For years, the specter of overdiagnosis — the detection of cancers that would never harm their host — cast a long shadow over breast cancer screening, with some estimates suggesting it affected nearly half of all mammography-detected cases. A major reanalysis of eight landmark randomized trials, led by researchers in Denmark and London, now suggests those figures were artifacts of incomplete data and methodological blind spots, placing the true rate below 5%. The findings invite medicine to revisit not only its screening guidelines but the harder question of how uncertainty, once embedded in c