Sir.—Weisman et al1 erroneously concluded that suppositories given to healthy term newborns to speed the evacuation of meconium were without effect. This conclusion has been cited elsewhere.2 The data seem to show a reliable statistical difference between a control group and a group treated with suppositories when six measures of serum bilirubin were taken between 12 and 72 hours of age. The practical effect was to only lower serum bilirubin levels by 1 mg/dL on the average in the treated group.
The authors performed multiple t test comparisons between the groups at different ages and then inappropriately extended this to a global conclusion regarding the effect of the treatment. To assess the treatment effect all ages should simultaneously be incorporated in the chosen analysis. A global, nonparametric (even somewhat insensitive) test such as the Wilcoxon signed rank test shows the effect of the treatment to be significant (P