Active treatment vs. expectant management of patent ductus arteriosus in preterm infants: critical analysis of a meta-analysis

A recent meta-analysis on the active versus expectant management of patent ductus arteriosus in preterm infants, including 10 randomized control trials involving 2035 infants, found that the incidence of death at 36 weeks or moderate to severe bronchopulmonary dysplasia (BPD) and of death was higher in the active treatment group. The accompanying editorial suggested that “the use of early routine pharmacologic treatment no longer appears to be justifiable”. However, this meta-analysis is based on highly heterogeneous studies due to different diagnostic criteria used to define hemodynamically significant PDA, postnatal age at which the echocardiography was performed, pharmacological treatment, high frequency of open label treatment in the placebo/expectant management group. Moreover, authors did not explain the reduction in mortality that they reported. We concluded that clinical and methodological heterogeneity among studies may not be detected by statistical heterogeneity tests, reducing the interpretability of the results and weakening causal inferences. The high occurrence of open-label treatment indicates the need to perform a per-protocol analysis, which would have been possible if the authors had performed an individual participant data network meta-analysis. Furthermore, the current lack of pathophysiological plausibility suggests caution in interpreting the results of this meta-analysis.