Social factors and prostate cancer survival. Free

B34

Background: Racial variation in comorbidity at diagnosis and cause-specific mortality among men with prostate cancer suggests socio-demographic factors influence prognosis. Objective: Our objective was to examine the relation of concentrated disadvantage at the census tract level with comorbidity and mortality from prostate cancer vs. other causes following a diagnosis of prostate cancer. Methods: We collected clinical and demographic data on 833 patients diagnosed between 1986 and 1990 at two VA and two private sector hospitals in the Chicago area. The Charlson index measured comorbidity at diagnosis; hospital and state records established vital status; and underlying cause of death assigned based on death certificate review. We calculated concentrated disadvantage (CD) of the patient's census tract of residence at the time of diagnosis as 0.85 x (% in poverty + %unemployed + %female-headed households) + 0.85 x (1 - %college educated) using 1990 U.S. census data. Its impact on comorbidity and survival was analyzed using linear regression and Cox proportional hazards models adjusted for patient demographics (age, race/ethnicity), tumor characteristics (stage, differentiation), treatments, and other clinical factors. Results: Concentrated disadvantage (range 37-236) increased the risk of death from prostate cancer among patients diagnosed in the private sector (2^nd thru 4^th quartile vs. 1^st quartile [lowest], multivariable hazard rate ratio [HRR] = 1.78, 95% confidence interval [CI] [1.64,2.71], P = 0.008). Concentrated disadvantage positively associated with comorbidity score (range 0 - 14) among whites (β_CD = 0.0187, P = < 0.0001) and patients in the private sector (β_CD = 0.0096, P = 0.011). It also increased all-cause mortality and risk of death from causes other than prostate cancer among whites (HRR = 1.39 [1.01, 1.89], P = 0.04 and 1.71 [1.11, 2.64], P = 0.016, respectively); and in the private sector (HRR = 1.99 [1.50, 2.63], P < 0.0001 and 2.11 [1.45, 3.08], P = 0.0001, respectively). However, concentrated disadvantage was not a risk factor for comorbidity or mortality from prostate cancer or from other causes among blacks and among patients diagnosed in the VA. Conclusions: Concentrated disadvantage at the census tract level associates with higher comorbidity at diagnosis, increases risk of death from prostate cancer and from other causes following a diagnosis of prostate cancer, but the associations vary by race and by healthcare setting.