A0858
Title: Leveraging drug therapeutic class to identify adverse events from drug-drug interactions: A decision-theoretic approach
Authors: Massimiliano Russo - the Ohio State University (United States) [presenting]
Abstract: The concurrent prescription of multiple drugs can increase the likelihood of adverse drug events due to potential interactions. Identifying these adverse events is complex, especially when considering rare events and infrequently prescribed drug combinations. A novel Bayesian framework is introduced, designed to detect potential adverse events resulting from drug-drug interactions by incorporating information about each drug's therapeutic class (THERCL) to enhance signal detection. Using a decision-theoretic framework facilitates the identification of potentially harmful drug combinations while explicitly controlling the proportion of false-positive results. It is shown that the proposed framework outperforms the state-of-the-art approaches in an extensive simulation study and application involving a cohort study of older adults. Results indicate that leveraging THERCL in Bayesian modeling can effectively reduce false positives in detecting adverse events from drug-drug interactions, potentially improving safety in medication prescribing.
