Project Overview +

This study translates a paper-based preference-screening tool into a web-based preference-tailored intervention that is effective for increasing informed decision making (IDM) and compliance with colorectal cancer (CRC) screening. The computer-based preference tool is used in clinical settings to help low-risk individuals decide which of five CRC screening tests best fits their preferences. By helping them choose which test to take, we hope to increase CRC screening rates.

Related Media +

Related Media:

Aims +

Aim 1. Translate a paper-based CRC screening information sheet and preference assessment tool currently being used among primary care patients to a more user-friendly, easy-interface medium, specifically a tablet PC with Web capability.

Aim 2. Pilot test the new tool/format (preference-based tool) among 20 primary care patients with diverse race/ethnicity and literacy levels.

Aim 3. Prepare an NCI R01 application to conduct a randomized controlled trial of the new preference-based tool versus usual care for increasing: a) IDM about CRC screening; and b) completion of CRC screening.

Participants +

Eleven individuals who are part of the Ann Arbor Veteran Affairs (VA) Healthcare System, between 50-79 years old, have never been screened for CRC or are out of date for CRC screening, have no personal history of CRC or colon polyps, and have no family history of CRC.

Intervention +

Participants arrive at the VA 40 minutes before their scheduled appointment. An RA reviews the study with them and takes them to the computer set up with the program. Participants read an overview about CRC, descriptions of four screening tests available to them, and definitions of key screening test terms used in the study's conjoint analysis exercise.

Each intervention participant then sees 20 screens, each with a hypothetical screening test scenario. Each screen shows participants two choices with different characteristics. For example:

Option 1: test every 5 to 10 years, follow-up tests needed.

Option 2: yearly testing, no follow-up tests.

Option 1 words and images are on the left side of the screen, while Option 2 is on the right. Participants choose one point on a 9-point scale ("I strongly prefer Option 1" to "I strongly prefer Option 2") and their responses are stored in a database for later conjoint analysis.

After completing the scenarios, participants review a page that lists their "Top 3 test attributes" (e.g., frequency, accuracy, preparation). They also see the screening test that is most consistent with their top attributes (e.g., FOBT) and basic information about this test. They may view more information if they choose. They also see a list of other test options and may view more information about each one. Finally, they choose the test they would prefer from a list of all four test options. A final screen lists their preference-matched test, determined by the conjoint analysis exercise, and their preferences and their final choice. A printout of this screen is provided to the participant, with a brief description of each of the screening options included on the back of the printout.

Findings +

Of the 11 Veterans who participated, 7 were recommended a colonoscopy, 3 were recommended fecal occult blood tests, and 1 was recommended sigmoidoscopy after completion of the exercise. The chosen preference at the end of the intervention matched the preference generated by conjoint analysis for 85% of them.

Eighty percent indicated they wanted to discuss CRC screening with their provider at their upcoming visit, and 73% indicated that they intended to get tested in the next year.

The intervention took an average of 16 minutes (+/- 5 minutes) to view, including completing the preference elicitation exercise.

Of the 11 Veterans, only 1 had a home computer yet all could navigate the program and felt the exercise was easy to complete. Three of the 11 Veterans had difficulty reading; in these cases, the information was read to them by the P.I. or study coordinator.

All 11 provided positive comments about the look, feel, and information provided in the website.

Conclusion +

A preference tailored tool that uses conjoint analysis to elicit preferences in real time was well received by Veterans in our pilot study. While we were not able to evaluate the impact of the tool on screening use, preliminary data suggest that using the tool has a positive impact on screening intentions and on intention to discuss CRC screening with providers.

The next step is to evaluate this tool in a larger sample of Veterans as well as across different primary care settings, which will be done in two upcoming projects (VA IIR-06-205 and NIH R01CA131041).