COMPARISON OF PREFERENCE ASSESSMENT METHODS BASED ON PROSTATE CANCER PATIENT CHARACTERISTICS

Purpose:   The objective of this study was to compare three methods of prostate cancer treatment preference assessment [time trade-off (TTO), rating scale (RS), and Likert-type scale] according to patient demographic and clinical characteristics.

Method:   A cross-sectional sample of men diagnosed with localized prostate cancer (n = 410) was recruited from two Veterans Health Administration and university-based urology clinics. Consented participants, diagnosed within five years of recruitment completed a self-report TTO, RS, and the Values Insight and Balance Evaluation scales (VIBEs). The TTO and RS elicited preferences on four domains including urinary, erectile and bowel symptoms, and prostate-related problems. The VIBEs is a 46-item scale measuring the importance of 11 different domains including urinary function, sexual function, gastrointestinal (GI) symptoms, hormonal symptoms, radiation therapy, surgery, survival, personal and family responsibilities, well-being, self-image, and mood. Socio-demographic and clinical characteristics included in the hierarchical regression analysis were age, relationship status, ethnicity, education level, employment status, household income level, comorbidities, Gleason score, and PSA level.

Result:   Most of the participants were Caucasian (83%), living with a significant other (61%), had some college education or less (67%), were unemployed or retired (73%), and had an income above 30K (58%). The regression analyses revealed that age was a significant independent predictor of sexual function (b=-0.04; p<0.0001), surgery (b=0.03; p<0.001), TTO (b=0.004; p=0.03) and RS (b=0.95; p=0.007) erectile problems, and RS prostate-related problems (b=0.60; p<0.001). Ethnicity was a significant predictor of sexual function (b=-0.69; p<0.001), GI symptoms (b=-0.41; p=0.001), responsibilities (b=-0.33; p=0.04), self-image (b=-0.55; p<0.001), and surgery (b=-0.36; p=0.03). Education level was a significant independent predictor of self-image (b=-0.55; p<0.001) and survival (b=-0.22; p=0.03).

Conclusion:   Men diagnosed with prostate cancer report a wide range of considerations in making decisions about treatment. Concerns range from specific treatment outcomes to impact on family responsibilities and self-image. All three types of preference assessment measures were sensitive to differences in age; however, ethnicity was only associated with GI symptoms on the VIBEs, which was not captured by the bowel problems TTO or RS measure. Compared to the TTO and RS, the VIBEs captured a broader range of patient treatment outcome concerns with its expanded domains of treatment considerations.