We read with interest the recent report on the Regional versus General Anesthesia for Promoting Independence after Hip Fracture (REGAIN) study of 1-yr survival with general versus spinal anesthesia in proximal femur fracture repair. Those authors reported that the choice “did not affect survival at up to 1 yr after hip fracture surgery.” This sentiment was echoed in an accompanying editorial : “anesthetic type makes no difference” as the “outcomes do not differ”. Unfortunately, the study’s findings do not support this inference of equivalence. The 95% CI for the primary outcome (hazard of death during 1 yr) with spinal versus general anesthesia was 0.81 to 1.44. Spinal anesthesia thus may be 19% safer than general anesthesia, or 44% more dangerous, or anywhere in between. The high incidence of hip fracture surgery combined with a significant mortality risk means that the 95% CI encompasses thousands of lives that may be saved or lost. It does not follow that the difference in mortality hazard is therefore somehow negligible. The study was not designed as an equivalence trial, and its results do not support that inference. All that can be said is that the result is inconclusive. In other words, failing to observe a difference is not the same as demonstrating equivalence.

To be fair, this inferential problem is not uncommon and this result is unsurprising. The investigators excluded thousands of otherwise eligible patients: ultimately eight patients were excluded for each one included. Specifically, more than a thousand were excluded “if they were considered to be unsuitable for randomization by the surgeon or anesthesiologist on the basis of the physician’s clinical assessment” (p. 2026). This significant physician input in the inclusion/exclusion decision seems to have yielded a hyperfocused sample in which no one expected a substantial difference in outcomes, and then the observed difference was small. Perhaps the main finding of this study, rather than equivalence, should be that physicians are becoming more adept at distinguishing patients likely to be better served by one form of anesthesia or another.