During a bitterly cold Minnesota weekend, I holed up at home with two dogs and hot tea while reading and rereading the paper that “everyone is talking about.” The findings of Grief and coworkers now have received world-wide exposure. In consequence, anesthesiologists, surgeons, and infection control workers may be pondering how soon to start “putting everybody on 80% oxygen in the OR and recovery room.” The article is not perfect—it has some warts—yet it may stimulate timely reconsideration of four intertwined topics: the difficulty of detecting an incremental infection prevention effect in a clinical trial; the importance of designing such a trial so that the control-group treatment represents current best practice; the bittersweet vagaries of inductive inference that are inescapable whenever we generalize to our practices (non-research conditions) techniques reported from clinical trials (research conditions); and the absence of logical rules of engagement to follow at our own hospitals when a new infection control gimmick or gadget arrives on the scene.

Of course, I have already submitted my e-mail letter to the editor of the New England Journal of Medicine. What I could not cover within the 400-word limit of that forum inspired this essay.

The apparent effect of only a few hours' supplemental oxygen administration for colorectal surgical-incision infection prevention was astounding. The 50% reduction in infection risk seems almost too good to be true. There is a seductive charm when a single, simple, relatively cheap and safe process change gives that much outcome improvement.