Sarah T. Hawley, Ph.D. Bob Volk, Ph.D., Maria Jibaja-Weiss, Ed.D.,
Elaine Wendt, M.D., Stephen Spann, M.D.
Baylor College of Medicine, Dept. of Family & Community Medicine

This project will assess preferences for colorectal cancer screening (CRCS) options among ethnically-diverse primary care patients using conjoint analysis (CA). Low screening rates and limited effectiveness of interventions to improve them has focused recent attention on understanding more about variation in preferences for one CRCS test over another. Little information exists about how racial/ethnically diverse patients regard CRCS, or whether certain tests are more appealing or less desirable to individuals of different minority groups. The specific aims of this 3-part study are to: 1) develop an educational module and accompanying assessment instrument to elicit preferences for CRCS among diverse patient populations using CA; 2) assess ethnic-group differences in preferences for CRCS options using CA with a large, urban practice-based research network serving as the laboratory; and 3) determine racial/ethnic (gender and age) differences in preferences for CRCS test attributes that can be used in development of future interventions. Conjoint analysis has 3 strengths for assessment of preferences for CRCS in a diverse population: 1) it is the best method when certain product attributes have to be traded off against one another, such as trading test accuracy for invasiveness; 2) potential "what if" situations, such as a more accurate fecal occult blood test or virtual colonoscopy, can be easily evaluated using CA; and 3) the approach identifies characteristics of populations that are more or less likely to prefer a particular service, or test. The first part of the proposed study will develop the educational modules needed to assess CRCS preferences using CA, the implementation phase will then use these materials to assess preferences among 225 white, African American and Hispanic primary care patients (75 per group). The evaluation of the results will reveal the most preferred CRCS test attributes (e.g., accuracy, low cost), and whether preferences vary according to patient characteristics. The proposed study will lay the groundwork for a study that will replicate the CA methods in a larger population, and use the results to develop and test a preference-based intervention, or decision aid, to improve compliance with CRCS among minority populations.

NIH CRISP Database