AI System Stabilizes Oxygen Levels More Reliably Than Standard Care

Maintaining safe oxygen levels in hospitalized adults is a constant challenge. Manual titration can allow dangerous dips in oxygenation, known as hypoxemia, or excessive oxygen levels, known as hyperoxemia, between checks while adding to clinician workload. These fluctuations can affect recovery across emergency, surgical, and inpatient care. A new multicenter randomized trial shows that an artificial intelligence-enabled oxygen delivery system can stabilize oxygenation more reliably than standard care.

In the SAVE-O2 AI trial led by the University of Colorado Anschutz School of Medicine, investigators evaluated autonomous oxygen titration using the O2matic PRO100. The system integrates continuous pulse oximetry with real-time algorithms to adjust oxygen flow without waiting for clinician input. Its goal is to maintain patients within a predefined target oxygen saturation range throughout day and night.

The randomized trial enrolled 300 adults at four U.S. hospitals, including UCHealth University of Colorado Hospital. Participants were hospitalized with acute respiratory illness, traumatic injury, or burns, or were recovering from surgery, and were receiving new supplemental oxygen. Patients were assigned to clinician-managed oxygen therapy or to automated titration with the O2matic PRO100.

Patients managed by the automated system spent 85% of their time in the target oxygen range, compared with 63% under standard care. The automated approach reduced time spent in hypoxemia and hyperoxemia and decreased the number of manual adjustments required by clinical staff. Serious adverse events did not increase with automation.

The results were published in JAMA Internal Medicine and presented at the Military Health System Research Symposium. Collaborating institutions included Vanderbilt University Medical Center, Oregon Health & Science University, Wake Forest University School of Medicine, and UCHealth University of Colorado Hospital.

Investigators noted that real-time automation may reduce workload for nurses and respiratory therapists by limiting frequent bedside adjustments while improving the precision of oxygen delivery. The team plans further evaluation of the technology in emergency transport and prehospital settings.

“Oxygen is one of the most widely used therapies in medicine, yet even in 2026, it is still managed largely through repeated manual adjustments made by clinicians. Our findings show that autonomous oxygen titration can help patients remain in their target range more consistently while reducing both underoxygenation and overoxygenation. This technology has the potential to fundamentally change how supplemental oxygen is delivered in both civilian and military medicine,” said Adit Ginde, MD, MPH, senior associate dean for clinical research and professor of emergency medicine at the University of Colorado Anschutz School of Medicine and principal investigator of the study.

Related Links
University of Colorado Anschutz School of Medicine