Value of recruitment strategies used in a primary care practice-based trial

https://doi.org/10.1016/j.cct.2006.08.009Get rights and content

Abstract

Purpose

“Physicians-recruiting-physicians” is the preferred recruitment approach for practice-based research. However, yields are variable; and the approach can be costly and lead to biased, unrepresentative samples. We sought to explore the potential efficiency of alternative methods.

Methods

We conducted a retrospective analysis of the yield and cost of 10 recruitment strategies used to recruit primary care practices to a randomized trial to improve cardiovascular disease risk factor management. We measured response and recruitment yields and the resources used to estimate the value of each strategy. Providers at recruited practices were surveyed about motivation for participation.

Results

Response to 6 opt-in marketing strategies was 0.40% (53 / 13290), ranging from 0% to 2.86% by strategy; 33.96% (18 / 53) of responders were recruited to the study. Of those recruited from opt-out strategies, 8.68% joined the study, ranging from 5.35% to 41.67% per strategy. A strategy that combined both opt-in and opt-out approaches resulted in a 51.14% (90 / 176) response and a 10.80% (19 / 90) recruitment rate. Cost of recruitment was $613 per recruited practice. Recruitment approaches based on in-person meetings (41.67%), previous relationships (33.33%), and borrowing an Area Health Education Center's established networks (10.80%), yielded the most recruited practices per effort and were most cost efficient. Individual providers who chose to participate were motivated by interest in improving their clinical practice (80.5%); contributing to CVD primary prevention (54.4%); and invigorating their practice with new ideas (42.1%).

Conclusions

This analysis provides suggestions for future recruitment efforts and research. Translational studies with limited funds could consider multi-modal recruitment approaches including in-person presentations to practice groups and exploitation of previous relationships, which require the providers to opt-out, and interactive opt-in approaches which rely on borrowed networks. These approaches can be supplemented with non-relationship-based opt-out strategies such as cold calls strategically targeted to underrepresented provider groups.

Introduction

Randomized trials in which the primary care practice is the unit of randomization are important tools with which to determine the best ways to improve the quality of medical care [1] and to accelerate the translation of research into practice. However, the methodological underpinnings of translational research of this type are only recently garnering attention and little work has been done to understand how to design these resource-intensive studies in the most cost effective and efficient way. While recruitment of healthcare practices to such research has been studied, relatively little is known about which strategies for recruiting primary care practices for such studies are both effective and economical [2], [3], [4], [5], [6], [7], [8], [9], [10].

Research to date has focused on recruiting individual providers either to single studies or to practice-based research networks [11] using either convenience samples or via random selection through larger sampling frames [5], [7]. The methodological limitations of convenience and volunteer samples of practices have been acknowledged [2], [7], [12], [13], [14] and strategies to assess representativeness and to reduce bias have been employed [7]. Yet, the time frames and funding limitations of most translational research pose challenges for achieving these standards. Still other research has sought to identify the most effective strategies. Accordingly, recruitment strategies have been studied in a number of ways to describe the most effective methods. For example, the degree of personal relationship of the recruiter and recruit have been compared, as well as the method of communication (phone or in person) [15], [16]. Personal contact with providers and exploitation of existing relationships, have been reported to be effective [2]. The “physicians-recruiting-physicians” method, in which local physician leaders are utilized to recruit practices, is highly regarded, producing response rates ranging from 39–91% [2], [3], [4], [5], [6], [7]. The strategies which appear to be successful also seem to be subject to the greatest potential for bias, and moreover, could potentially be the most resource intensive [8]. Physician time is expensive; and even a personal contact strategy utilizing less expensive personnel would be potentially more costly than a mail-based approach.

Incentives are thought to be important in provider recruitment, but little research confirms assumptions regarding why healthcare providers might be motivated to participate in research. Resource limitations further constrain the types of incentives which can be offered and continuing education credits are sometimes assumed to be an acceptable incentive to motivate providers to participate in research, in contrast to cash payments and other economic incentives such as reductions in malpractice premiums for participation [5].

In a retrospective analysis of 10 passive and active recruitment strategies used in a campaign to recruit a diverse sample of primary care providers to a randomized trial of cardiovascular disease management tools, we assessed the relative success and cost of each strategy and elicited participating providers' reasons for joining the study. We hypothesized that recruitment through an existing educational network of healthcare providers would be efficient and produce a diverse sample of practices and that the technological tools used in the study, coupled with continuing education credit opportunities would be sufficient incentive for providers to participate.

Section snippets

Methods

In an effort to minimize physician burden during recruitment, reduce costs, and maximize representation of minority providers, we utilized the established relationships of the local Northwest Area Health Education Center (AHEC) network of primary care practices to recruit primary care providers, but ultimately augmented this strategy with additional approaches to reach recruitment targets. After recruiting 68 primary care practices in central North Carolina, we conducted a retrospective

Results

Ultimately we recruited 68 practices. The majority of practices were Family Medicine (74%). The mean number of providers per practice was 4, ranging from 1 to 14. Nineteen percent of the practices had a solo provider. For 15% of the recruited practices, the majority of the providers in the practice were minorities. In 34% of the practices the majority of the providers were female. Table 1 summarizes practice characteristics.

Discussion

Although we hypothesized that opt-in recruitment strategies would be a more efficient use of study resources, requiring less staff time and effort, we found that the greatest yield was in those strategies that required the practice to opt-out. These strategies appear to be resource intensive, and often require the practice manager to act as an intermediary on behalf of the study. However, by putting the onus on healthcare providers to actively respond to recruitment materials through the opt-in

Conclusions

Recruiting healthcare practices to research studies is an arduous process and little is known about how best to apply limited resources. Our study supports the use of physician networks and face-to-face strategies, as has been previously reported, but also suggests the need for further research on other opt-out strategies such as cold calling and supplemental approaches to target ethnically and geographically diverse participants. With all of these strategies, careful collection and analysis of

Acknowledgements

The authors wish to acknowledge the contributions of the GLAD Heart Research staff, Virginia Burnette and Vanessa Duren-Winfield, for their recruitment efforts. Results from these efforts have been presented at the AHA 5th Scientific Forum on Quality of Care and Outcomes Research in Cardiovascular Disease and Stroke, May 17–19, 2004; the 5th International Heart Health Conference, July 13–16, 2004; and the North Carolina AHEC Statewide Meeting, October 6–8, 2004. The study was funded by the

References (27)

  • P.E. McBride et al.

    Recruitment of private practices for primary care research: experience in a preventive services clinical trial

    J Fam Pract

    (1996)
  • A. Pearl et al.

    Randomised trials in general practice—a New Zealand experience in recruitment

    NZ Med J

    (2003)
  • D. Wetzel et al.

    Participation in a quality of care study and consequences for generalizability of general practice research

    J Fam Pract

    (2005)
  • Cited by (57)

    • The Effect of a Personalized Newsletter to Physical Therapists on Patient Recruitment: A Cluster Randomized Trial in Primary Physiotherapy Care

      2020, Journal of Manipulative and Physiological Therapeutics
      Citation Excerpt :

      Therefore, increasing the extrinsic motivation of therapists, and thereby possibly reducing the number of inactive therapists, should be considered in future studies. Several clinician characteristics have been described that can influence the success rate of patient recruitment, such as sex, age, and specialization.2,5,6,24,40 We found no differences in these baseline characteristics between both groups.

    View all citing articles on Scopus
    View full text