Diabetic ketoacidosis — formative SAQs

SAQ 1 — New-onset DKA in a seven-year-old (10 marks)

A 7-year-old boy weighing 22 kg presents with a two-week history of polyuria, polydipsia and weight loss, now vomiting and drowsy with deep sighing breathing. Blood glucose is 32 mmol/L, venous pH 7.08, bicarbonate 6 mmol/L, and blood beta-hydroxybutyrate 5.8 mmol/L. He is 6 per cent dehydrated but not shocked. (a) State the diagnosis and grade the severity. (b) Outline your fluid and insulin plan for the first hour. (c) Explain the potassium logic. (d) Name the complication you most fear and how you would recognise it. [1] [3]

Diagnosis and severity

This is diabetic ketoacidosis, meeting all three arms of the ISPAD triad: hyperglycaemia with a glucose of 32 mmol/L (above 11 mmol/L), acidosis with a venous pH of 7.08 and a bicarbonate of 6 mmol/L, and ketosis with a beta-hydroxybutyrate of 5.8 mmol/L. By the pH and bicarbonate this is severe DKA, because the pH is below 7.1 and the bicarbonate is below 5 to 10 mmol/L, and severe DKA in a young child mandates high-dependency or intensive-care nursing with hourly monitoring. [1]

Fluid and insulin plan for the first hour

He is dehydrated but not shocked, so he does not need a fluid bolus; a bolus is reserved for the shocked child at 10 mL/kg of isotonic saline. I would begin steady rehydration with isotonic saline, giving his maintenance plus an estimated deficit of around 5 to 7 per cent evenly over 48 hours, and I would not chase a precise clinical dehydration percentage because over-generous fluids are linked to cerebral oedema. Insulin is started only after fluids have begun, an hour or two in, as a continuous infusion of 0.05 to 0.1 units/kg/h, and never as a bolus. [1] [2]

The potassium logic

His serum potassium at presentation is likely to be normal or high despite a large total-body deficit, because the acidosis and the lack of insulin have driven potassium out of the cells and it has been lost in the urine. The moment insulin and fluids begin, potassium will move back into the cells and the serum level will fall, so I would add potassium to the fluids at 40 mmol/L once the level is below 5.5 mmol/L and he has passed urine, and monitor it hourly. Insulin should not be started in a child who is frankly hypokalaemic until the potassium is corrected. [1]

The feared complication

Cerebral oedema is the complication I most fear, because it is the leading cause of death in paediatric DKA and this young child with severe acidosis is high-risk. It typically appears four to twelve hours into treatment, often as the biochemistry improves, and I would recognise it clinically from a headache, a falling conscious level, irritability, incontinence, or a rising blood pressure with a slowing heart rate. I would treat immediately with hypertonic saline or mannitol, reduce the fluid rate, elevate the head of the bed, and call intensive care — without waiting for a scan, because outcome depends on the speed of recognition. [3] [9]

SAQ 2 — The insulin and monitoring detail (10 marks)

A 12-year-old girl with known type 1 diabetes is being treated for moderate DKA (pH 7.15, bicarbonate 9 mmol/L). Two hours into treatment her glucose has fallen from 24 to 13 mmol/L but her venous pH is still 7.18 and her beta-hydroxybutyrate remains 3.4 mmol/L. (a) Why should you not stop the insulin? (b) What change do you make now? (c) What insulin rate does the evidence support and why? (d) What resolution criteria will let you transition to subcutaneous insulin? [1] [6]

Why not to stop insulin

The insulin's job is to switch off ketogenesis and clear the acidosis, not merely to lower the glucose, and her persisting acidosis with a beta-hydroxybutyrate of 3.4 mmol/L shows the ketoacidosis is not yet resolved. Stopping the insulin because the glucose has fallen would let ketogenesis resume and stall her recovery. The correct response is to keep the insulin running and support the glucose from the other side. [1]

The change to make now

Because her glucose has fallen to 13 mmol/L, I would add 5 per cent dextrose to her intravenous fluids so that the insulin infusion can continue at a rate that clears the ketones without causing hypoglycaemia. This is the standard manoeuvre once the glucose falls below roughly 14 to 17 mmol/L, and it allows the acidosis to keep resolving. I would also confirm her potassium is being replaced and check the pace of her glucose fall is controlled. [1]

The evidence on insulin rate

Insulin runs at 0.05 to 0.1 units/kg/h, and the Nallasamy randomised trial found that a lower rate of 0.05 units/kg/h resolved the acidosis just as effectively as the standard 0.1 units/kg/h while causing fewer episodes of hypoglycaemia and hypokalaemia. This supports using the lower end of the range in many children, particularly younger children in whom the risk of a rapid osmolar shift is greatest. The infusion is fixed and never given as a bolus. [6] [1]

Resolution criteria and transition

I would transition her to subcutaneous insulin only once DKA has resolved biochemically: a venous pH above 7.3, a bicarbonate above 15 mmol/L, and a closed anion gap with cleared ketones, together with her being able to eat and drink. The subcutaneous dose is given before a meal and the intravenous insulin is continued for a short overlap so there is no gap in insulin cover. Resolution is defined by the acidosis clearing, not by the glucose alone. [1]