Abstract 4139502: Projected Impact of GLP-1 adoption on 5-year coronary stent rates among Medicare beneficiaries with cardiovascular disease

Background: Semaglutide reduced 5-year rates of coronary revascularization by 23% in SELECT. The impact of broader adoption of glucagon-like peptide 1 (GLP-1) agonists on coronary stent use among elderly patients with cardiovascular (CV) disease is unclear. Methods: To project the impact of GLP-1 adoption on 5-year rates of coronary stent use, we analyzed claims data for the 5% random sample of fee-for-service Medicare and Medicare Advantage beneficiaries identified in 2016 (n=1,847,213). Patient risk for CV events was derived using the REGARDS study administrative CV risk algorithm. Five-year observed coronary stent use was based on treatment in 2017–21. The unadjusted association between derived CV event risk and observed stent use was modeled as a flexible linear spline. Projected impact of increased GLP-1 use on patients’ 5-year risk of stent use was calculated by applying SELECT estimates of coronary revascularization reduction (HR 0.77 95% CI 0.68–0.87) to patients’ REGARDS-derived CV event risk and estimating stent use at the GLP-1-driven reduced derived CV event risk level. We projected impact on 5-year stent use as GLP-1 use increased from 0% to 100% of beneficiaries under 4 CV risk-based adoption scenarios: highest CV risk first; lowest CV risk first; median CV risk patients first (base case); and random adoption. Results: Sixty percent adoption of GLP-1 (current statin use) by median CV risk patients first would reduce coronary stent use overall by 6.9% (95% CI 3.6%–9.8%). Impact varied by role of CV risk in GLP-1 adoption, with 60% use in highest CV risk patients first resulting in 21.5% (95% CI 14.3%–25.8%) reduction as opposed to 5.9% (95% CI 3.3%–8.3%) if driven by population with 60% lowest CV risk (Figure). Adoption by 25% highest CV risk would achieve 70% benefit of complete population use. Reductions varied more than two-fold across hospital referral regions (HRRs) in the base case, ranging from 4.5% (95% CI 2.4%–6.1%) to 9.2% (95% CI 4.7%–13.3%); the highest CV-risk-first approach impact ranged from 13.6% (95% CI 9.0%–16.8%) to 30.7% (95% CI 20.7%–35.6%). Conclusion: GLP-1 adoption by Medicare beneficiaries comparable to current statin use could reduce coronary stent use nationally by 6.9% if focused on median CV risk, but up to 21.5% if focused on highest CV risk patients first. Impact varies by HRR, with high GLP-1 use among highest CV risk patients reducing stent use as much as >30%.