Colonoscopic screening and surveillance is routinely recommended for patients at risk for colorectal cancer (CRC) and these endoscopic procedures are a major source of healthcare costs in the developed world. Screening and surveillance guidelines vary greatly usually depending on the estimates of risk.1–4 But, what is the logic and evidence for determining colonoscopy frequency as a function of risk? Biologically, risk can relate to initiation of colon carcinogenesis (the risk of an initiating event occurring or number of initiating events that have already occurred) or to increased velocity of carcinogenesis (how long it takes to develop a new adenoma and for that adenoma to progress to a serious lesion, for example). Alternatively, future risk of colonic neoplasia may be due to a pseudo-risk factor, but the chance of missing adenomas or cancers at screening, due to procedural or patient identifiable factors that are not related to the biology of carcinogenesis. Elevated risk of initiation, without acceleration of carcinogenesis, should not per se dictate increased frequency, as the lesions, when initiated (even if frequently), will not develop rapidly. Similarly, if colonoscopy and polypectomy were perfect (all polyps seen and totally removed), the pseudo-risk factors would not be relevant and the interval between screening or surveillance examinations could be based solely on an understanding of the velocity of colon carcinogenesis. Unfortunately, colonoscopy is not perfect and we do not have ways of predicting the velocity of colon carcinogenesis in individual patients.

Currently, colonoscopy surveillance intervals are stratified based on age, prior colonic neoplasia, family history and known hereditary colon cancer syndromes (see table 1). For the non-hereditary syndrome groups, which make up the large majority of people undergoing colonoscopic screening and surveillance, very little is known about the velocity of carcinogenesis. There is no clear evidence that it varies among individuals …