BREAST DENSITY LEGISLATION AND BREAST CANCER STAGE AT DIAGNOSIS

Purpose: 22 states now mandate that physicians notify their patients when mammography reveals radiographically dense breasts. Some states additionally require that physicians inform patients with dense breasts that they may benefit from supplemental screening studies such as breast MRI or ultrasound.  Critics have raised concerns that this may encourage low yield screening. We sought to evaluate the effect of this legislation on the incidence of early and late stage cancer at diagnosis as a proxy for screening yield.

Method: We used data from the Surveillance, Epidemiology, and End Results (SEER) program from 2005-2012. Our main outcome was stage specific, age-adjusted incidence among women ages 40-49, the population with the highest prevalence of dense breasts. We examined incidence in counties in Connecticut, a state that enacted breast density legislation in 2009, compared to counties in states without such legislation. We used a generalized linear model with a difference in differences framework to estimate the policy effect. We clustered standard errors by state.

Results: Our study included data from 209 counties in 9 SEER geographies, including 8 counties in Connecticut over an 8 year period.  Breast density legislation was associated with an estimated 11% increase in early stage diagnoses in Connecticut compared to other states in the same time period (95% CI 3.2-19, p=0.005).  Breast density legislation was not associated with a change in in-situ diagnoses (-1.1%, 95% CI -15.4-15.6), regional-stage diagnoses (11%, 95% CI -0.83-25.7, p=0.068) or metastatic diagnoses (-7.7%, 95% CI -44%-48%, p=0.75).

Conclusion: We found that breast density legislation in Connecticut is associated with an increase in early stage breast cancer diagnosis but did not result in a change in late stage diagnosis.  Longer follow up is needed to evaluate whether breast density legislation eventually results in a decrease in diagnosis at later stages.