Clinical Question: Are women who have a false-positive mammography result less likely to follow-up with additional screening mammography in the future?

Setting: Outpatient (any)

Study Design: Cohort (prospective)

Synopsis: Women who have screening mammography yearly for 10 years have a 50 percent chance of receiving a false-positive result. This study evaluated the effect of a false-positive result on subsequent screening, using a statewide mammography registry in Vermont. The registry includes all women in the state who receive mammography, independent of their insurance plan or health care system. The mammogram used as the starting point (the index mammogram) was not necessarily the first mammogram the women had, but was the one that occurred during the initial enrollment period (May 1996 through May 1997).

Of the 41,844 women older than 40 years who had a mammogram, 3,982 (9.6 percent) had a false-positive result (a number that is striking in itself). These women were younger than those in the true-negative group. False-positive results were more likely to occur if it was the woman's first mammogram; if a woman had a previous mammogram, the results of that one also were more likely to have been false-positive (all P <.01). Women who had a false-positive result were more likely to return for a follow-up mammogram 18 months (odds ratio [OR]: 1.4; 95 percent confidence interval [CI], 1.30 to 1.51) and 30 months (OR: 1.3; 95 percent CI, 1.18 to 1.44) after the index mammogram.

Bottom Line: False-positive mammography results actually increase a woman's likelihood of returning for subsequent mammography screening. Rather than feeling angry about medical care that induces unnecessary anxiety, increases medical costs, and results in additional work-up, women seem relieved that they “dodged a bullet” after the false-positive result is refuted, although they also become more, rather than less, worried about breast cancer. While in this case poor performance is actually good for business, we must get better at providing accurate mammography screening. (Level of Evidence: 1b)