Understanding preferences and how they influence health care utilization and cost-effectiveness is critical. However, there are three major gaps: 1. Many cost-effectiveness analyses assume “perfect” utilization. 2. Methods have not been fully developed to incorporate data from stated preference surveys (willingness-to-pay surveys using contingent valuation or conjoint analysis/discrete choice experimentation surveys) into cost-effectiveness analyses. 3. Although “informed decision-making” has been advocated for health care decisions where patient preferences are important, there is a lack of preference-based interventions and instruments that are efficient enough to be used in clinical practice.

Methods: Data come from three projects: 1. A conjoint analysis survey of patient and physician preferences for colorectal cancer screening tests (N=2000) 2. A cost-effectiveness analysis of utilization of colorectal cancer screening 3. A pilot study to develop a preference-based instrument that can be used in clinic settings to facilitate informed decision-making on choice of colorectal cancer screening methods.

Results: Cost, sensitivity, and how the test was conducted were key factors in determining preferences for colorectal cancer screening tests. Surprisingly, we found that many patients do not prefer any of the existing screening methods and thus would obtain negative utility from screening. Our cost-effectiveness analyses assuming variation in preferences and resultant utilization of screening found that estimates do change when this variation is considered. We also found that developing a screening method that is more preferred by patients (e.g., a highly sensitive test using a CT scan that is low cost) would provide greater utility to patients and likely lead to greater utilization and more cost-effective screening programs. Lastly, we found that we can use a predictive model derived from population preferences data to develop an abbreviated preference survey for clinical use.

Conclusions: Preference and utilization data can be incorporated into cost-effectiveness analyses and used to improve clinical practice and health policy, thereby providing a means for cost-effectiveness analyses to be more useful in the “real world”. This requires the development of approaches to integrate variations in preferences and utilization into cost-effectiveness analyses and the development of approaches to simplify preference surveys so that they can be easily used in clinical settings.