The Objective Use of Pulse Oximetry to Predict Respiratory Support Transition in Preterm Infants: An Observational Pilot Study

BACKGROUND: Preterm infants often require some form of respiratory support with supplemental oxygen and are monitored by continuous pulse oximetry (S-pO2). The study objective was to determine whether the histogram distribution of Spot over a 24-h period will predict readiness for weaning respiratory support in preterm infants. We hypothesize that infants with >= 15% of time spent with S-pO2 <86% before transitioning from CPAP or high-flow nasal cannula (HFNC) to low-flow nasal cannula, oxyhood, or room air are more likely to fail transitioning. METHODS: The S-pO2 histograms were downloaded daily for 31 infants, 24-32 weeks gestational age, before transition from CPAP or HFNC to low-flow nasal cannula, oxyhood, or room air. The S-pO2 histogram downloads were continued for each infant for 1 week after transition or when the infant reached 36 weeks postmenstrual age or when S-pO2 monitoring was discontinued. Failure was defined as an increase in respiratory support within 72 h of transitioning. We compared the percentage of time for each S-pO2 quintile for the 24-h periods before and immediately following CPAP/HFNC transitioning between groups. RESULTS: Twenty-four subjects transitioned successfully, but 7 subjects failed. Two of 7 subjects (28.6%) who failed transition experienced S-pO2 <86% >= 15% of the time pretransition compared with none in the success group (P = .045). The failure group experienced S-pO2 <86% 10.7 +/- 11.9% of time pre-wean compared with 3.3 +/- 4.7% of time in the success group (P = .02). In contrast, infants who were successfully weaned tended to experience a greater percentage of time with S-pO2 >95% compared with the failure group, both pre-wean (54.3 +/- 36.1% vs 33 +/- 27.7%, P = .16) and post-wean (52 +/- 35.4% vs 27.4 +/- 27.7%, P = .10). CONCLUSIONS: These data suggest that pulse oximetry histograms may be useful in assessing CPAP/HFNC support transition readiness.