Abstract

AIM: To compare the diagııostic efficacy of CT colonography and conventional colonoscopy in detecting colorectal lesions in adult patients w ith high risk of colorectal neoplasia. MATERIAL AND METHODS: ln all patients, CT colonography was performed by a 1.6-detector multi-siice scanner, followed by conventional colonoscopy. lmages were obtained with 1.25 mm slice thickness both in supine and prone positions, the latter with administration of Lv. Iodinated contrast materiaL Virtual colonoscopic views were obtained through "Shaded Surface Display" software. Taking the conventional colonoscopy as the gold Standard, sensitivity, specifıcity and Cohen's Kappa correlation coeffıcient were estimated for all lesions and polyps, seperately. FINDINGS: When all lesions are evaluated together, the sensitivity and spesifity of virtual colonography were 81,5% and 76,9%, whereas the sensitivity and spesifity calculated for polyps were 83,3% and 16.9%, respectively. In polyps with the diameter of 4 mm and over, sensitivity of CT colonography was 100%. With regard to the correlation between virtual colonoscopy and conventional colonoscopy, Cohen's Kappa coefficient was found to be fairly good (K=0.50-0.60). CONCLUSION: Due to the ineffıciency in detecting small polyps and fairly good results in Kappa correlation analysis, CT colonography is unsuitable as a screening method. In our opinion, the use of CT colonography should be reserved to follow up of the morphologic features of the lesions which were primarily detected by conventional colonoscopy.