Inpatient Perioperative Euglycemic Diabetic Ketoacidosis Due to Sodium-Glucose Cotransporter-2 Inhibitors - Lessons From a Case Series and Strategies to Decrease Incidence

Endocr Pract. 2022 Sep;28(9):884-888. doi: 10.1016/j.eprac.2022.06.006. Epub 2022 Jun 23.

Abstract

Objective: To identify clinical characteristics and factors associated with the development of euglycemic diabetic ketoacidosis (eDKA), and develop suitable strategies to reduce such events.

Methods: Electronic health record (EHR) data were extracted to identify all patients between December 1, 2013, and March 30, 2021, who underwent surgical procedures and had been prescribed a sodium-glucose cotransporter 2 inhibitor (SGLT2i) before these procedures. The resulting list was streamlined to a subset of patients who either had diabetic ketoacidosis (DKA) listed as a hospital diagnosis, postoperative serum bicarbonate ≤ 16 mmol/L, or postoperative serum pH ≤ 7.20. Clinical documentation and laboratory data were reviewed to determine the patients with eDKA.

Results: A total of 2183 procedures conducted on 1307 patients, met the inclusion criteria with the majority (1726, 79.1%) being nonemergent patients. Among 1307 patients, 625 (47.8%) were prescribed empagliflozin, 447 (34.2%) canagliflozin, 214 (16.4%) dapagliflozin, and 21 (1.6%) ertugliflozin, respectively. A total of 8 incidences pertaining to eDKA were noted for 8 unique patients; 5 had undergone emergency surgery whereas 3 had undergone nonemergent procedures. In the 3 nonemergent cases, only 1 patient had received counseling to stop the SGLT2i 3 days before the procedure. In perioperative patients who were prescribed an SGLT2i over 6 years, the incidence of eDKA was 0.17% and 1.1% for nonemergent and emergent procedures, respectively.

Conclusion: Euglycemic DKA was rare in patients undergoing nonemergent procedures, likely because of preoperative instructions to stop their SGLT2i 3 days before the procedure. Euglycemic DKA was more likely to occur in patients undergoing emergency surgery when the SGLT2i could not be prophylactically stopped.

Keywords: SGLT2 inhibitors; diabetes mellitus; euglycemic DKA; inpatient glucose.

MeSH terms

  • Diabetes Mellitus, Type 2* / complications
  • Diabetes Mellitus, Type 2* / drug therapy
  • Diabetic Ketoacidosis* / diagnosis
  • Diabetic Ketoacidosis* / epidemiology
  • Diabetic Ketoacidosis* / prevention & control
  • Glucose
  • Humans
  • Incidence
  • Inpatients
  • Sodium
  • Sodium-Glucose Transporter 2 Inhibitors* / adverse effects

Substances

  • Sodium-Glucose Transporter 2 Inhibitors
  • Sodium
  • Glucose