Open-label glucocorticoids modulate dexamethasone trial results in preterm infants.

CONTEXT: Open-label glucocorticoids (OLGs) were often used in trials that investigated postnatal dexamethasone treatment in ventilated preterm infants. OBJECTIVE: To determine if OLG use modulates the dexamethasone treatment effect on mortality, bronchopulmonary dysplasia (BPD), and neurodevelopmental outcome. METHODS: Electronic databases, abstracts from the Pediatric Academic Societies, and results of manual reference searches were used as data sources. Fifteen randomized controlled trials comparing dexamethasone with placebo in 721 ventilated preterm infants older than 7 days were identified. The interaction between dexamethasone treatment effect and OLG use was assessed by meta-regression analysis and subgroup meta-analysis according to the percentage of OLG use in the placebo group. Trials with a moderately early (7- to 14-day) or delayed (>3-week) treatment onset were analyzed separately. RESULTS: Moderately early, but not delayed, dexamethasone treatment significantly reduced mortality rates in trials with OLG use at <30% in the placebo arm. Meta-regression analysis revealed that this reduction was inversely related to OLG use. Increasing OLG use strengthened the positive effect of dexamethasone on BPD in the moderately early trials but attenuated the effect in the delayed-treatment trials. In trials with <30% OLG use, dexamethasone increased the risk for cerebral palsy in the delayed, but not the moderately early, treatment trials. CONCLUSIONS: When OLG use is taken into account moderately early dexamethasone treatment reduced mortality rates and the combined outcome mortality and BPD without increasing the risk of adverse neurodevelopmental outcome in ventilated preterm infants. A large randomized controlled trial is needed to confirm or refute these findings. Pediatrics 2010;126:e954-e964