I Congreso de Postgrado fcfm: ingeniería, ciencias e innovación

146 Santiago, 10 al 12 de agosto, 2022 COMBINING PRE-APPROVAL CLINICAL TRIALS AND POST-APPROVAL SPONTANEOUS ADVERSE EVENT REPORTING FOR IMPROVED SAFETY SIGNALING Fernanda Bravo¹*, Lawrence Chen², John Silberholz³ ¹UCLA Anderson School of M nagement, Los Angeles, California, USA. ²UC Berkeley Department of Industrial Engineering and Operations Research, Berkeley, California, USA. ³University of Michigan, Ross School of Business, Ann Harbor, Michigan, CA. *Email: fernanda.bravo@anderson.ucla.edu ABSTRACT Negative side effects from taking a drug, termed adverse drug reactions (ADRs), cause thousands of deaths a year in the U.S. alone. Regulators such as the U.S. Food and Drug Administration (FDA) have a critical role monitoring the safety of drugs on the market. Ideally, regulators would detect all safety issues prior to a drug’s marketing approval on the basis of clinical trial results. However, trials are often too small or too short to detect rare or slow-developing ADRs. As a result, regulators rely on post-approval surveillance from observational data sources such as spontaneous adverse event reporting systems to detect potential safety issues. These systems (such as the FDA’s FAERS system) contain large volumes of safety reports voluntarily submitted by patients and their doctors, and regulators generate hypotheses about potential safety issues (termed safety signals) by identifying side effects that occur at a disproportionately high rate in patients taking a drug versus patients taking other drugs for the same condition. Reliance on biased observational data — due to selection and reporting differences among patients — can result in regulators f lagging safety issues that are not truly present or in missing real safety issues. In this work, we seek to enhance the hypothesis generation step of safety signaling based on spontaneous ADR reporting systems via a Bayesian methodology that combines pre-approval clinical trial and post-approval observational data for multiple ADRs. We use data from more common adverse events to quantify the direction and magnitude of bias in observational data as compared to clinical trial data and use it to debias the observational data for more rare adverse events. We quantify the benefits of the proposed approach to regulators via both analytical modeling with a stylized model as well as via a detailed numerical evaluation using real-world clinical trial and FAERS data. Numerical results suggest our approach could improve expected benefit of signal detection from roughly 1% to 10% depending on the rare ADR or drug class. MOD E L AM I E N TO MAT EMÁT I CO 13