Abstract B89: The influence of race/ethnicity on facility use and postsurgical hospitalization among Medicare beneficiaries with breast cancer

Abstract

Abstract Background: Racial disparities in treatment and mortality among cancer patients are well documented, although some literature suggests that African Americans and Caucasians have similar mortality when treated in similar facilities. Racial differences in the types of facilities used for cancer care and the how intermediate outcomes may influence mortality are not well understood. The main goal of the project is to understand how women with breast cancer differ by race/ethnicity in their utilization of surgical facilities and to examine whether post-surgical hospitalization differs by race/ethnicity and facility characteristics. Methods: This retrospective cohort study included women with an incident breast cancer found in SEER-Medicare linked data (1998-2006). We examined race/ethnicity in relation to the following breast cancer surgical facility characteristics : nurse-bed ratio, breast cancer volume, rural/urban, teaching status, Disproportionate Share Hospital (DSH), status as a National Cancer Institute (NCI) Cancer Center, and membership in Cooperative Oncology Groups. We also developed logistic regression models to test the relations between race/ethnicity, facility characteristics, and hospitalization within 60 days of breast cancer surgery, while adjusting for other factors. Results: A total of 72,198 women with breast cancer attended an average of 1,046 facilities for breast cancer surgery from 1998-2006 and included 14,096 with a post-surgical hospitalization. Significant racial/ethnic differences in types of facilities used for breast surgery were seen. More African American women attended facilities with a nurse: bed ratio <1 compared to Caucasian women (50% v. 61%, respectively; p<.0001), a teaching facility (71% v. 54%; p<.0001), an NCI Cancer Center (7% v. 3%; p<.0001), and a DSH hospital facility (89% v. 72%; p<.0001). Caucasian women experienced the fewest post-surgical hospitalizations (19.22%) while African American and Native American women had the most (23.84% and 23.45%, respectively). Facility characteristics significantly associated with greater post-surgical hospitalization included rural location, non-teaching status, no membership in cooperative oncology groups, and a lower nurse: bed ratio. In adjusted logistic regression models of post-surgical hospitalization, African American women had 30% greater odds than Caucasians (OR=1.30; 95% CI1.19-1.41). Lower odds were significantly associated with post-surgical hospitalization among women attending facilities with the following characteristics: urban location, cooperative oncology group membership, NCI Cancer Center, nurse: bed ratio≥1, DSH status, and higher breast cancer volume. Conclusions: Significant differences by race are seen in the types of facilities used by women for breast cancer surgery. Both race/ethnicity and facility characteristics are associated with increased odds of post-surgical complications following breast cancer surgery. Differential use of certain types of facilities in relation to race, is likely to contribute to disparities in outcomes among women with breast cancer. Citation Information: Cancer Epidemiol Biomarkers Prev 2010;19(10 Suppl):B89.