(1393) Occult Hypoxemia in Critically Ill Patients: Is There a Delay to Respiratory Interventions?

Introduction: Occult hypoxemia is defined by arterial oxygen saturation less than 88% while pulse oximetry indicates a saturation of greater than or equal to 92%. Previous studies have shown that occult hypoxemia occurs more frequently in people of color and is associated with higher incidence of organ dysfunction and mortality. Whether the presence of occult hypoxemia impacts time to respiratory interventions has not previously been studied.

Methods: This is a retrospective study of adult patients diagnosed with respiratory failure secondary to COVID-19, admitted to the intensive care unit at Community Regional Medical Center in Fresno, California from February 1, 2020 through February 1, 2022. We compared pulse oximetry readings with arterial blood gas samples to identify the incidence of occult hypoxemia. We evaluated the number and timing of respiratory interventions performed during the patient’s first 24 hours of ICU care. Respiratory interventions were defined as an escalation of the amount of oxygen provided, an increase in the non-invasive or ventilator settings or a higher method of respiratory support. Our primary outcome was time to respiratory intervention. Secondary outcomes were ICU mortality, ICU length of stay, and organ dysfunction by scores.

Results: We identified a cohort of 358 adult patients with hypoxemic respiratory failure secondary to Covid-19 admitted to the ICU during the study period. The incidence of occult hypoxemia in our patient population was 6.1%. Similar to prior studies, we found an increased incidence of occult hypoxemia in patients of color (p = 0.009). We did not find a statistically significant difference in time to respiratory intervention between the occult hypoxemia and non occult hypoxemia groups (average 5.4 hours vs 7.0 hours, p = 0.644). There were also no statistically significant differences in any of our evaluated secondary outcomes.

Conclusions: We did identify occult hypoxemia in our patient population, and similarly to previous studies, this phenomenon was more common in patients of color. However, we were not able to identify a statistically significant difference in time to respiratory interventions or admission to the ICU. This may be due to a smaller than expected study size, and further investigation is warranted.