DKA = hyperglycaemia (BG >11 mmol/l) pH < 7.3 bicarbonate < 15 mmol/l
There is minimal information on DKA just in children, but plenty of information about DKA in general. 10 children a year die from DKA. There is an excellent module from the BMJ about type 1 diabetes in children, DKA vs HONK on BMJ, a case and a doctors.net module.
Any two of the three are needed to make the diagnosis.
The pathogenesis of DKA is nicely illustrated by the Calgary Guide. DKA can happen at any age, in type 1 and type 2 diabetes. It happens when the pancreas doesn't produce any insulin, so the body starts burning fatty acids instead, producing ketones. It is rare in Type 2 diabetes. Euglycaemic DKA is a recognised feature, but quite rare.
There is no cause for DKA identified in 40% of cases. It may rarely be precipitated by sepsis, and fever is not part of DKA. DKA is most commonly caused by a lack of insulin - non compliance, un - diagnosed diabetes, or concurrent illness. The "sick day rules" are nicely summarized on DFTB.
The commonest age for diabetes diagnosis is children between 10 and 14. The number of babies being diagnosed is increasing. There is another peak at around 40.
Polyuria, polydipsia. Abdominal pain, vomiting or drowsiness.
In secondary enuresis (bedwetting in children who have been dry) always think ?DKA.
Associated with coeliac disease, thyroid disease and vitiligo.
Low CO2 levels seen on capnography can help diagnose DKA!
- Blood gas
- Near patient ketones if available (superior to urine ketones)
- VBG to assess degree of acidosis
- As indicated - leucocytosis is common in DKA and does not necessarily indicate sepsis
- CXR, throat swab, blood culture, urinalysis, culture and sensitivity etc as indicated
- Elevated anion gap (to 25-35 mmol/L) - anion gap: Na+ - (Cl-+HCO3-) >15 mmol/L
- U+Es 2 hours after resuscitation begun, then 4 hourly
- Assess degree of dehydration. There is an excellent calculator from BSPED that lets you put in the patient's weight, estimated dehydration, and fluid given already - then calculates remaining fluid needed.
- Fluid resuscitation only if clinically shocked to maximum of 30ml/kg
- Requirement = maintenance + deficit - fluid already given
- Deficit = % dehydration x body weight
- Use NaCl + 20mmol KCl in 500ml. If BM <14mmol/L add glucose to the fluid. (After 12 hours, if Na stable or increasing - change to 500ml bags of 0.45% saline/5% dex/ 20mmol KCl)
- Insulin only after IV fluids have been running for at least an hour.
- Weight based insulin dosing at 0.1units/kg/hour- once you know the potassium as supplementation likely.
- Bicarbonate is rarely necessary but might help improve cardiac contractility in severe shock. Consider bicarb in profoundly acidotic children only.
- May continue normal long acting insulin (eg. glargine). Stop s/c insulin pumps.
- Phosphate levels are always low, but there is no evidence in adults or children that replacement has any clinical benefit and phosphate administration may lead to hypocalcaemia.
Complicates about 1% of cases of DKA in children and is lethal in 20% to 50% of victims
Signs and Symptoms
- Headache, irritability, slowing pulse, rising blood pressure, reducing conscious level
- Papilloedema is a late sign.
Inform senior staff
- Exclude hypoglycaemia as a possible cause of any behaviour change
- Hypertonic (2.7%) saline (5mls/kg over 5-10 mins) or Mannitol 0.5 – 1.0 g/kg stat (= 2.5 - 5 ml/kg Mannitol 20% over 20 minutes).
- restrict IV fluids to 1/2 maintenance and replace deficit over 72 rather than 48 hours
- Move to PICU and discuss with senior staff