Racial/Ethnicity Differences When Endorsing Influential Factors For Prostate Cancer Treatment Choice: An Analysis Of Data From The Personal Patient Profile-Prostate (P3p) I And Ii Trials

351 Background: Race and ethnicity impact the type of treatment received for localized prostate cancer in American men. We hypothesized that there may be differences in men’s influential values and preferences related to treatment decisions. Methods: We analyzed samples from two multicenter, randomized trials of the Patient Profile-Prostate (P3P) decision aid, first comparing the trial groups on demographic and decisional variables using Chi-square tests. Stratified (P3P I vs II) logistic regression was then used to assess the univariate association between race/ethnicity and endorsement of moderate-or-strong influence of 14 lifestyle factors, current or future symptoms, or important others on the decision. A multivariable stratified logistic regression with backward variable selection was used to further estimate the association between influential factors and race/ethnicity. Results: Data from 494 and 392 participants in P3PI and P3PII, respectively, were analyzed for 40 Hispanic, 168 non-Hispanic black, 637 non-Hispanic white, 19 others and 6 missing. Age (p=.0001), education (p<.0001), marital status (p<.0001), income (p<.0001), Internet use for information (p<.0001) and decisional control preference were significantly different across racial/ethnic groups. In adjusted analyses, racial/ethnic differences existed for influence of age (Non Hispanic Black (NHB) vs. Non Hispanic White (NHW) OR: 0.56 95%CI 0.38-0.85p=.002), religion/spirituality (NHB vs. NHW OR: 3.2095%CI1.95-5.26, p<.0001), future bladder function (NHB vs. NHW OR: 0.5795%CI0.35-0.90, p=.04), future ability to engage in recreation (NHB vs. NHW OR: 0.5495%CI 0.34-0.86, p=0.02), and a story of a famous person with prostate cancer (NHB vs. NHW OR: 2,11 95%CI1.30-3.43, p=.007). Conclusions: Our results suggest racial/ethnic differences for influences underlying treatment choice. Better understanding these influences may help us present salient information about treatment options to patients and address disparities.