Reminders EMbedded IN PCI Reports to Optimize Discharge Diabetes Mellitus Care (REMIND-DM): A Cluster Randomized Quality Improvement Implementation Study in the BMC2 PCI Collaborative

Background Despite strong guideline recommendations supporting SGLT2i and GLP-1RA for patients with T2D and established cardiovascular disease, implementation in routine clinical practice remains suboptimal. Percutaneous coronary intervention (PCI) clinical encounters may present an opportunity to improve implementation of cardioprotective therapies among high-risk patients. Methods REMIND-DM was a multi site, embedded, pragmatic, prospective, cluster randomized quality improvement implementation study conducted at 23 PCI sites (contributing 6,965 eligible PCI encounters: 3,539, baseline and 3,426, evaluation) in the BMC2 Collaborative. After a 6-month baseline period, sites were randomized to usual care or an intervention consisting of a templated PCI-report reminder of SGLT2i/GLP-1RA eligibility linked to a decision-support tool, followed by a 6-month evaluation period. The primary outcome was new discharge prescription of SGLT2 inhibitor or GLP-1RA in high-risk eligible participants. Results The intervention was not associated with a substantial change in new discharge prescription compared to usual care (aOR 0.90[0.66-1.21] p = 0.47). Among 9 of 11 intervention sites responding to a post-implementation survey, 3 reported reminder incorporation in ≥50% of elective PCI encounters and 2 in ≥50% of non-elective PCI encounters. In exploratory analyses, the intervention effect appeared to differ between elective and non-elective PCI, with a numerically greater likelihood of new discharge prescription among patients undergoing elective PCI (Pinteraction=0.023). Conclusions A low touch reminder-based QI intervention provided no evidence of substantial increase in new discharge prescription of SGLT2i or GLP-1RA in high-risk patients with T2D undergoing PCI.