Work

Response: Re: Cost-Effectiveness of Computed Tomographic Colonography Screening for Colorectal Cancer in the Medicare Population

Year: 2010

Type: article

Abstract: Pickhardt et al. criticize our natural history assumptions for small (<6 mm) adenomas. The natural history of adenomas of different sizes is unknown. Our three independently developed models differ widely with respect to the natural history of small adenomas, namely whether and how quickly they progress to colorectal cancer. Despite these differences, our models generated consistent conclusions about the cost-effectiveness of computed tomographic colonography (CTC) screening. We welcome the opportunity to engage in direct comparisons that may shed light on reasons for differences between our findings and those of Pickhardt et al. Our analysis was performed from the perspective of Centers for Medicare and Medicaid Services, whose payment rates may differ from private sector costs. Because charges are typically very different from costs or reimbursements, we do not agree with Pickhardt et al. that they are the appropriate economic metric. Nonetheless, we agree that our relative cost thresholds of CTC vs colonoscopy could be applied to other settings to examine whether CTC would be cost-effective. We are not convinced that current evidence supports the claim by Pickhardt et al. that including CTC as a screening option for the Medicare population would appreciably improve screening adherence. Two of the studies they cited ( 1 , 2 ) were based on small patient surveys (with 68 and 250 respondents), and each was performed at a single institution, thus bringing their generalizability into question. The response rate to one survey was only 39%, further raising concerns about the conclusions ( 1 ). The other survey was conducted among patients who had already chosen CTC screening ( 2 ), thus the findings are not generalizable to a broader unscreened population. The study ( 3 ) that reported colorectal cancer screening rates of less than 30% among the Medicare population was based on claims data, which are likely to underestimate rates of fecal occult blood testing because this inexpensive test is often underreported. Furthermore, that study was for the period 1991–1997 ( 3 ). Colorectal cancer screening rates have increased substantially since that time period. Data from the 2005 National Health Interview Survey suggest that 57% of Medicare-eligible individuals are adherent with colorectal cancer screening ( 4 ). It may not be as easy to achieve a 25% CTC-related increase in adherence from this higher level (ie, raising adherence to 71%). An ongoing study in the Netherlands ( 5 ) may provide information about whether adherence with CTC is higher than adherence with colonoscopy. In this population-based study, 7500 screening-eligible individuals have been randomly assigned to receive an invitation for either CTC or colonoscopy. The main outcome measure is the screening participation rate. Although the absolute test-specific adherence rates may not be generalizable to the US Medicare population, the relative adherence with CTC vs colonoscopy will be informative. In summary, our study demonstrated the conditions under which CTC could be a cost-effective screening strategy for the Medicare population. Pickhardt et al. claim that each of these conditions has been met. We do not agree that observations from small single-institution studies provide a convincing body of evidence to support their claim. What is needed are more compelling data that demonstrate the impact of CTC on overall rates of colorectal cancer screening and detail the costs of CTC and colonoscopy in different settings.

Source: JNCI Journal of the National Cancer Institute

DOI: 10.1093/jnci/djq383