Screening for Colorectal Cancer: A Targeted, Updated Systematic Review for the U.S. Preventive Services Task Force

You Are Here: AHRQ Home > Clinical Information > U.S. Preventive Services Task Force > Topic Index > Screening: Colorectal Cancer > Article

Table 2. Accuracy of Computed Tomographic Colonography and Estimated Rates of Referral to Colonoscopy

Variable	Colonoscopy: Pickhardt et al., 2003⁴⁹	CT Colonography Pickhardt et al., 2003⁴⁹	CT Colonography Johnson et al., 2008⁵⁵	CT Colonography Kim et al., 2007^54,a	CT Colonography Johnson et al., 2007^53,a
Study aim	To evaluate performance characteristics of CT colonography screening		To assess the accuracy of CT colonography in multicenter screening setting	To compare 3D vs. 2D interpretation of CT colonography	To compare 3D vs. 2D interpretation of CT colonography using 2.5-mm and 1.25-mm slice thickness
Patients, n	1233		2531	96	452
Population	50–79 y; 41% female		≥50 y; 54% female	40–76 y; 42% female	41–82 y; 44% female
CT Colonography		Flythrough 3D imaging with 2D correlation of abnormality (Viatronix V3D 1.2, Stony Brook, New York); stool tagging; luminal fluid tagging; 6 trained radiologists	Randomly assigned primary 2D or 3D flythrough analysis (5 software packages used); stool tagging; luminal fluid tagging; 15 trained and certified radiologists	3D virtual colon dissection (Perspective Filet View) and 2D display (Rapidia); intravenous contrast agent for extracolonic findings; 2 very experienced radiologists	3D virtual dissection (Voxtool 5.4.46, GE Healthcare, Milwaukee, Wisconsin); no contrast agent; 3 very experienced radiologists
Reference standard	Same-day colonoscopy by 1 of 17 experienced colonoscopists using segmental unblinding		Same-day blinded colonoscopy conducted or supervised by unspecified number of experienced endoscopists, with unblinded second colonoscopy for CT-detected lesions ≥10 mm not detected on initial colonoscopy	Same-day colonoscopy by 1 of 5 experienced gastroenterologists using segmental unblinding	Same-day videotaped colonoscopy conducted or supervised by 1 of 50 experienced endoscopists; repeat colonoscopy in 6 cases of large lesion on CT colonography
Study quality	Good: Use of enhanced reference standard allows distinguishing false-positive CT colonography results from false-negative optical colonoscopy results; interobserver agreement checked on subset of cases		Fair: Colonoscopy reference standard by community operators without clear quality guidelines; incomplete follow-through on second-look colonoscopies (15 of 27); test performance based on 5-mm CT colonography threshold	Fair: Retrospective analysis comparing types of CT colonography and reader reliability	Fair: Limited power because of multiple analyses comparing readers, displays, and collimation thicknesses; colonoscopy reference standard not high quality
Applicability	Predominantly average-risk screening population, 3% with family history; may represent best-case estimates because of technology used and limited number of experienced readers		Multicenter study of primarily average-risk participants (9% with family history; 2% with personal history of polyps or cancer); use of 15 trained, qualified readers, with range of sensitivity (67%–100%) for large adenomas and CRC	Uncertain because of setting, small study size, limited number of very experienced radiologists compared with endoscopists	Small number of more skilled radiologists; unusually high yield of CRC and low prevalence of polyps compared with other screening populations, possibly due to not excluding patients with previous colonic resections
Sensitivity (per patient) (95% CI), %				Range of 3D and 2D	Range of 3D and 2D
CRC	1 of 2 CRC cases detected	2 of 2 CRC cases detected	6 of 7 CRC cases detected	None detected	5 of 5 CRC cases detected
Adenoma ≥10 mm	87.5 (74.8–95.3)	93.8 (82.8–98.7)	90 (84–96)^b	100^c	50–83
Adenoma ≥6 mm	92.3 (87.1–95.8)	88.7 (82.9–93.1)	78, (71–85)^b	59–77^c	NR
Specificity (per patient) (95% CI), %
Lesions ≥10 mm	NA	96.0 (94.8–97.1)	86 (81.3–90.0)	99–100	97–99
Lesions ≥6 mm	NA	79.6 (77.0–82.0)	88 (84.0–92.0)	89–99	NR
Referral for colonoscopy
Lesions ≥10 mm	NA	1 in 13	NR	1 in 10	Not calculated^d
Lesions ≥6 mm	NA	1 in 3	1 in 6–8^e	1 in 5	Not calculated^d

2D = 2-dimensional; 3D = 3-dimensional; CRC = colorectal cancer; CT = computed tomography; NA = not applicable; NR = not reported.
a. Point estimates and CIs are calculated from multiple measurements provided in the studies. Methods can be found in reference 25.
b. Detection of adenoma or cancer in Johnson et al. (55) on CT colonography– detected lesions 5 mm or greater.
c. Detection of polyp in Kim et al. (54).
d. Polyp prevalence significantly different from those reported in other similar studies.
e. Range of estimates: 1 in 6 referred for colonoscopy is based on the referral threshold for 5-mm lesions on which sensitivity and specificity calculations are based; 1 in 8 is based on a colonoscopy referral threshold for lesions ≥6 mm.

Return to Document