(882) Probability of Hyperoxia Based on Oxygen Saturation in Critically Ill Children

Introduction: Hyperoxia, a consequence of excessive oxygen administration, has been associated with increased morbidity and mortality rates in critically ill patients. While the definition of hyperoxia remains inconsistent, it is clear that the degree of hyperoxia correlates with worse patient outcomes. In the pediatric intensive care unit (PICU), identifying hyperoxia poses challenges due to limited arterial blood gas measurements. Non-invasive Saturation of Peripheral Oxygen (SpO2) monitoring offers a valuable alternative, but the correlation between SpO2 and hyperoxia remains understudied.

The objective of our study is to explore the likelihood of hyperoxia at various SpO2 values among critically ill children and identify a specific SpO2 cutoff point where the risk of hyperoxia is minimized

Methods: We conducted a retrospective study of 316 mechanically ventilated children in the PICU. SpO2 and PaO2 data were collected, and PaO2-SpO2 couplets were clustered based on SpO2 values for probability analysis. We calculated the probability of hyperoxia for three PaO2 cutoff values: >100, >125, and >250 mmHg.

Results: Seven hundred and eighty-three SpO2 and PaO2 data pairs were analyzed. The average age was 7.6 years, and 58% were male. The average duration of mechanical ventilation was 12.4 days. The probability of PaO2 > 250 was low except at SpO2 values of 99 or 100 (5% and 2%, respectively). With SpO2 in the high range (97-100%), the probability of PaO2 > 100 was 26-77%, and >125 was 12-55%. To avoid hyperoxia in more than 90% of the time, SpO2 of 97-100% should be avoided, and the 88-94% range limits hyperoxia to less than 5%.

Conclusions: Our study supports maintaining oxygen saturation in the low normal range to avoid supraphysiologic PaO2 levels. The findings align with studies in adults, highlighting the consistency of results across patient populations.To minimize the likelihood of hyperoxia, we recommend setting SpO2 goals for critically ill and mechanically ventilated children at or below 96%. Understanding the SpO2-PaO2 relationship will aid in optimizing oxygen therapy and improving patient outcomes in the PICU. Future research should investigate the dose-dependent effects of hyperoxia and identify high-risk patient populations to refine oxygen therapy practices