Evaluating a risk-based model for mammographic screening of women in their forties.

Background: A risk-based model was developed for recommendations for mammography screening of women in their 40s. We determined retrospectively the proportion of women already diagnosed in their 40s with breast cancer (BC) who would have been recommended for mammography screening by this model.

Methods: Information was obtained from 404 women who were diagnosed with BC in their 40s from 1990 to 1998 at a large midwestern hospital. Data on BC detection method, mammography history, and BC risk factors existing before diagnosis were obtained from 353 of these women by mailed questionnaires. Data for an additional 51 women who had died were obtained by chart review. Two algorithms, the Exact Age Procedure (EAP) and the Grouped Age Procedure (GAP), were used to calculate the proportion of women who would have been recommended for mammography screening based on age, race, and BC risk factors.

Results: Of women diagnosed with BC in their 40s, 74.5% would have been recommended for screening by the EAP and 70.5% by the GAP. The model would have recommended screening for approximately one half of the women who had no conventional risk factors (54% by EAP, 48% by GAP).

Conclusions: Although use of the risk-based model could result in more recommendations for screening than individual recommendations based on patient-physician discussions about the woman's risk factors, one in four women diagnosed with breast cancer in their 40s would not have been recommended for mammography screening.