The development of a systematic approach to harness real-world evidence for the evaluation of medication safety and effectiveness in children

点击翻译按钮获取中文摘要

Approximately 20% of pediatric outpatient visits result in a prescription. Half of these prescriptions are off label, a number that rises to 80% among infants. The safe and effective prescribing of medications is essential to providing the highest quality care, and yet many children are receiving medications that have not been evaluated in their age group or in their disease area by rigorous randomized controlled trials (RCTs). There is a tremendous amount of data collected in the routine clinical care of patients via electronic health records (EHRs), claims data, and patient-reported data. These “real-world data” (RWD) can be harnessed with advanced epidemiologic methods focused on causal inference to generate “real-world evidence” (RWE) of medication safety and effectiveness. RWE studies, if conducted with scientific rigor, have the potential to complement RCTs and address important evidence gaps in the use of medications in children. This project aims to develop a framework for the rigorous conduct of RWE studies of pediatric medication use through the following specific aims: (1) to emulate 10 RCTs of medication effects in children using routinely collected healthcare data, (2) to evaluate alternative approaches to minimize bias when emulating RCTs in pediatric populations, and (3) to conduct 3 pediatric RWE studies on medication effects in clinical areas that lack RCT evidence. We will use RCTs as our gold-standard and use advanced causal inference and pharmacoepidemiology methods to design RWE studies to emulate published RCTs as closely as possible. Ten model RCTs will provide the scaffold for the evaluation of which methods perform best for the emulation of RCTs using RWE, and under which conditions. We will identify and summarize the methods that lead to the most successful emulation of RCTs. We will use best practice approaches to conduct three novel RWE studies. We will use three large healthcare utilization databases, including publicly and commercially insured children, as sources of RWD to conduct the RWE studies. These databases have detailed diagnostic and prescription information, making them well suited to studying medication use in children. The Division of Pharmacoepidemiology and Pharmacoeconomics at Brigham and Women’s Hospital provides the ideal setting for this endeavor. I will work with a team of mentors with expertise in pharmacoepidemiology, causal inference, comparative safety and effectiveness, pharmacology, and pediatric research to successfully complete the proposed project. This project provides the ideal framework to develop expertise in advanced pharmacoepidemiology and causal inference methods and support the overarching goal of my research: to generate high-quality evidence of medication safety and effectiveness in pediatric patients when no RCT data is available. Hands-on training and direct mentorship will be complemented by coursework at the Harvard T.H. Chan School of Public Health. This endeavor will provide the necessary foundation for me to become an independent investigator in pediatric pharmacoepidemiology.