Adjunctive corticosteroid therapy in pediatric severe sepsis: observations from the RESOLVE study.

OBJECTIVE To assess whether corticosteroids, used as adjunctive therapy for pediatric severe sepsis, is associated with improved outcomes. DESIGN Retrospective cohort study examining the clinical database derived from the RESOLVE (REsearching severe Sepsis and Organ dysfunction in children: a gLobal perspective, F1K-MC-EVBP) trial of activated protein C for pediatric severe sepsis. SETTING A total of 104 pediatric centers in 18 countries from which data were originally gathered. SUBJECTS Children with severe sepsis (n = 477), requiring both vasoactive-inotropic infusions and mechanical ventilation. Within this cohort, 193 children received corticosteroids during their septic episode and 284 did not. INTERVENTIONS None. MEASUREMENTS AND MAIN RESULTS Baseline summary characteristics demonstrated that children receiving or not receiving corticosteroids had similar demographics and disease severity as indicated by age, gender, mean Pediatric Risk of Mortality scores, and mean number of organ dysfunctions. Use of adjunctive corticosteroids increased during the F1K-MC-EVBP trial. Indications for corticosteroid prescription were therapeutic (89%, mostly shock) and prophylactic (13%). All cause 28-day mortality among children receiving and not receiving corticosteroids was 15.1% and 18.8%, respectively, p = .30. There was no difference in mean vasoactive-inotropic infusion days between the corticosteroid and no corticosteroid groups, 4.5 days vs. 4.3 days, respectively, p = .59. Similarly there was no difference in mean ventilator days between the corticosteroid and no corticosteroid groups, 8.3 days vs. 7.7 days, respectively, p = .38. CONCLUSIONS Children with severe sepsis who received adjunctive corticosteroid therapy exhibited similar illness severity compared with those who did not. No definitive improvement in outcomes can be attributable to adjunctive corticosteroid therapy in the largest pediatric sepsis trial conducted to date.