RESUMO
OBJECTIVE: To assess the clinical applicability of the current definitions of bronchopulmonary dysplasia (BPD) and their ability to predict subsequent respiratory outcomes. STUDY DESIGN: In this retrospective cohort study, data were collected from 104 preterm infants (mean gestational age, 25.8 weeks) diagnosed with BPD between 2010 and 2018 at a single tertiary care center. The infants were classified according to the following definitions: 2001 National Institute of Child Health and Human Development (NICHD), 2017 Canadian Neonatal Network (CNN), 2018 NICHD, and 2019 Neonatal Research Network (NRN). Logistic regression and the area under the receiver operating characteristic curve (AUC) were used to assess the predictive ability of each definition on mortality or severe respiratory morbidity at 18-24 months of age. RESULTS: All patients could be adequately classified by each definition, except for the 2001 NICHD definition, in which 11.4% were unclassifiable. The prevalence of BPD was 49% by the 2017 CNN definition and 70% by the 2018 NICHD and 2019 NRN definitions. The best regression model was the one that included the 2018 NICHD definition, which had an accuracy of 85.6% and a significantly higher AUC compared with the 2001 NICHD (0.891 vs 0.824; P = .015) and 2017 CNN (0.891 vs 0.811; P = .036) definitions, but not compared with the 2019 NRN definition (0.891 vs 0.833; P = .09). CONCLUSIONS: The current definitions of BPD showed a good predictive ability for mid-term respiratory outcomes, with the highest ability for the 2018 NICHD definition. Further studies are needed to establish the most appropriate definition of BPD.
