The goal of this study was to evaluate racial patterns in treatment for men with prostate cancer in an effort to explain the higher mortality rate for African American men with this disease compared with white men. The subjects, recruited from the North Carolina Central Cancer Registry, were men 50 to 74 years of age who were diagnosed with prostate cancer during 1994 and 1995. Sampling procedures aimed to identify equal numbers of African Americans and white men with prostate cancer and equal numbers of patients at local, regional, and distant disease stages. Telephone surveys were conducted; the participation rate was higher for whites (79% of African Americans versus 92% of whites).

Subjects differed with regard to educational attainment, annual income, and marital status (African Americans were less frequently married/in a stable union than whites). There were few racial differences in the types of treatment options discussed with patients; the only statistically significant pattern was in discussion about hormonal therapy (53% of African Americans versus 66.7% of whites). Among only the men with distant disease, clinicians were more likely to discuss either hormonal therapy or orchiectomy with white patients than with African American patients. The influence of the physician on the patient's treatment decision did not differ between racial groups. There were no overall differences in the treatments received, even after stratifying by disease stage. Finally, no racial patterns were observed with regard to treatment delay or patient satisfaction with treatment/treatment options.

One potentially important noted difference was that surgical removal of the prostate was contraindicated for distant disease, yet 7 of 22 African American and 2 or 22 white men staged as having distance disease reported this treatment. The authors comment that "although racial differences were statistically insignificant after controlling for stage, the fact that such differences exist is of concern." It is possible that small sample size precluded the study's ability to detect clinically important differences on this treatment domain.

The authors note that this study had different results from a larger previous study and suggested the localized geographic region (versus a national sample), the truncated age distribution (versus including older men), the study design (which included a telephone interview component), and treatment changes since 1990 (when the early study was conducted) may have contributed to the differences shown in the results.