A new study has found that using artificial intelligence (AI) to automate oxygen delivery can increase the amount of time hospital patients spend receiving the appropriate level of oxygen.
Researchers reported at the Military Health System Research Symposium and in the Journal of the American Medical Association (JAMA) Internal Medicine that patients in the automated-system group spent 85 per cent of their time within the target oxygen range, compared with 63 per cent among those receiving standard care.
Dr Adet Ginde of the University of Colorado School of Medicine at the Anschutz Medical Campus, who led the study, said in a statement: “Oxygen is one of the most widely used treatments in medicine, yet even in 2026, it is still largely managed through frequent manual adjustments by clinicians.”
The researchers enrolled 300 adults receiving supplemental oxygen at four US hospitals. Participants received either standard oxygen therapy managed by clinicians or automated oxygen titration using the Pro 100 system, developed by Danish company O2matic. The system continuously monitors oxygen levels through a small sensor placed on the fingertip and adjusts oxygen flow from moment to moment.
The researchers said patients in the AI group spent less time with oxygen levels below the target range and less time with oxygen levels above it. They also required fewer manual oxygen adjustments by medical staff, with no increase in serious adverse effects.
They added that the research contributes to ongoing military-supported efforts to improve oxygen delivery in high-pressure environments.
Dr Vic Pivetta, a co-author of the study and director of the University of Colorado’s Combat Medicine Research Centre at the Anschutz Medical Campus, as well as a colonel in the US Air Force Reserve, said in a statement: “Oxygen runs out, and so does a medic’s attention, when caring for an injured soldier hours away from a hospital.”
He added: “A device that automatically adjusts oxygen takes an urgent task off their plate, allowing them to focus on everything else the patient needs.”
