Published trial estimates of breast cancer overdiagnosis range from near zero to about 50%, differences the authors attribute to how exit screening in control arms and post-screening compensatory drops in incidence are handled. Reanalysing trial arms within a single framework, and comparing observed relative proportions of excess cumulative incidence with the temporal pattern seen in the population-based Funen programme in Denmark, 73 observed values across 52 timepoints matched the reference pattern for invasive cancer with and without DCIS. The authors read this as compatible with overall overdiagnosis below 5%.
Why it is interesting: Overdiagnosis is the central contested quantity in breast screening policy, and this reanalysis puts the trial evidence at the low end of a range that has spanned an order of magnitude.