JNCI: CAD Does Not Improve Film Mammogram Accuracy
A study published Friday in the Journal of the National Cancer Institute (JNCI) found that computer-aided detection (CAD) software used to help analyze and interpret mammography images does not improve the accuracy of mammograms in detecting breast cancer. In an accompanying editorial, Donald Berry, Ph.D., of the M.D. Anderson Cancer Center in Houston writes that CAD software developers should continue working on making these programs more useful, but that this research "should happen in an experimental setting and not while exposing millions of women to a technology that may be more harmful than it is beneficial.” According to JNCI, CAD software is used in the reading of three out of four mammograms performed in the United States, and the study authors expressed concern that this widespread use without firm data establishing the effectiveness of CAD may be driving up health care costs by leading to more false positives that require additional diagnostic testing to rule out. Study author Joshua J. Fenton, M.D., at the University of California, Davis, and his colleagues reviewed data from 684,956 patients. The data set included more than 1.6 million film screening mammograms carried out on at 90 facilities in seven states from 1998 to 2006. The facilities were all participants in the federally supported Breast Cancer Surveillance Consortium, a network set up to help monitor quality data. The researchers did not review CAD used with digital mammography. According to the researchers analysis of this data, women whose mammograms were reviewed with the help of CAD software had more false positives for tumors. CAD also did not improve detection of invasive cancers, and the cancers that were detected using CAD were no more likely to be early detections (smaller tumors, lower stage or less lymph node involvement) than those detected without CAD. The results were the same after adjusting for patient age, breast density, use of hormone replacement therapy, and other factors that might influence mammography findings. Read the study's abstract. Read Dr. Berry's accompanying commentary.