Home Oxygen Use and 1-Year Readmission among Infants Born Preterm with Bronchopulmonary Dysplasia Discharged from Children's Hospital Neonatal Intensive Care Units

Abstract

Objective: To determine associations between home oxygen use and 1-year readmissions for preterm infants with bronchopulmonary dysplasia (BPD) discharged from regional neonatal intensive care units. Study design: We performed a secondary analysis of the Children's Hospitals Neonatal Database, with readmission data via the Pediatric Hospital Information System and demographics using ZIP-code-linked census data. We included infants born <32 weeks of gestation with BPD, excluding those with anomalies and tracheostomies. Our primary outcome was readmission by 1 year corrected age; secondary outcomes included readmission duration, mortality, and readmission diagnosis-related group codes. A staged multivariable logistic regression was adjusted for center, clinical, and social risk factors; at each stage we included variables associated at P < .1 in bivariable analysis with home oxygen use or readmission. Results: Home oxygen was used in 1906 of 3574 infants (53%) in 22 neonatal intensive care units. Readmission occurred in 34%. Earlier gestational age, male sex, gastrostomy tube, surgical necrotizing enterocolitis, lower median income, nonprivate insurance, and shorter hospital-to-home distance were associated with readmission. Home oxygen was not associated with odds of readmission (OR, 1.2; 95% CI, 0.98-1.56), readmission duration, or mortality. Readmissions for infants with home oxygen were more often coded as BPD (16% vs 4%); readmissions for infants on room air were more often gastrointestinal (29% vs 22%; P < .001). Clinical risk factors explained 72% of center variance in readmission. Conclusions: Home oxygen use is not associated with readmission for infants with BPD in regional neonatal intensive care units. Center variation in home oxygen use does not impact readmission risk. Nonrespiratory problems are important contributors to readmission risk for infants with BPD.