SGLT2 Inhibitors Are Associated with Euglycemic Ketoacidosis in Diabetic Patients After Anesthesia
Sodium-glucose cotransporter-2 (SGLT2) inhibitors are the medication class most commonly associated with euglycemic ketoacidosis (euDKA) in diabetic patients after anesthesia. 1
Mechanism of Euglycemic Ketoacidosis
SGLT2 inhibitors cause euDKA through a distinct pathophysiologic mechanism that differs from typical diabetic ketoacidosis:
These medications lower blood glucose concentrations while simultaneously altering the insulin/glucagon ratio, creating a hormonal environment that promotes ketogenesis even when glucose levels remain normal (< 11.0 mmol/L or 200 mg/dL). 1, 2
Surgical stress amplifies this effect by triggering counterregulatory hormone release, which further drives hyperketonemia and can rapidly elevate ketone concentrations above 3.0 mmol/L with pH dropping below 7.3. 1
The clinical effects of SGLT2 inhibitors persist far beyond their plasma half-life, continuing for 3-4 days after discontinuation, which explains why euDKA can occur even when medications are stopped preoperatively. 2, 3
Epidemiology and Risk Stratification
The perioperative risk profile for SGLT2 inhibitor-associated ketoacidosis is well-defined:
Patients taking SGLT2 inhibitors have a significantly higher risk of perioperative DKA compared to those not taking them (1.02 vs. 0.69 per 1000 patients, OR 1.48,95% CI 1.02-2.15). 1, 4
Emergency surgery carries substantially higher risk than elective procedures (1.1% vs. 0.17% incidence of ketoacidosis). 1, 2, 4
Importantly, euDKA can occur in non-diabetic patients taking SGLT2 inhibitors for heart failure or chronic kidney disease, challenging the previous assumption that adequate endogenous insulin prevents significant ketosis. 1, 2
Clinical Presentation and Diagnostic Challenges
Recognizing euDKA requires high clinical suspicion because the presentation differs from typical DKA:
Classic symptoms include nausea, vomiting, tachypnea, and abdominal pain, but these are nonspecific and common in the postoperative period. 5, 6
Laboratory findings reveal anion gap metabolic acidosis (pH < 7.3, anion gap > 12 mmol/L) with paradoxically normal or near-normal glucose levels (< 14 mmol/L), elevated serum and urine ketones, and decreased bicarbonate (< 18 mEq/L). 2, 5
The diagnosis is frequently delayed because clinicians do not expect DKA with normal glucose levels, potentially leading to misdiagnosis and inappropriate treatment. 6, 7
Presentation timing varies widely, from hours to 6 weeks postoperatively, with some cases occurring despite withholding SGLT2 inhibitors for more than 72 hours preoperatively. 2, 7
Evidence-Based Perioperative Management
Current guidelines provide specific recommendations for SGLT2 inhibitor management:
Preoperative Discontinuation
The American College of Cardiology/American Heart Association 2024 guidelines recommend withholding SGLT2 inhibitors 3-4 days before elective surgery: canagliflozin, dapagliflozin, and empagliflozin should be stopped ≥3 days before, and ertugliflozin ≥4 days before. 8, 2
UK multidisciplinary consensus (2025) recommends a less conservative approach, suggesting omission only the day before and day of the procedure. 1, 8
The longer discontinuation period (3-4 days) is supported by evidence showing that omitting the drug >2 days preoperatively prevents DKA occurrence, and that SGLT2 inhibitor cessation reduces the risk of high anion gap acidosis. 1
Risk Mitigation Strategies
Maintain adequate hydration and avoid prolonged fasting periods to reduce ketone production during the perioperative period. 8, 2, 4
Monitor both glucose AND ketone levels during high-risk periods, as glucose monitoring alone will miss euDKA. 8, 4, 6
Consider glucose-containing intravenous fluids in cases of unavoidable prolonged fasting to mitigate ketone generation. 8, 2
Postoperative Resumption
Do not restart SGLT2 inhibitors until the patient is eating and drinking normally (usually 24-48 hours after surgery) and capillary ketones are <0.6 mmol/L. 2
Provide written sick-day rules to patients at discharge emphasizing the need to stop SGLT2 inhibitors and seek medical attention if symptoms of ketoacidosis develop. 2
Special Clinical Considerations
Heart Failure Patients
Cessation of SGLT2 inhibitors in patients taking them for heart failure may worsen cardiac function, creating a clinical dilemma that requires careful risk-benefit assessment. 1, 2
The cardiovascular benefits must be weighed against the risk of euDKA, particularly in patients undergoing cardiac surgery. 5
Bariatric Surgery
Thirteen cases of euDKA occurred after bariatric surgery, with 10 cases associated with very-low-calorie diet regimens as a precipitating factor. 7
Do not initiate SGLT2 inhibitors in patients on very low energy/liver reduction diets post-surgery, as these diets already induce ketosis and the combination could lead to severe ketoacidosis. 2
Treatment Protocol
When euDKA is diagnosed or suspected:
Initiate standard DKA protocol with insulin and dextrose infusions, which typically results in rapid resolution of metabolic abnormalities. 5, 9
Continue monitoring ketones, not just glucose, as ketonemia may persist longer than expected based on the drug's half-life. 3
Consider basal insulin if ketones persist or rebound after stopping insulin infusion. 3
Critical Pitfalls to Avoid
Do not rely on glucose levels alone to rule out DKA in patients who have taken SGLT2 inhibitors perioperatively—always check ketones if metabolic acidosis is present. 6, 7
Do not assume that stopping SGLT2 inhibitors 1-2 days preoperatively is sufficient protection—the American guidelines recommend 3-4 days for good reason. 8, 2
Do not dismiss nonspecific postoperative symptoms (nausea, vomiting, tachypnea) in patients with SGLT2 inhibitor exposure—maintain high suspicion for euDKA. 5, 6
Do not restart SGLT2 inhibitors too early postoperatively—wait until normal oral intake is established and ketones are documented to be low. 2, 6