**5. Reader experience and accuracy**

abdominal discomfort at the time of CTC. Therefore, the rate of occult colonic perforation

CTC is being increasingly used for the radiological evaluation of colorectal symptoms. In symptomatic patients, CTC is equivalent to OC for diagnosing colon cancer and clinically significant polyps [65]. In a retrospective study of 1,177 older symptomatic patients, 59 inva‐ sive CRC were detected [66]. Three small colorectal cancers were missed by CTC. CTC has a high sensitivity (95%) and negative predictive value (99.7%) in excluding a CRC in patients

CTC may be useful for diagnosing and managing patients with inflammatory bowel disease (IBD) [67]. CTC correctly identified acute and chronic IBD in 63.6%, and 100% of cases, re‐ spectively [68]. CTC was also helpful in assessing post-op strictures in Crohn's disease pa‐ tients [69]. Perianastomotic narrowing or stenosis was detected by CTC in 11 of 15 patients. The sensitivity and specificity for perianastomatic narrowing were 73% and 100% respec‐ tively [69]. The risk of perforation, especially in patients with severe active colitis is a poten‐ tial worry. Currently there is not enough data to measure the true risk in patients with

Examination of the colon is usually necessary after an adequate rest period for evaluation of patients with diverticulitis. CTC appears comparable to OC in the evaluation of these pa‐ tients and is a reasonable alternative in follow-up of patients with symptomatic diverticular disease [71]. Diverticulosis may however, increase the chance of having a false positive test for polyps on CTC due to the appearance of inverted diverticula and fecoliths[72]. On the other hand, CTC may be helpful in diagnosing complications of diverticular disease and in‐

CTC is very useful in detecting colon cancer after incomplete colonoscopy and also for eval‐ uating potential metastases [74-76]. CTC can help localize polyps or cancer prior to laparo‐ scopic surgery and detect synchronous lesions beyond the reach of OC due to obstructing lesions [77, 78]. In fact, CTC is superior to OC in the localization of colonic tumors prior to surgery [79]. CTC is also a safe and useful method for preoperative examination of the prox‐ imal colon after metallic stent placement in patients with acute colon obstruction caused by

CTC is useful in surveillance after surgery for colo-rectal cancer, detecting local recurrence and metastasis [81-84]. In patients with ovarian cancer CTC may be helpful in detecting rec‐ tosigmoid wall involvement wall and predict the need for rectosigmoid resection [85]. The sensitivity, specificity, PPV and negative predictive value of CTC for the prediction of recto‐

after incomplete colonoscopy may warrant a spot CT prior to full examination.

**4.2. For symptoms**

with colorectal symptoms [66].

severe active disease [70].

**4.4. Detection of tumor for surgery**

cancer [80].

**4.3. Inflammatory bowel disease and diverticulitis**

258 Colonoscopy and Colorectal Cancer Screening - Future Directions

flammatory bowel disease, such as colo-vesicular fistulae [73].

Individual accuracy of reading polyps with CTC is highly variable among radiologists and depends largely on training and experience [86-88]. There is a significant learning curve in‐ volved in the interpretation of CTC studies, with performance improving with operator ex‐ perience [89, 90]. Radiologists working in nonacademic centers may have less accurate results than would be expected from published data originating from experienced academic centers [91]. The steep learning curve involved with reading CTC has led some thought leaders to advise against widespread colorectal cancer screening programs with CTC out‐ side of academic centers [29].

False negatives are a major concern, i.e. missing significant lesions. It appears that many false negatives are due to observer error and not due to the technical capabilities of CTC. For instance in one study, 53% of missed polyps (60 of 114) were attributed to observer-related errors, and 26% were attributed to errors classified as technical [92]. This implies that with improvements in reader skill the sensitivity of finding significant lesions would be accepta‐ ble and comparable to OC [90]. Technical factors that appear to be associated with higher accuracy include meticulous bowel preparation and inflation, multidetector CT, combined two and three-dimensional visualization [28, 93].
