Alterations in glucose homeostasis in the pediatric intensive care unit: Hyperglycemia and glucose variability are associated with increased mortality and morbidity

Abstract

Objectives: Critically ill patients with alterations in glucose equilibrium may experience adverse outcomes. We sought to describe the distribution of blood glucose values in the absence of insulin therapy and to evaluate the association of hyperglycemia, hypoglycemia, and glucose variability with mortality and morbidity of critically ill children.

Design: Retrospective cohort analysis.

Setting: University-affiliated children's hospital pediatric intensive care unit (PICU).

Patients: All children admitted to the PICU for >24 hrs with at least one blood glucose level recorded from a 1-yr period. Patients were excluded if >18 yrs of age, if insulin was administered during their PICU stay, and if the PICU admitting diagnosis included diabetes mellitus or hypoglycemia.

Interventions: None.

Measurements and main results: Patients were categorized with isolated hyperglycemia (blood glucose >or=150 mg/dL, [>or=8.3 mmol/L]), isolated hypoglycemia (blood glucose <or=60 mg/dL, [<or=3.3 mmol/L]), and glucose variability (both hyper- and hypoglycemia), and the associations with mortality, hospital length of stay, and nosocomial infections were assessed. Fisher's exact test, Kruskal-Wallis test, and logistic and linear regression were used to test for associations. Hyperglycemic and hypoglycemic measurements occurred in 56.1% and 9.7% of all patients, respectively. Glucose variability occurred in 6.8% of all patients. Glucose variability (odds ratio 63.6; 95% confidence interval, 7.8-512) and hyperglycemia (odds ratio 11.1; 95% confidence interval, 1.5-85.6) in the univariate analysis were associated with increased mortality. There were no deaths among patients with isolated hypoglycemia. Hyperglycemia and glucose variability were also associated with nosocomial infections (p = .01) and increased hospital length of stay (p < .001). Hypoglycemia and glucose variability occurred more commonly in younger patients (p < .001).

Conclusions: We found a relationship between blood glucose level and PICU patient outcomes. The relationship is similar to that found in adults and raises the question whether attention to control of blood glucose will improve outcomes in critically ill children.