Noninvasive Respiratory Severity Indices Predict Adverse Outcomes in Bronchopulmonary Dysplasia

Objective To test the hypothesis that elevated respiratory severity indices will identify patients with severe bronchopulmonary dysplasia (BPD) at the greatest risk for adverse in-hospital outcomes. Study design This was a retrospective cohort study. A modified respiratory severity score (mean airway pressure x fraction of inspired oxygen) and a modified pulmonary score (respiratory support score x fraction of inspired oxygen + sum of medication scores) were calculated in a consecutive cohort of patients 36 weeks of post-menstrual age with severe BPD admitted to a referral center between 2010 and 2018. The association between each score and the primary composite outcome of death/prolonged length of stay (>75th percentile for cohort) was assessed using area under the receiver operator characteristic curve (AUROC) analysis and logistic regression. Death and the composite outcome death/tracheostomy were analyzed as secondary outcomes. Results In 303 patients, elevated scores were significantly associated with increased adjusted odds of death/prolonged length of stay: aOR 1.5 (95% CI 1.3-1.7) for the modified respiratory severity score and aOR 11.5 (95% CI 5.5-24.1) for the modified pulmonary score. The modified pulmonary score had slightly better discrimination of death/prolonged length of stay when compared with the modified respiratory severity score, AUROC 0.90 (95% CI 0.85-0.94) vs 0.88 (95% CI 0.84-0.93), P = .03. AUROCs for death and death/tracheostomy did not differ significantly when comparing the modified respiratory severity score with the modified pulmonary score. Conclusions In our referral center, the modified respiratory severity score or the modified pulmonary score identified patients with established severe BPD at the greatest risk for death/prolonged length of stay, death, and death/tracheostomy.