Efficacy of pharmacologic therapies for reducing bronchopulmonary dysplasia and mortality in preterm infants: a network meta-analysis.

backgroundBronchopulmonary dysplasia (BPD) is a major complication of prematurity and is associated with substantial respiratory morbidity, impaired growth, and increased mortality in severe cases. Pharmacologic prevention strategies have been evaluated in randomized trials, but their comparative effects remain uncertain.

aimThis network meta-analysis aimed to summarize the comparative evidence for pharmacologic interventions evaluated for reducing BPD incidence and pre-discharge mortality in preterm infants.

methodsWe searched PubMed, Embase, Cochrane Library, and Web of Science for randomized controlled trials (RCTs) investigating pharmacological therapies for BPD prevention, with a search cutoff of September 1, 2025. Data analysis was conducted using R 4.2.3 software. The primary outcome was BPD incidence, and the secondary outcome was pre-discharge mortality. BPD definitions and assessment timepoints were extracted from individual studies and interpreted considering variation across NIH 2001, NIH 2018, Shennan, or study-specific criteria.

resultsThirty-six randomized controlled trials involving 11,414 preterm infants were included in the systematic review. Of these, 24 trials contributed to the BPD incidence network, and 19 trials contributed to the pre-discharge mortality network. Several interventions showed comparative signals for lower odds of BPD incidence or pre-discharge mortality when compared with placebo; however, the strength and precision of these signals varied across comparisons. In the BPD incidence network, some estimates excluded the null, but several comparisons were informed partly by indirect evidence or sparsely connected nodes. In the pre-discharge mortality network, favorable estimates were less certain because mortality events were sparse and several 95% credible intervals (CrIs) were wide. The certainty of evidence was moderate for BPD incidence and low for pre-discharge mortality. SUCRA rankings were therefore interpreted only as probabilistic summaries within the network and not as evidence of clinical superiority.

conclusionThis network meta-analysis identified several comparative signals for pharmacologic prevention of BPD and reduction of pre-discharge mortality in preterm infants. However, these findings are constrained by low-to-moderate certainty of evidence, indirect comparisons, sparse network nodes, and imprecision in several estimates. The results should therefore be interpreted as evidence-mapping and hypothesis-generating rather than as treatment recommendations. Future studies should prioritize direct head-to-head comparisons, phenotype- or severity-specific analyses, and long-term respiratory and neurodevelopmental outcomes.