Radiology, Vol 200, 357-360, Copyright 1996 by Radiological Society of North America



Positive predictive value of breast biopsy performed as a result of mammography: there is no abrupt change at age 50 years

DB Kopans, RH Moore, KA McCarthy, DA Hall, CA Hulka, GJ Whitman, PJ Slanetz and EF Halpern
Department of Radiology, Massachusetts General Hospital, Harvard Medical School, Boston 02114, USA.

PURPOSE: To determine if the positive predictive value (PPV) of a biopsy initiated because of an abnormal mammogram changes abruptly at age 50 years. MATERIALS AND METHODS: The PPV and its variation with age was analyzed for 4,778 women who underwent biopsy for a clinically occult abnormality detected at mammography. The relationship of the results to the patient's age was analyzed with age represented as a continuous and two-categorized (< 50, > 50) measure. The latter measure represented an abrupt change, which distinguished those aged 49 years and younger from those aged 50 years and over. With this measure, the patients in each of the two age groups were statistically indistinguishable. RESULTS: The results were consistent with a steady increase in PPV and the yield of cancers with age, and there was no abrupt change at age 50 years. The modeled PPV for all cancers for these 4,778 patients was approximately 12% for women aged 40 years and increased to 46% by age 79 years. CONCLUSION: The PPV did not change abruptly at any age for women aged 40-79 years but increased steadily, which reflects the prior probability of breast cancer at each age. Inappropriate grouping of data can lead to misinterpretation of results. Screening guidelines should not be predicated on the false assumption that this variable changes at age 50 years.