Cancer of the proximal colon after a "normal" colonoscopy

In common with other diagnostic tests, colonoscopy has a false negative rate which is infrequently assessed. The available literature suggests that lesion miss rate is higher for proximal colonic tumors. A total of 367 patients were diagnosed with cancer of the colon and rectum over a period of 2 years. Ninety-two of these patients had tumors proximal to the splenic flexure. Their 5-year pre-diagnosis colonoscopic exposure was analyzed. The primary end-point of this study was to confirm the false negative colonoscopy rate in patients subsequently diagnosed with cancer of the proximal colon. The secondary endpoint was to assess the effects of diagnostic delay on tumor stage and presentation. In the group of patients with proximal colon cancer (n = 92) we identified 10 patients (11%) who, as a result of incomplete (2 cases) or falsely negative (8 cases) colonoscopies, suffered a median diagnostic delay of 17 months (range 3-60). At diagnosis, 4 of these patients had Dukes' D caecal cancer, 4 had Dukes' C caecal cancer and 2 had Dukes' B transverse colon cancer; 3 presented with perforated tumours and 1 with intestinal obstruction. In this small subgroup of patients therefore 40% presented with emergency complications compared to 8% in the rest of the group with proximal cancers (p < 0.01). Missed cancers are more likely to present with complications. This study highlights the importance of recognition of an incomplete examination and the adverse impact of missed diagnosis on subsequent presentation.