Search

ObjectivesTo demonstrate the application of the Large-scale Evidence Generation and Evaluation across a Network of Databases (LEGEND) principles described in our companion article to hypertension treatments and assess internal and external validity of the generated evidence.Materials and methodsLEGEND defines a process for high-quality observational research based on 10 guiding principles. We demonstrate how this process, here implemented through large-scale propensity score modeling, negative and positive control questions, empirical calibration, and full transparency, can be applied to compare antihypertensive drug therapies. We assess internal validity through covariate balance, confidence-interval coverage, between-database heterogeneity, and transitivity of results. We assess external validity through comparison to direct meta-analyses of randomized controlled trials (RCTs).ResultsFrom 21.6 million unique antihypertensive new users, we generate 6 076 775 effect size estimates for 699 872 research questions on 12 946 treatment comparisons. Through propensity score matching, we achieve balance on all baseline patient characteristics for 75% of estimates, observe 95.7% coverage in our effect-estimate 95% confidence intervals, find high between-database consistency, and achieve transitivity in 84.8% of triplet hypotheses. Compared with meta-analyses of RCTs, our results are consistent with 28 of 30 comparisons while providing narrower confidence intervals.ConclusionWe find that these LEGEND results show high internal validity and are congruent with meta-analyses of RCTs. For these reasons we believe that evidence generated by LEGEND is of high quality and can inform medical decision-making where evidence is currently lacking. Subsequent publications will explore the clinical interpretations of this evidence.

Evidence derived from existing health-care data, such as administrative claims and electronic health records, can fill evidence gaps in medicine. However, many claim such data cannot be used to estimate causal treatment effects because of the potential for observational study bias; for example, due to residual confounding. Other concerns include P hacking and publication bias. In response, the Observational Health Data Sciences and Informatics international collaborative launched the Large-scale Evidence Generation and Evaluation across a Network of Databases (LEGEND) research initiative. Its mission is to generate evidence on the effects of medical interventions using observational health-care databases while addressing the aforementioned concerns by following a recently proposed paradigm. We define 10 principles of LEGEND that enshrine this new paradigm, prescribing the generation and dissemination of evidence on many research questions at once; for example, comparing all treatments for a disease for many outcomes, thus preventing publication bias. These questions are answered using a prespecified and systematic approach, avoiding P hacking. Best-practice statistical methods address measured confounding, and control questions (research questions where the answer is known) quantify potential residual bias. Finally, the evidence is generated in a network of databases to assess consistency by sharing open-source analytics code to enhance transparency and reproducibility, but without sharing patient-level information. Here we detail the LEGEND principles and provide a generic overview of a LEGEND study. Our companion paper highlights an example study on the effects of hypertension treatments, and evaluates the internal and external validity of the evidence we generate.

ObjectivesThis study sought to evaluate the association and burden of coronary artery calcium (CAC) with long-term, cause-specific mortality across the spectrum of baseline risk.BackgroundAlthough CAC is a known predictor of short-term, all-cause mortality, data on long-term and cause-specific mortality are inadequate.MethodsThe CAC Consortium cohort is a multicenter cohort of 66,636 participants without coronary heart disease (CHD) who underwent CAC testing. The following risk factors (RFs) were considered: 1) current cigarette smoking; 2) dyslipidemia; 3) diabetes mellitus; 4) hypertension; and 5) family history of CHD.ResultsDuring the 12.5-years median follow-up, 3,158 (4.7%) deaths occurred; 32% were cardiovascular disease (CVD) deaths. Participants with CAC scores ≥400 had a significantly increased risk for CHD and CVD mortality (hazard ratio [HR]: 5.44; 95% confidence interval [CI]: 3.88 to 7.62; and HR: 4.15; 95% CI: 3.29 to 5.22, respectively) compared with CAC of 0. Participants with ≥3 RFs had a smaller increased risk for CHD and CVD mortality (HR: 2.09; 95% CI: 1.52 to 2.85; and HR: 1.84; 95% CI: 1.46 to 2.31, respectively) compared with those without RFs. Across RF strata, CAC added prognostic information. For example, participants without RFs but with CAC ≥400 had significantly higher all-cause, non-CVD, CVD, and CHD mortality rates compared with participants with ≥3 RFs and CAC of 0.ConclusionsAcross the spectrum of RF burden, a higher CAC score was strongly associated with long-term, all-cause mortality and a greater proportion of deaths due to CVD and CHD. Absence of CAC identified people with a low risk over 12 years of follow-up, with most deaths being non-CVD in nature, regardless of RF burden.

BACKGROUND:Uncertainty remains about the optimal monotherapy for hypertension, with current guidelines recommending any primary agent among the first-line drug classes thiazide or thiazide-like diuretics, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, dihydropyridine calcium channel blockers, and non-dihydropyridine calcium channel blockers, in the absence of comorbid indications. Randomised trials have not further refined this choice. METHODS:We developed a comprehensive framework for real-world evidence that enables comparative effectiveness and safety evaluation across many drugs and outcomes from observational data encompassing millions of patients, while minimising inherent bias. Using this framework, we did a systematic, large-scale study under a new-user cohort design to estimate the relative risks of three primary (acute myocardial infarction, hospitalisation for heart failure, and stroke) and six secondary effectiveness and 46 safety outcomes comparing all first-line classes across a global network of six administrative claims and three electronic health record databases. The framework addressed residual confounding, publication bias, and p-hacking using large-scale propensity adjustment, a large set of control outcomes, and full disclosure of hypotheses tested. FINDINGS:Using 4·9 million patients, we generated 22 000 calibrated, propensity-score-adjusted hazard ratios (HRs) comparing all classes and outcomes across databases. Most estimates revealed no effectiveness differences between classes; however, thiazide or thiazide-like diuretics showed better primary effectiveness than angiotensin-converting enzyme inhibitors: acute myocardial infarction (HR 0·84, 95% CI 0·75-0·95), hospitalisation for heart failure (0·83, 0·74-0·95), and stroke (0·83, 0·74-0·95) risk while on initial treatment. Safety profiles also favoured thiazide or thiazide-like diuretics over angiotensin-converting enzyme inhibitors. The non-dihydropyridine calcium channel blockers were significantly inferior to the other four classes. INTERPRETATION:This comprehensive framework introduces a new way of doing observational health-care science at scale. The approach supports equivalence between drug classes for initiating monotherapy for hypertension-in keeping with current guidelines, with the exception of thiazide or thiazide-like diuretics superiority to angiotensin-converting enzyme inhibitors and the inferiority of non-dihydropyridine calcium channel blockers. FUNDING:US National Science Foundation, US National Institutes of Health, Janssen Research & Development, IQVIA, South Korean Ministry of Health & Welfare, Australian National Health and Medical Research Council.

BackgroundThe relative prevalence and clinical importance of monogenic mutations related to familial hypercholesterolemia and of high polygenic score (cumulative impact of many common variants) pathways for early-onset myocardial infarction remain uncertain. Whole-genome sequencing enables simultaneous ascertainment of both monogenic mutations and polygenic score for each individual.MethodsWe performed deep-coverage whole-genome sequencing of 2081 patients from 4 racial subgroups hospitalized in the United States with early-onset myocardial infarction (age ≤55 years) recruited with a 2:1 female-to-male enrollment design. We compared these genomes with those of 3761 population-based control subjects. We first identified individuals with a rare, monogenic mutation related to familial hypercholesterolemia. Second, we calculated a recently developed polygenic score of 6.6 million common DNA variants to quantify the cumulative susceptibility conferred by common variants. We defined high polygenic score as the top 5% of the control distribution because this cutoff has previously been shown to confer similar risk to that of familial hypercholesterolemia mutations.ResultsThe mean age of the 2081 patients presenting with early-onset myocardial infarction was 48 years, and 66% were female. A familial hypercholesterolemia mutation was present in 36 of these patients (1.7%) and was associated with a 3.8-fold (95% CI, 2.1-6.8; P3-fold increased odds of early-onset myocardial infarction. However, high polygenic score has a 10-fold higher prevalence among patients presents with early-onset myocardial infarction.Clinical trial registrationURL: https://www.clinicaltrials.gov . Unique identifier: NCT00597922.

AimsThe European Society of Cardiology (ESC) guideline on cardiovascular risk assessment considers coronary artery calcium a class B indication for risk assessment. We evaluated the degree to which coronary artery calcium can change the recommendation for individuals based on a change in estimated risk.Methods and resultsWe stratified 5602 MESA participants according to the ESC recommendation as: no lipid-lowering treatment recommended ( N = 2228), consider lipid-lowering treatment if uncontrolled ( N = 1686), or lipid-lowering treatment recommended ( N = 1688). We evaluated the ability of coronary artery calcium to reclassify cardiovascular risk. Among the selected sample, 54% had coronary artery calcium of zero, 25% had coronary artery calcium of 1-100 and 21% had coronary artery calcium greater than 100. In the lipid-lowering treatment recommended group 31% had coronary artery calcium of zero, while in the lipid-lowering treatment if uncontrolled group about 50% had coronary artery calcium of zero. The cardiovascular mortality rate was 1.7%/10 years in the lipid-lowering treatment if uncontrolled, and 7.0%/10 years in the lipid-lowering treatment recommended group. The absence of coronary artery calcium was associated with 1.4%/10 years in the lipid-lowering treatment if uncontrolled group and 3.0%/10 years in the lipid-lowering treatment recommended group. Compared with coronary artery calcium of zero, any coronary artery calcium was associated with significantly higher cardiovascular mortality in the lipid-lowering treatment recommended group (9.0%/10 years), whereas only coronary artery calcium greater than 100 was significantly associated with a higher cardiovascular mortality in the lipid-lowering treatment if uncontrolled group (3.2%/10 years).ConclusionThe absence of coronary artery calcium is associated with a low incidence of cardiovascular mortality or coronary heart disease events even in individuals in whom lipid-lowering therapy is recommended. A significant proportion of individuals deemed to be candidates for lipid-lowering therapy might be reclassified to a lower risk group with the use of coronary artery calcium.

BackgroundPhysicians "purchase" many health care services on behalf of patients yet remain largely unaware of the costs of these services. Electronic health record (EHR) cost displays may facilitate cost-conscious ordering of health services.ObjectiveTo determine whether displaying hospital lab and imaging order costs is associated with changes in the number and costs of orders placed.DesignQuasi-experimental study.ParticipantsAll patients with inpatient or observation encounters across a multi-site health system from April 2013 to October 2015.InterventionDisplay of order costs, based on Medicare fee schedules, in the EHR for 1032 lab tests and 1329 imaging tests.Main measuresOutcomes for both lab and imaging orders were (1) whether an order was placed during a hospital encounter, (2) whether an order was placed on a given patient-day, (3) number of orders placed per patient-day, and (4) cost of orders placed per patient-day.Key resultsDuring the lab and imaging study periods, there were 248,214 and 258,267 encounters, respectively. Cost display implementation was associated with a decreased odds of any lab or imaging being ordered during the encounter (lab adjusted odds ratio [AOR] = 0.97, p = .01; imaging AOR = 0.97, p