Efficacy and safety of a tight glucose control protocol in critically ill term neonates.

BACKGROUND: A large single-center randomized trial showed that treating hyperglycemia in critically ill children improved outcome, despite an increased incidence of hypoglycemia, especially in infants.

OBJECTIVES: We evaluated the efficacy and incidence of hypoglycemia using a tight glucose protocol in critically ill term neonates.

METHODS: Term hyperglycemic (>8 mmol·l(-1); >144 mg·dl(-1)) neonates treated with a tight glucose protocol during a 3.5-year period in a tertiary pediatric intensive care unit were retrospectively analyzed.

RESULTS: Seventy-three term hyperglycemic neonates [age 0 days (0-6), weight 3.2 ± 0.8 kg, PRISM 16 (11-20)] were included for analysis. Eighteen neonates died (25%). The initial mean (range) glucose level was 11.1 mmol·l(-1) [9.6-15.2; 200 mg·dl(-1) (173-274)], and normoglycemia (<8 mmol·l(-1); <144 mg·dl(-1)) was reached within 5.3 h (1-25) with an overall treatment duration of 27 h (10-57). Seven hypoglycemic incidents (5 times ≤2.2 mmol·l(-1); 40 mg·dl(-1), and 2 times <1.7 mmol·l(-1); 31 mg·dl(-1)) occurred in 5 (6.7%) infants, without severe clinical signs. Three hypoglycemic incidents were directly explained due to a protocol violation. One hypoglycemic incident occurred with the onset of sepsis, while no apparent cause was identified for three hypoglycemic incidents.

CONCLUSIONS: Our glucose protocol was effective, but hypoglycemia occurred more frequently than in older children reported previously. Potential differences in glucose and insulin metabolism in term neonates appear to justify additional safety approaches, while awaiting further studies assessing the benefits of tight glucose protocols in this population. Meanwhile, we have decreased the initial insulin starting doses in our protocol.