Objective Our unit protocol for lower alarm settings is 87% and upper is 93% saturation. We audited compliance with these and how frequently spot saturation observations were within these values.

Background The BOOST trial (1) showed that lower oxygen saturation targets of 91–94% versus 95–98%, significantly decreased the numbers with BPD, with NNT = 5.

Method Twice daily recordings were taken on infants of gestational age 32 weeks or less, with a birth weight less than 1500 gm, on supplemental oxygen. Among the recorded measurements were spot oxygen saturations, oxygen saturation alarm settings, mode of ventilation, and fraction of oxygen administered.

Results 16 infants were studied and 165 recordings were taken. The mean GA was 28.4 weeks (SD 2.25), and the mean CGA was 33.3 weeks (SD 4.61). The upper limit alarms were correct in 27%, and the lower in 83.6%. The upper was set at 100% in 38%. The saturations were within the unit’s target range 44% of the time, and were too high in 49%. There were 63 recordings in the subgroup with a CGA less than 34 weeks. The upper limit was correct in 46%, and high in 54%, the lower was correct in 84%. The upper was set at 100% in 30%. The saturations were within the unit’s target range in 74%, and too high in 18%.

Conclusion The audit showed our compliance with the upper alarm settings is similar to that published in 2007 by Clucas (23.3%) (2). Our time within the target range was in keeping with that described in the AVIOx study (3).