Mammography computer-aided detection (CAD) adoption accelerated after Congress established national Medicare coverage in 2000, noted a review the June 14 edition of the Archives of Internal Medicine (AIM). However, an interview with lead author Joshua J. Fenton, MD, and an accompanying editorial, questioned its clinical value.
Prior to 2000, diffusion of CAD into practice was limited. “Despite uncertainties about the safety and effectiveness of CAD in real world practice, its use skyrocketed after Congress mandated coverage in 2000,” said Fenton, an assistant professor, department of family and community medicine at the University of California, Davis.
The researchers linked the Surveillance, Epidemiology and End Results (SEER) and Medicare data to estimate the annual prevalence of CAD use with screening mammography in Medicare enrollees aged 68 to 89 from 2001 to 2003.
Using logistic regression and generalized equations, the authors estimated the 2003 prevalence of CAD. Combining the estimated prevalence with the 2003 Medicare reimbursement rate of $19.13, Medicare CAD costs reached an approximate $19.5 million, wrote Fenton and colleagues.
The authors also found the prevalence of CAD use increased from 4.8 percent in 2001 to 26.9 percent in 2003, which suggests Medicare coverage can be associated with rapid diffusion of preventive technology.
The researchers identified several factors behind the rapid uptake of mammography CAD. Manufacturers viewed Medicare coverage as critical to early marketing and lobbied Congress for coverage. Members of Congress may have been sensitive to women’s citation of breast cancer as a top health concern and sought to avoid potential political ramifications from the perception of denying women access to advances in breast cancer screening. Finally, with the elimination of the Office of Technical Assessment in 1995, Congress lacked a body to provide independent assessments of medical evidence.
The Medicare Improvements for Patients and Providers Act (MIPPA), which empowers the Centers of Medicare & Medicaid Services (CMS) to add Medicare coverage for preventive services, could instigate change, suggested Fenton and colleagues. CMS demonstrated its ability to develop and promulgate evidence-based national coverage secisions when it decided not to cover virtual colonoscopy. However, Congress retains the authority to add benefits for preventive services, so advocates of new preventive technologies may try to bypass MIPPA by pursing Medicare coverage through Congress.
The burden of proof
The accompanying AIM editorial urged physicians to demand scientific evidence that demonstrates benefits derived from new technology outweigh potential harms.
Karla Kerlikowske, MD, of the departments of epidemiology and biostatistics, medicine, and section of general medicine at University of California, San Francisco, warned of technology creep and cited CAD as a case in point.
She referred to the 2007 New England of Medicine study linking CAD to substantial increases in the number of exams with abnormal results, increased numbers of biopsies and a nonsignificant increase in cancer detection. In addition, meta-analyses suggest CAD may increase both cancer detection and false positives, wrote Kerlikowske.
Based on meta-analytic summary estimates, if 100,000 women had CAD used in conjunction with screening mammography, CAD would yield an additional 50 breast cancer diagnoses (0.05 percent of women screened), 1,190 false positives exams (1.2 percent of women screened) and result in 80 biopsies of healthy women (0.08 percent of women screened).
Healthcare providers need to demand scientific evidence beyond FDA approval before adopting new technologies, recommended Kerlikowske, and they should assess the strength and consistency of evidence and weigh clinical benefits and harms of new technologies such as CAD, Fenton and colleagues agreed.
Fenton concurred: “We need to further study the benefits, harms and costs of new technologies such as CAD as they are disseminated in the community.”