Keeping the Wire Hot with Calls? Recruiting Family Practices for a Health Care Research Project
DOI: 10.3238/zfa.2012.0061-0068german translation / full article
Introduction: Successful practice recruitment is crucial when conducting projects in the field of primary health care research. Although family practices are a key target group there is little data available about the process of recruitment and the yield of family practices. This paper reports about systematic practice recruitment methods, their results and recruitment efficiency using a cluster randomised controlled intervention study PräfCheck as an example. We investigate whether it makes sense to follow up non-responding practices, and elaborate on the problem of practice representativeness.
Methods: To recruit practices for the PräfCheck project, all family practices within a specified post code area providing health care for patients with statutory health insurance received information letters. A predefined recruitment plan was used with the aim of receiving a definite positive or negative reply from each practice. Initially, 167 family practices were addressed in two recruitment waves followed by a third wave targeting 43 teaching practices of the Institute of Family Medicine (Hannover Medical School).
Results: Only 23% (38/167) of all informed practices from the first two waves responded to the invitation. The remaining 77% (129/167) had to be contacted again via telephone calls. 14% (24/167) of all practices agreed to participate. The participation rate of the „responders“ was 47% (18/38), compared to 5% (6/129) of the „non-responders“. The third recruitment wave with teaching practices resulted in a higher participation rate of 35% (15/43). Concerning representativeness, we observed a selection effect towards younger family practitioners which was enhanced by preselected teaching practices.
Conclusions: In the intervention trial PräfCheck, response and participation rates of family practices are quite low. The aim of receiving definitive answers of all practices is not worth the effort. Practices who do not respond are in fact not interested in study participation. Selection effects related to practice recruitment are caused by low participation rates or pre-selected samples and should be considered when transferring results into practice. Establishing research in German family practices seems to be crucial since evidence closely tied to practice reality is strongly needed.