Access, utilization, and mortality in relation to specialized cancer care for African American and Caucasian Medicare cancer patients in the United States (1998-2002)

A66

Background: Emerging evidence suggests that racial disparities in cancer care and outcomes may largely be a function of where care is received, but no studies have addressed this idea in relation to the degree of specialization in cancer care settings.
 Methods: Using SEER-Medicare linked data we identified individuals with a primary incident cancer of the lung, breast, colon/rectum, or prostate diagnosed between 1998 and 2002. Attendance at an NCI Cancer Center was defined as two or more claim-days at an NCI-designated Cancer Center in the first twelve months following diagnosis, and was determined based on claims from 1998-2003. NCI Cancer Center utilization was examined as a function of demographic, clinical, and geographic factors in multilevel models. Race/ethnicity was categorized as Caucasian or African American based on Medicare enrollment files. We examined the relation of NCI Cancer Center (and non-NCI-CC) attendance with 1- and 3-year all-cause mortality using multilevel logistic regression models accounting for cancer site, stage, and receipt of surgery. We compared models between African American and Caucasian individuals.
 Results: The study population included 201,305 beneficiaries, 18,008 (8.9%) of whom were African American. Over 12% of African American cancer patients attended an NCI Cancer Center, compared with 7.4% of Caucasian cancer patients. For both Caucasians and African Americans, travel time to the nearest NCI Cancer Center was a strong determinant of attendance, with each 10 minutes of increased travel time associated with an ~10% decrease in attendance. (OR=0.90, 95% CI 0.89-0.90, OR=0.88, 95% CI 0.86-0.91, respectively). Relative to their Caucasian counterparts, urban/suburban African American patients were more likely to attend (OR=1.66, 95% CI 1.55-1.77), but non-urban African Americans were less likely to attend ( OR=0.55, 95% CI 0.37-0.85) an NCI-CC. Overall, mortality at 1- and 3-years following diagnosis was higher for African American cancer patients compared with Caucasian cancer patients in both unadjusted and fully-adjusted models (unadjusted 1-year mortality: OR=1.29, 95% CI 1.24-1.34; adjusted 1-year mortality: OR=1.17, 95% CI 1.11-1.23; unadjusted 3-year mortality : OR=1.38, 95% CI 1.34-1.43; adjusted 3-year mortality: OR=1.23, 95% CI 1.18-1.28). Later stage at diagnosis accounted for 16% of the excess mortality among African Americans. NCI Cancer Center attendance was associated with decreased mortality among African American and Caucasian cancer patients in stratified, adjusted models (1-year mortality: African Americans OR=0.63, 95% CI 0.54-0.74; Caucasians OR=0.74, 95% CI 0.69-0.78, 3-year mortality: African Americans OR= OR=0.70, 95% CI 0.61-0.80; Caucasians OR=0.89, 95% CI 0.84-0.93). Conclusions: Utilization of NCI Cancer Centers is proportionately greater in African Americans than in Caucasians. Although NCI Cancer Center attendance confers a mortality benefit, 1- and 3-year all cause mortality is still higher for African Americans.