A 12-year-old with diabetic ketoacidosis and a rising heart rate during treatment
A 12-year-old boy with known type 1 diabetes (diagnosed 4 years ago, on basal-bolus insulin) is admitted with 2 days of vomiting and abdominal pain. He has been using his insulin but has not been able to keep food down. On arrival: weight 40 kg, HR 130, BP 95/60, RR 30 and deep, temperature 37.2 C, capillary refill 3 seconds, dry mucous membranes, and drowsiness (GCS 14). Capillary glucose is 28 mmol/L. Blood gas: pH 7.10, pCO2 2.8 kPa, bicarbonate 8 mmol/L, base excess −18. Capillary ketones 5.8 mmol/L. Sodium 132 mmol/L, potassium 5.4 mmol/L, chloride 98 mmol/L, urea 9 mmol/L, creatinine 90 micromol/L. He has been started on IV 0.9% saline and an insulin infusion at 0.05 units/kg/hour. Two hours later his HR is 150, BP 85/50, and he is more drowsy. Repeat glucose is 14 mmol/L, ketones 3.5 mmol/L, and potassium 4.6 mmol/L.
1. Classify his DKA severity and identify the most likely reason for his deterioration at 2 hours.
Read the worked answer
Severity: This is severe DKA. Using the ISPAD/BSPED criteria:
- pH less than 7.1 - met (7.10).
- Bicarbonate less than 5 mmol/L - not met (8), but pH alone qualifies.
- Ketones greater than 3 mmol/L - met (5.8).
- Severe dehydration and drowsiness - present.
So: severe DKA with clinical dehydration (estimated 5–7% deficit) and a GCS of 14.
Most likely reason for deterioration at 2 hours: The pattern is tachycardia, hypotension, and increasing drowsiness with a falling glucose and improving ketones. The differential is:
- Inadequate fluid resuscitation / ongoing hypovolaemia - he was started on 0.9% saline but the rate may have been too low for a 40 kg child with severe dehydration. The falling BP and rising HR fit this.
- Cerebral oedema - the classic concern in paediatric DKA. Warning signs include headache, falling GCS, bradycardia, hypertension, and pupillary changes. He has a rising heart rate and falling BP, which is less typical of cerebral oedema (which usually causes hypertension and bradycardia), but the drowsiness is a red flag and must be taken seriously.
- Hypokalaemia - his potassium has fallen from 5.4 to 4.6 with insulin; insulin drives potassium intracellularly. This is expected but needs replacement.
- Sepsis - a precipitant; he has a normal temperature but tachycardia and hypotension could reflect sepsis.
The most likely explanation is under-resuscitation with fluid (he has not received enough isotonic fluid), with cerebral oedema as the diagnosis that must be excluded urgently because it is the leading cause of death in paediatric DKA.
2. What are your immediate actions, including fluid and insulin adjustments, and what monitoring do you put in place?
Read the worked answer
Immediate actions:
- Reassess ABC and GCS. If GCS is falling or there are signs of raised intracranial pressure (headache, bradycardia, hypertension, unequal pupils), treat for cerebral oedema: elevate the head of the bed, give 3% saline 2.5–5 mL/kg or mannitol 0.5–1 g/kg over 10–15 minutes, and call for senior/ICU help. Do not wait for imaging.
- Fluid resuscitation. He is hypotensive and tachycardic with delayed capillary refill. Give a 10 mL/kg bolus of 0.9% saline (400 mL) over 30–60 minutes and reassess. Repeat if still shocked. In DKA, boluses should be given cautiously and reassessed between each; the aim is to restore perfusion without causing rapid osmotic shifts.
- Continue the insulin infusion at 0.05–0.1 units/kg/hour. Do not stop it because the glucose is falling; instead, add glucose to the fluids when glucose falls below 14–15 mmol/L (e.g. 0.9% saline with 5% dextrose, or 0.45% saline with 5% dextrose).
- Potassium replacement. His potassium is 4.6 mmol/L and he is on insulin; add 40 mmol/L potassium chloride to the fluids once he is passing urine and the potassium is below 5.5 mmol/L. Monitor closely - insulin will continue to drive it down.
- Continue 0.9% saline as the maintenance/background fluid, calculating the deficit (5–7% of 40 kg = 2–2.8 L) plus maintenance over 48 hours, minus the boluses already given.
Monitoring:
- Hourly capillary glucose, ketones, heart rate, blood pressure, respiratory rate, and neurological observations (GCS, pupils).
- Hourly fluid balance and urine output (catheter if needed).
- 2–4 hourly electrolytes, urea, creatinine, and venous blood gas until stable.
- Continuous cardiac monitoring for arrhythmia (potassium).
- Neurological checks every hour - document GCS and any headache.
- Avoid bicarbonate (not indicated at pH 7.10 unless pH less than 6.9 with shock), and avoid rapid fluid boluses.
Escalation: involve the paediatric intensive care team early; this child is at high risk of cerebral oedema.
3. The parents ask why he became so unwell when he 'was taking his insulin'. Explain the pathophysiology and the precipitant, and what could have been done differently.
Read the worked answer
Why he became unwell despite taking insulin:
- DKA is not simply 'not enough insulin'. It is a combination of absolute or relative insulin deficiency plus counter-regulatory hormone excess (glucagon, cortisol, catecholamines, growth hormone).
- The precipitant here is almost certainly vomiting from an intercurrent illness (gastroenteritis or viral illness). Illness raises counter-regulatory hormones, which increase insulin resistance and lipolysis. Even with the same insulin dose, this is not enough to suppress ketogenesis.
- Vomiting also causes dehydration, which reduces insulin absorption from subcutaneous sites and worsens the picture.
- Sick-day rules are the key: during illness, children should check glucose and ketones more frequently (every 2–4 hours), continue basal insulin, and give additional rapid-acting insulin for ketones. They should never stop basal insulin.
- The history suggests he continued his usual insulin but did not increase it for the illness, and he could not keep fluids down.
What could have been done differently:
- A sick-day plan should be in place for every child with type 1 diabetes, with written instructions and ketone testing strips at home.
- Early contact with the diabetes team when vomiting starts.
- Never stop basal insulin during illness.
- If vomiting prevents oral intake, the family should seek urgent review rather than wait.
Explain to the parents:
- "Taking insulin is necessary but not always sufficient during illness. The body's stress hormones rise and make the insulin less effective, so the dose often needs to be increased."
- Reassure them that this is a recognised complication and that the team will review his sick-day plan before discharge.
- Emphasise that the priority now is safe correction of the dehydration and ketosis, and that he will be monitored closely.