Screening remains ultimate hurdle for breast CAD

Despite the increasing availability of commercial computer-assisted detection packages for breast screening, the technology remains unproven in the screening setting, according to a leading U.K. breast radiologist. Prof. Fiona Gilbert, a professor of radiology at the University of Aberdeen, is calling for a Europe-wide randomized control trial to prove the case for CAD as a second reader.

The use of an automated prompt system holds great appeal for organizers of large-scale, population-based breast screening services. While radiologists are in short supply, an aging population means that screening programs must serve an increasing number of eligible women. Many regional and national programs are also extending their services to younger and older women. Finding sufficient trained personnel to provide a double-reading service is a substantial challenge.

Although a number of published studies have demonstrated CAD's clinical validity, most contained an inappropriate case mix, Gilbert said. And radiologists participating in those retrospective trials may have been influenced by the knowledge that their decisions would have no genuine clinical impact. A time series analysis of more than 115,000 screening mammograms that found that CAD made no discernible difference to radiologists' performance has further muddied the waters.

"What we would really like to do is a very robust evaluation of CAD in a prospective setting," Gilbert said. "We believe that we should randomize patients to be double-read, as is standard practice in the U.K., against a single reader using CAD."

The study design should permit the use of arbitration, so that the single reader using CAD could also benefit from a radiologist's opinion if required, she said.

"If another radiologist is allowed to look at the few cases that the single reader with CAD wants to bring back, we can actually reduce our recall rates," she said.

The cost-effectiveness of using breast CAD and the value of trained radiographers as second readers should also be evaluated, Gilbert said.

One problem for organizers of Europe-wide trial might be establishing appropriate recall parameters. Radiologists working for the U.K. National Breast Screening Service are expected to recall no more than 7.5% of women undergoing screening for the first time and no more than 5% of those presenting for subsequent screens. Expected recall rates in the Netherlands, for example, are several percentage points lower.