From Evidence to Policy: Tailoring GLP-1RA Coverage for Optimal Health Outcomes

About This Grant

GLP-1 receptor agonists (GLP-1RAs), initially approved for diabetes management, have demonstrated efficacy in reducing cardiovascular and renal diseases, reversing non-alcoholic fatty liver disease (NAFLD), and enabling significant weight loss in individuals with or without diabetes. With rising obesity rates, these developments offer significant benefits to over half of the US adult population. However, the high cost of GLP-1RAs (> $15,000 annually) has led major healthcare payers to limit coverage, particularly for weight management. Policymakers have historically aimed to maximize treatment uptake among high-benefit user groups to improve health outcomes. However, identifying 'high-benefit users' of GLP-1RAs is challenging due to substantial variability in treatment benefits, influenced by a complex interplay of clinical characteristics and genetic factors. There is a significant lack of robust data to support precision policy design, necessitating targeted research that integrates clinical and genetic data. Preliminary analysis of NIH’s All-of-Us (AoU) data shows significant variability in GLP-1RA benefits linked to genetic factors. We now propose to expand the use of AoU data for more intricate analyses that comprehensively examine obesity-related cardiometabolic diseases. The OBJECTIVE of this study is to use clinical and genetic data that inform precision health policy design for GLP-1RA Coverage. Aim 1: To develop an agent-based microsimulation model for obesity-related cardiometabolic complications. Aim 2: To develop predictive models for the magnitude of GLP-1RA treatment response in preventing various cardiometabolic complications. Aim 3: Integrate the models from Aim 1 and Aim 2 to conduct large-scale simulation experiments on a nationally weighted sample and develop a scoring system using the simulation-proliferated data to inform precision health policy for GLP-1RA coverage. Dissemination Aim: To develop a large language model for presenting the scoring system.  This project is significant because it fills in a critical knowledge gap that impacts half of the US population, generating critical information to inform both clinical treatment choices and policy-level decisions. This project is also novel due to its state-of-the-art microsimulation approach and the application of ML and generative AI.