Association between intermittent hypoxemia and neurodevelopmental outcomes in extremely premature infants: A single-center experience

Objective: To describe the association between intermittent hypoxemic events (IHEs) and severe neurodevelopmental impairment (SNDI) or death in extremely premature infants.Study design: Retrospective study of extremely premature infants 230/7-276/7 weeks gestational age (GA) and birthweight (BW) <= 1250 grams (g) admitted to a level IV neonatal intensive care unit (NICU) from 2013 to 2017. IHEs, defined as events with SpO(2) <= 80 % lasting 10 s to 5 min, were algorithmically identified using data extracted from bedside monitors at 2 s intervals (0.5 Hz). The primary outcome was SNDI at 18-24 months corrected age (CA), defined as a Bayley-III motor, language or cognitive composite score <= 69, or death before discharge while the secondary outcome was SNDI alone. We used mixed-effects regression models to evaluate the relationship between mean daily IHE rate per postnatal week of life for the first 12 weeks and the outcomes, and logistic regression models to assess the association between outcomes and summary measures of hypoxic burden for the entire NICU hospitalization.Results: The mortality rate was 7 % (18/249) during NICU hospitalization. Of 249 infants born during this time period, IHE and neurodevelopmental outcome data were fully available for 65 infants (mean GA 26 +/- 1.4 weeks, mean birth weight (BW) 738 +/- 199 g. The outcome of SNDI alone occurred in 34 % (22/65) with a majority demonstrating motor or language delay on the Bayley-III. Although mean daily IHE rate/week was not associated with SNDI or death, total IHE duration was associated with increased odds of SNDI (OR (95 % CI) 1.03 (1.01, 1.05), p = 0.008) in models adjusted for GA.Conclusions: In a cohort of extremely premature infants 23-27 weeks GA, each hour of total IHE duration (SpO(2) <= 80 %) was associated with a 2.7 % (0.7 %, 4.8 %) increase in the odds of SNDI at 18-24 months CA.