Mode of Respiratory Support at 36 Weeks Predicts Weaning Timeline for Infants With Severe Bronchopulmonary Dysplasia

Bronchopulmonary dysplasia (BPD) is the most common complication of preterm birth, and its incidence is rising due to advances in neonatal care that have improved survival for extremely premature infants. Respiratory support is a cornerstone of management for these patients, yet the expected timeline for weaning in those with moderate-to-severe (grade 2/3) BPD is not well defined. A clearer understanding of in-hospital respiratory trajectories and the time required to achieve liberation from respiratory support (LRS) could refine the characterization of this heterogeneous patient population, improve parental counseling, and guide future research priorities.

“Since BPD is defined at 36 weeks by the treatment a baby is receiving, the point at which a patient can finally come off that support represents a major clinical milestone and real lung healing,” said attending neonatologist Matthew Kielt, MD, who practices in the Comprehensive Center for Bronchopulmonary Dysplasia at Nationwide Children’s Hospital.

Dr. Kielt and colleagues conducted a large multicenter retrospective cohort study, published in The Journal of Pediatrics, examining the association between respiratory support mode at 36 weeks’ postmenstrual age (PMA) and time to LRS in infants with grade 2/3 BPD. The analysis included 3,483 infants born at <32 weeks’ gestation and admitted to one of 41 U.S. and Canadian level IV NICUs participating in the Children’s Hospitals Neonatal Consortium between 2017 and 2022.

At 36 weeks’ PMA, 17% of infants were on high-flow nasal cannula (HFNC >2 L/min), 36% on continuous positive airway pressure (CPAP), 16% on non-invasive positive pressure ventilation (NIPPV), and 32% on invasive mechanical ventilation (MV). LRS was defined as weaning to <2 L/min nasal cannula or room air for >2 days.

After excluding infants who died, required tracheostomy, or were transferred before achieving LRS, the median time to LRS was shortest for HFNC (1.3 weeks after 36 weeks’ PMA), followed by CPAP (2.9 weeks), NIPPV (5.1 weeks), and MV (7.7 weeks). These differences persisted after adjustment for gestational age, small-for-gestational age status, tracheobronchomalacia, BPD-associated pulmonary hypertension, and center. Overall, 80.7% achieved LRS by 66 weeks’ PMA, with rates ranging from 95% for HFNC to 59% for MV.

“On one hand, the results weren’t surprising—babies on higher levels of support at 36 weeks took longer to wean. But what was striking was the heterogeneity within grade 2 BPD, where different non-invasive modes showed very different timelines. That tells us we may need to better describe distinct subgroups within this population,” added Dr. Kielt, who is also an associate professor of Clinical Pediatrics at The Ohio State University College of Medicine.

Reintubation after initial LRS was common, often for surgical procedures, but most infants were successfully weaned again without requiring long-term invasive support.

Notably, time-to-LRS varied by as much as 10 weeks across centers, independent of patient and clinical characteristics, underscoring the potential influence of institutional practices, resources, and care protocols. The authors suggest that where an infant is treated may be as influential as their baseline clinical status in determining the pace of respiratory recovery.

“We were able to leverage a very high-resolution, multicenter dataset that tracked daily changes in respiratory support. That allowed us to model time to liberation and even build a prediction calculator, which could eventually help neonatologists counsel families about what to expect,” said Dr. Kielt.

From a clinical standpoint, the study provides a quantifiable benchmark for anticipating respiratory progress in infants with severe BPD. For neonatologists, knowing the likely time to LRS based on support mode at 36 weeks could facilitate more precise counseling for families, better planning for developmental and feeding interventions, and more informed discharge preparation. The authors also propose that time-to-LRS could serve as a pragmatic, objective inpatient outcome metric for severe BPD in future research and quality improvement initiatives.

The authors call for further investigation into the drivers of inter-center variation, refinement of LRS definitions, and evaluation of the relationship between time-to-LRS and long-term neurodevelopmental outcomes. Ultimately, understanding these trajectories could help optimize care strategies and improve outcomes for this vulnerable population.

“We now have some good clinical predictors of when babies might come off support, but they’re not perfect. What I’m working on now is whether biologic data, such as inflammatory biomarkers from tracheal aspirates and blood samples, can strengthen those models and give us a clearer picture of which babies will recover faster,” noted Dr. Kielt. “That prospective study is underway, and we hope it will add another dimension to predicting outcomes for infants with severe BPD.”

Reference

Kielt MJ, Zaniletti I, Lagatta JM, Padula MA, Grover TR, Porta NFM, Wymore EM, Jensen EA, Leeman KT, Levin JC, Evans JR, Yallapragada S, Nelin LD, Vyas-Read S, Murthy K; Children’s Hospitals Neonatal Consortium Severe BPD Focus Group. Liberation from Respiratory Support in Bronchopulmonary Dysplasia. Journal of Pediatrics. 2025 Jul;282:114390.