No standard definition, but typically glucose >11mmol/L, with pH <7.3 and/or bicarb <15. Associated with glycosuria and ketonuria. It is unusual but possible to present with near normal glucose values. Usual triggers are inadequate insulin, and/or intercurrent illness.
RHSC policy very generous, esp if in ICU! Plus corrects over 36h cf 48hr (ESPE guideline). Boluses for cardiovascular instability should be treated as surplus to deficit.
It is best to keep the child nil by mouth for the first 12 hours at least, and in a sick
child admitted to ITU, probably nearer 24 hours.
It is important to recheck the electrolytes and osmolality every 2 hours initially in a
severely ill child, but unless there is established renal failure the child's fluid
management is usually proceeding smoothly by 4 hours of starting fluids and the
frequency of electrolyte derangement can be adjusted accordingly
RHSC policy is to start insulin 60-90mins after fluids, presumably to control fall in glucose (which drops with fluid replacement alone, hence no need for bolus), evidence?
Initially intravenous insulin is given by continuous infusion using a 1 unit/ml solution (eg 50 units soluble insulin to 49.5 ml of 0.9% saline. The aim is to reduce the blood glucose steadily but gradually and to this end we recommend no initial insulin bolus. Institute a continuous IV infusion of insulin, via a Y-connector, according to the initial glucose:
If the rate of fall of glucose (measured by Glucostix) is greater than 5 mmol/l hour the rate should be decreased accordingly. When the blood sugar is below 14 mmol/l the infusion is changed to Normal or 1/2 normal Saline and 5% dextrose according to the serum sodium, as outlined above, and the insulin infusion rate adjusted according to 1-2 hourly Glucostix measurement. Once the Glucostix are in the normal range a crude sliding scale is useful and can be constructed according to the current infusion rate. For example, if patient stable on 1.0U/hr of insulin:
Note that the sliding scale is deliberately crude so that you don't chase the glucose. Institution of any sliding scale should not occur until the patient is clinically stable and is NO SUBSTITUTE for regular review by the MEDICAL STAFF.
Although it is standard practice to change to 8 hourly subcutaneous injections of soluble insulin once the patient is drinking, there is a good case for persisting with the insulin infusion until all signs of dehydration or acidosis have disappeared. At that stage the child can go straight on to his/her assigned insulin at 8 am the following morning. Continue with IV fluids until the child is drinking well and able to tolerate food. Do not expect ketones to have disappeared completely before changing to subcutaneous insulin.
The insulin infusion should be discontinued half an hour after subcutaneous insulin has been given thus avoiding rebound hyperglycaemia.
Do not stop the insulin infusion while dextrose is being infused as insulin is required to switch off ketone production. If the blood glucose falls below 7 mmol/L, consider adding extra glucose to the infusion. If the blood glucose rises out of control, consult senior medical staff, re-evaluate (? sepsis or other condition) and consider starting whole protocol again.
Ill children should be given nil by mouth until at least 12 hours after admission and gradually allowed to eat and drink, starting with sips.
Most DKA have subclinical (ie CT abnormal) cerebral oedema! But most asymptomatic. Presents typically 4-12 hours after initiation of treatment! But can be there at time of presentation. Progressive neurological deterioration esp drowsiness, confusion, focal signs. Headache. Preterminal is Cushing's triad: bradycardia, hypertension, irregular respiration.
Young & Asian more at risk. Acidosis more important than glucose level in predicting. Increased risk if insulin given in first 60 mins hence policy to delay. Mechanism? Neurones adapt to dehydration by retaining fluid then challenged by fluid therapy. Change in vascular permeability (seen on MRI - would explain cases presenting before therapy).
Apart from neuro obs, a warning sign is a static or falling sodium concentration despite rehydration. Expect a Na rise of 2mmol/l per 5.5 fall in glucose.
Intubation at presentation for coma will mask signs of oedema (so scan immediately?).
Manage by intubation and ventilation, control pCO2 carefully, 45deg head up position, head in line, aggressive treatment of seizures. Aggressive hyperventilation harmful (as for other causes RICP) - may acutely reduce ICP but at expense of brain perfusion.
| Blood glucose < 2.5 mmol/L | IV glucose, bolus dose 0.5g/Kg (5ml/Kg of 10%) | Start 10% dextrose infusion at 6ml/Kg/hr | If IV access not available: give Glucagon IM (>6yr 1mg, <6yr 0.5mg) NB - may cause vomiting |
| Blood glucose 2.5 - 4mmol/L | Oral glucose unless vomiting then 10% dextrose as above | ||
| Blood glucose > 4mmol/L | If coma persists monitor blood glucose frequently and infuse glucose as required | ||
Restlessness may be due to a full bladder. Consider other causes of decreased consciousness.
This work
is licensed under a Creative
Commons Attribution-Noncommercial-Share Alike 2.5 UK: Scotland License.