Individualized Decision-Making For GLP-1RA Therapy and Improving Obesity Care Using Real World Data - PROJECT ABSTRACT Obesity affects more than 40% of U.S. adults (more than 100 million) and contributes to substantial morbidity, mortality, and over $170 billion in annual healthcare expenditures. Glucagon-like peptide-1 receptor agonists (GLP-1RAs) offer a transformative opportunity to improve obesity-related outcomes, including significant weight loss, cardiovascular risk reduction, and emerging effects on metabolic comorbidities. However, real-world responses to GLP-1RAs vary widely. Many patients experience suboptimal weight loss and adverse effects. Current prescribing relies on body mass index (BMI) thresholds, failing to account for individual heterogeneity driven by biological, genetic, behavioral, social, and environmental factors yet to be uncovered. This contributes to trial-and-error prescribing, avoidable harm, inappropriate resource allocation, and unnecessary spending. This K23 proposes to develop and validate clinically actionable tools to individualize GLP-1RA prescribing and improve obesity care delivery. In Aim 1, machine learning will be applied to electronic health record (EHR) data (N=25,212) to develop and validate predictive models that estimate the likelihood of (1) achieving 12-month weight loss and (2) experiencing treatment-related adverse events, categorized as intolerances (e.g., nausea, constipation) or complications (e.g., gallbladder disease). In Aim 2, the expected value of individualized care (EVIC) framework will be used with a validated microsimulation model to quantify the clinical and economic value of individualized GLP-1RA prescribing versus a treat-all approach. In Aim 3, a conversation aid will be designed to communicate individualized risks and benefits, followed by usability testing to optimize its effectiveness. The overarching goal is to build tools that support scalable, individualized obesity treatment strategies grounded in real-world evidence. The candidate will receive mentored training in (1) predictive analytics and machine learning, (2) economic evaluation using the EVIC framework, and (3) user-centered design and usability testing. Completion of this training will position the candidate to lead an independent research program focused on the implementation and evaluation of cost-effective, individualized obesity care.
