Search

With the widespread implementation of electronic health records (EHRs), there has been significant progress in developing learning health systems (LHSs) aimed at improving health and health care delivery through rapid and continuous knowledge generation and translation. To support LHSs in achieving these goals, implementation science (IS) and its frameworks are increasingly being leveraged to ensure that LHSs are feasible, rapid, iterative, reliable, reproducible, equitable, and sustainable. However, 6 key challenges limit the application of IS to EHR-driven LHSs: barriers to team science, limited IS experience, data and technology limitations, time and resource constraints, the appropriateness of certain IS approaches, and equity considerations. Using 3 case studies from diverse health settings and 1 IS framework, we illustrate these challenges faced by LHSs and offer solutions to overcome the bottlenecks in applying IS and utilizing EHRs, which often stymie LHS progress. We discuss the lessons learned and provide recommendations for future research and practice, including the need for more guidance on the practical application of IS methods and a renewed emphasis on generating and accessing inclusive data., (©Katy E Trinkley, Anna M Maw, Cristina Huebner Torres, Amy G Huebschmann, Russell E Glasgow. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 07.10.2024.)

Background: Drug-induced long-QT syndrome (diLQTS) is a major concern among patients who are hospitalized, for whom prediction models capable of identifying individualized risk could be useful to guide monitoring. We have previously demonstrated the feasibility of machine learning to predict the risk of diLQTS, in which deep learning models provided superior accuracy for risk prediction, although these models were limited by a lack of interpretability., Objective: In this investigation, we sought to examine the potential trade-off between interpretability and predictive accuracy with the use of more complex models to identify patients at risk for diLQTS. We planned to compare a deep learning algorithm to predict diLQTS with a more interpretable algorithm based on cluster analysis that would allow medication- and subpopulation-specific evaluation of risk., Methods: We examined the risk of diLQTS among 35,639 inpatients treated between 2003 and 2018 with at least 1 of 39 medications associated with risk of diLQTS and who had an electrocardiogram in the system performed within 24 hours of medication administration. Predictors included over 22,000 diagnoses and medications at the time of medication administration, with cases of diLQTS defined as a corrected QT interval over 500 milliseconds after treatment with a culprit medication. The interpretable model was developed using cluster analysis (K=4 clusters), and risk was assessed for specific medications and classes of medications. The deep learning model was created using all predictors within a 6-layer neural network, based on previously identified hyperparameters., Results: Among the medications, we found that class III antiarrhythmic medications were associated with increased risk across all clusters, and that in patients who are noncritically ill without cardiovascular disease, propofol was associated with increased risk, whereas ondansetron was associated with decreased risk. Compared with deep learning, the interpretable approach was less accurate (area under the receiver operating characteristic curve: 0.65 vs 0.78), with comparable calibration., Conclusions: In summary, we found that an interpretable modeling approach was less accurate, but more clinically applicable, than deep learning for the prediction of diLQTS. Future investigations should consider this trade-off in the development of methods for clinical prediction., (©Steven T Simon, Katy E Trinkley, Daniel C Malone, Michael Aaron Rosenberg. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 01.12.2022.)

Patients with congestive heart failure (CHF) require complex medical management across the continuum of care. Electronic health records (EHR) are currently used for traditional tasks of documentation, reviewing and managing test results, computerized order entry, and billing. Unfortunately many clinicians view EHR as merely digitized versions of paper charts, which create additional work and cognitive burden without improving quality or efficiency of care. In fact, EHR are revolutionizing the care of chronic diseases such as CHF. This review describes how appropriate use of technologies offered by EHR can help standardize CHF care, promote adherence to evidence-based guidelines, optimize workflow efficiency, improve performance metrics, and facilitate patient engagement. This review discusses a number of tools including documentation templates, telehealth and telemedicine, health information exchange, order sets, clinical decision support, registries, and analytics. Where available, evidence of their potential utility in management of CHF is presented. Together these EHR tools can also be used to enhance quality improvement, patient management, and clinical research as part of a learning health care system model. This review describes how existing EHR tools can support patients, cardiologists, and care teams to deliver consistent, high-quality, coordinated, patient-centered, and guideline-concordant care of CHF., (Copyright © 2020 American College of Cardiology Foundation. Published by Elsevier Inc. All rights reserved.)