A caries-prevention trial for Cree children: challenges of subject recruitment

Objectives: The goal of this cluster-randomized clinical trial was to test effectiveness of one-on-one counseling to control caries in Cree children. Phase 1 objectives were recruitment of subjects and implementation of intervention in the 9 communities. Methods: The trial's proposed protocol was to hire local women as project staff, “dental health representatives" (DHRs), from the 5 test/4 control communities. However, the territorial Health Board decided that existing community health representatives (CHRs) should do the project's work. To realize this goal, CHRs attended in-service workshops and had extensive support from the Project Manager (PM). Results: For numerous reasons, most CHRs were unsuccessful in subject recruitment. In 2/5 test-communities and 2/4 control-communities, recruitment was eventually done by the PM. A DHR achieved success in one test-community and eventually took over the work of a retiring CHR in another test-community. A CHR, aided by the PM, completed and continued the project's work in the 5th test-community. Local DHRs completed recruitment in the remaining 2 control-communities. After 2-years, 272/309 mothers were recruited. Whereas the projected sample size of 309 would have yielded 82% power to detect a reduction in caries prevalence from 0.86 to 0.69, power calculations using the actual sample size showed that power will still be 79% to detect this treatment effect. If an attrition fraction of 4% per year is maintained, power will be 80%. Power will still be >90% to detect a 25% reduction in caries prevalence. Conclusions: Administrators often expect research to be implemented within their organization using existing personnel. When existing human resources prove insufficient, the research trial's human and financial resources need to be efficiently redistributed to ensure success of the research. Study supported by CIHR-RCT Grant FRN 67817.