Routine Oxygen at End of Life Is Typically Unhelpful

Abstract

DENVER –The routine administration of oxygen to terminally ill patients near death is unwarranted, a randomized, double-blind trial indicates. If oxygen reduces a patient’s distress, “then that patient should have oxygen, but if it does not – if there’s no change in patient distress – then that oxygen can be discontinued, or certainly not initiated in the first place,” Mary L. Campbell, PhD, declared at the annual meeting of the American Academy of Hospice and Palliative Care Medicine. Without ever having been subjected to scientific scrutiny, oxygen administration has become routine for patients who are near death, asserted Dr. Campbell of Wayne State University, Detroit.“Oxygen has become almost an iconic intervention at the end of life.” It is not benign, she said. It’s expensive, intrusive, and causes nasal drying and nosebleeds as well as feelings of suffocation. To assess the value of routine oxygen administration, she conducted a crossover study involving 32 terminally ill patients. None were in respiratory distress at baseline, but all were at high risk for distress because of underlying conditions. All had a Palliative Performance Scale score of 30 or less. Each patient received a nasal cannula that also captured exhaled carbon dioxide. Next, randomly rotating 10-minute intervals of oxygen, medical air, and no flow were administered for 90 minutes. The key finding: 29 of 32 patients experienced no distress during the 90-minute protocol, indicating that they didn’t need the oxygen. Yet, at enrollment, 27 patients had oxygen flowing. The remaining three patients rapidly became hypoxemic and distressed when crossed over from oxygen to no flow. They were returned to baseline oxygen and respiratory comfort. As many of the study participants were unconscious or cognitively impaired and couldn’t self-report their distress, the Respiratory Distress Observation Scale was assessed at baseline and for 10 minutes after every flow change. A score of 4 or less on the 0-16 scale indicates little or no distress; the average baseline score was 1.47, and it didn’t vary significantly during the different flow conditions. The average oxygen saturation at baseline was 93.6%, and it didn’t change significantly during the 90-minute protocol. Dr. Campbell said that she determines the need for oxygen in an end-of-life patient by taking the person off oxygen for 10 minutes and watching for distress. Several audience members predicted family objections to this approach because oxygen has become an expected part of end-of-life care. “I think if you explain to families that this is a treatment that can be helpful but has side effects, and we always take away the things that aren’t helping when they’re no longer helping, you won’t have pushback from family members,” Dr. Campbell responded. She reported no financial conflict. The study was funded by the Blue Cross/ Blue Shield of Michigan Foundation. CfA