Diabetic ketoacidosis (DKA)

This page works the case through in full, answer included. If you would rather reason it out yourself: Play this case

How the diagnosis is reached

Thirst, polyuria and vomiting tipping into confusion over two days, in a nineteen-year-old with no medical history, is type 1 diabetes announcing itself. The deep laboured breathing is Kussmaul respiration, a deliberate attempt to blow off carbon dioxide and buy back some pH, and the fruity smell is exhaled acetone. The panel then confirms all three things DKA requires at once: glucose of 486, a venous pH of 7.14 with bicarbonate of 9, and large ketones, giving an anion gap of 24. Hyperglycaemia alone would not be enough, which is what separates this from a hyperosmolar state. The potassium is the trap, and the reason the treatment order matters. It reads 5.7, apparently high, while total body potassium is in fact severely depleted; insulin deficiency and acidosis have driven it out of the cells and the kidneys have been excreting it for two days. Give insulin first and it moves straight back intracellularly, and the number falls off a cliff. So fluids come first, then a fixed-rate insulin infusion, with potassium replaced alongside and checked often. Insulin is withheld entirely below 3.3 mmol/L until potassium is replaced. Dextrose is added once glucose drops under 200 mg/dL, not because the sugar needs support but because the insulin has to keep running to clear the ketones, and the acidosis, not the glucose, is what is actually being treated.

Key clinical clues

• Polyuria, polydipsia, and vomiting progressing to altered mental status
• Kussmaul breathing with a fruity breath odor
• Hyperglycemia with a high anion-gap metabolic acidosis and ketosis
• Potassium elevated at presentation despite total-body depletion
• No infectious trigger identified, consistent with new-onset type 1 diabetes

The case in full