Search

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., (© The Author(s) 2019. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.)

Background/aims: Pragmatic clinical trials embedded within health care systems provide an important opportunity to evaluate new interventions and treatments. Networks have recently been developed to support practical and efficient studies. Pragmatic trials will lead to improvements in how we deliver health care and promise to more rapidly translate research findings into practice., Methods: The National Institutes of Health (NIH) Health Care Systems Collaboratory was formed to conduct pragmatic clinical trials and to cultivate collaboration across research areas and disciplines to develop best practices for future studies. Through a two-stage grant process including a pilot phase (UH2) and a main trial phase (UH3), investigators across the Collaboratory had the opportunity to work together to improve all aspects of these trials before they were launched and to address new issues that arose during implementation. Seven Cores were created to address the various considerations, including Electronic Health Records; Phenotypes, Data Standards, and Data Quality; Biostatistics and Design Core; Patient-Reported Outcomes; Health Care Systems Interactions; Regulatory/Ethics; and Stakeholder Engagement. The goal of this article is to summarize the Biostatistics and Design Core's lessons learned during the initial pilot phase with seven pragmatic clinical trials conducted between 2012 and 2014., Results: Methodological issues arose from the five cluster-randomized trials, also called group-randomized trials, including consideration of crossover and stepped wedge designs. We outlined general themes and challenges and proposed solutions from the pilot phase including topics such as study design, unit of randomization, sample size, and statistical analysis. Our findings are applicable to other pragmatic clinical trials conducted within health care systems., Conclusion: Pragmatic clinical trials using the UH2/UH3 funding mechanism provide an opportunity to ensure that all relevant design issues have been fully considered in order to reliably and efficiently evaluate new interventions and treatments. The integrity and generalizability of trial results can only be ensured if rigorous designs and appropriate analysis choices are an essential part of their research protocols., (© The Author(s) 2016.)

Background: Multivariable cardiovascular disease (CVD) risk calculators, such as the Framingham risk equations, can be used to identify populations most likely to benefit from treatments to decrease risk., Purpose: To determine the proportion of adults within an electronic health record (EHR) for whom Framingham CVD risk scores could be calculated using cholesterol (lab-based) and/or BMI (BMI-based) formulae., Methods: EHR data were used to identify patients aged 30-74 years with no CVD and at least 2 years continuous enrollment before April 1, 2010, and relevant data from the preceding 5-year time frame. Analyses were conducted between 2010 and 2011 to determine the proportion of patients with a lab- or BMI-based risk score, the data missing, and the concordance between scores., Results: Of 122,270 eligible patients, 59.7% (n=73,023) had sufficient data to calculate the lab-based risk score and 84.1% (102,795) the BMI-based risk score. Risk categories were concordant in 78.2% of patients. When risk categories differed, BMI-based risk was almost always in a higher category, with 20.3% having a higher and 1.4% a lower BMI- than lab-based risk score. Concordance between lab- and BMI-based risk was greatest among those at lower estimated risk, including people who were younger, female, without diabetes, not obese, and those not on blood pressure- or lipid-lowering medications., Conclusions: EHR data can be used to classify CVD risk for most adults aged 30-74 years. In the population for the current study, CVD risk scores based on BMI could be used to identify those at low risk for CVD and potentially reduce unnecessary laboratory cholesterol testing., Trial Registration: This study is registered at clinicaltrials.gov NCT01077388., (Copyright © 2012 American Journal of Preventive Medicine. All rights reserved.)

Purpose: Many challenges arise when conducting a sequentially monitored medical product safety surveillance evaluation using observational electronic data captured during routine care. We review existing sequential approaches for potential use in this setting, including a continuous sequential testing method that has been utilized within the Vaccine Safety Datalink (VSD) and group sequential methods, which are used widely in randomized clinical trials., Methods: Using both simulated data and preliminary data from an ongoing VSD evaluation, we discuss key sequential design considerations, including sample size and duration of surveillance, shape of the signaling threshold, and frequency of interim testing., Results and Conclusions: All designs control the overall Type 1 error rate across all tests performed, but each yields different tradeoffs between the probability and timing of true and false positive signals. Designs tailored to monitor efficacy outcomes in clinical trials have been well studied, but less consideration has been given to optimizing design choices for observational safety settings, where the hypotheses, population, prevalence and severity of the outcomes, implications of signaling, and costs of false positive and negative findings are very different. Analytic challenges include confounding, missing and partially accrued data, high misclassification rates for outcomes presumptively defined using diagnostic codes, and unpredictable changes in dynamically accessed data over time (e.g., differential product uptake). Many of these factors influence the variability of the adverse events under evaluation and, in turn, the probability of committing a Type 1 error. Thus, to ensure proper Type 1 error control, planned sequential thresholds should be adjusted over time to account for these issues., (Copyright © 2012 John Wiley & Sons, Ltd.)

Electronic medical records (EMRs) allow for real-time access to blood pressure (BP) information on a population basis and improve identification and treatment of individuals with hypertension. Despite the potential uses of the data available from EMRs, relatively little research has examined the reliability of this data. To address this gap, the authors examined the reliability of BP taken at primary care visits and recorded in an EMR with those taken at a research study visit at which standard protocols were used to measure BP among all adults as well as by sex and age. Systolic BP (SBP) averaged 3.7 (17.3) points and diastolic BP (DBP) was 2.8 (10.6) points lower in the EMR than in the study visit across age and sex groups, with all differences statistically significant. For this cohort of patients with a diagnosis of hypertension, there was moderate correlation between BP measurements taken in clinic and at research. However, BP control for individuals, as defined by an SBP <140 mm Hg and DBP 90 mm Hg, differed by almost 25%. Known variability of BP and clinic procedures for measuring and recording BP may account for these differences., (© 2011 Wiley Periodicals, Inc.)

Background: The US Preventive Services Task Force recommends measuring blood pressure (BP) outside of clinic/office settings. While various options are available, including home devices, BP kiosks, and 24-h ambulatory BP monitoring (ABPM), understanding patient acceptability and adherence is a critical factor for implementation. Objective: To compare the acceptability and adherence of clinic, home, kiosk, and ABPM measurement. Design: Comparative diagnostic accuracy study which randomized adults to one of three BP measurement arms: clinic, home, and kiosk. ABPM was conducted on all participants. Participants: Adults (18–85 years) receiving care at 12 Kaiser Permanente Washington primary care clinics (Washington State, USA) with a high BP (≥ 138 mmHg systolic or ≥ 88 mmHg diastolic) in the electronic health record with no hypertension diagnosis and on no hypertensive medications and with high BP at a research screening visit. Measures: Patient acceptability was measured using a validated survey which was used to calculate an overall acceptability score (range 1–7) at baseline, after completing their assigned BP measurement intervention, and after completing ABPM. Adherence was defined based on the pre-specified number of BP measurements completed. Key Results: Five hundred ten participants were randomized (mean age 59 years), with mean BP of 150/88. Overall acceptability score was highest (i.e. most acceptable) for Home BP (mean 6.2, SD 0.7) and lowest (least acceptable) for ABPM (mean 5.0, SD 1.0); scores were intermediate for Clinic (5.5, SD 1.1) and Kiosk (5.4, SD 1.0). Adherence was higher for Home (154/170, 90.6%) and Clinic (150/172, 87.2%) than for Kiosk (114/168, 67.9%)). The majority of participants (467/510, 91.6%) were adherent to ABPM. Conclusions: Participants found home BP measurement most acceptable followed by clinic, BP kiosks, and ABPM. Our findings, coupled with recent evidence regarding the accuracy of home BP measurement, further support the routine use of home-based BP measurement in primary care practice in the US. Trial registration: ClinicalTrials.gov NCT03130257 https://clinicaltrials.gov/ct2/show/NCT03130257 [ABSTRACT FROM AUTHOR]

In the presence of unmeasured spatial confounding, spatial models may actually increase (rather than decrease) bias, leading to uncertainty as to how they should be applied in practice. We evaluated spatial modelling approaches through simulation and application to a big data electronic health record study. Whereas the risk of bias was high for purely spatial exposures (e.g. built environment), we found very limited potential for increased bias for individual‐level exposures that cluster spatially (e.g. smoking status). We also proposed a novel exposure‐penalized spline approach that selects the degree of spatial smoothing to explain spatial variability in the exposure. This approach appeared promising for efficiently reducing spatial confounding bias. [ABSTRACT FROM AUTHOR]

We consider the critical problem of pharmacosurveillance for adverse events once a drug or medical product is incorporated into routine clinical care. When making inference on comparative safety using large‐scale electronic health records, we often encounter an extremely rare binary adverse outcome with a large number of potential confounders. In this context, it is challenging to offer flexible methods to adjust for high‐dimensional confounders, whereas use of the propensity score (PS) can help address this challenge by providing both confounding control and dimension reduction. Among PS methods, regression adjustment using the PS as a covariate in an outcome model has been incompletely studied and potentially misused. Previous studies have suggested that simple linear adjustment may not provide sufficient control of confounding. Moreover, no formal representation of the statistical procedure and associated inference has been detailed. In this paper, we characterize a three‐step procedure, which performs flexible regression adjustment of the estimated PS followed by standardization to estimate the causal effect in a select population. We also propose a simple variance estimation method for performing inference. Through a realistic simulation mimicking data from the Food and Drugs Administration's Sentinel Initiative comparing the effect of angiotensin‐converting enzyme inhibitors and beta blockers on incidence of angioedema, we show that flexible regression on the PS resulted in less bias without loss of efficiency, and can outperform other methods when the PS model is correctly specified. In addition, the direct variance estimation method is a computationally fast and reliable approach for inference. [ABSTRACT FROM AUTHOR]

Background: Pragmatic clinical trials often use automated data sources such as electronic health records, claims, or registries to identify eligible individuals and collect outcome information. A specific advantage that this automated data collection often yields is having data on potential participants when design decisions are being made. We outline how this data can be used to inform trial design. Methods: Our work is motivated by a pragmatic clinical trial evaluating the impact of suicide-prevention outreach interventions on fatal and non-fatal suicide attempts in the 18 months after randomization. We illustrate our recommended approaches for designing pragmatic clinical trials using historical data from the health systems participating in this study. Specifically, we illustrate how electronic health record data can be used to inform the selection of trial eligibility requirements, to estimate the distribution of participant characteristics over the course of the trial, and to conduct power and sample size calculations. Results: Data from 122,873 people with patient health questionnaire (PHQ) responses, recorded in their electronic health records between 1 July 2010 and 31 March 2012, were used to show that the suicide attempt rate in the 18 months following completion of the questionnaire varies by response to item nine of the PHQ. We estimated that the proportion of individuals with a prior recorded elevated PHQ (i.e. history of suicidal ideation) would decrease from approximately 50% at the beginning of a trial to about 5%, 50 weeks later. Using electronic health record data, we conducted simulations to estimate the power to detect a 25% reduction in suicide attempts. Simulation-based power calculations estimated that randomizing 8000 participants per randomization arm would allow 90% power to detect a 25% reduction in the suicide attempt rate in the intervention arm compared to usual care at an alpha rate of 0.05. Conclusions: Historical data can be used to inform the design of pragmatic clinical trials, a strength of trials that use automated data collection for randomizing participants and assessing outcomes. In particular, realistic sample size calculations can be conducted using real-world data from the health systems in which the trial will be conducted. Data-informed trial design should yield more realistic estimates of statistical power and maximize efficiency of trial recruitment. [ABSTRACT FROM AUTHOR]

Introduction: The large-scale assembly of electronic health care data combined with the use of sequential monitoring has made proactive postmarket drug- and vaccine-safety surveillance possible. Although sequential designs have been used extensively in randomized trials, less attention has been given to methods for applying them in observational electronic health care database settings. Existing Methods: We review current sequential-surveillance planning methods from randomized trials, and the Vaccine Safety Datalink (VSD) and Mini-Sentinel Pilot projects--two national observational electronic health care database safety monitoring programs. Future Surveillance Planning: Based on this examination, we suggest three steps for future surveillance planning in health care databases: (1) prespecify the sequential design and analysis plan, using available feasibility data to reduce assumptions and minimize later changes to initial plans; (2) assess existing drug or vaccine uptake, to determine if there is adequate information to proceed with surveillance, before conducting more resource-intensive planning; and (3) statistically evaluate and clearly communicate the sequential design with all those designing and interpreting the safety- surveillance results prior to implementation. Plans should also be flexible enough to accommodate dynamic and often unpredictable changes to the database information made by the health plans for administrative purposes. Conclusions: This paper is intended to encourage dialogue about establishing a more systematic, scalable, and transparent sequential design-planning process for medical-product safety-surveillance systems utilizing observational electronic health care databases. Creating such a framework could yield improvements over existing practices, such as designs with increased power to assess serious adverse events. [ABSTRACT FROM AUTHOR]

We consider the critical problem of pharmacosurveillance for adverse events once a drug or medical product is incorporated into routine clinical care. When making inference on comparative safety using large-scale electronic health records, we often encounter an extremely rare binary adverse outcome with a large number of potential confounders. In this context, it is challenging to offer flexible methods to adjust for high-dimensional confounders, whereas use of the propensity score (PS) can help address this challenge by providing both confounding control and dimension reduction. Among PS methods, regression adjustment using the PS as a covariate in an outcome model has been incompletely studied and potentially misused. Previous studies have suggested that simple linear adjustment may not provide sufficient control of confounding. Moreover, no formal representation of the statistical procedure and associated inference has been detailed. In this paper, we characterize a three-step procedure, which performs flexible regression adjustment of the estimated PS followed by standardization to estimate the causal effect in a select population. We also propose a simple variance estimation method for performing inference. Through a realistic simulation mimicking data from the Food and Drugs Administration's Sentinel Initiative comparing the effect of angiotensin-converting enzyme inhibitors and beta blockers on incidence of angioedema, we show that flexible regression on the PS resulted in less bias without loss of efficiency, and can outperform other methods when the PS model is correctly specified. In addition, the direct variance estimation method is a computationally fast and reliable approach for inference. [ABSTRACT FROM AUTHOR]

Background: Patient-shared electronic health records provide opportunities for care outside of office visits. However, those who might benefit may be unable to or choose not to use these resources, while others might not need them. Objective: Electronic Communications and Home Blood Pressure Monitoring (e-BP) was a randomized trial that demonstrated that Web-based pharmacist care led to improved blood pressure (BP) control. During recruitment we attempted to contact all patients with hypertension from 10 clinics to determine whether they were eligible and willing to participate. We wanted to know whether particular subgroups, particularly those from vulnerable populations, were less willing to participate or unable to because they lacked computer access. Methods: From 2005 to 2006, we sent invitation letters to and attempted to recruit 9298 patients with hypertension. Eligibility to participate in the trial included access to a computer and the Internet, an email address, and uncontrolled BP (BP = 140/90 mmHg). Generalized linear models within a modified Poisson regression framework were used to estimate the relative risk (RR) of ineligibility due to lack of computer access and of having uncontrolled BP. Results: We were able to contact 95.1% (8840/9298) of patients. Those refusing participation (3032/8840, 34.3%) were significantly more likely (P < .05) to be female, be nonwhite, have lower levels of education, and have Medicaid insurance. Among patients who answered survey questions, 22.8% (1673/7354) did not have computer access. Older age, minority race, and lower levels of education were risk factors for lack of computer access, with education as the strongest predictor (RR 2.63, 95% CI 2.30-3.01 for those with a high school degree compared to a college education). Among hypertensive patients with computer access who were willing to participate, African American race (RR 1.22, 95% CI 1.06-1.40), male sex (RR 1.28, 95% CI 1.18-1.38), and obesity (RR 1.53, 95% CI 1.31-1.79) were risk factors for uncontrolled BP. Conclusion: Older age, lower socioeconomic status, and lower levels of education were associated with decreased access to and willingness to participate in a Web-based intervention to improve hypertension control. Failure to ameliorate this may worsen health care disparities. [ABSTRACT FROM AUTHOR]

Background: Pharmacist- or nurse-led team care decreases patient blood pressure (BP) and cardiovascular disease (CVD) risk. Purpose: To evaluate whether a Web-based dietitian-led (WD) team care intervention was feasible and resulted in decreased BP, CVD risk, and weight compared to usual care (UC). Methods: Electronic health record (EHR) data identified patients aged 30−69 years with BMI >26, elevated BP, and 10%−25% 10-year Framingham CVD risk who were registered patient website users. Patients with uncontrolled BP at screening were randomized to UC or WD, which included a home BP monitor, scale, and dietitian team care. WD participants had a single in-person dietitian visit to obtain baseline information and create a plan to reduce CVD risk. Planned follow-up occurred via secure messaging to report BP, weight, and fruit and vegetable intake and receive ongoing feedback. If needed, dietitians encouraged patients and their physicians to intensify antihypertensive and lipid-lowering medications. Primary outcomes were change in systolic BP and weight loss ≥4 kg at 6 months. Feasibility outcomes included intervention utilization and satisfaction. Results: Between 2010 and 2011, a total of 90 of 101 participants completed 6-month follow-ups. The WD group had higher rates of secure messaging utilization and patient satisfaction. The WD group lost significantly more weight than the UC group (adjusted net difference=−3.2 kg, 95% CI=−5.0, −1.5, p<0.001) and was more likely to lose ≥4 kg (adjusted relative risk [RRadj]=2.96, 95% CI=1.16, 7.53). BP control and CVD risk reduction were greater in WD than UC, but differences were not statistically significant. Conclusions: WD intervention was feasible and resulted in decreased weight, BP, and CVD risk. A larger trial is justified. Trial Registration Number: Trial Registration Number: NCT01077388. [Copyright &y& Elsevier]

High-dose, long-term opioid therapy (LtOT) is associated with risk for serious harms. Rapid opioid discontinuation may lead to increased pain, psychological distress, and illicit opioid use, but gradual, supported opioid taper may reduce these risks. We previously demonstrated that an opioid taper support and pain coping skills training intervention reduced opioid dose more than usual care (43% vs 19% dose reduction from baseline), with no increase in pain intensity and a significant reduction in activity interference. We aim to adapt and test this intervention in the Kaiser Permanente Washington healthcare system with STRategies to Improve Pain and Enjoy life (STRIPE, NCT03743402), a pragmatic, randomized trial. Our goal was to randomize 215 participants on moderate-high dose (≥40 morphine milligram equivalent/day) LtOT to either cognitive-behavioral therapy-based pain coping skills training involving 18 telephone sessions over 52 weeks with optional opioid taper support or usual care. Data are collected from electronic health records, claims, and self-report. The primary outcomes are mean daily opioid dose and the pain intensity, interference with enjoyment of life, and interference with general activity (PEG) score at 12 months (primary time point) and 6 months (secondary time point). Secondary outcomes include having ≥30% opioid dose reduction from baseline, and patient-reported problem opioid use, opioid-related difficulties, pain self-efficacy, opioid craving, global impression of change, and anxiety and depressive symptoms at 6 and 12 months. If effective, this treatment could reduce opioid exposure and associated risks to patients, families, and communities while offering patients an alternative for managing pain. [ABSTRACT FROM AUTHOR]