Colonic cancer surveillance in ulcerative colitis is not essential for every patient.

It is generally recognised that there is an increased risk of colonic cancer in patients with long-standing extensive colitis, and regular annual or biennial colonoscopic surveillance protocols have been recommended in order to detect early cancer. There is, however, little evidence to suggest that these protocols are of value. There have been no properly conducted controlled trials in this area, and the studies that have been reported are flawed by selection bias, the inclusion of patients with "pseudo disease" and protocol violators. Many studies have not distinguished between "screening colonoscopy" and "colonoscopic surveillance". Some have not drawn attention to the failures in the surveillance, i.e. patients with Dukes' grade C or worse, and overall the conclusions drawn have been unrealistically optimistic. The diagnosis of low grade dysplasia which has been accorded importance is insensitive, non-specific and is subject to gross interobserver error. It is of little clinical value. Colonoscopic surveillance using currently available techniques is of only marginal benefit to patients included within the protocol. It is not cost-effective and cannot be made to be so.