Whose Utilities for Decision Analysis?

Abstract

The goal of this study was to examine sources of variation in the utilities assigned to health states. The authors selected a common clinical problem, carcinoma of the rectum, and examined the utilities assigned to colostomy, a common outcome of treatment for that disease. After preparing and validating a description of colostomy and its effects on patients' lives, utilities for the state were obtained from five groups of individuals. These comprised two groups of patients who received treatment for rectal cancer, a group of physicians and surgeons specializing in the treatment of this disease, and two groups of healthy subjects, none of whom were health professionals. Of the patients who had been treated for rectal cancer, one group had been treated surgically with the formation of colostomies and the other had been treated with radiotherapy and none had a colostomy. Utilities for colostomy were elicited using the standard gamble, category rating, and a treatment choice questionnaire. The groups differed substantially in the utilities assigned to colostomy. In general, patients with colostomies and physicians assigned significantly higher utilities than did patients who did not themselves have a colostomy. The clinical significance of these differences was examined in a simplified clinical decision problem that compared surgery (with colostomy) and radiotherapy (without colostomy) as primary treatment. The expected clinical value of these treatment alternatives was substantially influenced by the differences observed in utilities for colostomy. These results emphasize the importance of patient utilities in clinical decision making and the need to gain greater understanding of the factors that influence the utilities that patients assign to health states.