[Validity of Recommendations for Secondary Prevention of Colon Cancer]

Prevention programmes should only be recommended if they achieve what they promise to achieve. Therefore, we checked the variation and validity of recommendations for screening for colorectal cancer of nine organisations.

We analysed the information concerning recommended screening age, guaiac faecal occult blood test (gFOBT), faecal immunological test (FIT), faecal DNA test, sigmoidoscopy, colonoscopy, double-contrast examination/double-contrast barium enema, and virtual colonoscopy/CT colonography in the following three steps: 1) we gathered the references quoted by the nine organisations; 2) references were categorised according to mortality, incidence and sensitivity/specificity; 3) the validity of references that reported reduced mortality attributed to screening were evaluated.

Evidence of occult faecal blood was the only screening method recommended by all nine organisations. Colonoscopy was recommended by seven organisations. Fifteen of the 33 references used endpoints other than mortality to justify screening. One publication was a meta-analysis. Eleven of 17 publications evaluated the gFOBT, three evaluated sigmoidoscopy, one FIT, one coloscopy, and one general diagnosis of the intestine. On average, two of nine validity criteria were completely fulfilled, five only partially, and two were not fulfilled. In two publications, none of the validity criteria were completely met.

Analysis of screening for colorectal cancer revealed that nine organisations had different goals and different recommendations. Scrupulous and thorough evaluation of the scientific studies in relation to mortality, upon which these recommendations are based, revealed numerous shortcomings and therefore could not sufficiently substantiate the international recommendations for screening for colorectal cancer. It would be useful to establish a consensus about which data have to be collected to provide a reliable basis for health-care decisions.