Determination of preferences and utilities for the treatment of prostate cancer

Resource allocation for programs for the early detection and treatment of prostate cancer is under much debate. Several recent cost-utility analyses have shown little or no benefit for the cost of screening, early detection, or treatment of asymptomatic prostate cancer. Utility is derived from individual preferences for a condition or health state. Individual preferences are summed and the ensuing value is used to weight survival or other health outcome. Determination of utilities is itself under debate regarding which method and whose preferences to use. Using welfare economics as a framework, this study aimed to: (1) assess preferences for radiation therapy, surgery, hormones, and observation for the treatment of prostate cancer; (2) evaluate sociodemographic and clinical factors as predictors of subject preferences; and (3) calculate utility values for stated preferences. Fifty men, 25 at-risk for prostate cancer, and 25 previously treated with radiation therapy for prostate cancer were interviewed, using the Time Trade-Off technique, to elicit preferences and utilities for disease states and treatments. The men in this study, regardless of group (at-risk vs. cancer), showed an increased preference and slightly higher utility for surgery versus radiation therapy or hormonal therapy, as compared to observation. Since surgery was presented with twice the risk of impotence and three times the risk of incontinence than radiation therapy, either the technique for eliciting utilities, the Time Trade-Off, is flawed, or there is a utility to cancer removal separate from any particular therapy or risk of QOL impairments. Higher education and being married were the only factors identified that predicted a preference for any prostate cancer treatment that prolonged survival, despite QOL impairments, over observation. Utility values elicited from these participants were closer in value to previous samples of men with cancer than healthy men or physicians. This indicates that men at-risk for prostate cancer have preferences more similar to those affected than healthy men, and therefore cannot represent a societal perspective for use in cost-utility analysis. These findings have relevance to outcomes research for program evaluation and health policy, and to clinical nursing practice