Euglycemic DKA on an SGLT2 inhibitor

Clinical stem

A 54-year-old with type 2 diabetes, CKD stage 3, and obesity on empagliflozin presents two days after elective surgery with nausea, malaise, and deep breathing. Glucose is 172 mg/dL; venous pH 7.24, bicarbonate 14 mmol/L, anion gap 22, with elevated beta-hydroxybutyrate.

The trap

Because the glucose looks reassuring, DKA is not considered and the acidosis is misattributed to the CKD. SGLT2-inhibitor euglycemic ketoacidosis is missed precisely because the glucose is near-normal.

Diagnostic reasoning

A normal glucose can hide a ketoacidosis — highest risk during fasting, acute illness, the perioperative period, dehydration, or low-carbohydrate intake. Whenever a patient on an SGLT2 inhibitor is acidotic or unwell, measure ketones (beta-hydroxybutyrate) and a blood gas regardless of the glucose.

Safety pearl

Hold the SGLT2 inhibitor during acute illness and around surgery (per label, roughly 3–4 days before a major procedure); check ketones when the patient is unwell; treat euglycemic DKA with fluids, insulin, and carbohydrate. The 2024 multisociety hyperglycemic-crises consensus report explicitly covers euglycemic ketoacidosis.

What not to overclaim

Euglycemic DKA is uncommon and is not a reason to withhold SGLT2 inhibitors, whose cardiorenal benefit is well proven. The point is recognition plus sick-day and perioperative holds — not avoidance.