Background: Although colorectal cancer is preventable through regular screening, utilization rates for each of the four recommended modalities are low. A better understanding of preference variation for existing colorectal cancer screening (CRCS) tests is needed to increase patient compliance with screening recommendations.

Purpose: To use conjoint analysis (CA) to understand variation in preferences for different attributes of CRCS tests (e.g., test accuracy, cost) among low literacy and minority primary care patients.

Methods: This study has two phases that correspond to stages of CA. Phase I involves development of CA tools and addresses the first 3 stages of CA. We conducted in-depth interviews with 74 patients (25 white, 27 African American, 22 Hispanic), and found that accuracy, preparation, frequency, discomfort, and cost were the top 5 attributes relating to CRCS. These data were used to create and pilot-test an initial CA assessment instrument. Phase II addresses the final 2 stages of CA and involves determination of the utilities (i.e., “part-worths”) related to specific attribute/level combinations obtained from rating and ranking of attributes by respondents. Preliminary analysis entails assessing the proportion of respondents willing to tradeoff between different attribute/level combinations.

Results: Phase II results indicate that patients trade off attributes of CRCS presented in CA scenario format. The leading tradeoff was a willingness to pay for a more accurate test, but was modified by patient education. Among pilot interviews with individuals with “some college or more” 80% were willing to pay more money for a more accurate test, while 50% of individuals with “high school or lower” were willing to make this tradeoff, but were not willing to pay more than $25. Fifty percent of respondents were willing to trade “some discomfort” for a more accurate test and 75% were willing to trade “some discomfort” for a less frequent test. The extent of these tradeoffs varied among racial/ethnic groups.

Conclusions: Using CA to assess preferences for CRCS provides a good understanding of how attribute/level combinations are valued among low literacy and minority primary care patients. Further evaluation of the relationship between CRCS preferences and patient characteristics will be important in tailoring interventions designed to increase compliance with CRCS in diverse populations.