Prospective Risk Stratification Identifies Healthcare Utilization Associated with Home Oxygen Therapy for Infants with Bronchopulmonary Dysplasia

Abstract

To test whether prospective classification of infants with bronchopulmonary dysplasia identifies lower-risk infants for discharge with home oxygen who have fewer rehospitalizations by 1year after neonatal intensive care unit discharge. This is a prospective single-center cohort that included infants from 2016 to 2019 with bronchopulmonary dysplasia, defined as receiving respiratory support at 36weeks of postmenstrual age. "Lower-risk" infants were receiving ≤2L/min nasal cannula flow, did not have pulmonary hypertension or airway comorbidities, and had blood gas partial pressure of carbon dioxide <70mm Hg. We compared 3 groups by discharge status: lower-risk room air, lower-risk home oxygen, and higher-risk home oxygen. The primary outcome was rehospitalization at 1 year postdischarge, and the secondary outcomes were determined by the chart review and parent questionnaire. Among 145 infants, 32 (22%) were lower-risk discharged in room air, 49 (32%) were lower-risk using home oxygen, and 64 (44%) were higher-risk. Lower-risk infants using home oxygen had rehospitalization rates similar to those of lower-risk infants on room air (18% vs 16%, P=.75) and lower rates than higher-risk infants (39%, P=.018). Lower-risk infants using home oxygen had more specialty visits (median 10, IQR 7-14 vs median 6, IQR 3-11, P=.028) than those on room air. Classification tree analysis identified risk status as significantly associated with rehospitalization, along with distance from home to hospital, inborn, parent-reported race, and siblings in the home. Prospectively identified lower-risk infants discharged with home oxygen had fewer rehospitalizations than higher-risk infants and used more specialty care than lower-risk infants discharged in room air.