Powering cancer screening for overall mortality

Historically, we have made assumptions. If a screening test for cancer decreases death from the target disease, and there is no obvious signal that any other type of death is increased, we have assumed that overall mortality (OM) is improved. In other words, the heuristic goes, if the trial’s power were larger, we would see this benefit. This reasoning buttresses nearly all widely used screening tests, but remains unproven in all but the rarest cases. Neither mammography, colonoscopy, sigmoidoscopy, fecal occult blood testing, prostate specific antigen screening, nor ultrasonography for abdominal aortic aneurysm (AAA) have shown overall mortality benefit in randomized controlled trials.