Assessment of validity and utility of a bronchopulmonary dysplasia outcome estimator

AbstractObjectiveTo assess the validity and utility of the bronchopulmonary dysplasia (BPD) outcome estimator developed by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD).MethodsThis retrospective study correlated the BPD and mortality outcomes of 23–28‐week prematurely born infants with those predicted by the 2022 NICHD BPD outcome estimator. In addition, evidence‐based recommended postnatal steroid administration using risk predictions by the BPD estimator was compared with actual postnatal steroid use.ResultsAmong the 223 infants enrolled in the study, 139 did not develop BPD, 64 babies developed BPD, and 20 babies died. BPD estimator predicted outcomes correctly in 74% of cases who did not develop BPD, 48% of cases with Grade 1 BPD, and 35% of deaths. Only 6% of cases with Grade 2 BPD and none of the babies with Grade 3 BPD were identified correctly. The predicted and actual steroid use for BPD prevention/treatment were 37.2% and 29.6%, respectively (p = 0.18). The BPD estimator had a sensitivity of 72.7%, specificity of 77.7%, positive predictive value of 57.8%, and negative predictive value of 87.1% to predict actual postnatal steroid use.ConclusionsIn this cohort, the 2022 NICHD BPD estimator underestimated probabilities for babies who did not develop BPD, may overestimate probabilities for babies who develop BPD and had low sensitivity to predict mortality. In addition, the application of the BPD estimator may result in an overestimation of postnatal steroid use.