Recruiting blacks to the Adventist health study: Do follow-up phone calls increase response rates?

PURPOSE: To determine whether follow-up phone calls improve response rates to a long questionnaire among black and white subjects. METHODS: Forty black and 39 white Seventh-day Adventist churches were randomized to experimental or control status in a 2:1 ratio favoring the intervention, which is a follow-up phone call to certain church members. Subjects selected from each church were those who had signed up for the Adventist Health Study-2 but not returned a questionnaire 3 months after promotion began. Further returns from a church over the next 3 months, and this increment as a proportion of baseline response, were assessed using t-tests and Poisson regression, respectively. RESULTS: Comparing black experimental and control churches, the mean difference was 5.5 returned questionnaires per church (p < 0.01). Among white churches the mean difference was 3.0 (ns). The baseline-adjusted increment, however, was greater by a factor of 3.37 (95% confidence interval, 1.92, 5.93) in the black experimental relative to control churches, but among white experimental churches was 13% (ns) lower than controls. This difference in response by ethnic group was statistically significant (p < 0.01). CONCLUSION: Follow-up phone calls improved response rates among black subjects only.