Although survival of very low birth weight (VLBW, <1500 g) infants has improved, many of these infants sustain chronic lung injury that is a significant health burden. Previous studies within the NICHD Neonatal Research Network have shown substantial differences in the incidence of BPD between centers. These differences were not explained by birth weight, gestational age, race, frequency of antenatal steroid use, or incidence of respiratory distress syndrome. Practice differences may contribute to BPD incidence variation. This study will evaluate the efficacy of a Benchmarking Initiative to modify clinical care practices and decrease incidence of BPD in VLBW infants.

Benchmarking is a method involving detailed comparisons of processes between similar organizations. For this study, three Network centers with the lowest rates of BPD have been identified as benchmark centers. During a 6-month pre-intervention period, details of care practices and management style at these three centers will be carefully assessed. The baseline rate of BPD at each center will be measured for a 1-year period using a physiologic definition of BPD designed for this study. This definition of BPD is based on a measurement of room air oxygen saturation.

After the pre-intervention period, 14 study sites within the Network will be randomized to either the benchmarking intervention or usual care practices. The benchmarking intervention will consist of data feedback, training in continuous quality improvement techniques, evidence review, site visits to the 3 benchmark centers, and identification of practice changes. The intervention will last 1 year. Change in rate of survival without BPD at 36 weeks corrected age will be compared between the intervention sites and the control sites. The rate of BPD at each site in the post-intervention period will also be compared to its pre-intervention rate.