Overdiagnosis is an often underappreciated harm of screening. In the context of cancer screening, it refers to the detection of cancers that appear histopathologically to be invasive malignant tumors but grow so slowly that they never would have become clinically evident during a usual lifetime or occur in a person who dies of another cause before the cancer symptoms appear.^1,2 The causes of overdiagnosis include more sensitive screening tests, increasing biopsy rates, and lower thresholds for reporting abnormal-appearing cells in biopsy specimens as malignant.^3-5

The optimal way to determine the percentage of screen-detected cancers that are overdiagnosed would be to randomize patients to screening or no screening and then follow up these patients until death. The excess number of cancers detected in the screened group represents overdiagnosis because the counterpart malignant tumors in the unscreened group never caused symptoms. The excess cancers divided by the number of screen-detected (or alternately total) cancers provides the prevalence of overdiagnosis.⁶ However, such studies do not exist, particularly for lung cancer, for which the duration of follow-up after screening is typically less than 10 years.^7-9