Purpose: The usefulness of computed tomographic colonography (CTC) as a noninvasive alternative to standard colonoscopy (SC) for colorectal cancer screening depends on an accurate assessment of its sensitivity. In some studies comparing CTC with SC, the initial blinded SC is followed by a reexamination of the colon with unblinding to the CTC results (Unblinded SC), but in others, the initial SC is assumed to be the gold standard (Blinded SC). We examined the effect of this work up intensity bias (whether the colon was reexamined) on the sensitivity of CTC. Method: We searched Medline 1996-2006 for prospective studies comparing CTC with SC. We classified studies by their work up intensity and calculated pooled sensitivity of the initial SC and the sensitivity CTC depending on the workup intensity using the DerSimonian and Laird random effects model for patients with ³10 mm polyps. Results: We identified a total of 37 studies (8016 patients). Reexamination of the colon with increased work up intensity detected additional polyps (sensitivity<100% for initial SC) and consistently yielded a lower sensitivity for CTC (Unblinded vs Blinded SC) in the patient and polyp analyses, regardless of type of polyp. See table below.
Conclusions: CTC has a lower sensitivity than SC but both the increased work up intensity and the choice of disease criteria (any polyp or only adenomatous polyps) alter the sensitivity of CTC. However, individuals may be willing to accept the lower sensitivity to avoid colonoscopy.