, heart failure, renal failure, cerebrovascular disease, peripheral vascular disease, or hypertension. Those with heart failure, renal failure, or cerebrovascular disease had the worst outcomes (39.5 [95% CI 39.0-40.0], 38.2 [27.7-26.8], and 26.6 [25.2-26.4] deaths per 100 person-years, respectively). Latent class analysis revealed 3 multimorbidity phenotype clusters: (1) a high multimorbidity class, with concomitant heart failure, peripheral vascular disease, and hypertension, (2) a medium multimorbidity class, with peripheral vascular disease and hypertension, and (3) a low multimorbidity class. Patients in class 1 were less likely to receive pharmacological therapies compared with class 2 and 3 patients (including aspirin, 83.8% versus 87.3% and 87.2%, respectively; β-blockers, 74.0% versus 80.9% and 81.4%; and statins, 80.6% versus 85.9% and 85.2%). Flexible parametric survival modelling indicated that patients in class 1 and class 2 had a 2.4-fold (95% CI 2.3-2.5) and 1.5-fold (95% CI 1.4-1.5) increased risk of death and a loss in life expectancy of 2.89 and 1.52 years, respectively, compared with those in class 3 over the 8.4-year follow-up period. The study was limited to all-cause mortality due to the lack of available cause-specific mortality data. However, we isolated the disease-specific association with mortality by providing the loss in life expectancy following AMI according to multimorbidity phenotype cluster compared with the general age-, sex-, and year-matched population.ConclusionsMultimorbidity among patients with AMI was common, and conferred an accumulative increased risk of death. Three multimorbidity phenotype clusters that were significantly associated with loss in life expectancy were identified and should be a concomitant treatment target to improve cardiovascular outcomes.Trial registrationClinicalTrials.gov NCT03037255.

[This corrects the article DOI: 10.1371/journal.pone.0199026.].

BACKGROUND:The time a patient spends with blood pressure at target level is an intuitive measure of successful BP management, but population studies on its effectiveness are as yet unavailable. METHOD:We identified a population-based cohort of 169,082 individuals with newly identified high blood pressure who were free of cardiovascular disease from January 1997 to March 2010. We used 1.64 million clinical blood pressure readings to calculate the TIme at TaRgEt (TITRE) based on current target blood pressure levels. RESULT:The median (Inter-quartile range) TITRE among all patients was 2.8 (0.3, 5.6) months per year, only 1077 (0.6%) patients had a TITRE ≥11 months. Compared to people with a 0% TITRE, patients with a TITRE of 3-5.9 months, and 6-8.9 months had 75% and 78% lower odds of the composite of cardiovascular death, myocardial infarction and stroke (adjusted odds ratios, 0.25 (95% confidence interval: 0.21, 0.31) and 0.22 (0.17, 0.27), respectively). These associations were consistent for heart failure and any cardiovascular disease and death (comparing a 3-5.9 month to 0% TITRE, 63% and 60% lower in odds, respectively), among people who did or did not have blood pressure 'controlled' on a single occasion during the first year of follow-up, and across groups defined by number of follow-up BP measure categories. CONCLUSION:Based on the current frequency of measurement of blood pressure this study suggests that few newly hypertensive patients sustained a complete, year-round on target blood pressure over time. The inverse associations between a higher TITRE and lower risk of incident cardiovascular diseases were independent of widely-used blood pressure 'control' indicators. Randomized trials are required to evaluate interventions to increase a person's time spent at blood pressure target.

To evaluate unmet clinical need in unselected hypertrophic cardiomyopathy (HCM) patients to determine the risk of a wide range of subsequent cardiovascular disease endpoints and safety endpoints relevant for trial design.Population based cohort (CALIBER, linked primary care, hospital and mortality records in England, period 1997-2010), all people diagnosed with HCM were identified and matched by age, sex and general practice with ten randomly selected people without HCM. Random-effects Poisson models were used to assess the associations between HCM and cardiovascular diseases and bleeding.Among 3,290,455 eligible people a diagnosis of hypertrophic cardiomyopathy was found in 4 per 10,000. Forty-one percent of the 1,160 individuals with hypertrophic cardiomyopathy were women and the median age was 57 years. The median follow-up was 4.0 years. Compared to general population controls, people with HCM had higher risk of ventricular arrhythmia (incidence rate ratio = 23.53, [95% confidence interval 12.67-43.72]), cardiac arrest or sudden cardiac death (6.33 [3.69-10.85]), heart failure (4.31, [3.30-5.62]), and atrial fibrillation (3.80 [3.04-4.75]). HCM was also associated with a higher incidence of myocardial infarction ([MI] 1.90 [1.27-2.84]) and coronary revascularisation (2.32 [1.46-3.69]).The absolute Kaplan-Meier risks at 3 years were 8.8% for the composite endpoint of cardiovascular death or heart failure, 8.4% for the composite of cardiovascular death, stroke or myocardial infarction, and 1.5% for major bleeding.Our study identified major unmet need in HCM and highlighted the importance of implementing improved cardiovascular prevention strategies to increase life-expectancy of the contemporary HCM population. They also show that national electronic health records provide an effective method for identifying outcomes and clinically relevant estimates of composite efficacy and safety endpoints essential for trial design in rare diseases.

OBJECTIVES:Electronic health records (EHR) might be a useful resource to study the risk factors and clinical care of people with dementia. We sought to determine the diagnostic validity of dementia captured in linked EHR. METHODS AND FINDINGS:A cohort of adults in linked primary care, hospital, disease registry and mortality records in England, [CALIBER (CArdiovascular disease research using LInked Bespoke studies and Electronic health Records)]. The proportion of individuals with dementia, Alzheimer's disease, vascular and rare dementia in each data source was determined. A comparison was made of symptoms and care between people with dementia and age-, sex- and general practice-matched controls, using conditional logistic regression. The lifetime risk and prevalence of dementia and mortality rates in people with and without dementia were estimated with random-effects Poisson models. There were 47,386 people with dementia: 12,633 with Alzheimer's disease, 9540 with vascular and 1539 with rare dementia. Seventy-four percent of cases had corroborating evidence of dementia. People with dementia were more likely to live in a deprived area (conditional OR 1.26;95%CI:1.20-1.31 most vs least deprived), have documented memory impairment (cOR = 11.97;95%CI:11.24-12.75), falls (cOR = 2.36;95%CI:2.31-2.41), depression (cOR = 2.03; 95%CI:1.98-2.09) or anxiety (cOR = 1.27; 95%CI:1.23-1.32). The lifetime risk of dementia at age 65 was 9.2% (95%CI:9.0%-9.4%), in men and 14.9% (95%CI:14.7%-15.1%) in women. The population prevalence of recorded dementia increased from 0.3% in 2000 to 0.7% in 2010. A higher mortality rate was observed in people with than without dementia (IRR = 1.56;95%CI:1.54-1.58). CONCLUSIONS:Most people with a record of dementia in linked UK EHR had some corroborating evidence for diagnosis. The estimated 10-year risk of dementia was higher than published population-based estimations. EHR are therefore a promising source of data for dementia research.

Prognostic modelling is important in clinical practice and epidemiology for patient management and research. Electronic health records (EHR) provide large quantities of data for such models, but conventional epidemiological approaches require significant researcher time to implement. Expert selection of variables, fine-tuning of variable transformations and interactions, and imputing missing values are time-consuming and could bias subsequent analysis, particularly given that missingness in EHR is both high, and may carry meaning. Using a cohort of 80,000 patients from the CALIBER programme, we compared traditional modelling and machine-learning approaches in EHR. First, we used Cox models and random survival forests with and without imputation on 27 expert-selected, preprocessed variables to predict all-cause mortality. We then used Cox models, random forests and elastic net regression on an extended dataset with 586 variables to build prognostic models and identify novel prognostic factors without prior expert input. We observed that data-driven models used on an extended dataset can outperform conventional models for prognosis, without data preprocessing or imputing missing values. An elastic net Cox regression based with 586 unimputed variables with continuous values discretised achieved a C-index of 0.801 (bootstrapped 95% CI 0.799 to 0.802), compared to 0.793 (0.791 to 0.794) for a traditional Cox model comprising 27 expert-selected variables with imputation for missing values. We also found that data-driven models allow identification of novel prognostic variables; that the absence of values for particular variables carries meaning, and can have significant implications for prognosis; and that variables often have a nonlinear association with mortality, which discretised Cox models and random forests can elucidate. This demonstrates that machine-learning approaches applied to raw EHR data can be used to build models for use in research and clinical practice, and identify novel predictive variables and their effects to inform future research.

ABSTRACT Objectives We used data from UK-Biobank that were linked with Hospital Episode Statistics and Office for National Statistics to assess the relationship between low-density lipoprotein cholesterol (LDL-C) and atrial fibrillation (AF). In this study, we applied Mendelian randomization in order to find out whether there is a causal effect of LDL-C to AF. Approach We used data from the UK Biobank (~500,000 subjects) which is linked with electronic health records. At baseline (2006-2010), participants from across the UK took part in this project. They have undergone measures, provided blood, urine and saliva samples for future analysis, detailed information about themselves and agreed to have their health followed. Information in relation to the development of atrial fibrillation was derived from a) the enrollment of the participants (self-reported events), b) their hospitalization before and after their recruitment to UK-Biobank (confirmed events from Hospital Episode Statistics) and c) the death certificates [confirmed events from Office for National Statistics]. We also used genetic data from the analyses of the participants’ blood sample that have been stored. We used Mendelian randomization to capture the effect of LDL-C to AF. As instruments, we used a genetic predisposition risk score (GPRs) for LDL-C, which was created as a weighted sum of the 18 most significant SNPs related to LDL-C, as there were documented in Global Lipid Consortium, in 18 out of 22 chromosomes. We ran a logistic regression model, using AF as outcome and GPRs as exposure. Results Our final sample consisted of 144,092 individuals, for which we have valid information for their genetic data. The AF cases in this sample were 3207, most of which were identified from Hospital Episode Statistics (hospitalization of the participants). From the Mendelian randomization study, from our preliminary results, we found a weak positive relationship between GPRs and AF, when we did not adjust for any covariate [OR per one unit increase of GPRs=1.08, 95% CI= (0.95-1.22)] and results remained practically the same when we adjusted for age and sex [OR=1.09, 95% CI= (0.96-1.24 )]. Conclusion We observed a weak positive association between LDL-C and AF in this study. This is the first Mendelian randomization approach that focuses on this relationship. More Mendelian randomization studies should be performed in order to identify the causal effect of LDL-C to AF. The use of electronic health records will facilitate the conduction of similar studies.

ABSTRACT Aims We aimed (i) to create algorithms for identifying stroke and acute myocardial infarction (MI) cases in a large prospective population-based study (UK Biobank), using linked data from national hospital admissions data and death registers, combined with self-report data from the baseline assessment; and (ii) to assess validity by examining associations with risk factors and mortality. Methods UK Biobank is a prospective study of 503,000 participants, aged 40-69 years when recruited in 2006-2010 from centres across England, Scotland and Wales. Participants provided extensive questionnaire data on lifestyle, environment and medical history (with confirmation of self-reported medical conditions during a brief interview with a trained research nurse), had physical measures, and provided biological samples. Follow-up is principally through linkages to national health-related datasets, integrated from different data providers for each country and including cohort-wide data for ICD-coded hospital admissions and registered deaths, with follow-up to 2011 for this report. We used expert opinion and systematic reviews to identify baseline self-report items and ICD codes maximising positive predictive value for identification of stroke, MI and their subtypes. We classified participants with at least one relevant code as being first affected either before (prevalent cases) or after recruitment (incident cases). We compared cases with non-cases (controls) using logistic regression (prevalent cases) and survival analysis (incident cases) to examine associations with vascular risk factors, defined using data from the baseline assessment. We used survival analysis to compare vascular, non-vascular and all-cause mortality for cases versus controls post recruitment (prevalent cases) and post hospitalization (incident cases). Results We identified 8654 stroke cases and 13,479 MI cases. 90% of both were prevalent at recruitment, of which 29% (stroke) and 54% (MI) were identified through both research nurse-confirmed self-report and hospital admission records. During ≈1 million person years of follow-up in those without a prevalent record, we identified 871 incident strokes and 1387 incident MIs, of which 8% (stroke) and 2.9% (MI) were identified in both hospital and death records. Male sex, low socio-economic status, smoking and increased body mass index were all positively associated with both stroke and MI. Compared with age and sex-matched controls, mortality for stroke and MI cases was increased both after recruitment (prevalent cases) and after hospitalization (incident cases). Conclusion Information from linked coded datasets from different data providers can be combined with information collected at recruitment to identify and validate prevalent and incident acute MI and stroke.

ABSTRACT Objectives Electronic health records (EHR) across primary, secondary, and tertiary care are increasingly being linked for research at a population level. The increasing volume, variety, velocity, and veracity of big biomedical data makes research reproducibility challenging. Research reproducibility and replicability is essential for the external validity and generalizability of scientific findings and the lack of standardized approaches and tools and relative opaqueness of data manipulation methods is detrimental to their integrity. The objective of this study was to explore, evaluate and propose methods, tools and approaches for addressing some of the challenges associated with reproducibility when using linked national electronic health records for research. Approach We systematically searched literature and internet resources for well-established and appropriate methods, tools, and approaches used in related scientific disciplines. The identified techniques were systematically evaluated in terms of their capacity to facilitate reproducible research in routinely collected health data across the life course of a research project: from protocol creation and raw data curation to data transformation and statistical analysis though to finding dissemination and impact. Most importantly, the identified techniques were tested and applied in a contemporary database of linked electronic health records. CALIBER is a research data platform of linked national electronic health records from primary care (Clinical Practice Research Datalink), secondary care (Hospital Episode Statistics), acute coronary syndrome disease registry (Myocardial Ischaemia National Audit Project) and cause-specific mortality (Office for National Statistics) for roughly 2 million adults. Results Firstly, we present the review of methods and approaches which we identified through our search. Secondly, we propose a set of recommendations for applying them within the context of research projects making use of linked routinely collected health data. Focal interests included: a) documentation of data (attributes, relationships, and interpretation), b) data processing (source code, instructions, and parameters), c) results (visualizations, figures), and any supplementary material. Thirdly, we present approaches around a) raw data curation using international metadata standards, b) study protocol encoding, c) provenance and sharing of data transformation and statistical analysis operations, d) public and private data retention, and e) computable EHR-driven phenotypes. Conclusion The complexity and size of routinely collected health data is increasing through linkages across distributed data sources. The scientific community benefits from findings which can be replicated. This study presents a number of methods, tools and approaches across the project life course for ensuring that their research studies are reproducible and replicable from the wider scientific community.

While the association of ethnic group with individual cardiovascular diseases has been studied, little is known about ethnic differences in the initial lifetime presentation of clinical cardiovascular disease in contemporary populations.We studied 1,068,318 people, aged ≥30 years and free from diagnosed CVD at baseline (90.9% White, 3.6% South Asian and 2.9% Black), using English linked electronic health records covering primary care, hospital admissions, acute coronary syndrome registry and mortality registry (CALIBER platform). During 5.7 years median follow-up between 1997-2010, 95,224 people experienced an incident cardiovascular diagnosis. 69.9% (67.2%-72.4%) of initial presentation in South Asian

Background: Electronic health records (EHRs), when linked across primary and secondary care and curated for research use, have the potential to improve our understanding of care quality and outcomes. Objective: To evaluate new opportunities arising from linked EHRs for improving quality of care and outcomes for patients at risk of or with coronary disease across the patient journey. Design: Epidemiological cohort, health informatics, health economics and ethnographic approaches were used. Setting: 230 NHS hospitals and 226 general practices in England and Wales. Participants: Up to 2 million initially healthy adults, 100,000 people with stable coronary artery disease (SCAD) and up to 300,000 patients with acute coronary syndrome. Main outcome measures: Quality of care, fatal and non-fatal cardiovascular disease (CVD) events. Data platform and methods: We created a novel research platform [ClinicAl disease research using LInked Bespoke studies and Electronic health Records (CALIBER)] based on linkage of four major sources of EHR data in primary care and national registries. We carried out 33 complementary studies within the CALIBER framework. We developed a web-based clinical decision support system (CDSS) in hospital chest pain clinics. We established a novel consented prognostic clinical cohort of SCAD patients. Results: CALIBER was successfully established as a valid research platform based on linked EHR data in nearly 2 million adults with > 600 EHR phenotypes implemented on the web portal (see https://caliberresearch.org/portal). Despite national guidance, key opportunities for investigation and treatment were missed across the patient journey, resulting in a worse prognosis for patients in the UK compared with patients in health systems in other countries. Our novel, contemporary, high-resolution studies showed heterogeneous associations for CVD risk factors across CVDs. The CDSS did not alter the decision-making behaviour of clinicians in chest pain clinics. Prognostic models using real-world data validly discriminated risk of death and events, and were used in cost-effectiveness decision models. Conclusions: Emerging ‘big data’ opportunities arising from the linkage of records at different stages of a patient’s journey are vital to the generation of actionable insights into the diagnosis, risk stratification and cost-effective treatment of people at risk of, or with, CVD. Future work: The vast majority of NHS data remain inaccessible to research and this hampers efforts to improve efficiency and quality of care and to drive innovation. We propose three priority directions for further research. First, there is an urgent need to ‘unlock’ more detailed data within hospitals for the scale of the UK’s 65 million population. Second, there is a need for scaled approaches to using EHRs to design and carry out trials, and interpret the implementation of trial results. Third, large-scale, disease agnostic genetic and biological collections linked to such EHRs are required in order to deliver precision medicine and to innovate discovery. Study registration: CALIBER studies are registered as follows: study 2 – NCT01569139, study 4 – NCT02176174 and NCT01164371, study 5 – NCT01163513, studies 6 and 7 – NCT01804439, study 8 – NCT02285322, and studies 26–29 – NCT01162187. Optimising the Management of Angina is registered as Current Controlled Trials ISRCTN54381840. Funding: The National Institute for Health Research (NIHR) Programme Grants for Applied Research programme (RP-PG-0407-10314) (all 33 studies) and additional funding from the Wellcome Trust (study 1), Medical Research Council Partnership grant (study 3), Servier (study 16), NIHR Research Methods Fellowship funding (study 19) and NIHR Research for Patient Benefit (study 33).

The original version of this Article contained an error in the spelling of the author Harry Hemingway, which was incorrectly given as Harry Hemmingway. This has been corrected in both the PDF and HTML versions of the Article.