3275 Nonclinical factors associated with contralateral prophylactic mastectomy among breast cancer patients in the Surveillance, Epidemiology, and End Results (SEER) database

Abstract

OBJECTIVES/SPECIFIC AIMS: The study aims to measure the associations between nonclinical factors and the likelihood of electing contralateral prophylactic mastectomy [(CPM) i.e. bilateral mastectomy for unilateral cancer] among women with breast cancer, with a focus on the roles of race, relationship status, and geographic location. The outcome of interest is a dichotomized surgery type variable (i.e. CPM versus other surgery). METHODS/STUDY POPULATION:. The Surveillance, Epidemiology, and End Results registry was queried to identify female breast cancer patients diagnosed at stage IA through IIIC from 2010 through 2015 and received surgery as part of their primary treatment (n=174,776). A multilevel logistic regression was used to model likelihood of CPM versus less aggressive surgical treatment (i.e. breast conserving surgery or unilateral mastectomy). Fixed-effects included age at diagnosis, race, relationship status, insurance type, county-level median income, county-level population density, stage at diagnosis (low-stage, IA-IIB; advanced-stage, IIIA-IIIC), an interaction term between race and stage at diagnosis, and breast tumor subtype. County of residence was used as a random-effect. RESULTS/ANTICIPATED RESULTS: Among women with low-stage cancer, compared to the reference class of white women, black women had 0.57 times lower odds of CPM (p<0.0001), Hispanic women had 0.69 times lower odds of CPM (p<0.0001), and Asian women had 0.60 times lower odds of CPM (p<0.0001). Among women with advanced-stage cancer, compared to white women, black women had 0.42 times lower odds of CPM (p<0.0001), Hispanic women had 0.51 times lower odds of CPM (p<0.0001), and Asian women had 0.45 times lower odds of CPM (p<0.0001). Compared to the reference class of single, never-married women, divorced/separated women had 1.25 times greater odds of CPM (p<0.0001), widowed women had 1.11 times greater odds of CPM (p=0.009), and married/partnered women had 1.18 times greater odds of CPM (p<0.0001). County-level variation from the random-effect (MOR, 1.49; p<0.001) had a greater influence on CPM election than fixed-effects for insurance class, breast tumor subtype, county median income, county population density, and year of surgery. DISCUSSION/SIGNIFICANCE OF IMPACT: The nonclinical factors associated with variation in breast cancer surgical decision-making suggest patients and providers both may benefit from further education about surgical treatment options. Providers may also benefit from educational materials that highlight treatment selection disparities within specific contexts, such as surgery for primary unilateral breast cancer. To more narrowly tailor future policy interventions, an additional mixed-methods exploration is recommended to clarify how relationship status and location serve as mechanisms for breast cancer decision-making.