Rarely do more than a few months go by without us posting yet another pearl about the dangers of hypoxemia or hyperoxemia. The data in general is strong that patients with significant exposure to either very high or very low oxygen levels do worse than patients with normoxia. But if the first step intervention is simple – if the SpO2/PaO2 is high, turn the oxygen support down, and if the SpO2/PaO2 is low, turn the oxygen support up - then in this day and age can we improve this with a closed-loop system? Click the Additional Information to find out!
Additional Information
Douin et al recently published the SAVE-O2 AI trial. It used the O2matic PRO100 (of note, it is approved in Europe and Australia/New Zealand, but NOT currently FDA approved in the US), which autotitrates the oxygen flow rate to maintain a goal SpO2 (usually of 90-96%). This study was not powered to look at patient outcomes, but they did find statistically significant greater time in goal SpO2 range for the autonomous system vs standard care. Their primary outcome was time under an SpO2 of 87%, but I was even more interested in time above 96%, which dropped from 29% in the standard care group to 9% in the autonomous device group, because we know often times we are not good about turning the oxygen DOWN when the saturation is high.
Bottom Line: Autonomous closed-loop oxygen titration devices likely increase time in goal SpO2 range. Whether they impact the outcomes of hospitalized patients remains unknown.