Bronchopulmonary dysplasia: its relationship to invasive modes of ventilation and the magnitude of respiratory support.

OBJECTIVE: To determine whether the magnitude of respiratory support mediates the association between intubation/invasive ventilation and the incidence of bronchopulmonary dysplasia (BPD). STUDY DESIGN: Observational study of BPD (diagnosed at 40 weeks' postmenstrual age) using Poisson regression and Mplus mediation analyses in 544 consecutively admitted inborn infants <28 weeks' gestation. RESULTS: A change in guidelines emphasizing noninvasive ventilation led to a significant decrease in intubation rates. However, neither the magnitude of support nor the incidence of BPD-40 was decreased. While invasive ventilation at 1, 7 and 14 days was significantly associated with BPD-40 in multivariable regression and mediation analyses, its association appeared to be mediated by the magnitude of support (especially mean airway pressure), with no direct effect from the intubation/invasive ventilation. CONCLUSION: Reducing intubation/invasive ventilation may not impact the rate of BPD-40 unless the infant's need for high levels of respiratory support (even if delivered noninvasively) is also minimized.