Introduction: The importance of colonic polyp is increasing because colorectal cancer (CRC) screening is based on their identification. However, consensus is lacking on which technique should be used in CRC screening, whether colonoscopy or other methods such as sigmoidoscopy, which shows good sensitivity only in distal segments of the colon. We studied the characteristics and localization of polyps in our environment to verify their tendency to change localization and determine the implications of this in screening.

Material and methods: We reviewed the prevalence and characteristics of colonic polyps in 3604 patients who underwent colonoscopy for various causes in our service, as well as their distribution in 837 patients with a diagnosis of polyps who underwent complete colonoscopy. Localization proximal or distal to the splenic flexure and polyp size (less than or more than 1 cm) were analyzed.

Results: Polyps were found in 28.3% of 3604 patients (11.7% showed polyps > or = 1 cm). Of the 837 patients who underwent complete colonoscopy, proximal polyps were found in 44.9% (14.6% showed polyps > or = 1 cm), while distal polyps were found in 79.9% (30.3% showed polyps > or = 1 cm). Both proximal and distal polyps were found in 24.8%. The probability of finding both proximal and distal lesions was 31%. Proximal lesions without distal lesions were found in 20% of the patients (5.6% presented polyps > or = 1 cm), and consequently sigmoidoscopy would miss 15.6% of polyps > or = 1 cm.

Conclusions: A relatively high percentage of patients with proximal polyps > or = 1 cm would remain undiagnosed if complete colonoscopy were only performed in patients with distal polyps. Because of the well-known increase in the frequency of proximal polyps and because a considerable number of these present without distal sentinel lesions, there are increasing arguments for considering colonoscopy as the basic technique for CRC screening.