S01E56 SAVE-O2 AI: Can Oxygen Titrate Itself Better Than We Can?

In this episode of PRISM Rounds, we discuss the SAVE-O2 AI randomized clinical trial, which tested autonomous closed-loop oxygen titration versus usual manual oxygen adjustment in 300 hospitalized adults receiving supplemental oxygen. The autonomous system kept patients in the target SpO2 range much more often, 85% vs 63%, while also reducing time spent in hypoxemia and, even more strikingly, hyperoxemia. We break down what the trial shows, what it does not show, the importance of pulse oximetry accuracy and skin pigmentation, and whether improved oxygen control is enough to justify broader implementation. We also discuss the accompanying editorial by Weingart and Matthay, which explores the potential impact on clinician workload, oxygen stewardship, resource-limited settings, and the next generation of implementation and outcomes studies. SAVE-O2 AI trial, JAMA Internal Medicine https://jamanetwork.com/journals/jama... DOI: https://doi.org/10.1001/jamainternmed... Invited Commentary: Potential Transformative Impact of Autonomous Oxygen Delivery in Hospitalized Patients https://jamanetwork.com/journals/jama... DOI: https://doi.org/10.1001/jamainternmed... Find more episodes at www.prismrounds.com (http://www.prismrounds.com) and subscribe on Apple Podcasts or Spotify. Tags #PRISMRounds #SAVE02AI #CriticalCare #ICU #PulmonaryCriticalCare #OxygenTherapy #Hypoxemia #Hyperoxemia #PulseOximetry #ClosedLoopControl #ArtificialIntelligence #DigitalHealth #ClinicalTrials #RandomizedTrial #JAMAInternalMedicine #RespiratoryCare #FOAMcc #MedEd