AI-Driven Oxygen Therapy Outperforms Manual Care in Clinical Trial

The study, led by the University of Colorado Anschutz and published in JAMA Internal Medicine, evaluated 300 adults across four U.S. hospitals. Participants were assigned to either traditional manual titration—where nurses or therapists adjust oxygen flow during periodic checks—or the O2matic PRO100, an autonomous device that monitors blood oxygen via pulse oximetry and adjusts flow in real time. Beyond increasing time within the target range, the automated group experienced fewer instances of hypoxemia and hyperoxemia, all while requiring fewer manual interventions from clinical staff.

Dr. Adit Ginde, the study’s principal investigator, noted that oxygen management remains largely manual despite being one of medicine’s most common therapies. By automating the process, hospitals can reduce the risk of oxygen levels drifting between assessments. The research team is now looking toward broader applications, including prehospital and military settings. Dr. Vik Bebarta, director of the CU Anschutz Combat Medicine Research Center, highlighted the potential for such technology to assist medics in high-stress environments, effectively removing a constant, urgent task from their workload without compromising patient safety.