Replicating cardiovascular outcome trials of medications used to treat type 2 diabetes using real-world data: A systematic review of observational studies

Background: Cardiovascular outcome trials (CVOTs) are mandated by the U.S. Food and Drug Administration to assess the cardiovascular safety of new antidiabetic medications before entering the market. However, CVOTs often involve highly selective populations and results may not generalize to real-world settings. Methods: Our study aimed to synthesize observational studies to assess the generalizability of CVOTs to real-world settings. We systematically reviewed observational studies that emulated previous CVOTs for dipeptidyl peptidase-4 (DPP-4) inhibitors, glucagon-like peptide 1 (GLP-1) receptor agonists, and sodium glucose cotransporter-2 (SGLT-2) inhibitors among patients with type 2 diabetes. We searched the MEDLINE, EMBASE and Cochrane databases for observational studies that focused on trial emulation or cross-sectional studies that reported the proportion of real-world patients eligible for completed CVOTs. Two independent reviewers screened articles, extracted data, and assessed study concordance with randomized controlled trial (RCT) results. Results: Nineteen studies were included in our systematic review, including four cohort studies that emulated previous RCTs and 15 cross-sectional studies that evaluated trial eligibility. Results between RCTs and real-world data (RWD) were concordant for all drug classes in finding non-inferiority. The median eligibility percentage ranged from 13% to 31% for SGLT-2 inhibitor trials and 12% to 43% for GLP-1 receptor agonist trials. Conclusions: These results suggest that, while RCTs and RWD are concordant in their estimates, the trials lack representativeness. More research is needed on the replication of CVOTs using RWD to understand how different replication methods may impact findings.