PREFERENCES FOR COLORECTAL CANCER SCREENING AND UTLIZATION IN A LARGE MULTI-SPECIALTY PRACTICE

Objective: To assess factors associated with preferences for different colorectal cancer (CRC) screening tests and to evaluate the association between preferences and screening use 6 months later.
Methods: Participants in a behavioral intervention trial to increase CRC screening were members of a large multi-specialty medical practice, aged 50-70, at average CRC risk, and overdue for CRC screening. Multinomial logistic regression assessed preference for CRC screening tests from baseline surveys. Correlates included patient age, race, education, CRC test knowledge, CRC test barriers, perceived CRC risk, physician recommendation, prior CRC use, stage of readiness to get tested, and intervention group. Contingency tables were used to compare baseline preferences to CRC test use 6 months later using the clinic’s administrative database. 
Results: A total of 1,245 participants completed the 6-month follow-up: 37% had a baseline preference for fecal occult blood test (FOBT), 43% for colonoscopy (COL), 14% for sigmoidoscopy (SIG) and 6% for barium enema (BE). Preference for FOBT was negatively associated with stage of readiness (P<0.05). Preference for COL vs. FOBT or COL vs. SIG/BE was associated with greater knowledge (P<0.01). Physician recommendation and prior screening were not consistently associated with preferences. There was no significant association between baseline preferences and 6-month screening. The majority of participants (70%) were not screened by 6 months; however most who were screened received COL regardless of baseline preference.
Conclusions: Results confirm the existence of variation in preferences for CRC screening tests in a large, insured patient population. They suggest that the more knowledgeable and ready patients are to get screened, the more they prefer COL over other tests. The lack of association between preferences and use, combined with the low rate of use overall, suggests that further work to match screening recommendations to preferences may be warranted.