Predictors of home oxygen duration in chronic neonatal lung disease.

Abstract

In infants with chronic neonatal lung disease (CNLD), we aimed to identify predictors of home oxygen duration, predictors of discharge oxygen flow rates, and the association of oxygen flow rates with respiratory outcomes. Infants with CNLD requiring home oxygen in 2016 and 2017 were retrospectively reviewed. Hazard ratios (HR) were estimated from Cox proportional hazards regression models in the cohort. A multinomial logistic regression model examined the effects of maternal and infant variables on discharge oxygen flow rates. Kruskal-Wallis test with univariate linear regression and Fisher's exact test with binomial univariate logistic regression were used to examine associations between oxygen flow groups and post-discharge clinical variables. One hundred and forty-nineinfants were included. Median corrected gestational age (CGA) at oxygen cessation was 6.8 months (interquartile range, 4.4) with 87.2% of infants weaned by 12 months CGA. Shorter initial neonatal intensive care unit (NICU) stay predicted faster oxygen weaning at 9 months (HR, 0.99; 95% confidence interval [CI], 0.98-1.00, p = .02) and 12 months (HR, 0.99; 95% CI, 0.98-1.00, p = .02). Infants with hypercarbia at discharge or discharged from NICU at higher CGA had higher odds of requiring ≥ 200 ml/min relative to ≤ 125 ml/min oxygen. Infants discharged with > 250 ml/min oxygen were more likely to have a respiratory-related admission before 2 years chronological age. Shorter initial NICU stay was the best predictor of earlier home oxygen cessation. At NICU discharge, infants with hypercarbia or a higher CGA may require more home oxygen and experience more respiratory-related hospital admission in the first 2 years of chronological age.