Background Five trials in infants <28 weeks' gestation are investigating outcomes after randomisation to SpO2 targets of 85–89% vs 91–95%. In the SUPPORT trial, the high target increased severe ROP (p<0.001) and survival to discharge (Relative Risk (RR) 1.27, 95% CI 1.01 to 1.60, p=0.04). The DMCs of the UK, Australian and New Zealand BOOST II trials found no reason to stop recruiting. These trials used similar oximeters except that by early 2009, a revised calibration algorithm was installed that improved SpO2 targeting.

Objective In December 2010, the UK and ANZ DMCs performed an interim safety analysis of survival to 36 weeks gestation in 2315 infants in the UK and ANZ trials and the 1316 infants in SUPPORT.

Design/methods Guidelines specified that investigators be unblinded if a difference between groups in survival exceeded three standard errors (99.73% CI, equivalent to p<0.0027) for all infants or for those recruited after revising the oximeter calibration.

Results The high SpO2 target was associated with higher survival in all 3631 infants (RR 1.21 (99.73% CI, 0.96 to 1.52, p=0.015), test for interaction for RRs for new vs old algorithm: p=0.006) and in the 1055 UK and ANZ infants enrolled after revising the oximeters (RR 1.65 (99.73% CI 1.09 to 2.49, p=0.0003, risk difference 8.6%)). The UK and Australian trials closed recruitment.

Conclusions Survival to 36 weeks gestation is higher when SpO2 is targeted at 91–95%. Until the five trials report disability-free survival to 2 years, we consider it prudent not to target SpO2 85–89%.