Diagnostic accuracy and interobserver agreement of CT colonography (virtual colonoscopy)

Background and aims-Computed tomographic (CT) colonography or virtual colon oscopy (VC) is a non-invasive imaging method proposed for screening patient s with colorectal neoplasias. Our aims were to study the diagnostic accurac y and interobserver agreement of VC for correct patient identification comp ared with conventional colonoscopy (CC). Methods-This was a prospective study of 50 patients successively undergoing VC and CC. Multiplanar two dimensional CT images and three dimensional VC were constructed using surface rendering software and interpreted by two in dependent investigator teams. VC findings were compared with those of CC. I nterobserver agreement was determined using kappa statistics. Results-CC found 65 polyps in 24 patients. For identification of patients w ith polyps greater than or equal to 10 mm, the sensitivity of VC was 38% an d 63%, and specificity was 74% and 74% for teams 1 and team 2. Interobserve r agreement was good (kappa 0.72). For patients with polyps of any size, th e sensitivity of VC was 75% and 71%, and specificity was 62% and 69% for te ams 1 and 2. Interobserver agreement was fair (kappa 0.56). Accuracy improv ed when comparing the results of the first 24 with the last 26 patients. Conclusions-In our experience, VC had a low diagnostic value for identifica tion of patients with colorectal neoplasias. Interobserver agreement for VC interpretation was fair. These results may be explained by software imperf ections and a learning curve effect.