Cecal photodocumentation: Reliable when you are there, unreliable when you are not

Endoscopists have sought a reliable way to document complete colonoscopy to the cecum. Photographs of various cecal landmarks have been used for this purpose. This study was designed to answer the question whether cecal photographs can accomplish this goal. METHODS: NYSGE endoscopists were asked to submit pictures which were their "best" efforts to demonstrate the cecum; to include up to 4 views of the ileocecal valve, appendiceal orifice, and cecal strap (caput). 9 endoscopists submitted 170 photographs, including 11 photographs that appeared to be the cecum but were not (true negatives). The best 89 photographs of the cecum were used as true positives. The 100 photographs were then submitted to a panel of 3 expert reviewers who were independently asked to decide whether each of the 100 photographs reliably demonstrated the cecum, or were "not cecum". Errors in choosing a photo as the cecum when it truly was not, were considered as "false positives"; and choosing "not cecum" when it truly was the cecum, were considered "false negatives". RESULTS: Table shows each reviewer's scores for the packet of 100 photos. Total true pos (cecum) = 89; total true neg (not cecum) = 11 Reviewer True Pos True Neg False Pos False Neg 1 88 7 4 1 2 87 4 7 2 3 76 6 5 13 The sensitivity of picking out the positive cecum photographs was 99%, 98%, and 85% for the 3 reviewers for a mean positive predictive value of 94%. Specificity (correctly choosing photos that were not the cecum) was 88%, 67%, and 32% for the 3 reviewers. The mean specificity was 52%. CONCLUSION: When given a set of cecal photographs with clear landmarks, experts can reliably and accurately verify that the cecum was reached 94% of the time. However, experts could not reliably identify photographs that were not the cecum. Cecal photodocumentation has limitations, even when the best efforts of a select group of endoscopists are studied.