placebo-controlled RCT testing dutasteride vs placebo on PC risk. Eligible men were 50-75 yrs old, had a single, negative, pre-study bi- opsy with a PSA¼2.5-10. Men were mandated to undergo a protocol- dependent biopsy at 2 & 4 yrs regardless of PSA. Race was self-re- ported. PC grade was classified as low (Gleason<7) or high (Gleason>7). Baseline clinical data were compared by race using rank- sum & chi-square. Uni- & multivariable logistic & multinomial regression were used to test if race predicts PC, grade (low vs high vs no PC) & biopsy compliance. Multivariable models were adjusted for age, PSA, BMI, DRE, prostate volume, region, treatment arm, PC family history & smoking. RESULTS: Black & white men had similar age, prostate vol- ume, treatment arm, smoking status & PC family history. Black men were more likely to have abnormal DRE (p<0.001), had lower pre-study PSA (p¼0.05) & came largely from North America (p<0.001). On uni- & multivariable analyses, black race was linked with lower risk of having any on-study, 2 & 4yr biopsies (all p¼0.001). Of 1,393 PC diagnoses, there were no differences in distribution between black (2.3%) & white men (97.7%, p¼0.08). There was no link between race & risk of PC overall or at 4yr biopsy on uni- or multivariable analysis; however, black race was linked with higher PC risk on 2yr biopsy in uni- (OR 1.55, p¼0.05) & multivariable analyses (OR 1.60, p¼0.04). Black men had higher risk of high-grade PC overall (unadjusted RRR 1.76, p¼0.05; adjusted RRR 1.91, p¼0.03) & on 2yr biopsy (unadjusted RRR 2.44, p¼0.004; adjusted RRR 2.53, p¼0.005). There was no association between race & low-grade PC. CONCLUSIONS: Among men with a negative biopsy, black men had greater non-compliance with study mandated biopsy. While black men had greater PC risk, especially high-grade, on the 2yr biopsy, this effect was lost overall (except high-grade which remained but was attenuated) due to lower compliance leading to less opportunity for PC detection. These data suggest black race is linked with increased PC risk but the effect may be attenuated by poor compliance. If true, pop- ulation-level estimates of excess PC burden in black men may under- estimate the degree of PC disparity. Source of Funding: none MP34-05 ROLE OF CONTINUITY OF CARE IN RACIAL ETHNIC DISPARITY IN PROSTATE CANCER CARE AND OUTCOMES Ravishankar Jayadevappa*, S Bruce Malkowicz, J Sanford Schwartz, Sumedha Chhatre, Philadelphia, PA INTRODUCTION AND OBJECTIVES: Assessment of care continuity is important in evaluating the racial and ethnic disparity in the quality of care for men with prostate cancer. In this study we analyzed the contribution of continuity of care on racial and ethnic disparity in process of care and outcomes among prostate cancer patients. METHODS: This was a population-based cohort study. Sur- veillance, Epidemiological, and End Results-Medicare (SEER-Medi- care) data from 1995- 2013 was linked with AMA and AHA data. Eligible patients were men 66 years or older with localized or advanced prostate cancer at diagnosis. Continuity of Care (COC) index was computed for the cohort. We considered pre-diagnosis, treatment, follow-up and ter- minal phases of care. Process of care measures were complications, ER admissions, readmissions, and treatment. Outcomes were overall and disease specific survival and cost. Propensity score and instru- mental variable approaches were used to minimize potential biases. We used Cox regression for survival, log-link GLM models for cost, and Poisson (zero inflated) models for count data. Additionally, difference in difference analysis was used to study the contribution of COC to racial and ethnic disparity in outcomes. RESULTS: We identified 668,510 prostate cancer patients. Of these, 73.7% were white, 12.7% were African American, 6.6% were Hispanic and 4.6% were Asian. Thirty four percent of these had surgery and 43% had radiation therapy as primary treatment with curative intent within three months of diagnosis. Level of continuity of care varied between racial and ethnic groups and higher level of continuity of care was associated with improved survival outcomes (OR¼1.12, CI 1.04, 1.34). Continuity of care made significant contribution to the racial and ethnic disparity in process care (number of procedures performed OR¼0.91, CI¼0.89-0.94) and outcomes (complications, OR¼0.89, CI¼0.83, 0.94) for localized and advanced prostate cancer patients. CONCLUSIONS: This study is first of its kind to identify the mediating effect of continuity of care on racial and ethnic disparity. Increasing continuity of care can help reduce some of the disparity in process of care and outcomes among prostate cancer patients. Source of Funding: This work was supported by Agency for Healthcare and Research Quality 1R01HS024106-01 MP34-06 SOCIOECONOMIC DISPARITIES IN RADICAL PROSTATECTOMY PROCEDURE UTILIZATION Olamide Omidele*, Mark Finkelstein, New York, NY; Michael Palese, New York, NY INTRODUCTION AND OBJECTIVES: Radical prostatectomy (RP) is the benchmark surgical treatment for men diagnosed with clin- ically organ-confined prostate cancer. Several studies have analyzed the effectiveness of radical prostatectomy in eradicating the disease. However, few studies have evaluated racial and socioeconomic char- acteristics of patients undergoing RP. Specifically, there is a paucity of data on patient characteristics based on physician level and facility level volume. METHODS: Data was extracted from the New York State Statewide Planning and Research Cooperative System (SPARCS) database from 2009 to 2014. ICD-9-CM procedure code 60.5 was used to identify all radical prostatectomies conducted in NYS during the study period. Patients’ diagnosis, age, race/ethnicity, primary payment method, severity of illness, length of stay, and hospital characteristics were included. All physicians were categorized into very-high-, high-, medium-, and low-volume groups. Hospitals were stratified into similar volume-based groupings. RESULTS: Low-volume centers and physicians were more likely to conduct surgeries in rural and non-teaching settings. RP per- formed by very high-volume physicians was largely conducted in NYC (p¼ <0.001). African-American patients made up a greater proportion of patients seen by low-volume physician when compared with very-high volume physicians, 22% versus 7% respectively (p ¼<0.001). Similar results were seen when evaluating facility level data, 25% versus 8% respectively (p¼<0.001). Medicaid patients were overrepresented in low-volume physicians and hospital groups at 5% compared to <1% in very high volume groups (p¼<0.001). The median income by zip code was highest for patients who were treated in very-high volume facilities and by very-high volume physicians, $75,459.00 and $70,124.06 respectively (p¼<0.001). CONCLUSIONS: The results showed significant differences in hospital characteristics, racial distribution, and primary payment methods between low- and high-volume categories. Specifically, Medicaid patients and African-American patients were more likely to see low-volume physicians or go to low-volume facilities. This suggests that disparities exists for disadvantaged groups and this may be attributable to accessibility to care from high-volume physicians and facilities. Vol. 199, No. 4S, Supplement, Saturday, May 19, 2018 THE JOURNAL OF UROLOGY â e439