Sensitivity vs. Specificity

"It's true that increased sensitivity is generally achieved at the expense of ...specificity," Lee said. "But if it's increased cancer detection at whatever cost, then clearly CAD has been demonstrated to do that."


Nonetheless, in an accompanying editorial to Fenton's study, Dr. Karla Kerlikowske, a professor of medicine, epidemiology and biostatistics at University of California, San Francisco, observed that matched design studies - where CAD and non-CAD interpretations are compared on a single mammogram - show better results than many community practice investigations.

One of the largest studies of CAD in a community practice, published in the New England Journal of Medicine in 2007 and cited by Kerlikowske, found the technology increased the number of women getting biopsies with no significant increase in the number of cancers found.

"However, to my knowledge, no study has had a sufficiently large sample size to determine whether CAD use is associated with significantly higher rates of breast cancer detection or early-stage invasive disease, and no study has examined whether CAD use would result in greater reduction in breast cancer mortality," Kerlikowske wrote.

Lee said a study showing mortality reduction would be difficult, requiring huge numbers of patients.

"Without a study situation doing it prospectively in a kind of randomized way, I don't know how you'd actually demonstrate the impact of CAD in terms of mortality reduction," she said.





Back to HCB News

More Industry Headlines