University of California, Irvine researchers have developed recommended practices for evaluating evidence generated from adverse drug event studies that use electronic health record databases. Their findings are published online in the Journal of the American College of Clinical Pharmacy.
The availability of extensive electronic health record (EHR) databases offers opportunities for big data analytics and machine learning to be applied in precision medicine, disease risk prediction and clinical decision support research, but there are associated limitations and caveats. The team conducted a systematic review of current practices for conducting adverse drug event studies utilizing EHR databases and developed a set of recommended practices to help improve quality.
“Our recommendations will be useful to evaluate the evidence generated from adverse drug event studies,” said lead author Quinton Ng, a student in UCI’s PhD in pharmacological sciences program who is supervised by Alexandre Chan, PharmD, Department Chair and Professor of Clinical Pharmacy. “Our review caters to a wide audience, including clinicians, health informaticians and other non-clinician scientists to improve their knowledge in using EHR data for research studies to improve patient outcomes. This study also aligns with UCI’s current strategic directions on health care analytics.”
The UCI School of Pharmacy & Pharmaceutical Sciences team members also included undergraduate students Emily Dang, Lijie Chen, Mary Nguyen, Michael Nguyen, Sarah Samman and Tiffany Nguyen; health sciences assistant clinical professor and corresponding author Christine Cadiz, PharmD; Lee Nguyen, PharmD, health sciences associate clinical professor; and Chan.
Recommended practices from this study are informing current research; faculty and student researchers from the UCI School of Pharmacy & Pharmaceutical Sciences, UCI School of Medicine, and Department of Informatics are collaborating on a project to characterize adverse drug events and develop a clinical prediction model using the National Institute of Health’s All of Us Research Program database.