Objective: To evaluate the differences in the application of an insulin infusion algorithm protocol for the treatment of hyperglycaemia in the paediatric intensive care unit (PICU) between nurses and physicians.

Design: Prospective observational study on two PICU.

Methods: All consecutively admitted critically ill children with hyperglycaemia needing insulin therapy were included. Hyperglycaemia >8 mmol/l was treated according to a strict insulin protocol. The adjustments in insulin infusion rate (decrease, increase and stop of insulin infusion) according to the protocol were executed either by nurses on unit 1 or by physicians on unit 2.

Results: In an 18-month period 136 children (77 male), aged 3.2 years (1 day–19 years) with mixed diagnoses were treated. 59 children were treated on unit 1 and 77 on unit 2. There were no significant differences in age, disease severity, glucose intake (3.7 vs 2.8 mg/kg per minute), steroid use and first glucose level at start of insulin treatment (11.2 vs 11.2 mmol/l) between both groups. There were significant differences between unit 1 and unit 2 concerning duration until normoglycaemia (5 vs 10.5 h), duration of insulin treatment (31 vs 34 h) and the occurrence of hypoglycaemia <2.6 mmol/l (5.1 vs 14.3% of the patients).

Conclusion: Treatment of hyperglycaemia in critically ill children by PICU nurses following a strict insulin algorithm protocol seems to give faster achievement of normoglycaemia and less occurrence of hypoglycaemia.