Precision Prescribing of SGLT2 Inhibitors in Individuals With Type 2 Diabetes for Primary Prevention of Heart Failure: Model Development and Validation Study

OBJECTIVE Sodium–glucose cotransporter 2 inhibitors (SGLT2i) reduce heart failure (HF) risk in type 2 diabetes (T2D) and are recommended for patients with T2D who have atherosclerotic cardiovascular disease (ASCVD), HF, or chronic kidney disease (CKD). However, most individuals with T2D do not have these conditions, and current guidelines for this group do not indicate which individuals may benefit most from SGLT2i. We aimed to develop and validate a model to predict the individual-level HF benefit of SGLT2i in individuals with T2D without ASCVD, HF, or CKD. RESEARCH DESIGN AND METHODS We developed the SGLT2i Absolute Benefit Response (SABRE) model, combining absolute HF risk from the validated QDiabetes-HF model with the SGLT2i-associated hazard ratio (HR) for HF hospitalization from a trial meta-analysis (HR 0.63) to estimate individual 5-year HF benefit. Model components and predictions were validated using U.K. primary care data with linked hospital and death records from 2013 to 2020. RESULTS Among 57,368 SGLT2i initiators and 111,673 comparator (dipeptidyl peptidase 4 inhibitor or sulfonylurea) initiators, SGLT2i use was associated with a 30% lower risk of new-onset HF (HR 0.70 [95% CI 0.63–0.78]), consistent with trial evidence. Relative HF benefit did not vary by baseline absolute HF risk (P = 0.82). The SABRE model–predicted 5-year absolute HF benefit with SGLT2i ranged from <0.1% to 14.1% (median 1.0% [interquartile range 0.6–1.8%]) and calibrated well against observed HF outcomes. SABRE provided more targeted HF prevention than current guidelines in those with T2D without ASCVD, HF, or CKD. CONCLUSIONS The SABRE model is an easily deployed clinical prediction model integrating trial evidence and allowing more precise targeting of SGLT2i for primary HF prevention in T2D.