Comparison of woman-specific versus breast-specific data for reporting screening mammography performance

Abstract

Rationale and Objectives.
Screening mammography data can be reported on a breast-specific or woman-specific level, and much mammography data available for research is woman-specific. The purpose of this study was to determine if woman-specific screening mammography data are sufficient for research and reporting by measuring and comparing the accuracy of screening mammography on a breast-specific and on a woman-specific level.

Materials and Methods.
Definitions for true-positive and false-positive mammography results were developed to distinguish between breast-specific and woman-specific calculations. The sensitivity, specificity, and positive predictive values of screening mammography were calculated on a breast-specific and on a woman-specific basis for the entire population of the Carolina Mammography Registry and for a randomly selected subset of the population.

Results.
Only small differences were found in breast-specific versus woman-specific calculations of sensitivity, specificity, and positive predictive values for both the entire population and the smaller subset population. For both populations, woman-specific sensitivity and positive predictive values were slightly higher than the same breast-specific values, and woman-specific specificity was slightly lower.

Conclusion.
For research and reporting, woman-specific data are sufficient.