Evaluation of surveillance programmes for colorectal cancer in ulcerative colitis patients by case-control studies: methodological considerations.

Abstract

The evaluation of the efficacy of colonoscopy screening in patients with ulcerative colitis for colorectal cancer is associated with methodological difficulties. Case-control studies can, however, be used to determine the efficacy of such a programme and the outline of the methodology in such a programme is presented. The randomised controlled trial provides perspective for case-control studies of screening efficacy. Cases are selected from persons who have ulcerative colitis with manifestations of colorectal cancer: for example, those who have died of colorectal cancer or have symptomatic metastases. Controls are selected from persons who have ulcerative colitis, who had been alive when the case died of colorectal cancer, and who had been subject to the risk of dying from, but had not had, colorectal cancer diagnosed when the case was diagnosed with colorectal cancer. The relevant screening history for cases begins with the case's diagnosis of ulcerative colitis and ends with the cases diagnosis of colorectal cancer; that for controls should be comparable to that for cases to avoid bias. Cases and controls are compared with respect to their "exposure" to colonoscopy during their screening histories: the occurrence of any screening, which took place during the period of time that an occult tumour (or an identifiable lesion) may plausibly have been present. The proposed methodology can evaluate the efficacy of a screening programme rapidly, practically, and ethically.