Clinical Practice Research Datalink

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection is associated with a range of persistent symptoms impacting everyday functioning, known as post-COVID-19 condition or long COVID. We undertook a retrospective matched cohort study using a UK-based primary care database, Clinical Practice Research Datalink Aurum, to determine symptoms that are associated with confirmed SARS-CoV-2 infection beyond 12 weeks in non-hospitalized adults and the risk factors associated with developing persistent symptoms. We selected 486,149 adults with confirmed SARS-CoV-2 infection and 1,944,580 propensity score-matched adults with no recorded evidence of SARS-CoV-2 infection. Outcomes included 115 individual symptoms, as well as long COVID, defined as a composite outcome of 33 symptoms by the World Health Organization clinical case definition. Cox proportional hazards models were used to estimate adjusted hazard ratios (aHRs) for the outcomes. A total of 62 symptoms were significantly associated with SARS-CoV-2 infection after 12 weeks. The largest aHRs were for anosmia (aHR 6.49, 95% CI 5.02-8.39), hair loss (3.99, 3.63-4.39), sneezing (2.77, 1.40-5.50), ejaculation difficulty (2.63, 1.61-4.28) and reduced libido (2.36, 1.61-3.47). Among the cohort of patients infected with SARS-CoV-2, risk factors for long COVID included female sex, belonging to an ethnic minority, socioeconomic deprivation, smoking, obesity and a wide range of comorbidities. The risk of developing long COVID was also found to be increased along a gradient of decreasing age. SARS-CoV-2 infection is associated with a plethora of symptoms that are associated with a range of sociodemographic and clinical risk factors.

Clinical Practice Research Datalink (CPRD) is a UK government, not-for-profit research service that has been supplying anonymized primary care data for public health research for more than 30 years. In October 2017 CPRD launched a new data resource called CPRD Aurum. CPRD Aurum is a database containing routinely collected data from primary care practices in England, capturing diagnoses, symptoms, prescriptions, referrals and tests for over 19 million patients as of September 2018 (Figure Primary care data in CPRD Aurum have been linked to national secondary care databases as well as deprivation and death registration data (Table

Several recent studies suggest that obesity may be a risk factor for fracture. The aim of this study was to investigate the association between body mass index (BMI) and future fracture risk at different skeletal sites. In prospective cohorts from more than 25 countries, baseline data on BMI were available in 398,610 women with an average age of 63 (range, 20-105) years and follow up of 2.2 million person-years during which 30,280 osteoporotic fractures (6457 hip fractures) occurred. Femoral neck BMD was measured in 108,267 of these women. Obesity (BMI ≥ 30 kg/m(2) ) was present in 22%. A majority of osteoporotic fractures (81%) and hip fractures (87%) arose in non-obese women. Compared to a BMI of 25 kg/m(2) , the hazard ratio (HR) for osteoporotic fracture at a BMI of 35 kg/m(2) was 0.87 (95% confidence interval [CI], 0.85-0.90). When adjusted for bone mineral density (BMD), however, the same comparison showed that the HR for osteoporotic fracture was increased (HR, 1.16; 95% CI, 1.09-1.23). Low BMI is a risk factor for hip and all osteoporotic fracture, but is a protective factor for lower leg fracture, whereas high BMI is a risk factor for upper arm (humerus and elbow) fracture. When adjusted for BMD, low BMI remained a risk factor for hip fracture but was protective for osteoporotic fracture, tibia and fibula fracture, distal forearm fracture, and upper arm fracture. When adjusted for BMD, high BMI remained a risk factor for upper arm fracture but was also a risk factor for all osteoporotic fractures. The association between BMI and fracture risk is complex, differs across skeletal sites, and is modified by the interaction between BMI and BMD. At a population level, high BMI remains a protective factor for most sites of fragility fracture. The contribution of increasing population rates of obesity to apparent decreases in fracture rates should be explored.

Machine learning, artificial intelligence, and other modern statistical methods are providing new opportunities to operationalise previously untapped and rapidly growing sources of data for patient benefit. Despite much promising research currently being undertaken, particularly in imaging, the literature as a whole lacks transparency, clear reporting to facilitate replicability, exploration for potential ethical concerns, and clear demonstrations of effectiveness. Among the many reasons why these problems exist, one of the most important (for which we provide a preliminary solution here) is the current lack of best practice guidance specific to machine learning and artificial intelligence. However, we believe that interdisciplinary groups pursuing research and impact projects involving machine learning and artificial intelligence for health would benefit from explicitly addressing a series of questions concerning transparency, reproducibility, ethics, and effectiveness (TREE). The 20 critical questions proposed here provide a framework for research groups to inform the design, conduct, and reporting; for editors and peer reviewers to evaluate contributions to the literature; and for patients, clinicians and policy makers to critically appraise where new findings may deliver patient benefit.

Background: We aimed to systematically review the evidence on adverse mental health outcomes in breast cancer survivors (≥1 year) compared with women with no history of cancer. Methods: Studies were identified by searching MEDLINE, PsycINFO, the Cumulative Index to Nursing and Allied Health Literature, and the Social Sciences Citation Index, and through backward citation tracking. Two researchers selected the studies, extracted data, and assessed the risk of bias. Results: Sixty studies were included. Of 38 studies of depression, 33 observed more depression in breast cancer survivors; this was statistically significant in 19 studies overall, including six of seven where depression was ascertained clinically, three of four studies of antidepressants, and 13 of 31 that quantified depressive symptoms. Of 21 studies of anxiety, 17 observed more anxiety in breast cancer survivors, statistically significant in 11 studies overall, including two of four with clinical/prescription-based outcomes, and in eight of 17 of anxiety symptoms. Breast cancer survivors also had statistically significantly increased symptoms/frequency of neurocognitive dysfunction (18 of 24 studies), sexual dysfunctions (5 of 6 studies), sleep disturbance (5 of 5 studies), stress-related disorders/PTSD (2 of 3 studies), suicide (2 of 2 studies), somatisation (2 of 2 studies), and bipolar and obsessive-compulsive disorders (1 of 1 study each). Studies were heterogeneous in terms of participants' characteristics, time since diagnosis, ascertainment of outcomes, and measures reported. Approximately one-half of the studies were at high risk of selection bias and confounding by socio-economic status. Conclusions: There is compelling evidence of an increased risk of anxiety, depression and suicide, and neurocognitive and sexual dysfunctions in breast cancer survivors compared with women with no prior cancer. This information can be used to support evidence-based prevention and management strategies. Further population-based and longitudinal research would help to better characterize these associations.

OBJECTIVE: To determine the completeness and diagnostic validity of myocardial infarction recording across four national health record sources in primary care, hospital care, a disease registry, and mortality register. DESIGN: Cohort study. PARTICIPANTS: 21 482 patients with acute myocardial infarction in England between January 2003 and March 2009, identified in four prospectively collected, linked electronic health record sources: Clinical Practice Research Datalink (primary care data), Hospital Episode Statistics (hospital admissions), the disease registry MINAP (Myocardial Ischaemia National Audit Project), and the Office for National Statistics mortality register (cause specific mortality data). SETTING: One country (England) with one health system (the National Health Service). MAIN OUTCOME MEASURES: Recording of acute myocardial infarction, incidence, all cause mortality within one year of acute myocardial infarction, and diagnostic validity of acute myocardial infarction compared with electrocardiographic and troponin findings in the disease registry (gold standard). RESULTS: Risk factors and non-cardiovascular coexisting conditions were similar across patients identified in primary care, hospital admission, and registry sources. Immediate all cause mortality was highest among patients with acute myocardial infarction recorded in primary care, which (unlike hospital admission and disease registry sources) included patients who did not reach hospital, but at one year mortality rates in cohorts from each source were similar. 5561 (31.0%) patients with non-fatal acute myocardial infarction were recorded in all three sources and 11 482 (63.9%) in at least two sources. The crude incidence of acute myocardial infarction was underestimated by 25-50% using one source compared with using all three sources. Compared with acute myocardial infarction defined in the disease registry, the positive predictive value of acute myocardial infarction recorded in primary care was 92.2% (95% confidence interval 91.6% to 92.8%) and in hospital admissions was 91.5% (90.8% to 92.1%). CONCLUSION: Each data source missed a substantial proportion (25-50%) of myocardial infarction events. Failure to use linked electronic health records from primary care, hospital care, disease registry, and death certificates may lead to biased estimates of the incidence and outcome of myocardial infarction. TRIAL REGISTRATION: NCT01569139 clinicaltrials.gov.

RATIONALE: Cohort evidence linking long-term exposure to outdoor particulate air pollution and mortality has come largely from the United States. There is relatively little evidence from nationally representative cohorts in other countries. OBJECTIVES: To investigate the relationship between long-term exposure to a range of pollutants and causes of death in a national English cohort. METHODS: A total of 835,607 patients aged 40-89 years registered with 205 general practices were followed from 2003-2007. Annual average concentrations in 2002 for particulate matter with a median aerodynamic diameter less than 10 (PM(10)) and less than 2.5 μm (PM(2.5)), nitrogen dioxide (NO(2)), ozone, and sulfur dioxide (SO(2)) at 1 km(2) resolution, estimated from emission-based models, were linked to residential postcode. Deaths (n = 83,103) were ascertained from linkage to death certificates, and hazard ratios (HRs) for all- and cause-specific mortality for pollutants were estimated for interquartile pollutant changes from Cox models adjusting for age, sex, smoking, body mass index, and area-level socioeconomic status markers. MEASUREMENTS AND MAIN RESULTS: Residential concentrations of all pollutants except ozone were positively associated with all-cause mortality (HR, 1.02, 1.03, and 1.04 for PM(2.5), NO(2), and SO(2), respectively). Associations for PM(2.5), NO(2), and SO(2) were larger for respiratory deaths (HR, 1.09 each) and lung cancer (HR, 1.02, 1.06, and 1.05) but nearer unity for cardiovascular deaths (1.00, 1.00, and 1.04). CONCLUSIONS: These results strengthen the evidence linking long-term ambient air pollution exposure to increased all-cause mortality. However, the stronger associations with respiratory mortality are not consistent with most US studies in which associations with cardiovascular causes of death tend to predominate.

BACKGROUND: Evidence based largely on US cohorts suggests that long-term exposure to fine particulate matter is associated with cardiovascular mortality. There is less evidence for other pollutants and for cardiovascular morbidity. By using a cohort of 836,557 patients age 40 to 89 years registered with 205 English general practices in 2003, we investigated relationships between ambient outdoor air pollution and incident myocardial infarction, stroke, arrhythmia, and heart failure over a 5-year period. METHODS: Events were identified from primary care records, hospital admissions, and death certificates. Annual average concentrations in 2002 for particulate matter with a median aerodynamic diameter <10 (PM10) and <2.5 microns, nitrogen dioxide (NO2), ozone, and sulfur dioxide at a 1 × 1 km resolution were derived from emission-based models and linked to residential postcode. Analyses were performed using Cox proportional hazards models adjusting for relevant confounders, including social and economic deprivation and smoking. RESULTS: While evidence was weak for relationships with myocardial infarction, stroke, or arrhythmia, we found consistent associations between pollutant concentrations and incident cases of heart failure. An interquartile range change in PM10 and in NO2 (3.0 and 10.7 µg/m, respectively) both produced a hazard ratio of 1.06 (95% confidence interval = 1.01-1.11) after adjustment for confounders. There was some evidence that these effects were greater in more affluent areas. CONCLUSIONS: This study of an English national cohort found evidence linking long-term exposure to particulate matter and NO2 with the development of heart failure. We did not, however, replicate associations for other cardiovascular outcomes that have been reported elsewhere.

OBJECTIVES: Overutilisation of antibiotics may contribute to the emergence of antimicrobial drug resistance, a growing international concern. This study aimed to analyse the performance of UK general practices with respect to antibiotic prescribing for respiratory tract infections (RTIs) among young and middle-aged adults. SETTING: Data are reported for 568 UK general practices contributing to the Clinical Practice Research Datalink. PARTICIPANTS: Participants were adults aged 18-59 years. Consultations were identified for acute upper RTIs including colds, cough, otitis-media, rhino-sinusitis and sore throat. PRIMARY AND SECONDARY OUTCOME MEASURES: For each consultation, we identified whether an antibiotic was prescribed. The proportion of RTI consultations with antibiotics prescribed was estimated. RESULTS: There were 568 general practices analysed. The median general practice prescribed antibiotics at 54% of RTI consultations. At the highest prescribing 10% of practices, antibiotics were prescribed at 69% of RTI consultations. At the lowest prescribing 10% of practices, antibiotics were prescribed at 39% RTI consultations. The median practice prescribed antibiotics at 38% of consultations for 'colds and upper RTIs', 48% for 'cough and bronchitis', 60% for 'sore throat', 60% for 'otitis-media' and 91% for 'rhino-sinusitis'. The highest prescribing 10% of practices issued antibiotic prescriptions at 72% of consultations for 'colds', 67% for 'cough', 78% for 'sore throat', 90% for 'otitis-media' and 100% for 'rhino-sinusitis'. CONCLUSIONS: Most UK general practices prescribe antibiotics to young and middle-aged adults with respiratory infections at rates that are considerably in excess of what is clinically justified. This will fuel antibiotic resistance.

BACKGROUND: Several case reports of the onset or exacerbation of multiple sclerosis or other demyelinating conditions shortly after vaccination have suggested that vaccines may increase the risk of demyelinating diseases. OBJECTIVE: To evaluate the association between vaccination and onset of multiple sclerosis or optic neuritis. DESIGN: Case-control study involving cases of multiple sclerosis or optic neuritis among adults 18 to 49 years of age. Data on vaccinations and other risk factors were obtained from computerized and paper medical records and from telephone interviews. SETTING: Three health maintenance organizations. PARTICIPANTS: Four hundred forty case subjects and 950 control subjects matched on health maintenance organization, sex, and date of birth. INTERVENTIONS: None. MAIN OUTCOME MEASURES: Onset of first symptoms of demyelinating disease at any time after vaccination and during specified intervals after vaccination (<1 year, 1-5 years, and >5 years). RESULTS: Cases and controls had similar vaccination histories. The odds ratios (95% confidence intervals), adjusted for potential confounding variables, of the associations between ever having been vaccinated and risk of demyelinating disease (multiple sclerosis and optic neuritis combined) were 0.9 (0.6-1.5) for hepatitis B vaccine; 0.6 (0.4-0.8) for tetanus vaccination; 0.8 (0.6-1.2) for influenza vaccine; 0.8 (0.5-1.5) for measles, mumps, rubella vaccine; 0.9 (0.5-1.4) for measles vaccine; and 0.7 (0.4-1.0) for rubella vaccine. The results were similar when multiple sclerosis and optic neuritis were analyzed separately. There was no increased risk according to timing of vaccination. CONCLUSION: Vaccination against hepatitis B, influenza, tetanus, measles, or rubella is not associated with an increased risk of multiple sclerosis or optic neuritis.

OBJECTIVES: The optimal method of identifying people with chronic obstructive pulmonary disease (COPD) from electronic primary care records is not known. We assessed the accuracy of different approaches using the Clinical Practice Research Datalink, a UK electronic health record database. SETTING: 951 participants registered with a CPRD practice in the UK between 1 January 2004 and 31 December 2012. Individuals were selected for ≥1 of 8 algorithms to identify people with COPD. General practitioners were sent a brief questionnaire and additional evidence to support a COPD diagnosis was requested. All information received was reviewed independently by two respiratory physicians whose opinion was taken as the gold standard. PRIMARY OUTCOME MEASURE: The primary measure of accuracy was the positive predictive value (PPV), the proportion of people identified by each algorithm for whom COPD was confirmed. RESULTS: 951 questionnaires were sent and 738 (78%) returned. After quality control, 696 (73.2%) patients were included in the final analysis. All four algorithms including a specific COPD diagnostic code performed well. Using a diagnostic code alone, the PPV was 86.5% (77.5-92.3%) while requiring a diagnosis plus spirometry plus specific medication; the PPV was slightly higher at 89.4% (80.7-94.5%) but reduced case numbers by 10%. Algorithms without specific diagnostic codes had low PPVs (range 12.2-44.4%). CONCLUSIONS: Patients with COPD can be accurately identified from UK primary care records using specific diagnostic codes. Requiring spirometry or COPD medications only marginally improved accuracy. The high accuracy applies since the introduction of an incentivised disease register for COPD as part of Quality and Outcomes Framework in 2004.

BACKGROUND: People with intellectual disabilities (IDs) die at younger ages than the general population, but nationally representative and internationally comparable mortality data about people with ID, quantifying the extent and pattern of the excess, have not previously been reported for England. METHOD: We used data from the Clinical Practice Research Datalink database for April 2010 to March 2014 (CPRD GOLD September 2015). This source covered several hundred participating general practices comprising roughly 5% of the population of England in the period studied. General practitioner (GP) records identified people diagnosed by their GP as having ID. Linked national death certification data allowed us to derive corresponding mortality data for people with and without ID, overall and by cause. RESULTS: Mortality rates for people with ID were significantly higher than for those without. Their all-cause standardised mortality ratio was 3.18. Their life expectancy at birth was 19.7 years lower than for people without ID. Circulatory and respiratory diseases and neoplasms were the three most common causes of death for them. Cerebrovascular disease, thrombophlebitis and pulmonary embolism all had standardised mortality ratios greater than 3 in people with ID. This has not been described before. Other potentially avoidable causes included epilepsy (3.9% of deaths), aspiration pneumonitis (3.6%) and colorectal cancer (2.4%). Avoidable mortality analysis showed a higher proportion of deaths from causes classified as amenable to good medical care but a lower proportion from preventable causes compared with people without ID. International comparison to areas for which data have been published in sufficient detail for calculation of directly standardised rates suggest England may have higher death rates for people with ID than areas in Canada and Finland, and lower death rates than Ireland or the State of Massachusetts in the USA. CONCLUSIONS: National data about mortality in people with ID provides a basis for public health interventions. Linked data using GP records to identify people with ID could provide comprehensive population-based monitoring in England, unbiased by the circumstances of illnesses or death; to date information governance constraints have prevented this. However, GPs in England currently identify only around 0.5% of the population as having ID, suggesting that individuals with mild, non-syndromic ID are largely missed. Notably common causes of death suggest control of cardiovascular risk factors, epilepsy and dysphagia, management of thrombotic risks and colorectal screening are important areas for health promotion initiatives.

INTRODUCTION: Recent developments in the care of patients with idiopathic pulmonary fibrosis have the potential to improve survival rates. Population-based estimates of the current disease burden are needed to evaluate the future impact of newly approved therapies. The objective of this study is to describe incidence, prevalence, and survival of idiopathic pulmonary fibrosis patients in the UK. METHODS: Between 2000 and 2012, a patient cohort (N = 9,748,108), identified from Clinical Practice Research Datalink primary care data, was used to identify incident and prevalent cases of idiopathic pulmonary fibrosis-clinical syndrome. Incident cases were followed up to identify deaths. Poisson and Cox regressions were used to calculate incidence rate ratios (IRR) and hazard ratios for mortality, respectively. Adjustments were made for age, gender, and strategic health authority. Survival from diagnosis was estimated using Kaplan-Meier analysis. RESULTS: In total 1491 and 4527 incident cases were identified using narrow and broad idiopathic pulmonary fibrosis-clinical syndrome definitions, respectively. Incidence and prevalence increased during the study. Compared with 2000, a near 80% increase in incidence was observed by 2012 [IRR 1.78 (95% CI 1.50-2.11; broad definition)], despite an observed decrease using the narrow definition [0.50 (0.38-0.65)]. Median survival was 3.0 years (95% CI 2.8-3.1) and 2.7 years (95% CI 2.5-3.0) in broad (n = 2168) and narrow case sets (n = 996), respectively. No significant changes in survival were observed. CONCLUSIONS: Idiopathic pulmonary fibrosis incidence rates have increased since 2000 and survival remains poor. These results provide a benchmark against which the effects of future treatment changes can be measured. FUNDING: InterMune UK and Ireland (now part of F. Hoffman La Roche).

Record linkage is increasingly used to expand the information available for public health research. An understanding of record linkage methods and the relevant strengths and limitations is important for robust analysis and interpretation of linked data. Here, we describe the approach used by Clinical Practice Research Datalink (CPRD) to link primary care data to other patient level datasets, and the potential implications of this approach for CPRD data analysis. General practice electronic health record software providers separately submit de-identified data to CPRD and patient identifiers to NHS Digital, excluding patients who have opted-out from contributing data. Data custodians for external datasets also send patient identifiers to NHS Digital. NHS Digital uses identifiers to link the datasets using an 8-stage deterministic methodology. CPRD subsequently receives a de-identified linked cohort file and provides researchers with anonymised linked data and metadata detailing the linkage process. This methodology has been used to generate routine primary care linked datasets, including data from Hospital Episode Statistics, Office for National Statistics and National Cancer Registration and Analysis Service. 10.6 million (M) patients from 411 English general practices were included in record linkage in June 2018. 9.1M (86%) patients were of research quality, of which 8.0M (88%) had a valid NHS number and were eligible for linkage in the CPRD standard linked dataset release. Linking CPRD data to other sources improves the range and validity of research studies. This manuscript, together with metadata generated on match strength and linkage eligibility, can be used to inform study design and explore potential linkage-related selection and misclassification biases.

OBJECTIVE: To evaluate the long term association between antidepressant prescribing and body weight. DESIGN: Population based cohort study. SETTING: General practices contributing to the UK Clinical Practice Research Datalink, 2004-14. PARTICIPANTS: 136 762 men and 157 957 women with three or more records for body mass index (BMI). MAIN OUTCOME MEASURES: The main outcomes were antidepressant prescribing, incidence of ≥5% increase in body weight, and transition to overweight or obesity. Adjusted rate ratios were estimated from a Poisson model adjusting for age, sex, depression recording, comorbidity, coprescribing of antiepileptics or antipsychotics, deprivation, smoking, and advice on diet. RESULTS: In the year of study entry, 17 803 (13.0%) men and 35 307 (22.4%) women with a mean age of 51.5 years (SD 16.6 years) were prescribed antidepressants. During 1 836 452 person years of follow-up, the incidence of new episodes of ≥5 weight gain in participants not prescribed antidepressants was 8.1 per 100 person years and in participants prescribed antidepressants was 11.2 per 100 person years (adjusted rate ratio 1.21, 95% confidence interval 1.19 to 1.22, P<0.001). The risk of weight gain remained increased during at least six years of follow-up. In the second year of treatment the number of participants treated with antidepressants for one year for one additional episode of ≥5% weight gain was 27 (95% confidence interval 25 to 29). In people who were initially of normal weight, the adjusted rate ratio for transition to overweight or obesity was 1.29 (1.25 to 1.34); in people who were initially overweight, the adjusted rate ratio for transition to obesity was 1.29 (1.25 to 1.33). Associations may not be causal, and residual confounding might contribute to overestimation of associations. CONCLUSION: Widespread utilisation of antidepressants may be contributing to long term increased risk of weight gain at population level. The potential for weight gain should be considered when antidepressant treatment is indicated.

OBJECTIVE: To assess the inclusion of patients as international research partners in Outcome Measures in Rheumatology (OMERACT) conferences and how this has influenced the scope and conduct of outcomes research in rheumatology. DESIGN: A thematic content analysis of OMERACT internal documents, publications and conference proceedings, followed by a responsive evaluation including 32 qualitative semistructured interviews. SETTING: The international, biannual research conference OMERACT 10 (Malaysia, 2010). PARTICIPANTS: Senior researchers (n=10), junior researchers (n=2), representatives of the pharmaceutical industry and regulators (n=2), conference staff (n=2), new patient delegates (n=8) and experienced patient delegates (n=8). RESULTS: The role of patients evolved over 10 years from a single patient focus group to full participation in all areas of the meeting and inclusion in research group meetings between conferences. Five main categories of impact emerged: widening the research agenda; including patient relevant outcomes in core sets; enhancing patient reported instruments; changing the culture of OMERACT and consequences outside OMERACT. Patient participants identified previously neglected outcome domains such as fatigue, sleep disturbances and flares which prompted collaborative working on new programmes of research. Specific benefits and challenges for patients and professionals were identified, such as personal fulfilment, widening of research interests, difficulties in establishing equal partnerships and concerns about loss of research rigour. CONCLUSIONS: Including patients as partners in OMERACT conferences has widened its focus and adjusted the way of working. It has resulted in new developments in the research agenda and the use of more patient-relevant outcomes in clinical trials. These collaborations have influenced perceptions and beliefs among many patients and researchers, and led to wider patient involvement as partners in research.

Background Pragmatic trials compare the effects of different decisions in usual clinical practice. Objectives To develop and evaluate methods to implement simple pragmatic trials using routinely collected electronic health records (EHRs) and recruiting patients at the point of care; to identify the barriers and facilitators for general practitioners (GPs) and patients and the experiences of trial participants. Design Two exemplar randomised trials (Retropro and eLung) with qualitative evaluations. Setting Four hundred and fifty-nine English and Scottish general practices contributing EHRs to a research database, of which 17 participated in the trials. Participants Retropro aimed to recruit 300 patients with hypercholesterolaemia and high cardiovascular risk and eLung aimed to recruit 150 patients with a chronic obstructive pulmonary disease exacerbation. Interventions Retropro randomised between simvastatin and atorvastatin and eLung between immediate antibiotics and deferred or non-use. eLung recruited during an unscheduled consultation using EHR flagging. Main outcome measure Successful trial completion with implementation of information technology (IT) system for flagging and data processing and documentation of operational and scientific experiences. Data sources EHR research database. Results The governance approval process took over 3 years. A total of 58.8% of the practices (n = 270) expressed interest in participating. The number of interested practices dropped substantially with each stage of the governance process. In Retropro, 6.5% of the practices (n = 30) were eventually approved and 3.7% (n = 17) recruited patients; in eLung, these numbers were 6.8% (n = 31) and 1.3% (n = 6) respectively. Retropro successfully completed recruitment (301 patients) whereas eLung recruited 31 patients. Retropro recruited 20.6% of all statin starters in recruiting practices and 1.1% in the EHR database; the comparable numbers for eLung were 32.3% and 0.9% respectively. The IT system allowed for complex eligibility criteria with central on and off control of recruitment and flagging at a practice. Good Clinical Practice guidelines, governance and consent procedures were found to have substantially affected the intended simple nature of the trials. One qualitative study of 13 clinicians found that clinicians were generally positive about the principle of computerised trial recruitment (flagging during consultation). However, trials which did not include patients with acute illness were favoured. The second qualitative process evaluation interviewed 27 GPs about their actual experiences, including declining, recruiting and non-recruiting GPs. Opportunistic patient recruitment during a routine GP consultation was found to be the most controversial element. The actual experiences of recruiting patients during unscheduled consultation were generally more positive than the hypothetical views of GPs. Several of the recruiting GPs reported the process took 5 minutes and was straightforward and feasible on most occasions. Almost all GPs expressed their strong support for the use of EHRs for trials. Ten eLung participants were interviewed, all of whom considered it acceptable to be recruited during a consultation and to use EHRs for trials. Conclusions EHR point-of-care trials are feasible, although the recruitment of clinicians is a major challenge owing to the complexity of trial approvals. These trials will provide substantial evidence on clinical effectiveness only if trial interventions and participating clinicians and patients are typical of usual clinical care and trials are simple to initiate and conduct. Recommendations for research include the development of evidence and implementation of risk proportionality in trial governance and conduct. Trial registration Current Controlled Trials ISRCTN33113202 and ISRCTN72035428. Funding This project was funded by the NIHR Health Technology Assessment programme and the Wellcome Trust and will be published in full in Health Technology Assessment; Vol. 18, No. 43. See the NIHR Journals Library website for further project information.

BACKGROUND: We evaluated the risk of pelvic inflammatory disease (PID), ectopic pregnancy, and infertility in women with a previous Chlamydia trachomatis (CT) diagnosis compared with women who tested negative for CT and CT untested women, considering both targeted and incidental (ie, prescribed for another indication) use of CT-effective antibiotics. METHODS: This was a retrospective study of women aged 12-25 years at start of follow-up within the Clinical Practice Research Datalink GOLD database linked to index of multiple deprivation quintiles, 2000-2013. CT test status and antibiotic use were determined in a time-dependent manner. Risk of PID, ectopic pregnancy, or female infertility were evaluated using of Cox proportional hazard models. RESULTS: We studied 857 324 women, contributing 6 457 060 person-years. Compared with women who tested CT-negative, women who tested CT-positive had an increased risk of PID (adjusted hazard ratio [aHR], 2.36; 95% confidence interval [CI], 2.01-2.79), ectopic pregnancy (aHR, 1.87; 95% CI, 1.38-2.54), and infertility (aHR, 1.85; 95% CI, 1.27-2.68). The PID risk was higher for women with 2 or more positive CT tests than those with 1 positive test. PID risk increased with the number of previous antibiotic prescriptions, regardless of CT test status. CONCLUSIONS: We showed an association between CT-positive tests and 3 adverse reproductive health outcomes. Moreover, this risk increased with repeat CT infections. CT-effective antibiotic use showed no decreased risks of subsequent PID regardless of CT history. Our results confirm the reproductive health burden of CT, which requires adequate public health interventions.

This study aimed to determine whether patients with statin-induced myopathy could be identified using the United Kingdom Clinical Practice Research Datalink, whether DNA could be obtained, and whether previously reported associations of statin myopathy with the SLCO1B1 c.521T>C and COQ2 rs4693075 polymorphisms could be replicated. Seventy-seven statin-induced myopathy patients (serum creatine phosphokinase (CPK) > 4× upper limit of normal (ULN)) and 372 statin-tolerant controls were identified and recruited. Multiple logistic regression analysis showed the SLCO1B1 c.521T>C single-nucleotide polymorphism to be a significant risk factor (P = 0.009), with an odds ratio (OR) per variant allele of 2.06 (1.32-3.15) for all myopathy and 4.09 (2.06-8.16) for severe myopathy (CPK > 10× ULN, and/or rhabdomyolysis; n = 23). COQ2 rs4693075 was not associated with myopathy. Meta-analysis showed an association between c.521C>T and simvastatin-induced myopathy, although power for other statins was limited. Our data replicate the association of SLCO1B1 variants with statin-induced myopathy. Furthermore, we demonstrate how electronic medical records provide a time- and cost-efficient means of recruiting patients with severe adverse drug reactions for pharmacogenetic studies.

PURPOSE: Primary care databases are increasingly used for researching pregnancy, eg, the effects of maternal drug exposures. However, ascertaining pregnancies, their timing, and outcomes in these data is challenging. While individual studies have adopted different methods, no systematic approach to characterise all pregnancies in a primary care database has yet been published. Therefore, we developed a new algorithm to establish a Pregnancy Register in the UK Clinical Practice Research Datalink (CPRD) GOLD primary care database. METHODS: We compiled over 4000 read and entity codes to identify pregnancy-related records among women aged 11 to 49 years in CPRD GOLD. Codes were categorised by the stage or outcome of pregnancy to facilitate delineation of pregnancy episodes. We constructed hierarchical rule systems to handle information from multiple sources. We assessed the validity of the Register to identify pregnancy outcomes by comparing our results to linked hospitalisation records and Office for National Statistics population rates. RESULTS: Our algorithm identified 5.8 million pregnancies among 2.4 million women (January 1987-February 2018). We observed close agreement with hospitalisation data regarding completeness of pregnancy outcomes (91% sensitivity for deliveries and 77% for pregnancy losses) and their timing (median 0 days difference, interquartile range 0-2 days). Miscarriage and prematurity rates were consistent with population figures, although termination and, to a lesser extent, live birth rates were underestimated in the Register. CONCLUSIONS: The Pregnancy Register offers huge research potential because of its large size, high completeness, and availability. Further validation work is underway to enhance this data resource and identify optimal approaches for its use.