Utilities play a key role in the evaluation of treatments that, in addition to cost and survival, have quality of life implications (Gold 96). However, considerable dissatisfaction exists regarding the quality of utility instruments available (Froberg-III, 89, Gold 96), and how utility scores are used in cost-effectiveness analyses (CEA; Neumann 1997 MDM, Fryback). The extent to which utility assessment, and method for incorporating utilities in to the cost-effectiveness (CE) models, actually affects the CE model results in a significant and meaningful way is largely unstudied. In the proposed study, we will use a model for ovarian cancer screening to address these issues.
Our specific aims are: (1) Compare six utility questionnaires: (2) Determine the impact of frame of reference on current health and hypothetical health state utilities; and (3) Determine the sensitivity of cost- effectiveness models to assessment method, sample characteristics, including experience with the disease, age, race/ethnicity and gender, and to techniques for incorporating the utility scores in to the cost-effectiveness model. We will conduct simulations to determine both the range of screening parameters for which any screen for ovarian cancer might be considered cost-effective, and to fit our ovarian cancer screening model to screening for other cancer diagnoses, to determine whether cost-effectiveness models for screening for other cancer diagnoses are also sensitive to choice of utility score. We will specifically administer six different utility questionnaires for current health and multiple hypothetical health states in ovarian cancer and primary care patients, and in women enrolled in the Northwestern Ovarian Cancer Early Detection Program (n=1800). A sampling strategy will be used where each patient responds to four of the six utility questionnaires for three health states to minimize respondent burden. This methodological study will focus on the improvement of cost- effectiveness methods through both the better measurement and use of utilities in CFAs, providing information that is generalizable across all cancers and chronic illnesses, and will also provide the framework for future studies of the CE of ovarian cancer screening.
Showing the most recent 10 out of 15 publications
