Objective:

To initiate continuous subcutaneous insulin infusion (CSII) in a 1.9kg (-2.97 z-score) term baby where intravenous insulin infusion and subcutaneous insulin produced dangerous and unacceptable glucose levels. This study aims to determine the starting dose of insulin and appropriate infusion set, suitable hypoglycaemia management and achieve catch-up.

Method:

CSII commenced on day 14 of life. Starting insulin dose was 0.5units/kg/day (1 unit/day). Insulin aspart diluted with normal saline to 1:5. Cannula insertion sites were sparse due to limited subcutaneous fat. Bottle feeds initiated at 1.3concentration and given 2-3 hourly with no pre-feed boluses. Growth monitored weekly aiming for ~150-200g/week. Hypoglycaemia treated at <3.5mmol/l by 7gms of carbohydrates provided from 15ml glucose polymer solution. Child discharged at 5 weeks and parents taught set changes and insulin dilution prior.

Result:

One mild hypoglycaemic event of 3.2mmol/l was noted and treated with good effect. Cannula choice was limited with two breakages resulting in silica remaining in buttocks. Pre-feed boluses (1 unit per 50gms carbohydrates) started on day 7. Feeds remained at 1.3concentration up to 200kcal/kg and 200ml/kg/day. Weight gain initially slow, averaging 135g/week. Weight SDS improved from -3.65 at 1 month to -1.73 at 3 months, averaging 261g/week. CSII insulin requirements gradually reduced and ceased at 3 months.

Conclusion:

The management of neonatal diabetes is complex due to the very small insulin dose requirements and paucity of subcutaneous fat. In our experience CSII, using diluted insulin is safe, achieved close to target BGLS, and fewer hypoglycaemic events than Lantus insulin or insulin infusion. Added calories and within target BGLs required for adequate growth. The low birth weight was consistent with the genotype 6q24 transient neonatal diabetes.