Katharine S Gries1, Dean A Regier2,3, Scott D Ramsey4, Donald L Patrick5. 1. Evidera, 1417 4th Ave, Suite 510, Seattle, WA, 98101, USA. kt04@u.washington.edu. 2. Canadian Centre for Applied Research in Cancer Control, BC Cancer Agency Research Centre, Vancouver, BC, Canada. 3. School of Population and Public Health, University of British Columbia, Vancouver, BC, Canada. 4. Fred Hutchinson Cancer Research Center, Seattle, WA, USA. 5. Department of Health Services, University of Washington, Seattle, WA, USA.
Abstract
OBJECTIVE: To develop a statistical model generating utility estimates for prostate cancer specific health states, using preference weights derived from the perspectives of prostate cancer patients, men at risk for prostate cancer, and society. METHODS: Utility estimate values were calculated using standard gamble (SG) methodology. Study participants valued 18 prostate-specific health states with the five attributes: sexual function, urinary function, bowel function, pain, and emotional well-being. Appropriateness of model (linear regression, mixed effects, or generalized estimating equation) to generate prostate cancer utility estimates was determined by paired t-tests to compare observed and predicted values. Mixed-corrected standard SG utility estimates to account for loss aversion were calculated based on prospect theory. RESULTS: 132 study participants assigned values to the health states (n = 40 men at risk for prostate cancer; n = 43 men with prostate cancer; n = 49 general population). In total, 792 valuations were elicited (six health states for each 132 participants). The most appropriate model for the classification system was a mixed effects model; correlations between the mean observed and predicted utility estimates were greater than 0.80 for each perspective. CONCLUSIONS: Developing a health-state classification system with preference weights for three different perspectives demonstrates the relative importance of main effects between populations. The predicted values for men with prostate cancer support the hypothesis that patients experiencing the disease state assign higher utility estimates to health states and there is a difference in valuations made by patients and the general population.
OBJECTIVE: To develop a statistical model generating utility estimates for prostate cancer specific health states, using preference weights derived from the perspectives of prostate cancerpatients, men at risk for prostate cancer, and society. METHODS: Utility estimate values were calculated using standard gamble (SG) methodology. Study participants valued 18 prostate-specific health states with the five attributes: sexual function, urinary function, bowel function, pain, and emotional well-being. Appropriateness of model (linear regression, mixed effects, or generalized estimating equation) to generate prostate cancer utility estimates was determined by paired t-tests to compare observed and predicted values. Mixed-corrected standard SG utility estimates to account for loss aversion were calculated based on prospect theory. RESULTS: 132 study participants assigned values to the health states (n = 40 men at risk for prostate cancer; n = 43 men with prostate cancer; n = 49 general population). In total, 792 valuations were elicited (six health states for each 132 participants). The most appropriate model for the classification system was a mixed effects model; correlations between the mean observed and predicted utility estimates were greater than 0.80 for each perspective. CONCLUSIONS: Developing a health-state classification system with preference weights for three different perspectives demonstrates the relative importance of main effects between populations. The predicted values for men with prostate cancer support the hypothesis that patients experiencing the disease state assign higher utility estimates to health states and there is a difference in valuations made by patients and the general population.
Authors: Mark R Middleton; Michael B Atkins; Kaitlan Amos; Peter Feng Wang; Srividya Kotapati; Javier Sabater; Kathleen Beusterien Journal: BMC Cancer Date: 2017-10-17 Impact factor: 4.430
Authors: Y van Wijk; I Halilaj; E van Limbergen; S Walsh; L Lutgens; P Lambin; B G L Vanneste Journal: Biomed Res Int Date: 2019-06-06 Impact factor: 3.411
Authors: J M van Dongen; M L van Hooff; A P Finch; M W van Tulder; J E Bosmans; R W J G Ostelo; M de Kleuver Journal: Health Qual Life Outcomes Date: 2019-11-06 Impact factor: 3.186