Linking evidence-based program participant data with medicare data: the consenting process and correlates of...

As part of a nation-wide study of the Chronic Disease Self-Management Program (National Study), older participants were asked to consent to have their Medicare data matched with study data. This provided an opportunity to examine the consenting process and compare consenters, refusers, and non-responders. We compared the three groups on a large number of variables. These included demographic, National Study participation, health indicator, health behavior, and health-care utilization variables. We assessed differences in 6-month change scores for time-varying variables. We also examined whether asking participants to consent prior to the final questionnaire impacted completion of that questionnaire. Of 616 possible participants, 42% consented, 44% refused, and 14% failed to respond. Differences by ethnicity were found, with Hispanics more likely to consent. There was a consistent tendency for those who participated most in the National Study to consent. With the exception of number of chronic diseases, there was no evidence of health indicators or health behaviors being associated with consenting. Participants with more physician visits and more nights in the hospital were also more likely to consent. Those asked to consent before the 12-month follow-up questionnaire were less likely to complete that questionnaire than those who were asked after. Fewer than half consented to link to their Medicare data. The greater willingness to consent by those who participated most suggests that willingness to consent may be part of program engagement. Consenters had more diseases, more MD visits, and more nights in the hospital, suggesting that greater contact with the medical system may be associated with willingness to consent. This indicates that examinations of Medicare data based only on those willing to consent could introduce bias. Asking for consent appears to reduce participation in the larger study.