BACKGROUND: Antihyperglycemic agents with cardiovascular (CV) benefits, including SGLT-2i and GLP-1RA, are underused in clinical practice, particularly by cardiologists. Understanding the prescribing patterns of these agents by cardiologists may aid in implementation efforts.
METHODS: The COORDINATE-Diabetes trial enrolled participants with type 2 diabetes (T2D) and atherosclerotic cardiovascular disease (ASCVD) from US cardiology clinics and evaluated the impact of cluster randomization to a multifaceted implementation intervention versus usual care on proportional prescription of evidence-based therapies; the present analyses focus on SGLT2i and GLP-1 RA prescription. Participants prescribed an SGLT-2i were pooled between study arms and compared with those not initiating. A logistic regression model with random intercepts for the site was fitted with adjustment for treatment assignment (intervention vs. usual care), demographics, medical history, baseline medications and site location to determine factors associated with starting SGLT-2i. The same analysis was performed to determine factors associated with prescribing GLP-1RA. Reasons for not commencing either SGLT-2i or GLP-1RA in the intervention arm were aggregated and reported as counts.
RESULTS: Of all 1045 participants enrolled between July 2019 and May 2022 and followed for 6-12 months, 290 (27.8 %) were prescribed an SGLT-2i and 118 (11.3 %) a GLP-1RA; 8 (0.8 %) were prescribed both. Enrollment at an intervention site was an important predictor of SGLT-2i (OR 9.28, 4.82-17.89) and GLP-1RA prescription (OR 3.11, 1.32-7.37). Prior MI/coronary revascularization (OR 1.64, 1.01-2.67) was significantly associated with SGLT-2i prescription. A trend towards significance was observed for the association of preserved kidney function (OR 1.47, 0.99-2.19) and higher Charlson comorbidity index (OR 1.54, 0.97-2.44) with higher odds of SGLT-2i prescription. In contrast, T2D foot complications (OR 0.34, 0.15-0.80) were significantly associated with lower odds of SGLT-2i prescription. Older age was also directionally associated (OR 0.91, 0.81-1.02) with lower prescription of SGLT-2i. With respect to GLP-1RA, the presence of obesity (OR 1.70, 1.04-2.79) was associated with prescription, while increasing age (OR 0.72, 0.61-0.85) was associated with lower odds of prescription. The most common identifiable reason for not prescribing either SGLT-2i or GLP-1RA was related to now outdated guidance (i.e. permissible exclusion if metformin monotherapy and HbA1c <7 %). Contraindications to either agent and high cost were infrequently cited as reasons for not prescribing.
CONCLUSION: Consistent with the main trial results, participation in the COORDINATE-Diabetes intervention arm was an important determinant of higher odds for SGLT-2i and GLP-1RA prescription. Patient-level characteristics appeared to modestly influence the likelihood of prescription and may benefit from targeted education content.
