Sodium-based paracetamol: impact on blood pressure, cardiovascular events, and all-cause mortality.

BACKGROUND AND AIMS: Effervescent formulations of paracetamol containing sodium bicarbonate have been reported to associate with increased blood pressure and a higher risk of cardiovascular diseases and all-cause mortality. Given the major implications of these findings, the reported associations were re-examined. METHODS: Using linked electronic health records data, a cohort of 475 442 UK individuals with at least one prescription of paracetamol, aged between 60 and 90 years, was identified. Outcomes in patients taking sodium-based paracetamol were compared with those taking non-sodium-based formulations of the same. Using a deep learning approach, associations with systolic blood pressure (SBP), major cardiovascular events (myocardial infarction, heart failure, and stroke), and all-cause mortality within 1 year after baseline were investigated. RESULTS: A total of 460 980 and 14 462 patients were identified for the non-sodium-based and sodium-based paracetamol exposure groups, respectively (mean age: 74 years; 64% women). Analysis revealed no difference in SBP [mean difference -0.04 mmHg (95% confidence interval -0.51, 0.43)] and no association with major cardiovascular events [relative risk (RR) 1.03 (0.91, 1.16)]. Sodium-based paracetamol showed a positive association with all-cause mortality [RR 1.46 (1.40, 1.52)]. However, after further accounting of other sources of residual confounding, the observed association attenuated towards the null [RR 1.08 (1.01, 1.16)]. Exploratory analyses revealed dysphagia and related conditions as major sources of uncontrolled confounding by indication for this association. CONCLUSIONS: This study does not support previous suggestions of increased SBP and an elevated risk of cardiovascular events from short-term use of sodium bicarbonate paracetamol in routine clinical practice.

Sex-Specific Effects of Blood Pressure Lowering Pharmacotherapy for the Prevention of Cardiovascular Disease: An Individual Participant-Level Data Meta-Analysis.

BACKGROUND: Whether the relative effects of blood pressure (BP)-lowering treatment on cardiovascular outcomes differ by sex, particularly when BP is not substantially elevated, has been uncertain. METHODS: We conducted an individual participant-level data meta-analysis of randomized controlled trials of pharmacological BP lowering. We pooled the data and categorized participants by sex, systolic BP categories in 10-mm Hg increments from <120 to ≥170 mm Hg, and age categories spanning from <55 to ≥85 years. We used fixed-effect one-stage individual participant-level data meta-analyses and applied Cox proportional hazard models, stratified by trial, to analyze the data. RESULTS: We included data from 51 randomized controlled trials involving 358 636 (42% women) participants. Over 4.2 years of median follow-up, a 5-mm Hg reduction in systolic BP decreased the risk of major cardiovascular events both in women and men (hazard ratio [95% CI], 0.92 [0.89-0.95] for women and 0.90 [0.88-0.93] for men; P for interaction, 1). There was no evidence for heterogeneity of relative treatment effects by sex for the major cardiovascular disease, its components, or across the different baseline BP categories (all P for interaction, ≥0.57). The effects in women and men were consistent across age categories and the types of antihypertensive medications (all P for interaction, ≥0.14). CONCLUSIONS: The effects of BP reduction were similar in women and men across all BP and age categories at randomization and with no evidence to suggest that drug classes had differing effects by sex. This study does not substantiate sex-based differences in BP-lowering treatment.

Stratification of diabetes in the context of comorbidities, using representation learning and topological data analysis.

Diabetes is a heterogenous, multimorbid disorder with a large variation in manifestations, trajectories, and outcomes. The aim of this study is to validate a novel machine learning method for the phenotyping of diabetes in the context of comorbidities. Data from 9967 multimorbid patients with a new diagnosis of diabetes were extracted from Clinical Practice Research Datalink. First, using BEHRT (a transformer-based deep learning architecture), the embeddings corresponding to diabetes were learned. Next, topological data analysis (TDA) was carried out to test how different areas in high-dimensional manifold correspond to different risk profiles. The following endpoints were considered when profiling risk trajectories: major adverse cardiovascular events (MACE), coronary artery disease (CAD), stroke (CVA), heart failure (HF), renal failure (RF), diabetic neuropathy, peripheral arterial disease, reduced visual acuity and all-cause mortality. Kaplan Meier curves were plotted for each derived phenotype. Finally, we tested the performance of an established risk prediction model (QRISK) by adding TDA-derived features. We identified four subgroups of patients with diabetes and divergent comorbidity patterns differing in their risk of future cardiovascular, renal, and other microvascular outcomes. Phenotype 1 (young with chronic inflammatory conditions) and phenotype 2 (young with CAD) included relatively younger patients with diabetes compared to phenotypes 3 (older with hypertension and renal disease) and 4 (older with previous CVA), and those subgroups had a higher frequency of pre-existing cardio-renal diseases. Within ten years of follow-up, 2592 patients (26%) experienced MACE, 2515 patients (25%) died, and 2020 patients (20%) suffered RF. QRISK3 model's AUC was augmented from 67.26% (CI 67.25-67.28%) to 67.67% (CI 67.66-67.69%) by adding specific TDA-derived phenotype and the distances to both extremities of the TDA graph improving its performance in the prediction of CV outcomes. We confirmed the importance of accounting for multimorbidity when risk stratifying heterogenous cohort of patients with new diagnosis of diabetes. Our unsupervised machine learning method improved the prediction of clinical outcomes.

Spousal concordance in adverse childhood experiences and the association with depressive symptoms in middle-aged and older adults: findings across China, the US, and Europe.

Background: Adverse childhood experiences (ACEs) are associated with higher depressive risks in adulthood. Whether respondents' ACEs are associated with their own depressive symptoms in adulthood and whether this association extends to their spouses' depressive symptoms remain unexplored. Methods: Data were from China Health and Retirement Longitudinal Study (CHARLS), the Health and Retirement Study (HRS), and the Survey of Health, Ageing and Retirement in Europe (SHARE). ACEs were categorized into overall, intra-familial, and extra-familial ACEs. Correlations of couples' ACEs were calculated using Cramer's V and partial Spearman's correlation. Associations of respondents' ACEs with spousal depressive symptoms were assessed using logistic regression, and mediation analyses were conducted to explore the mediating role of respondents' depressive symptoms. Results: Significant associations between husbands' ACEs and wives' depressive symptoms, with odds ratios (ORs) and 95% confidence intervals (CIs) of 2.09 (1.36-3.22) for 4 or more ACEs in CHARLS, and 1.25 (1.06-1.48) and 1.38 (1.06-1.79) for 2 or more ACEs in HRS and SHARE. However, wives' ACEs were associated with husbands' depressive symptoms only in CHARLS and SHARE. Findings in intra-familial and extra-familial ACEs were consistent with our main results. Additionally, respondents' depressive symptoms mediated more than 20% of the effect of respondents' ACEs on spousal depressive symptoms. Conclusion: We found that ACEs were significantly correlated between couples. Respondents' ACEs were associated with spousal depressive symptoms, with respondents' depressive symptoms mediating the association. The bidirectional implications of ACEs on depressive symptoms should be considered within household and effective interventions are warranted.

Systolic blood pressure, chronic obstructive pulmonary disease and cardiovascular risk.

OBJECTIVE: In individuals with complex underlying health problems, the association between systolic blood pressure (SBP) and cardiovascular disease is less well recognised. The association between SBP and risk of cardiovascular events in patients with chronic obstructive pulmonary disease (COPD) was investigated. METHODS AND ANALYSIS: In this cohort study, 39 602 individuals with a diagnosis of COPD aged 55-90 years between 1990 and 2009 were identified from validated electronic health records (EHR) in the UK. The association between SBP and risk of cardiovascular end points (composite of ischaemic heart disease, heart failure, stroke and cardiovascular death) was analysed using a deep learning approach. RESULTS: In the selected cohort (46.5% women, median age 69 years), 10 987 cardiovascular events were observed over a median follow-up period of 3.9 years. The association between SBP and risk of cardiovascular end points was found to be monotonic; the lowest SBP exposure group of <120 mm Hg presented nadir of risk. With respect to reference SBP (between 120 and 129 mm Hg), adjusted risk ratios for the primary outcome were 0.99 (95% CI 0.93 to 1.05) for SBP of <120 mm Hg, 1.02 (0.97 to 1.07) for SBP between 130 and 139 mm Hg, 1.07 (1.01 to 1.12) for SBP between 140 and 149 mm Hg, 1.11 (1.05 to 1.17) for SBP between 150 and 159 mm Hg and 1.16 (1.10 to 1.22) for SBP ≥160 mm Hg. CONCLUSION: Using deep learning for modelling EHR, we identified a monotonic association between SBP and risk of cardiovascular events in patients with COPD.

Hi-BEHRT: Hierarchical Transformer-Based Model for Accurate Prediction of Clinical Events Using Multimodal Longitudinal Electronic Health Records.

Electronic health records (EHR) represent a holistic overview of patients' trajectories. Their increasing availability has fueled new hopes to leverage them and develop accurate risk prediction models for a wide range of diseases. Given the complex interrelationships of medical records and patient outcomes, deep learning models have shown clear merits in achieving this goal. However, a key limitation of current study remains their capacity in processing long sequences, and long sequence modelling and its application in the context of healthcare and EHR remains unexplored. Capturing the whole history of medical encounters is expected to lead to more accurate predictions, but the inclusion of records collected for decades and from multiple resources can inevitably exceed the receptive field of the most existing deep learning architectures. This can result in missing crucial, long-term dependencies. To address this gap, we present Hi-BEHRT, a hierarchical Transformer-based model that can significantly expand the receptive field of Transformers and extract associations from much longer sequences. Using a multimodal large-scale linked longitudinal EHR, the Hi-BEHRT exceeds the state-of-the-art deep learning models 1% to 5% for area under the receiver operating characteristic (AUROC) curve and 1% to 8% for area under the precision recall (AUPRC) curve on average, and 2% to 8% (AUROC) and 2% to 11% (AUPRC) for patients with long medical history for 5-year heart failure, diabetes, chronic kidney disease, and stroke risk prediction. Additionally, because pretraining for hierarchical Transformer is not well-established, we provide an effective end-to-end contrastive pre-training strategy for Hi-BEHRT using EHR, improving its transferability on predicting clinical events with relatively small training dataset.

Effects of the COVID-19 pandemic on secondary care for cardiovascular disease in the UK: an electronic health record analysis across three countries.

BACKGROUND: Although morbidity and mortality from COVID-19 have been widely reported, the indirect effects of the pandemic beyond 2020 on other major diseases and health service activity have not been well described. METHODS AND RESULTS: Analyses used national administrative electronic hospital records in England, Scotland, and Wales for 2016-21. Admissions and procedures during the pandemic (2020-21) related to six major cardiovascular conditions [acute coronary syndrome (ACS), heart failure (HF), stroke/transient ischaemic attack (TIA), peripheral arterial disease (PAD), aortic aneurysm (AA), and venous thromboembolism(VTE)] were compared with the annual average in the pre-pandemic period (2016-19). Differences were assessed by time period and urgency of care.In 2020, there were 31 064 (-6%) fewer hospital admissions [14 506 (-4%) fewer emergencies, 16 560 (-23%) fewer elective admissions] compared with 2016-19 for the six major cardiovascular diseases (CVDs) combined. The proportional reduction in admissions was similar in all three countries. Overall, hospital admissions returned to pre-pandemic levels in 2021. Elective admissions remained substantially below expected levels for almost all conditions in all three countries [-10 996 (-15%) fewer admissions]. However, these reductions were offset by higher than expected total emergency admissions [+25 878 (+6%) higher admissions], notably for HF and stroke in England, and for VTE in all three countries. Analyses for procedures showed similar temporal variations to admissions. CONCLUSION: The present study highlights increasing emergency cardiovascular admissions during the pandemic, in the context of a substantial and sustained reduction in elective admissions and procedures. This is likely to increase further the demands on cardiovascular services over the coming years.

Using national electronic health records for pandemic preparedness: validation of a parsimonious model for predicting excess deaths among those with COVID-19-a data-driven retrospective cohort study.

OBJECTIVES: To use national, pre- and post-pandemic electronic health records (EHR) to develop and validate a scenario-based model incorporating baseline mortality risk, infection rate (IR) and relative risk (RR) of death for prediction of excess deaths. DESIGN: An EHR-based, retrospective cohort study. SETTING: Linked EHR in Clinical Practice Research Datalink (CPRD); and linked EHR and COVID-19 data in England provided in NHS Digital Trusted Research Environment (TRE). PARTICIPANTS: In the development (CPRD) and validation (TRE) cohorts, we included 3.8 million and 35.1 million individuals aged ≥30 years, respectively. MAIN OUTCOME MEASURES: One-year all-cause excess deaths related to COVID-19 from March 2020 to March 2021. RESULTS: From 1 March 2020 to 1 March 2021, there were 127,020 observed excess deaths. Observed RR was 4.34% (95% CI, 4.31-4.38) and IR was 6.27% (95% CI, 6.26-6.28). In the validation cohort, predicted one-year excess deaths were 100,338 compared with the observed 127,020 deaths with a ratio of predicted to observed excess deaths of 0.79. CONCLUSIONS: We show that a simple, parsimonious model incorporating baseline mortality risk, one-year IR and RR of the pandemic can be used for scenario-based prediction of excess deaths in the early stages of a pandemic. Our analyses show that EHR could inform pandemic planning and surveillance, despite limited use in emergency preparedness to date. Although infection dynamics are important in the prediction of mortality, future models should take greater account of underlying conditions.

Systolic Blood Pressure and Cardiovascular Risk in Patients With Diabetes: A Prospective Cohort Study.

BACKGROUND: Whether the association between systolic blood pressure (SBP) and risk of cardiovascular disease is monotonic or whether there is a nadir of optimal blood pressure remains controversial. We investigated the association between SBP and cardiovascular events in patients with diabetes across the full spectrum of SBP. METHODS: A cohort of 49 000 individuals with diabetes aged 50 to 90 years between 1990 and 2005 was identified from linked electronic health records in the United Kingdom. Associations between SBP and cardiovascular outcomes (ischemic heart disease, heart failure, stroke, and cardiovascular death) were analyzed using a deep learning approach. RESULTS: Over a median follow-up of 7.3 years, 16 378 cardiovascular events were observed. The relationship between SBP and cardiovascular events followed a monotonic pattern, with the group with the lowest baseline SBP of <120 mm Hg exhibiting the lowest risk of cardiovascular events. In comparison to the reference group with the lowest SBP (<120 mm Hg), the adjusted risk ratio for cardiovascular disease was 1.03 (95% CI, 0.97-1.10) for SBP between 120 and 129 mm Hg, 1.05 (0.99-1.11) for SBP between 130 and 139 mm Hg, 1.08 (1.01-1.15) for SBP between 140 and 149 mm Hg, 1.12 (1.03-1.20) for SBP between 150 and 159 mm Hg, and 1.19 (1.09-1.28) for SBP ≥160 mm Hg. CONCLUSIONS: Using deep learning modeling, we found a monotonic relationship between SBP and risk of cardiovascular outcomes in patients with diabetes, without evidence of a J-shaped relationship.

Blood pressure and risk of venous thromboembolism: a cohort analysis of 5.5 million UK adults and Mendelian randomization studies.

AIMS: Evidence for the effect of elevated blood pressure (BP) on the risk of venous thromboembolism (VTE) has been conflicting. We sought to assess the association between systolic BP and the risk of VTE. METHODS AND RESULTS: Three complementary studies comprising an observational cohort analysis, a one-sample and two-sample Mendelian randomization were conducted using data from 5 588 280 patients registered in the Clinical Practice Research Datalink (CPRD) dataset and 432 173 UK Biobank participants with valid genetic data. Summary statistics of International Network on Venous Thrombosis genome-wide association meta-analysis was used for two-sample Mendelian randomization. The primary outcome was the first occurrence of VTE event, identified from hospital discharge reports, death registers, and/or primary care records. In the CPRD cohort, 104 017(1.9%) patients had a first diagnosis of VTE during the 9.6-year follow-up. Each 20 mmHg increase in systolic BP was associated with a 7% lower risk of VTE [hazard ratio: 0.93, 95% confidence interval (CI): (0.92-0.94)]. Statistically significant interactions were found for sex and body mass index, but not for age and subtype of VTE (pulmonary embolism and deep venous thrombosis). Mendelian randomization studies provided strong evidence for the association between systolic BP and VTE, both in the one-sample [odds ratio (OR): 0.69, (95% CI: 0.57-0.83)] and two-sample analyses [OR: 0.80, 95% CI: (0.70-0.92)]. CONCLUSION: We found an increased risk of VTE with lower BP, and this association was independently confirmed in two Mendelian randomization analyses. The benefits of BP reduction are likely to outweigh the harms in most patient groups, but in people with predisposing factors for VTE, further BP reduction should be made cautiously.

Blood pressure-lowering treatment for prevention of major cardiovascular diseases in people with and without type 2 diabetes: an individual participant-level data meta-analysis.

BACKGROUND: Controversy exists as to whether the threshold for blood pressure-lowering treatment should differ between people with and without type 2 diabetes. We aimed to investigate the effects of blood pressure-lowering treatment on the risk of major cardiovascular events by type 2 diabetes status, as well as by baseline levels of systolic blood pressure. METHODS: We conducted a one-stage individual participant-level data meta-analysis of major randomised controlled trials using the Blood Pressure Lowering Treatment Trialists' Collaboration dataset. Trials with information on type 2 diabetes status at baseline were eligible if they compared blood pressure-lowering medications versus placebo or other classes of blood pressure-lowering medications, or an intensive versus a standard blood pressure-lowering strategy, and reported at least 1000 persons-years of follow-up in each group. Trials exclusively on participants with heart failure or with short-term therapies and acute myocardial infarction or other acute settings were excluded. We expressed treatment effect per 5 mm Hg reduction in systolic blood pressure on the risk of developing a major cardiovascular event as the primary outcome, defined as the first occurrence of fatal or non-fatal stroke or cerebrovascular disease, fatal or non-fatal ischaemic heart disease, or heart failure causing death or requiring hospitalisation. Cox proportional hazard models, stratified by trial, were used to estimate hazard ratios (HRs) separately by type 2 diabetes status at baseline, with further stratification by baseline categories of systolic blood pressure (in 10 mm Hg increments from <120 mm Hg to ≥170 mm Hg). To estimate absolute risk reductions, we used a Poisson regression model over the follow-up duration. The effect of each of the five major blood pressure-lowering drug classes, including angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, ß blockers, calcium channel blockers, and thiazide diuretics, was estimated using a network meta-analysis framework. This study is registered with PROSPERO, CRD42018099283. FINDINGS: We included data from 51 randomised clinical trials published between 1981 and 2014 involving 358 533 participants (58% men), among whom 103 325 (29%) had known type 2 diabetes at baseline. The baseline mean systolic/diastolic blood pressure of those with and without type 2 diabetes was 149/84 mm Hg (SD 19/11) and 153/88 mm Hg (SD 21/12), respectively. Over 4·2 years median follow-up (IQR 3·0-5·0), a 5 mm Hg reduction in systolic blood pressure decreased the risk of major cardiovascular events in both groups, but with a weaker relative treatment effect in participants with type 2 diabetes (HR 0·94 [95% CI 0·91-0·98]) compared with those without type 2 diabetes (0·89 [0·87-0·92]; pinteraction=0·0013). However, absolute risk reductions did not differ substantially between people with and without type 2 diabetes because of the higher absolute cardiovascular risk among participants with type 2 diabetes. We found no reliable evidence for heterogeneity of treatment effects by baseline systolic blood pressure in either group. In keeping with the primary findings, analysis using stratified network meta-analysis showed no evidence that relative treatment effects differed substantially between participants with type 2 diabetes and those without for any of the drug classes investigated. INTERPRETATION: Although the relative beneficial effects of blood pressure reduction on major cardiovascular events were weaker in participants with type 2 diabetes than in those without, absolute effects were similar. The difference in relative risk reduction was not related to the baseline blood pressure or allocation to different drug classes. Therefore, the adoption of differential blood pressure thresholds, intensities of blood pressure lowering, or drug classes used in people with and without type 2 diabetes is not warranted. FUNDING: British Heart Foundation, UK National Institute for Health Research, and Oxford Martin School.

Targeted-BEHRT: Deep Learning for Observational Causal Inference on Longitudinal Electronic Health Records.

Observational causal inference is useful for decision-making in medicine when randomized clinical trials (RCTs) are infeasible or nongeneralizable. However, traditional approaches do not always deliver unconfounded causal conclusions in practice. The rise of "doubly robust" nonparametric tools coupled with the growth of deep learning for capturing rich representations of multimodal data offers a unique opportunity to develop and test such models for causal inference on comprehensive electronic health records (EHRs). In this article, we investigate causal modeling of an RCT-established causal association: the effect of classes of antihypertensive on incident cancer risk. We develop a transformer-based model, targeted bidirectional EHR transformer (T-BEHRT) coupled with doubly robust estimation to estimate average risk ratio (RR). We compare our model to benchmark statistical and deep learning models for causal inference in multiple experiments on semi-synthetic derivations of our dataset with various types and intensities of confounding. In order to further test the reliability of our approach, we test our model on situations of limited data. We find that our model provides more accurate estimates of relative risk least sum absolute error (SAE) from ground truth compared with benchmark estimations. Finally, our model provides an estimate of class-wise antihypertensive effect on cancer risk that is consistent with results derived from RCTs.

How Much Lowering of Blood Pressure Is Required to Prevent Cardiovascular Disease in Patients With and Without Previous Cardiovascular Disease?

PURPOSE OF REVIEW: To review the recent large-scale randomised evidence on pharmacologic reduction in blood pressure for the primary and secondary prevention of cardiovascular disease. RECENT FINDINGS: Based on findings of the meta-analysis of individual participant-level data from 48 randomised clinical trials and involving 344,716 participants with mean age of 65 years, the relative reduction in the risk of developing major cardiovascular events was proportional to the magnitude of achieved reduction in blood pressure. For each 5-mmHg reduction in systolic blood pressure, the risk of developing cardiovascular events fell by 10% (hazard ratio [HR] (95% confidence interval [CI], 0.90 [0.88 to 0.92]). When participants were stratified by their history of cardiovascular disease, the HRs (95% CI) in those with and without previous cardiovascular disease were 0.89 (0.86 to 0.92) and 0.91 (0.89 to 0.94), respectively, with no significant heterogeneity in these effects (adjusted P for interaction = 1.0). When these patient groups were further stratified by their baseline systolic blood pressure in increments of 10 mmHg from < 120 to ≥ 170 mmHg, there was no significant heterogeneity in the relative risk reduction across these categories in people with or without previous cardiovascular disease (adjusted P for interaction were 1.00 and 0.28, respectively). Pharmacologic lowering of blood pressure was effective in preventing major cardiovascular disease events both in people with or without previous cardiovascular disease, which was not modified by their baseline blood pressure level. Treatment effects were shown to be proportional to the intensity of blood pressure reduction, but even modest blood pressure reduction, on average, can lead to meaningful gains in the prevention of incident or recurrent cardiovascular disease.

The Blood Pressure Lowering Treatment Trialists' Collaboration: methodological clarifications of recent reports.

Epidemiological evidence has consistently shown that people with higher systolic or diastolic blood pressure are at greater risk of cardiovascular diseases. However, there has been limited randomized evidence to determine the role of blood pressure level at treatment initiation in the reduction of cardiovascular diseases risk. The extent to which other characteristics of individuals, such as prior disease history, age or sex, should be taken into account has also been controversial. Furthermore, effects on less commonly reported efficacy and safety outcomes remain underexplored. The Blood Pressure Lowering Treatment Trialists' Collaboration has collected individual-level participant data from 52 randomized clinical trials, with more than 360 000 participants, and is now the largest source of individual-level data from randomized clinical trials of blood pressure-lowering treatment. This resource provides an unprecedented opportunity to address major areas of uncertainty relating to stratified efficacy and safety of antihypertensive therapy. Recent reports have demonstrated the power of pooled analyses of the Blood Pressure Lowering Treatment Trialists' Collaboration dataset in filling long-standing gaps in our knowledge. However, there have been some misconceptions regarding the methods underpinning the recent reports, which we clarify in this article.

An Explainable Transformer-Based Deep Learning Model for the Prediction of Incident Heart Failure.

Predicting the incidence of complex chronic conditions such as heart failure is challenging. Deep learning models applied to rich electronic health records may improve prediction but remain unexplainable hampering their wider use in medical practice. We aimed to develop a deep-learning framework for accurate and yet explainable prediction of 6-month incident heart failure (HF). Using 100,071 patients from longitudinal linked electronic health records across the U.K., we applied a novel Transformer-based risk model using all community and hospital diagnoses and medications contextualized within the age and calendar year for each patient's clinical encounter. Feature importance was investigated with an ablation analysis to compare model performance when alternatively removing features and by comparing the variability of temporal representations. A post-hoc perturbation technique was conducted to propagate the changes in the input to the outcome for feature contribution analyses. Our model achieved 0.93 area under the receiver operator curve and 0.69 area under the precision-recall curve on internal 5-fold cross validation and outperformed existing deep learning models. Ablation analysis indicated medication is important for predicting HF risk, calendar year is more important than chronological age, which was further reinforced by temporal variability analysis. Contribution analyses identified risk factors that are closely related to HF. Many of them were consistent with existing knowledge from clinical and epidemiological research but several new associations were revealed which had not been considered in expert-driven risk prediction models. In conclusion, the results highlight that our deep learning model, in addition high predictive performance, can inform data-driven risk factor identification.

Antihypertensive drug effects on long-term blood pressure: an individual-level data meta-analysis of randomised clinical trials.

OBJECTIVE: Evidence from randomised trials of pharmacological treatments on long-term blood pressure (BP) reduction is limited. We investigated the antihypertensive drug effects on BP over time and across different participant characteristics. METHODS: We conducted an individual patient-level data meta-analysis of 52 large-scale randomised clinical trials in the Blood Pressure Lowering Treatment Trialists' Collaboration using mixed models to examine treatment effects on BP over 4 years of mean follow-up. RESULTS: There were 363 684 participants (42% women), with baseline mean age=65 years and mean systolic/diastolic BP=152/87 mm Hg, and among whom 19% were current smokers, 49% had cardiovascular disease, 28% had diabetes and 69% were taking antihypertensive treatment at baseline. Drugs were effective in lowering BP showing maximal effect after 12 months and gradually attenuating towards later years. Based on measures taken ≥12 months postrandomisation, mean systolic/diastolic BP difference (95% CI) between more and less intense BP-lowering treatment was -11.1 (-11.3 to -10.8)/-5.6 (-5.7 to -5.4) mm Hg; between active treatment and placebo was -5.1 (-5.3 to -5.0)/-2.3 (-2.4 to -2.2) mm Hg; and between active and control arms for drug comparison trials was -1.4 (-1.5 to -1.3)/-0.6 (-0.7 to -0.6) mm Hg. BP reductions were observed across different baseline BP values and ages, and by sex, history of cardiovascular disease and diabetes and prior antihypertensive treatment use. CONCLUSION: These findings suggest that BP-lowering pharmacotherapy is effective in lowering BP, up to 4 years on average, in people with different characteristics. Appropriate treatment strategies are needed to sustain substantive long-term BP reductions.

Validation of risk prediction models applied to longitudinal electronic health record data for the prediction of major cardiovascular events in the presence of data shifts.

Aims: Deep learning has dominated predictive modelling across different fields, but in medicine it has been met with mixed reception. In clinical practice, simple, statistical models and risk scores continue to inform cardiovascular disease risk predictions. This is due in part to the knowledge gap about how deep learning models perform in practice when they are subject to dynamic data shifts; a key criterion that common internal validation procedures do not address. We evaluated the performance of a novel deep learning model, BEHRT, under data shifts and compared it with several ML-based and established risk models. Methods and results: Using linked electronic health records of 1.1 million patients across England aged at least 35 years between 1985 and 2015, we replicated three established statistical models for predicting 5-year risk of incident heart failure, stroke, and coronary heart disease. The results were compared with a widely accepted machine learning model (random forests), and a novel deep learning model (BEHRT). In addition to internal validation, we investigated how data shifts affect model discrimination and calibration. To this end, we tested the models on cohorts from (i) distinct geographical regions; (ii) different periods. Using internal validation, the deep learning models substantially outperformed the best statistical models by 6%, 8%, and 11% in heart failure, stroke, and coronary heart disease, respectively, in terms of the area under the receiver operating characteristic curve. Conclusion: The performance of all models declined as a result of data shifts; despite this, the deep learning models maintained the best performance in all risk prediction tasks. Updating the model with the latest information can improve discrimination but if the prior distribution changes, the model may remain miscalibrated.

Blood pressure lowering and risk of new-onset type 2 diabetes: an individual participant data meta-analysis.

BACKGROUND: Blood pressure lowering is an established strategy for preventing microvascular and macrovascular complications of diabetes, but its role in the prevention of diabetes itself is unclear. We aimed to examine this question using individual participant data from major randomised controlled trials. METHODS: We performed a one-stage individual participant data meta-analysis, in which data were pooled to investigate the effect of blood pressure lowering per se on the risk of new-onset type 2 diabetes. An individual participant data network meta-analysis was used to investigate the differential effects of five major classes of antihypertensive drugs on the risk of new-onset type 2 diabetes. Overall, data from 22 studies conducted between 1973 and 2008, were obtained by the Blood Pressure Lowering Treatment Trialists' Collaboration (Oxford University, Oxford, UK). We included all primary and secondary prevention trials that used a specific class or classes of antihypertensive drugs versus placebo or other classes of blood pressure lowering medications that had at least 1000 persons-years of follow-up in each randomly allocated arm. Participants with a known diagnosis of diabetes at baseline and trials conducted in patients with prevalent diabetes were excluded. For the one-stage individual participant data meta-analysis we used stratified Cox proportional hazards model and for the individual participant data network meta-analysis we used logistic regression models to calculate the relative risk (RR) for drug class comparisons. FINDINGS: 145 939 participants (88 500 [60·6%] men and 57 429 [39·4%] women) from 19 randomised controlled trials were included in the one-stage individual participant data meta-analysis. 22 trials were included in the individual participant data network meta-analysis. After a median follow-up of 4·5 years (IQR 2·0), 9883 participants were diagnosed with new-onset type 2 diabetes. Systolic blood pressure reduction by 5 mm Hg reduced the risk of type 2 diabetes across all trials by 11% (hazard ratio 0·89 [95% CI 0·84-0·95]). Investigation of the effects of five major classes of antihypertensive drugs showed that in comparison to placebo, angiotensin-converting enzyme inhibitors (RR 0·84 [95% 0·76-0·93]) and angiotensin II receptor blockers (RR 0·84 [0·76-0·92]) reduced the risk of new-onset type 2 diabetes; however, the use of ß blockers (RR 1·48 [1·27-1·72]) and thiazide diuretics (RR 1·20 [1·07-1·35]) increased this risk, and no material effect was found for calcium channel blockers (RR 1·02 [0·92-1·13]). INTERPRETATION: Blood pressure lowering is an effective strategy for the prevention of new-onset type 2 diabetes. Established pharmacological interventions, however, have qualitatively and quantitively different effects on diabetes, likely due to their differing off-target effects, with angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers having the most favourable outcomes. This evidence supports the indication for selected classes of antihypertensive drugs for the prevention of diabetes, which could further refine the selection of drug choice according to an individual's clinical risk of diabetes. FUNDING: British Heart Foundation, National Institute for Health Research, and Oxford Martin School.