RANDOMIZED TRIAL OF A DECISION AID TO PROMOTE APPROPRIATE COLORECTAL CANCER SCREENING AMONG OLDER ADULTS
Method: A single-blinded, randomized controlled trial testing the effectiveness of a written decision support intervention targeted for older adults making a decision about whether to undergo CRC screening compared to an attention control about driving safety. Participants were recruited from 14 primary care practices affiliated with the Duke Primary Care Research Consortium. Eligible patients were ages 70 to 84, had upcoming appointments within 4 to 6 weeks, and were eligible for CRC screening. The primary outcome was appropriate CRC screening behavior six months after the index visit. Appropriate screening behavior was defined as screening test completion for participants in the good health state, discussion about CRC screening in the intermediate health state (preference sensitive), and no screening test completion in the poor health state. Health state classifications were derived from age and Charlson Comorbidity Index and were based on the life expectancy needed to expect a net benefit from CRC screening.
Result: Of 424 patients randomized, 412 (97%) had complete data for analysis. The proportion of participants who exhibited appropriate CRC screening behavior at six months was higher in the decision support group compared to control (55% vs. 45%; p=0.03). Consistent with our hypothesis, a higher proportion of participants in the good health state were screened in the intervention group compared to control group (27% vs 13%; p=0.01) and in the intermediate health state, a higher proportion of participants in the intervention group reported discussions compared to controls. (53% vs 38%; p=0.02). We did not find a difference in the proportion who did not undergo screening among participants the poor health state (92% PDA vs 95% control; p=0.56).
Conclusion: Overall, the decision aid promoted appropriate CRC screening, primarily in participants with good or intermediate health status. Low rates of screening in participants with poor health status limited our ability to test the effectiveness of the decision aid in decreasing overuse in this subgroup. Targeted patient decision aids may be an effective method to promote appropriate care in an older population along the continuum of health.