Prevalent Sodium-Glucose Co-Transporter 2 Inhibitor Use, Baseline Laboratory Profile, and Clinical Outcomes in a Real-World Heart Failure Cohort.

OBJECTIVE

To evaluate the prevalence of sodium-glucose co-transporter 2 inhibitor (SGLT2i) use, baseline laboratory profiles, and associated clinical outcomes in a real-world heart failure (HF) cohort.

METHOD: This retrospective, single-center cohort study included consecutive patients with HF who were evaluated between January 2022 and January 2024. Of the 609 patients, 581 with known SGLT2i use status constituted the analytic cohort. The primary outcome was HF hospitalization. A composite outcome comprising all-cause death, HF hospitalization, stroke, or peripheral embolism was assessed as an exploratory endpoint. Propensity score matching and overlap-weighted Cox regression models were used in sensitivity analyses.

RESULTS

Among the 581 patients, 349 were SGLT2i users and 232 were non-users. Users had a lower left ventricular ejection fraction, higher B-type natriuretic peptide concentrations, a higher prevalence of diabetes and ischemic etiology, and more intensive background therapy. In the unadjusted analyses, the rates of HF hospitalization and composite events were higher among users; however, these associations were attenuated after adjustment. In the overlap-weighted propensity score analysis, SGLT2i use was not associated with HF hospitalization (HR: 0.91; 95% CI: 0.57-1.47; P=0.700), the study-defined composite endpoint (HR: 0.95; 95% CI: 0.61-1.48; P=0.827), or exploratory all-cause mortality (HR: 1.04; 95% CI: 0.45-2.42; P=0.931).

CONCLUSION

SGLT2i users represented a treated population with more advanced clinical characteristics, and their higher unadjusted event burden did not persist after adjustment. These findings reflect treatment selection rather than therapeutic benefit or harm. Therapies proven to reduce mortality should be systematically implemented in all eligible patients with HF, regardless of their clinical status or estimated mortality risk.