Automated control of inspired oxygen in mechanically ventilated trauma and acute care surgery patients: A randomized controlled trial

BACKGROUND: Automated physiologic closed-loop control (PCLC) of inspired oxygen fraction (FIO2) may improve oxygen titration while reducing clinician workload. We conducted a randomized noninferiority trial comparing PCLC with protocolized manual control (MC) in mechanically ventilated trauma and acute care surgery patients. METHODS: In this randomized, controlled, multicenter trial, adult mechanically ventilated patients were randomized to PCLC or MC FIO2 titration targeting arterial oxygen saturation (SpO2) of 94%±2%. The primary effectiveness outcome was the percentage of time SpO2 remained within the target range. The primary safety outcome was hypoxemia (SpO2 <88% at any time). A noninferiority margin of 10% was prespecified. Secondary outcomes included oxygen utilization and clinician interventions. Analyses were performed using a modified intention-to-treat approach. RESULTS: One hundred ninety-five subjects completed the trial (98 MC and 97 PCLC). The median time within the SpO2 target range was 36.8% (IQR, 22.1-49.6) with MC and 30.1% (IQR, 18.4-44.7) with PCLC, meeting criteria for noninferiority [the mean±SD of the relative duration on the transformed scale was 0.53 (0.37) and 0.61 (0.36). The mean difference between the transformed data was -0.08 with a one-sided lower 97.5% CL of -0.18. Comparing this lower 95% CL with the prespecified noninferiority margin (NM) of -0.35, this difference lay within the NM. Hypoxemia occurred in fewer patients assigned to PCLC compared with MC (22% vs. 38%, p=0.04). PCLC was associated with lower FIO2 exposure, more frequent automated adjustments, and fewer manual FIO2 adjustments. CONCLUSIONS: In mechanically ventilated trauma and acute care surgery patients, PCLC of FIO2 was noninferior to protocolized MC for maintaining target oxygen saturation. PCLC was associated with fewer hypoxemic events, reduced oxygen usage, and more frequent automated FIO2 changes. (J Trauma Acute Care Surg. 2026;000:000-000. Copyright © 2026 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Association for the Surgery of Trauma.). LEVEL OF EVIDENCE: Randomized Controlled Trial with single blinding; Level II.