Outcomes of Extremely Premature Infants Comparing Patent Ductus Arteriosus Management Approaches.

OBJECTIVE: To evaluate the change in the proportion of deaths/bronchopulmonary dysplasia (BPD) among premature infants (born <26 and 26-29 weeks of gestational age) following a policy change to a strict nonintervention approach, compared with standard treatment. STUDY DESIGN: We examined 1249 infants (341 born <26 weeks of gestational age) at 2 comparable sites. Site 1 (control) continued medical treatment/ligation, and site 2 (exposed) changed to a nonintervention policy in late 2013. Using the difference-in-differences approach, which accounts for time-invariant differences between sites and secular trends, we assessed changes in death or BPD separately among infants born 26-29 weeks and <26 weeks of gestational age in 2 epochs (epoch 1: 2011-2013; epoch 2: 2014-2017). RESULTS: Baseline characteristics were similar across sites and epochs. Medical treatment/ligation use remained stable at site 1 but declined progressively to 0% at site 2, indicating adherence to policy. We saw no difference in death/BPD among infants born at 26-29 weeks of gestational age (12%, 95% CI -1% to 24%). However, incidence of death/BPD increased by 31% among infants born <26 weeks of gestational age (95% CI 10%-51%) in site 2, whereas there was no change in outcomes in site 1. The Score for Neonatal Acute Physiology-Version II, used as a control outcome, did not change in either site, suggesting that our findings were not due to changes in patients' severity. CONCLUSIONS: Adherence to a strict conservative policy did not impact death or BPD among 26 weeks but was associated with a significant rise in infants born <26 weeks.