Postoperative Outcomes Among Sodium-Glucose Cotransporter 2 Inhibitor Users

Importance Case reports and small retrospective studies have suggested that there is an increased risk of postoperative euglycemic ketoacidosis (eKA) and acute kidney injury (AKI) among patients using sodium-glucose cotransporter 2 inhibitors (SGLT2i) preoperatively. However, there has not been a representative assessment of the risks of these agents among patients undergoing surgery. Objective To evaluate the risk of postoperative eKA, AKI, and mortality within 30 days after surgery among preoperative long-term SGLT2i users compared with nonusers. Design, Settings, and Participants This is a multicenter, propensity-matched, retrospective case-control study from the Veterans Affairs Health Care System (VAHCS) National Registry performed from January 1, 2014, to December 31, 2022. Adult patients using SGLT2i preoperatively who underwent inpatient surgical procedures were compared with a 1:5 matched control group using propensity score matching, including the patient’s demographic characteristics, comorbidities, and surgical characteristics. Data analysis was performed from June 2023 to August 2024. Exposure Long-term use of SGLT2i, defined as having more than 3 fills of outpatient prescription or less than a 180-day gap of the last fill according to the VAHCS pharmacy registries. Main Outcomes and Measures The primary outcome was the rate of postoperative eKA among SGLT2i users vs control patients. Secondary outcomes included postoperative AKI and 30-day mortality after surgery. Results Among 462 968 patients undergoing surgery, 7448 SGLT2i users (mean [SD] age, 67.7 [8.1] years; 7204 [96.7%] male) and 455 520 nonusers (mean [SD] age, 65.8 [11.0] years; 424 785 [93.3%] male) were identified. After propensity score matching, 7439 patients were identified as SGLT2i users and compared with 33 489 control patients. SGLT2i use was associated with an increased risk of eKA (odds ratio [OR], 1.11; 95% CI, 1.05-1.17) but reduced risks of perioperative AKI (OR, 0.69; 95% CI, 0.62-0.78) and 30-day mortality (OR, 0.70; 95% CI, 0.55-0.88). The mortality rate 30 days after surgery was 1.1% among SGLT2i users vs 1.6% among control patients. The median hospital length of stay among the patients presenting with eKA increased by 3 days (median [IQR], 6 [3-10] days for those with eKA vs 3 [2-6] days for those without eKA). Conclusions and Relevance Patients treated with SGLT2i had a small but significantly higher risk of postoperative eKA but lower risks of postoperative AKI and 30-day mortality.