Does Virtual Colonoscopy Make Fiscal Sense?

Abstract & Commentary

Synopsis: Virtual colonoscopy refers to the newly developed imaging techniques using either computerized tomography or magnetic resonance, rapidly generated images to simulate colonoscopy. The technique has proven effective at detecting small colonic lesions and is safe and well-accepted by patients. In this report, a hypothetical population is used to calculate and compare (vs colonoscopy) the cost-effectiveness of this new technique as an initial screening tool for colorectal cancer. The data indicate that CT-colonography would be somewhat more expensive. It is suggested that, from a cost perspective, in order for it to replace colonoscopy as an initial screen, the procedural expenses would need to be reduced to a level 54% less than colonoscopy.

Source: Sonnenberg A, et al. Am J Gastroenterol 1999; 94:2268-2274.

Computed tomography (ct) or magnetic resonance (MR) colonography is a novel application of computerized imaging designed to generate three-dimensional representations of the colon. Some have suggested that the technique be applied to widespread screening for colorectal cancer.1,2 Both CT and MR techniques are reported to have high sensitivity (> 75%) and specificity (> 90%) in detecting colorectal cancer and polyps more than 10 mm in size.3-5 However, the question remains whether there is added value to this approach when compared to colonoscopy alone.

In this report, the issues of cost-effectiveness were raised and a model was developed to assess the cost of this procedure compared to colonoscopy or no screening. A hypothetical population of 100,000 individuals aged 50 years undergoing screening every 10 years was used. Suspicious lesions picked up by the CT technique were further worked up by colonoscopy. For those requiring polypectomy, colonoscopy was repeated every three years until no adenomatous polyps were found.

The technique involves the generation of endoluminal images at a fast rate (15-30/sec). Subjects receive the same type of bowel cleansing as with colonoscopy and the colon is insufflated with 2 L CO2 or room air. (For MR colonography, after the bowel cleansing the patient is given a single contrast enema of 2 L water containing 20 mL of gadolinium-DTPA). While a single breath is held, the subject is moved through a rotating x-ray beam of a helical CT scanner. With the rapid image generation, there is an illusion of traveling through the colon. The technique is reported to have a sensitivity of more than 75% and specificity more than 90% for detecting colorectal polyps or cancers of 10 mm size.3-5 If polyps or other abnormalities are observed, conventional colonoscopy is required for further evaluation, biopsy, and/or resection.

Applying the Markov technique for cost effective modeling,6 screening CT colonography costs $24,586 per life-year saved, compared with $20,930 spent on colonoscopy screening. The incremental cost-effectiveness ratios comparing CT colonography to no screening was $11,484, but comparing colonoscopy to CT colonography was $10,408. Thus, screening by colonoscopy is more cost effective. For the two procedures to become comparable, CT colonography needs to be associated with an initial compliance rate 15-20% better than colonoscopy or the procedural costs need to be more than 50% less than colonoscopy.


CT colonography has been advanced as a screening tool on the rationale that it will be well tolerated by patients, and be safe and effective at detecting small lesions.3,4 However, the technique is still under development and it is not clear that these issues alone warrant replacing conventional colonoscopy screening. It is likely that it will be better accepted by patients than colonoscopy but the procedure still calls for bowel cleansing and gas insufflation, which are the most unpleasant features of colonoscopy for many patients. No anesthesia is involved and theoretically this provides a safety advantage. Reported levels of detection are excellent but these have come from the centers that developed the technique. No doubt, there is a learning curve and it would seem the gold standard for sensitivity and specificity would remain colonoscopy.

For screening techniques to be widely applied, they must also prove to be fiscally feasible. Thus, the current study using the Markov model for determining cost effectiveness is justified. An estimate from a typical population for whom screening by colonoscopy has proven both effective and cost effective provides a useful comparison. When examined by this model, it turns out that the new technique, when applied as initial screening, adds significant expense. Yet, it still falls within a level that would be considered reasonable. For example, the estimate is similar to the cost per year-of-life saved by mammography in the 50-69 year-old age group ($21,400) and is four times less expensive than similar estimates for mammography, if the 40-49 year-old age group is included ($110,000).7,8

Thus, the question remains whether this new technique should replace conventional colonoscopy as an initial screen for colorectal cancer. Perhaps the debate that is likely to develop will stimulate new advances in technique and, perhaps, reduced procedural costs (of 1 or the other) that will allow a clear choice. For now, and until more experience is gained, in most communities conventional colonoscopy remains the standard.


1. Luboldt W, et al. Radiology 1998;207:59-65.

2. Royster AP, et al. Acad Radiol 1998;5:282-288.

3. Hara AK, et al. Radiology 1997;205:59-65.

4. Royster AP, et al. Am J Roentgen 1997;169:1237-1242.

5. Schoenenberger AW, et al. Gastroenterology 1997;112: 1863-1870.

6. Sonnenberg FA, et al. Med Decis Making 1993;13: 332-338.

7. Salzmann P, et al. Ann Intern Med 1997;127:955-965.

8. Ransohoff DF and Harris RP. Ann Intern Med 1997; 127:1029-1034.

Which of the following statements about applying CT colonography as an initial screening tool for colorectal cancer is true?

a. In terms of costs per year-of-life saved, it is less expensive than when colonoscopy is the initial screening tool but the sensitivity of colonoscopy is greater.

b. In terms of costs per year-of-life saved, it is less expensive than when colonoscopy is the initial screening tool and the sensitivity for detecting small lesions is greater than conventional colonoscopy.

c. In terms of costs per year-of-life saved, it is more expensive than when colonoscopy is the initial screening tool and the sensitivity of colonoscopy is greater.

d. In terms of costs per year-of-life saved, it is more expensive than when colonoscopy is the initial screening tool but the increased sensitivity for detecting small lesions makes it the screening test of choice.

e. It eliminates all of the features of colonoscopy that cause patient discomfort and inconvenience.