FRIDAY, April 23 (HealthDay News) -- Researchers have identified minimally acceptable performance levels for physicians who interpret screening mammography studies, and they recommend that those who don't meet the standards be considered for additional training, according to a study in the May issue of Radiology.
Patricia A. Carney, Ph.D., of the Oregon Health and Science University in Portland, and colleagues asked 10 expert radiologists to consider a hypothetical group of 100 interpreting physicians and identify cut points for minimally acceptable performance. They then illustrated the potential impact those cut points would have on radiology practice, and rescored until the group reached a consensus identifying minimally acceptable performance.
The researchers found that, among the cut points for low performance identified by the team were sensitivity below 75 percent; specificity below 88 percent or above 95 percent; recall rate of less than 5 percent or greater than 12 percent; cancer detection rates of less than 2.5 per 1,000 interpretations; positive predictive value of recall (PPV1) less than 3 percent or greater than 8 percent; and positive predictive value of biopsy recommendation (PPV2) less than 20 percent or greater than 40 percent. Using these cut points, an estimated 18 to 28 percent of interpreting physicians would likely be considered for additional training on the basis of sensitivity and cancer detection rate, and the cut points for specificity, PPV1, PPV2, and recall would likely affect 34 to 49 percent of interpreting physicians.
"This study identified minimally acceptable performance levels for interpreters of screening mammography studies. Interpreting physicians whose performance falls outside the identified cut points should be reviewed in the context of their specific practice settings and be considered for additional training," the authors write.
Abstract
Full Text (subscription or payment may be required)
