Radiology: Digital mammo ups cancer detection
Although multiple studies have demonstrated that digital mammography improves the cancer detection rate, 65 percent of U.S. facilities continue to rely on film-screen mammography. Multiple factors likely account for the slow uptick: economics may be a factor as digital mammography carries a high start-up cost. Also, a somewhat hefty learning curve may dissuade radiologists. However, mixed outcomes also may shoulder some of the blame.
To better understand clinical outcomes and help guide organizations through the decision-making process, researchers designed a retrospective study following the transition to digital mammography at Mallinckrodt Institute of Radiology at Washington University of St. Louis in Mo., in November 2006. They focused on the recall rate, cancer detection rate and positive predictive value (PPV) as the main outcomes.
The authors analyzed screening mammography data for three breast imaging radiologists from four time periods: baseline (2004-2005), digital year 1 (2007), digital year 2 (2008) and digital year 3 (2009). The radiologists read 32,600 mammograms in the baseline period; 11,358 mammograms in digital year 1; 7,924 mammograms in digital year 2; and 14,597 mammograms in digital year 3.
Over the course of the study, the recall rates significantly increased from 6 percent at baseline to 7.1 percent in digital year 1; 8 percent in digital year 2; and 8.5 percent in digital year 3, according to the study's senior author Charles F. Hildebolt, PhD, professor of radiology at Mallinckrodt, and colleagues
The authors surmised that the recall rate may have increased with the conversion to an entirely digital practice because the digital system improved visualization of calcifications. “Given that digital imaging is more sensitive than screen-film mammography in the identification of microcalcifications it is not surprising that there is a greater number of recalls for microcalcifications," they explained. "The increased recalls with digital mammography in our study were mainly due to microcalcifications.”
PPV, defined as the probability of cancer among patients undergoing biopsy after a BI-RADS assessment of 4 or 5, dropped from 44.5 percent at baseline to 30.3 percent by digital year 3. The researchers emphasized that the decline in PPV was most evident for calcifications.
Meanwhile, cancer detection rates also climbed from 3.34 percent at baseline to 5.28 percent in digital year 1, and 5.93 percent in digital year 2 before dipping to 4.52 percent in digital year 3, indicating that there may be a consistent increase in cancer detection rate with digital mammography over at least two years, according to Hildebolt and colleagues.
The researchers observed a slight increase in cancers detected at stage 0 and stage 1 after the digital conversion.
Hildebolt et al reported an 81.5 percent reduction in technical repeat cases after the digital conversion.
The authors noted a few limitations to the study, including the assumption of a stable screening population and its single-institution design.
They pointed to the need for further studies and a longer follow-up period to determine whether the trends observed in this study--increased cancer detection and recall rates coupled with reduced PPV particularly with respect to calcifications--persist over time.