Diabetes mellitus and perioperative glucose

Why examiners ask this

Diabetes sits in a large share of stems because it carries a hidden cardiac history, a renal history, a gastroparetic stomach and a stiff-joint airway all at once. The examiner uses it to test judgment: what glucose do you accept, what makes you cancel, and how do you run insulin without producing the hypoglycemia that is worse than the hyperglycemia you were treating. It stresses application of knowledge and organization, because the answer has a preoperative, intraoperative and postoperative part and the candidate who blurs them loses points.

The 60-second answer

Decision points and the "why"

• Basal insulin continues, bolus insulin stops — a type 1 patient without basal insulin is in ketoacidosis within hours regardless of glucose; I give 75–80% of the usual glargine the night before, half the morning NPH, and hold rapid-acting insulin with the missed meal.¹ I reduce further if the patient has a history of nocturnal hypoglycemia.
• Oral agents — metformin is traditionally held the day of surgery for lactic acidosis risk,¹ though SAMBA now considers it safe to continue unless renal function is a concern;³ sulfonylureas stop because they cause hypoglycemia while fasting;³ SGLT2 inhibitors stop at least 3 days before surgery (4 for ertugliflozin) because they cause euglycemic ketoacidosis.⁴ A patient still on an SGLT2 inhibitor with an anion gap gets a beta-hydroxybutyrate before I proceed.
• The number that cancels — I postpone elective surgery for ketoacidosis or a hyperosmolar state, because the problem is the metabolic derangement, not the number; SAMBA does not cancel ambulatory cases for hyperglycemia alone,⁵ while inpatient guidance suggests considering postponement above about 250 mg/dL.¹ The A1c alone does not cancel surgery;⁵ above 8% it prompts optimization only if control can be improved quickly, for example before elective joint or spine procedures.²
• Insulin infusion for long or major cases — subcutaneous absorption is unreliable under anesthesia with vasoconstriction and hypothermia; for cardiac surgery, transplant, or cases over about 4 hours with large fluid shifts and glucose above 180, I run an intravenous insulin infusion with hourly glucose and dextrose available.¹
• Targets differ by setting — tighter control raises hypoglycemia without improving outcomes. The ADA perioperative range is 100–180 mg/dL,² 140–180 in the critically ill;¹ SAMBA 2024 accepts a more liberal 180–250 for ambulatory surgery, treated with subcutaneous insulin.⁵
• Airway and stomach — stiff-joint syndrome limits atlanto-occipital extension; I check the prayer sign. Gastroparesis makes aspiration more likely, so I ask about early satiety and use a rapid-sequence induction when it is present.
• Autonomic neuropathy — resting tachycardia, orthostasis and loss of heart-rate variability predict hemodynamic instability on induction and silent ischemia; I place an arterial line early for major cases.

The complication they will inject

The patient does not wake at the end of a four-hour case. I check a glucose before anything else, because hypoglycemia is the fastest reversible cause and the anesthetic has hidden every sign of it. I treat below 70 mg/dL with 25 g of dextrose, 50 mL of D50, and recheck in 15 minutes. If the glucose is high rather than low, I look for hyperosmolar coma and check sodium, osmolality and an anion gap. If the glucose is normal I run the rest of the delayed-emergence differential: residual drug, hypothermia, hypercarbia, stroke. The alternative inject is a glucose of 420 with an anion gap on arrival for an urgent case: I start the DKA protocol, fluids first, then insulin at 0.1 units/kg/h once potassium is above 3.5 mmol/L, replace potassium as it falls,⁶ and delay the incision until the gap is closing unless the surgery is the source of the ketoacidosis.

Pitfalls that cost points

• Holding all insulin in a type 1 patient and then being asked why the ketones appeared.
• Quoting a single glucose cutoff for cancellation without asking about ketones, hydration and the urgency of the case.
• Treating a glucose of 190 with a large subcutaneous dose and ignoring the delayed absorption that produces hypoglycemia in recovery.
• Forgetting the organs: no mention of the airway, the stomach, the kidneys or silent ischemia.
• Missing the SGLT2 inhibitor in the medication list.

Rapid-why

Q: Why does hyperglycemia matter intraoperatively? A: It impairs neutrophil function and wound healing, worsens outcome after cardiac and neurologic injury, and causes osmotic diuresis with volume depletion.

Q: Why does stress hyperglycemia occur in non-diabetics? A: Catecholamines, cortisol and glucagon raise hepatic glucose output and cause insulin resistance; I treat it to the same target when it persists above 180.

Q: Why avoid lactated Ringer's? A: I do not avoid it; the lactate load raises glucose trivially and the evidence of harm is weak. I avoid dextrose-containing maintenance fluids unless insulin is running.

Q: Why does hypoglycemia present differently under anesthesia? A: Sweating, tachycardia and tremor are blunted by anesthesia and beta-blockade, so the first sign is often delayed emergence or a seizure.

Numbers to know

Links

• dka, Preoperative cardiac evaluation for non-cardiac surgery, Chronic kidney disease and the dialysis patient, Delayed emergence, aspiration
• ADA Standards of Care, diabetes care in the hospital; SAMBA consensus on perioperative glucose management