Quantifying the Potential Benefit of Sigmoidoscopic Rescreening for Colorectal Cancer

Abstract

After a person has tested as negative on their first screening test for a given cancer, how soon should a second test be performed? In a perfect world, the answer to this question, at least in terms of efficacy, would be obtained from the results of randomized trials that examine the occurrence of cancer mortality in persons assigned to different rescreening frequencies. However, trials of this type are difficult and not commonly conducted, so it is necessary to look for guidance from studies that seek to estimate the occurrence of clinically important lesions after an initially negative screen (1). Some relevant data can come from studies that document the incidence of clinically detected disease over time following a negative screening exam. For flexible sigmoidoscopy, such studies have been both cohort (2) and case–control (3–6) in design. The results of randomized trials of screening sigmoidoscopy, in which the incidence of distal colon and rectal cancers is monitored over time among initially screen-negative persons, are relevant as well (7,8). Each of these study designs has the virtue of focusing on an endpoint that generally is clinically relevant but has some potential limitations with respect to informing decisions regarding a screening interval. For example, some tumors are not detectable by screening before their clinical presentation. It is also possible that by the time they are clinically apparent, some malignancies may no longer be amenable to effective treatment; a screening interval determined by the period between a negative screening exam and the clinical diagnosis may be longer than the one that could best