DKA is the endocrine scenario most likely to come up, and it is marked on precision. The management is protocolised, so vagueness is obvious — the panel can hear immediately whether you have run one.
What are the diagnostic criteria for DKA?
All three are needed, and the Joint British Diabetes Societies define them explicitly.
- Ketonaemia — blood ketones 3.0 mmol/L or above, or significant ketonuria
- Acidosis — bicarbonate below 15.0 mmol/L and/or venous pH below 7.3
- Hyperglycaemia — blood glucose above 11.0 mmol/L, or known diabetes mellitus
Say the third clause out loud. Euglycaemic DKA is real, is more common with SGLT2 inhibitors, and a candidate who requires a high glucose to make the diagnosis will miss it.
How is DKA treated?
Fluid first, then insulin. Start intravenous 0.9% sodium chloride, then a fixed-rate intravenous insulin infusion at 0.1 units per kilogram per hour.
- Continue the patient's usual long-acting insulin throughout — stopping it is the classic error and causes rebound ketosis when the infusion stops
- Once blood glucose falls below 14.0 mmol/L, start 10% glucose alongside the saline and reduce the fixed rate from 0.1 to 0.05 units/kg/hr — you are treating ketosis, not glucose, so the insulin continues
- Replace potassium in the second and subsequent litres if potassium is below 5.5 mmol/L and the patient is passing urine — total body potassium is depleted even when the serum value looks normal
What tells you it has resolved?
Ketones below 0.6 mmol/L and venous pH above 7.3. Glucose is not a resolution criterion — that is the point of adding glucose rather than stopping insulin.
What else will the panel probe?
- The precipitant — infection, non-adherence, myocardial infarction, or a new diagnosis. Managing DKA without looking for why loses marks
- Cerebral oedema, particularly in younger patients and with rapid osmolar shifts
- Involving the diabetes specialist team early, and the safety issue of discharge before the patient understands sick-day rules