Further Considerations on Corticosteroid Use in Premature Infants

To the Editor Doyle et al conducted a meta-regression to analyze the association of cerebral palsy (CP) with postnatal corticosteroids administered to premature infants. Their findings suggest that dexamethasone may be beneficial for preventing bronchopulmonary dysplasia (BPD) but continue to caution its use due to an apparent association with CP in follow-up studies. However, model performance measures for these results were unreported. We conducted similar research using robust statistical methods that included multivariable meta-regression using the Sidik-Jonkman estimator and Knapp-Hartung adjustment, along with permutation tests using 10 000 iterations. We concluded that dexamethasone was effective for preventing BPD without a significant increase in CP. However, both analyses incurred important limitations as neither study could rule out congenital or other acquired causes of CP. Further, the overall rate of CP, in both steroid and control groups, was significantly higher in the study by Doyle et al compared to ours (32% vs 13.2%).