Abstract
Purpose :
Recently, a group of harmonized randomized trials (NeOProM) compared the effect of lower (85-89%) versus higher (91-95%) oxygen saturation (SpO2) targets on the risk of developing severe retinopathy of prematurity (ROP) and other outcomes. Pooled results showed a decreased risk of treated ROP in the lower SpO2 target group, but increased mortality at 18-24 months corrected age, which resulted in increases in SpO2 target policies at some institutions. In contrast to clinical trial data, observational “real world” data reflect clinical variability due to factors such as compliance that can influence outcomes when clinical trial results are translated to clinical practice. Using “real world” data, we evaluated the effects of institutional SpO2 target changes on the risk of developing severe ROP.
Methods :
Secondary analysis of data from 21 North American hospitals that participated in both of two studies completed by the Postnatal Growth and ROP (G-ROP) study group: G-ROP-1 (2006-2012) and G-ROP-2 (2015-2017). Hospitals were categorized into two groups based on whether or not SpO2 target ranges were increased between the two study periods. Severe ROP (ETROP Type 1 or 2 disease) rates were compared within and between groups. Mixed effect models with random intercepts were used to account for intra-center clustering and to calculate BW and GA-adjusted odds ratios (aOR).
Results :
8,119 infants underwent ROP examinations at 21 hospitals during the two study periods: 5,694 during 2006-2012 (mean (SD) BW 1109g (369), GA 28.0 (2.6); 11.6% severe ROP rate) and 2,425 during 2015-2017 (mean (SD) BW 1085g (346), GA 28.0 (2.8); 12.8% severe ROP rate). 13 hospitals increased SpO2 target lower and/or upper limits between 2012 and 2015, and 8 hospitals did not. Hospitals that increased SpO2 targets had a 3% increase in severe ROP (12% to 15%, aOR=1.25, 95% CI: 1.00-1.55, p=0.046), while hospitals without SpO2 target changes had a 2% decrease (11% to 9%, aOR=0.72, 95% CI: 0.52-1.00, p=0.049). The difference between groups was statistically significant (p=0.005).
Conclusions :
Increases in institutional SpO2 targets subsequent to the publication of recent randomized trial findings were associated with an increase in the rate of severe ROP. These “real world” data may help clinicians who are weighing this increase in ROP risk against potential benefits to mortality or other outcomes.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.