Objective: To evaluate an implementation of lower oxygen saturation targets with retinopathy of prematurity (ROP) as primary outcome, in infants at the lowest extreme of prematurity.
Study design: Retrospective cohort including infants born at 22-25 weeks of gestation in 2005-2015 (n = 325), comparing high (87-93%) and low (85-90%) targets; infants transferred early were excluded from the main analysis to avoid bias.
Results: Overall survival was 76% in high saturation era, and 69% in low saturation era (p = .17). Treatment-requiring ROP was less common in low saturation group (14% vs 28%, p < .05) with the most prominent difference in the most immature infants. Including deceased infants in the analysis, necrotizing enterocolitis was more frequent in low saturation era (21% vs 10%, p < .05).
Conclusions: Implementing lower saturation targets resulted in a halved incidence of treatment-requiring ROP; the most immature infants seem to benefit the most. An association between lower oxygenation and necrotizing enterocolitis cannot be excluded.