Search

Data collected from a validation substudy permit calculation of a bias-adjusted estimate of effect that is expected to equal the estimate that would have been observed had the gold standard measurement been available for the entire study population. In this paper, we develop and apply a framework for adaptive validation to determine when sufficient validation data have been collected to yield a bias-adjusted effect estimate with a prespecified level of precision. Prespecified levels of precision are decided a priori by the investigator, based on the precision of the conventional estimate and allowing for wider confidence intervals that would still be substantively meaningful. We further present an applied example of the use of this method to address exposure misclassification in a study of transmasculine/transfeminine youth and self-harm. Our method provides a novel approach to effective and efficient estimation of classification parameters as validation data accrue, with emphasis on the precision of the bias-adjusted estimate. This method can be applied within the context of any parent epidemiologic study design in which validation data will be collected and modified to meet alternative criteria given specific study or validation study objectives. [ABSTRACT FROM AUTHOR]

Distributed networks of health-care data sources are increasingly being utilized to conduct pharmacoepidemio-logic database studies. Such networks may contain data that are not physically pooled but instead are distributed horizontally (separate patients within each data source) or vertically (separate measures within each data source) in order to preserve patient privacy. While multivariable methods for the analysis of horizontally distributed data are frequently employed, few practical approaches have been put forth to deal with vertically distributed healthcare databases. In this paper, we propose 2 propensity score-based approaches to vertically distributed data analysis and test their performance using 5 example studies. We found that these approaches produced point estimates close to what could be achieved without partitioning. We further found a performance benefit (i.e., lower mean squared error) for sequentially passing a propensity score through each data domain (called the "sequential approach") as compared with fitting separate domain-specific propensity scores (called the "parallel approach"). These results were validated in a small simulation study. This proof-of-concept study suggests a new multivariable analysis approach to vertically distributed health-care databases that is practical, preserves patient privacy, and warrants further investigation for use in clinical research applications that rely on health-care databases. [ABSTRACT FROM AUTHOR]

We sought to determine whether machine learning and natural language processing (NLP) applied to electronic medical records could improve performance of automated health-care claims-based algorithms to identify anaphylaxis events using data on 516 patients with outpatient, emergency department, or inpatient anaphylaxis diagnosis codes during 2015–2019 in 2 integrated health-care institutions in the Northwest United States. We used one site's manually reviewed gold-standard outcomes data for model development and the other's for external validation based on cross-validated area under the receiver operating characteristic curve (AUC), positive predictive value (PPV), and sensitivity. In the development site 154 (64%) of 239 potential events met adjudication criteria for anaphylaxis compared with 180 (65%) of 277 in the validation site. Logistic regression models using only structured claims data achieved a cross-validated AUC of 0.58 (95% CI: 0.54, 0.63). Machine learning improved cross-validated AUC to 0.62 (0.58, 0.66); incorporating NLP-derived covariates further increased cross-validated AUCs to 0.70 (0.66, 0.75) in development and 0.67 (0.63, 0.71) in external validation data. A classification threshold with cross-validated PPV of 79% and cross-validated sensitivity of 66% in development data had cross-validated PPV of 78% and cross-validated sensitivity of 56% in external data. Machine learning and NLP-derived data improved identification of validated anaphylaxis events. [ABSTRACT FROM AUTHOR]

At the time medical products are approved, we rarely know enough about their comparative safety and effectiveness vis-à-vis alternative therapies to advise patients and providers. Postmarket generation of evidence on rare adverse events following medical product exposure increasingly requires analysis of millions of longitudinal patient records that can provide complete capture of data on patient experiences. In the accompanying article by Pradhan et al. (Am J Epidemiology. 2022;191(8):1352–1367), the authors demonstrate how observational database studies are often the most practical approach, provided these databases are carefully chosen to be "fit for purpose." Distributed data networks with common data models have proliferated in the last 2 decades in pharmacoepidemiology, allowing efficient capture of patient data in a standardized and structured format across disparate real-world data sources. Use of common data models facilitates transparency by allowing standardized programming approaches that can be easily reproduced. The distributed data network architecture, combined with a common data approach, supports not only multisite observational studies but also pragmatic clinical trials. It also helps bridge international boundaries and further increases the sample size and diversity of study populations. [ABSTRACT FROM AUTHOR]

To measure disparities in coronavirus disease 2019 (COVID-19) hospitalization and intensive care unit (ICU) transfer among racially/ethnically marginalized groups before and after implementation of the California statewide shelter-in-place (SIP) policy, we conducted a retrospective cohort study within a health-care system in California. COVID-19 patients diagnosed from January 1, 2020, to August 31, 2020, were identified from electronic health records. We examined hospitalizations and ICU transfers by race/ethnicity and pandemic period using logistic regression. Among 16,520 people with COVID-19 (mean age = 46.6 (standard deviation, 18.4) years; 54.2% women), during the post-SIP period, patients were on average younger and a larger proportion were Hispanic. In adjusted models, odds of hospitalization were 20% lower post-SIP as compared with the SIP period, yet all non-White groups had higher odds (odds ratios = 1.6–2.1) than non-Hispanic White individuals, regardless of period. Among hospitalized patients, odds of ICU transfer were 33% lower post-SIP than during SIP. Hispanic and Asian patients had higher odds than non-Hispanics. Disparities in hospitalization persisted and ICU risk became more pronounced for Asian and Hispanic patients post-SIP. Policy-makers should consider ways to proactively address racial/ethnic inequities in risk when considering future population-level policy interventions for public health crises. [ABSTRACT FROM AUTHOR]

Hydroxychloroquine (HCQ) was proposed as an early therapy for coronavirus disease 2019 (COVID-19) after in vitro studies indicated possible benefit. Previous in vivo observational studies have presented conflicting results, though recent randomized clinical trials have reported no benefit from HCQ among patients hospitalized with COVID-19. We examined the effects of HCQ alone and in combination with azithromycin in a hospitalized population of US veterans with COVID-19, using a propensity score–adjusted survival analysis with imputation of missing data. According to electronic health record data from the US Department of Veterans Affairs health care system, 64,055 US Veterans were tested for the virus that causes COVID-19 between March 1, 2020 and April 30, 2020. Of the 7,193 veterans who tested positive, 2,809 were hospitalized, and 657 individuals were prescribed HCQ within the first 48-hours of hospitalization for the treatment of COVID-19. There was no apparent benefit associated with HCQ receipt, alone or in combination with azithromycin, and there was an increased risk of intubation when HCQ was used in combination with azithromycin (hazard ratio = 1.55; 95% confidence interval: 1.07, 2.24). In conclusion, we assessed the effectiveness of HCQ with or without azithromycin in treatment of patients hospitalized with COVID-19, using a national sample of the US veteran population. Using rigorous study design and analytic methods to reduce confounding and bias, we found no evidence of a survival benefit from the administration of HCQ. [ABSTRACT FROM AUTHOR]

Genotype-phenotype association studies often combine phenotype data from multiple studies to increase statistical power. Harmonization of the data usually requires substantial effort due to heterogeneity in phenotype definitions, study design, data collection procedures, and data-set organization. Here we describe a centralized system for phenotype harmonization that includes input from phenotype domain and study experts, quality control, documentation, reproducible results, and data-sharing mechanisms. This system was developed for the National Heart, Lung, and Blood Institute's Trans-Omics for Precision Medicine (TOPMed) program, which is generating genomic and other -omics data for more than 80 studies with extensive phenotype data. To date, 63 phenotypes have been harmonized across thousands of participants (recruited in 1948–2012) from up to 17 studies per phenotype. Here we discuss challenges in this undertaking and how they were addressed. The harmonized phenotype data and associated documentation have been submitted to National Institutes of Health data repositories for controlled access by the scientific community. We also provide materials to facilitate future harmonization efforts by the community, which include 1) the software code used to generate the 63 harmonized phenotypes, enabling others to reproduce, modify, or extend these harmonizations to additional studies, and 2) the results of labeling thousands of phenotype variables with controlled vocabulary terms. [ABSTRACT FROM AUTHOR]

Methodological advancements in epidemiology, biostatistics, and data science have strengthened the research world's ability to use data captured from electronic health records (EHRs) to address pressing medical questions, but gaps remain. We describe methods investments that are needed to curate EHR data toward research quality and to integrate complementary data sources when EHR data alone are insufficient for research goals. We highlight new methods and directions for improving the integrity of medical evidence generated from pragmatic trials, observational studies, and predictive modeling. We also discuss needed methods contributions to further ease data sharing across multisite EHR data networks. Throughout, we identify opportunities for training and for bolstering collaboration among subject matter experts, methodologists, practicing clinicians, and health system leaders to help ensure that methods problems are identified and resulting advances are translated into mainstream research practice more quickly. [ABSTRACT FROM AUTHOR]

Improvements in data resources and computational power provide important opportunities to ensure the continued relevance and growth of observational epidemiology. To achieve that promise, rigorous statistical analyses are important but not sufficient. We must prioritize articulating relevant research questions and developing strong study designs. Relevance depends on designing observational research so it delivers actionable clinical or population health evidence. Expanding data sources, including administrative records and data from emerging technologies such as sensors, can potentially be leveraged to improve study design, statistical power, measurement, and availability of evidence on diverse populations. With these advantages, particularly evidence on the heterogeneity of treatment effects, observational research can better guide design of randomized trials. Evidence on the heterogeneity of treatment effects is also essential to extend the evidence from randomized trials beyond the narrow range of settings and populations for which trials have been conducted. Machine learning tools will likely grow in importance in observational epidemiology in coming years, although we need careful attention to the appropriate uses of prediction models. Despite the potential of these innovations, they will only be useful if embedded in theoretical frameworks motivated by applied clinical and population health questions. [ABSTRACT FROM AUTHOR]

The benefits of using electronic health records (EHRs) for disease risk screening and personalized health-care decisions are being increasingly recognized. Here we present a computationally feasible statistical approach with which to address the methodological challenges involved in utilizing historical repeat measures of multiple risk factors recorded in EHRs to systematically identify patients at high risk of future disease. The approach is principally based on a 2-stage dynamic landmark model. The first stage estimates current risk factor values from all available historical repeat risk factor measurements via landmark-age-specific multivariate linear mixed-effects models with correlated random intercepts, which account for sporadically recorded repeat measures, unobserved data, and measurement errors. The second stage predicts future disease risk from a sex-stratified Cox proportional hazards model, with estimated current risk factor values from the first stage. We exemplify these methods by developing and validating a dynamic 10-year cardiovascular disease risk prediction model using primary-care EHRs for age, diabetes status, hypertension treatment, smoking status, systolic blood pressure, total cholesterol, and high-density lipoprotein cholesterol in 41,373 persons from 10 primary-care practices in England and Wales contributing to The Health Improvement Network (1997-2016). Using cross-validation, the model was well-calibrated (Brier score = 0.041, 95% confidence interval: 0.039, 0.042) and had good discrimination (C-index = 0.768, 95% confidence interval: 0.759, 0.777). [ABSTRACT FROM AUTHOR]

The predictive value of the World Health Organization's Fracture Risk Assessment Tool (FRAX) was evaluated using real-world community data. A population-based cohort of 141,320 women aged 50-90 years (median age, 58 years; interquartile range, 54-67) in 2004 was extracted from the central database of a large Israeli health-care services provider and insurer. Retrospective FRAX scores were calculated using computerized health records and compared with actual incidence of major osteoporotic fractures (MOFs) during the following 10 years. Fracture proportions of 6.9% for MOFs and 2.2% for hip fractures were expected, as opposed to 13.5% and 2.9% observed. The area under the receiver operating characteristic curve (AUC) of FRAX scores calculated without the inclusion of bone mineral density (BMD) data was 0.65 (95% confidence interval: 0.65, 0.66) for MOF and 0.82 (95% confidence interval: 0.81, 0.82) for hip fracture. A total of 16,578 subjects had BMD data at the index date, and their risk estimates based solely on BMD exhibited lower predictive performance for both MOFs (AUC = 0.62 vs. 0.65; P = 0.003) and hip fractures (AUC = 0.78 vs. 0.84; P < 0.001) as compared with FRAX. FRAX scores based on electronic health records provided reasonable discrimination despite some underestimation of the absolute risk of nonhip fractures. Integration of FRAX with routine clinical systems could increase implementation in daily practice and improve risk detection, especially for patients without BMD data. [ABSTRACT FROM AUTHOR]

Electronic health records (EHRs) are an increasingly utilized resource for clinical research. While their size allows for many analytical opportunities, as with most observational data there is also the potential for bias. One of the key sources of bias in EHRs is what we term informed presence--the notion that inclusion in an EHR is not random but rather indicates that the subject is ill, making people in EHRs systematically different from those not in EHRs. In this article, we use simulated and empirical data to illustrate the conditions under which such bias can arise and how conditioning on the number of health-care encounters can be one way to remove this bias. In doing so, we also show when such an approach can impart M bias, or bias from conditioning on a collider. Finally, we explore the conditions under which number of medical encounters can serve as a proxy for general health. We apply these methods to an EHR data set from a university medical center covering the years 2007-2013. [ABSTRACT FROM AUTHOR]