Naik and Das claim that 5 of 8 trials included in our systematic review did not favor repeat doses for respiratory distress syndrome (RDS). In fact, in 7 trials the point estimate favored repeat doses, and there was no significant heterogeneity.

They also claim that only 1 of 5 trials reported a reduced risk of severe RDS. In fact, statistically significant reductions in severe RDS were found in 2 trials, and 2 more showed a similar point estimate.

Forest plots were not included because of the article size limits but are available in the Cochrane Library. A funnel plot is designed to assess reporting bias, not internal biases. We assessed the risk of internal bias to be low except for 1 small unblinded trial. That unblinded trial and the pilot study contributed little to the weighted average of intervention effects and thus have minimal impact on the meta-analyses.

The funnel plot for RDS does indicate a possible gap for small negative trials, but their simulated inclusion would not change the result. Furthermore, we extensively sought additional trials for our ongoing individual patient data metaanalysis that aims to explore some of the differences in treatment regimens such as dose and number of courses.

Naik and Das suggest that the findings in 2 trials of potential isolated adverse neurodevelopmental outcomes may be type 2 errors. The objective of systematic review is to help with robust interpretation of such isolated positive or negative results. The outcomes we reported from 3800 children (4 trials) are very reassuring in showing no differences between groups.