Objective: To study the relationship between serum glucose level and mortality in children with septic shock.
Design: Prospective cohort study.
Setting: Twelve-bed pediatric intensive care unit at the Hospital São Lucas da PUCRS, Porto Alegre, Brazil.
Patients: All children admitted with septic shock refractory to fluid therapy over a period of 32 months.
Measurements and main results: Serum glucose levels were measured in all children during the study period, and the highest value was assessed in relation to outcome. Fifty-seven of 1,053 children admitted to the intensive care unit were enrolled in the study. The peak glucose level in those with septic shock was 214 +/- 98 mg/dL (mean +/- SD), and the mortality rate was 49.1% (28/57). In nonsurvivors, the peak glucose level was 262 +/- 110 mg/dL, which was higher (p < .01) than that found in survivors (167.8 +/- 55 mg/dL). The area under the receiver operator curve for peak glucose level and mortality rate was 0.754. The best peak glucose level for predicting death in children with sepsis was 178 mg/dL (sensitivity, 0.714; specificity, 0.724), and the relative risk of death in patients with peak glucose levels of > or =178 mg/dL was 2.59 (range, 1.37-4.88).
Conclusion: In children with septic shock, a peak glucose level of >178 mg/dL is associated with an increased risk of death.