CT colonography for cancer screening: insight into the CMS-Medicare decision against coverage

In 2009, the US Centers for Medicare and Medicaid Services (CMS) decided against covering costs for colorectal cancer screening by CT colonography. This meant that it will not pay outpatient clinics for performing this test on persons with Medicare, the national medical insurance for the elderly. CMS's decision-making process had been guided by the Cochrane-Holland framework for evaluating screening methods (which considers simplicity, acceptability, accuracy, cost, precision, sensitivity and specificity), together with the impact on health outcomes. As detailed in a “ decision memo” [ 1 ], the CMS found that the “evidence [was] inadequate to conclude that CT colonography is an appropriate colorectal cancer screening test”. The evidence supporting this decision, including independent and ad hoc commissioned technology assessments, practice guidelines, consensus statements and public comments, is summarized in the memo.

One of the technology assessments commissioned by the CMS was a study to identify the threshold cost below which CT colonography could be considered cost-effective to currently covered tests, namely fecal occult blood test, sigmoidoscopy and colonoscopy (although barium enema is also covered, it was not included in the study). The study was done by the Cancer Intervention and Surveillance Modeling Network (CISNET), involving researchers across the US and in The Netherlands. Although research findings were available to the CMS last year, only this month has the study been published in the Journal of the National Cancer Institute [ 2 ].

Briefly, the newly published study used three different microsimulation models to describe the risk of colorectal cancer in the Medicare population and to assess the impact of screening; all three models used standardized data regarding each test's accuracy, complications profile and costs. The researchers evaluated the life-time costs and health effects (both risks and benefits) of each method, assuming that the screening population undergoes programmed testing from age 65 to 80 years with full compliance. They found that CT colonography every 5 years would afford 143-178 life-years gained per 1000 persons (depending on the simulation model), similar to that gained with 5-yearly sigmoidoscopy plus annual fecal testing but less than that of 10-yearly colonoscopy (affording 152-185 life-years). To be cost-effective, CT colonography would have to be billed at US $108-$205, a price less than half of that estimated from abdominopelvic CT examinations. This threshold cost would increase, however, if CT colonography improved compliance, but so far these data are lacking.

As noted in the accompanying editorial [ 3 ] by Harris, a former member of the US Preventive Services Task Force, this cost-effectiveness study is an important contribution to the colorectal cancer screening debate. At the same time, it illustrates the uncertainties in our understanding of the benefits and harms of both CT colonography and colonoscopy, and leaves one hoping for the emergence of a new technology that is simpler, safer and cheaper.

References

1. Syrek Jensen T, Salive ME, Larson W et al (2009). Decision memo for screening computed tomography colonography (CTC) for colorectal cancer. Centers for Medicare and Medicaid Services, Baltimore, administrative file CAG-00396N.
2. Knudsen AB, Lansdorp-Vogelaar I, Rutter CM et al (2010). Cost-effectiveness of computed tomographic colonography screening for colorectal cancer in the Medicare population. J Natl Cancer Inst [Epub ahead of print].
3. Harris R (2010). Speaking for the evidence: colonoscopy vs computed tomographic colonography. J Natl Cancer Inst [Epub ahead of print].

New research on colorectal cancer screening

Among the articles made available to MDCT.net users this month are two new and widely diverse studies pertinent to colorectal cancer screening.

Pickhardt et al. [ 1 ] focused on the positive predictive value (PPV) of CT colonography. As they noted, in a screening program, it is not possible to calculate sensitivity and specificity, since only positive results are further evaluated with second-line tests. Thus, important “quality metrics” for screening programs are the false-positive rate and PPV. To illustrate how this is done, they retrospectively evaluated data from over 5000 adults who had CT colonography for screening purposes; 639 persons were found to have at least one polyp ≥6 mm, and 479 of them (with 739 lesions) chose to have colonoscopy for follow-up. Colonoscopy confirmed 677 lesions, giving an overall per-lesion PPV of 91.6% and a per-patient PPV of 92.3%. Similarly high rates were obtained for small and large lesions and for sessile, pedunculated and mass-like lesions, but flat lesions had a PPV of 77.7%. The authors compared their excellent single-center results with published data and discussed how this type of analysis can be used for quality assessment.

In a completely different approach, Imaeda and colleagues [ 2 ] developed a tool to help patients understand the characteristics of colorectal screening tests and to choose the test that best matches their personal needs. They used the maximum differences scaling method to help patients prioritize their concerns about the advantages and disadvantages of fecal occult blood testing, sigmoidoscopy, colonoscopy, CT colonography and colon capsule endoscopy. In a pilot study, 92 adults ranked sensitivity, risk of colonic perforation, and need for a second test to remove polyps as the most important attributes to consider when choosing a screening method. They then expressed their preference for a screening method, choosing colonoscopy in 62% of cases (mostly for its reported sensitivity) and CT colonography in 10%. The authors suggested that this tool will facilitate shared, informed decision-making involving patients and their physicians.

References

1. Pickhardt PJ, Wise SM, Kim DH (2010). Positive predictive value for polyps detected at screening. CT colonography. Eur Radiol 20(7):1651-1656.
2. Imaeda A, Bender D, Fraenkel L (2010). What is most important to patients when deciding about colorectal screening?. J Gen Intern Med 25(7):688-693.

Automatic exposure control: one cause of accidental radiation overdose

One year after cases of accidental radiation overexposure during CT examinations were reported in the United States, still the scientific literature lacks a clear explanation of the causes. For this reason, a recent article in the New York Times [ 1 ] may be interesting to radiologists and clinicians who wish to avoid similar errors.

Through interviews with patients, hospital staff and state officials, investigative journalists uncovered a complex story of multiple human errors due to poor judgement and unskilled use of CT scanners. For example, technicians made mistakes due to inadequate training and, at one hospital, excessive doses were intentionally used to get better images. Another worrisome cause was attributed to improper use of the automatic exposure control (AEC) feature during brain perfusion CT, resulting in up to 8-times more radiation than necessary. Although one of the manufacturers, GE Healthcare, told reporters that AEC had limited value for perfusion CT, staff at two California hospitals were unaware that this feature would actually raise radiation dose when used with certain scanner settings. These serious medical errors can be avoided by better training, standardized dose reporting, and implementation of additional safety features on scanners.

References

1. Bogdanich W (2010). After stroke scans, patients face serious health risks. The New York Times (New York Ed.), 1 August; Sect. A:1.

Pediatric chest and heart imaging: a course in Florence

The European Society of Paediatric Radiology will hold its 19th “European course” this coming November in Florence, Italy. The course, in English, will be directed by Claudio Fonda, head of the pediatric radiology department at Meyer Children's Hospital of Florence. The 3-day event will address both fetal and child pathologies, and will discuss applications of plain radiography, MDCT, ultrasonography and MRI. The program offers numerous lectures in chest and lung imaging followed by a panel discussion, additional lectures in cardiac imaging, short workshops, manufacturers' symposia, and a social dinner. The course is to be accredited as a continuing medical education event by the European Union of Medical Specialists. A detailed program and registration information are available at the course's website, www.aimgroup.it/2010/ecpr/index.html.