Abstract
The risk of individuals who had a previous history of normal findings in diagnostic examinations using total colonoscopy to be newly diagnosed with early colorectal cancer or adenoma in colorectal cancer screening was assessed. The odds ratio was calculated using a logistic regression model, for individuals with early colorectal cancer or adenoma as a case group and others with normal findings as a control group. The past history of a normal colonoscopy result was sought in a database for mass screening programs at the workplace and was considered as the exposure factor. The odds ratios for early cancer detection within 1, 2, and 3 years were 0.17 (95% confidence interval: 0.04-0.70), 0.13 (95% confidence interval: 0.04-0.42), and 0.10 (95% confidence interval: 0.03-0.31), respectively. This suggests that the risk of an individual with a history of diagnostic examination within the previous 1 to 3 years for being newly diagnosed with early cancer is reduced by 83 to 90%, as compared with an individual with no such history. The odds ratio was lowest for within 3 years, which indicated that false negative cases in the previous examinations were included in the analysis. If the detection of early or more advanced cancer is defined as an objective of the mass screening programs, an individual who has had a normal colonoscopic examination result should take part in mass screening tests every two to three years. To achieve this goal, however, the accuracy of colonoscopy should be evaluated more strictly on a clinical basis.
