Ailawadhi, Sikander

e18061 Background: Despite a favorable genetic profile, African-Americans (AAs) with mm have poorer outcomes secondary to inferior treatment. NCI-CCs provide the highest-quality of care and attendance has been associated with better outcomes in many cancers. AAs have greater access to NCI-CCs proximally; however, they attend these facilities at lower rates than their white peers. The impact of attendance at NCI-CCs on mediating racial disparities in mm outcomes has not been reported on to date. Methods: We reviewed cases of mm the Surveillance, Epidemiology, and End Results (SEER)-Medicare linked database from 2000-2011 who were enrolled in Medicare Part A and B > 1 year prior to diagnosis, excluding cases enrolled prior to age 65 and those where mm diagnosis occurred post-mortem. Any center designated a NCI-clinical or -comprehensive center in 2002, 2005, or 2010 was considered a NCI-CC; attendance was defined as 2 or more claims on separate dates (MEDPAR and Outpatient files) from a NCI-CC in the 12 months following mm diagnosis. Logistic regression was performed to determine if race was associated with attendance, Cox regression to determine the association of attendance with survival. Results: 21,843 cases were analyzed; the median age was 77 years; 80% were white, 15% AA/Black. Overall NCI-CC attendance was low, only 11% of the population. Compared to white patients, black patients had a 13% decreased odds (aOR 0.87, 95% CI 0.75-0.99) of NCI-CC attendance after controlling for age, gender, socioeconomic status, geographic, and overall health variables. Attendance was associated with a 28% decrease risk for death (aHR 0.72, 95% CI 0.68-0.76), but had little impact on black-white outcome disparities. Black patients had a 9% increase in risk (aHR 1.09, 95% CI 1.04-1.13) after controlling for NCI-CC attendance. Conclusions: Black patients with mm have lower NCI-CC attendance. This may be related to referral bias and/or patient declining referral. NCI-CC attendance was associated with superior outcomes; however, controlling for attendance did not mediate black-white outcome disparities suggesting that racial treatment disparities pervade beyond access to NCI-CCs.