Recruiting participants for focus groups in health research: a meta-research study | BMC Medical Research Methodology

The aim of our meta-research study was to investigate what information published focus group studies in health research report on the planning and conduct of recruitment of study participants, and what this information reveals about the comprehensiveness and thoroughness of the actual recruitment practices. We undertook this study because focus groups are currently a widely used research method, but to our knowledge there is little insight into their participant recruitment practices. To do this, we randomly selected 80 publications from a systematic literature search and used a descriptive scoring matrix that included key aspects of the recruitment process. Our findings can be used to improve the reporting of this part of conducting (qualitative) research, and can also help health researchers to think carefully about the different steps of recruitment, thereby improving the quality and validity of research.

Reporting of recruiting

Overall, our findings suggest that to date there is neither structured nor consistent reporting of recruitment plans and procedures in published FG studies in health research — with considerable variation in reporting across publications. Firstly, this stands in contrast to the available information in reporting guidelines as well as recruitment literature, which detail crucial aspects that may be considered in terms of ‘best practice’ in reporting, but also actually conducting the recruitment: study participant characteristics (e.g. clearly specifying and delimiting a target group), research team characteristics (e.g. considering whether the research team has been sufficiently trained in recruitment and if not, providing insights into fundamental aspects), recruitment planning (e.g. summarizing previous recruitment strategies that proved successful and how they relate to the intended study), recruitment conduct (e.g. providing clear arguments to potential participants for how their participation will be valuable), quality assurance (e.g. involving members of the target group in the recruitment planning and conduct).

While aspects related to the characteristics of FG participants and information on who recruited them, where and how are fairly often available, a number of other aspects are regularly missing. These include details of the people involved in approaching potential study participants, the use of literature and guidelines for (better) planning, the skills and experience of the research team in recruitment, the duration of recruitment, details of sampling objectives, and the use of incentives to attract participants.

The use of guidelines for reporting qualitative research, such as COREQ and SRQR, is also rarely mentioned in the studies we analysed, despite their widespread recognition and relevance in the scientific community. Both of these well-known guidelines in this field cover questions of participant selection — which indeed should not be confused with actual guidance or instructions about which steps to take — to some extent, particularly regarding which participants were selected, how many, and with what methods. However, our results show that there are several additional elements that could be considered when planning and conducting the recruitment, such as who is responsible for each task, and how potential participants are motivated to take part in a study. As noted above, these aspects are comparatively underreported, suggesting areas where reporting practices could be improved to provide a more comprehensive view of recruitment strategies. Reporting guidelines such as COREQ and SRQR could address this by detailing such aspects more explicitly, and/or incorporating findings from studies that have summarized good practices / evidence for successful recruitment. Additionally, actors such as journal editors and peer reviewers may point authors more explicitly towards its use. For instance, journals could integrate respective information as part of the requirements for information on the research methods in ‘preparing your manuscript’ sections. However, as we found, reporting guidelines are not yet applied rigorously, so recruitment would not necessarily improve as long as this is the case. On the other hand, the various considerations and options to plan and implement recruitment comprehensively may require separate instructions. Hence, while some attempts have been made, e.g. by umbrella organizations such as Cochrane [3] or individual research groups [12], a widely known and approved/agreed recruitment guideline does not seem to exist in health research. Apart from creating such overarching documents, we suggest that key recruiting principles could be incorporated in the training of researchers at their host institutions, for instance in clinical research courses or workshops on qualitative research methods often provided by health science faculties. In addition, our paper can serve as a key resource guiding researchers on essential recruitment considerations to report in future focus group studies, thereby improving the consistency and comprehensiveness of recruitment reporting in health research.

Methods used for recruiting

Beyond simply reporting the above information, our findings suggest that the measures taken to recruit FG participants vary considerably in terms of methodological detail and depth. For example, while there is a range of recruitment channels and methods overall, many studies tended to recruit participants in a relatively unilateral and ‘obvious’ way, e.g. through personal contact with patients during appointments at a doctor’s surgery or hospital. Furthermore, the main responsibility and workload seemed to lie with the researchers — and, where relevant, with the health care staff. As a result, there was comparatively little collaboration with other potentially helpful actors, such as local councils or social care agencies. Finally, the studies we analysed often did not consider the different ways of increasing the interest and willingness of potential participants to participate in a FG: financial remuneration is only one reason among many for people to participate in research [5, 45]. However, for example, carefully weighing up distinct recruitment channels for their effectiveness or cooperating with people and institutions who naturally have direct relationships/contact with the study’s target group may contribute to conducting research both more efficiently (in terms of the available resources) and effectively (finding the ‘right’ participants).

Previous research on participant recruitment in health research has focused on identifying common barriers and facilitators, such as the experience and skills of research teams (e.g. [46]). On the one hand, our study is likely to be consistent with key findings, such as the fact that recruitment is more successful when researchers engage more personally with potentially interested study participants [3] and when the reasons for and aims of a particular study are described comprehensibly [47]. On the other hand, while much seems to be known in theory about what can facilitate or hinder recruitment, our study suggests that — at least in relation to recruitment for FGs — this knowledge is not yet being thoroughly applied in recruitment practice. It is also be possible that researchers may not even be aware of the various conclusions and/or recommendations that have been made. If this is the case, then future research or practice initiatives should focus on efforts to communicate and translate findings on this topic to those responsible for planning and conducting recruitment. This would also include the need to generate evidence on how researchers actually approach the relevant steps. Further, it seems important to understand what kind of support they would welcome in practice, which could relate to more knowledge, guidelines on how to conduct recruitment, or better conditions to cover tasks such as approaching a wide range of diverse potential study participants.

In addition to focusing on the practices and needs of researchers, in a previous study we argued for the need to focus on the perspective of study participants, e.g. the factors that influence the decision to participate in a study or not [2]. For instance, providing a clear description in the study invitation about how one’s own participation will either be beneficial for oneself or for others has been suggested as an important aspect when approaching potential participants. Given that FGs are a specific research setting due to their ‘group’ nature, with which potential study participants may often be unfamiliar, research teams could benefit from a better understanding of how and where to approach candidates in this regard.

A particular challenge here is the aspect of engaging ‘hard-to-reach’ individuals and communities in research [47,48,49]. In our study, we found little or no evidence that researchers consider the need for specific strategies or additional efforts to reach such individuals, e.g. those who are typically more reluctant to consider research participation — even though this may be very beneficial to increase diversity in the perspectives of study participants, thereby uncovering important aspects necessary for providing comprehensive answers for a certain research question. This may constitute a good example of the importance of providing recruitment training for researchers, which could draw researchers’ attention to developing strategies for collaborating for instance with local community organizations during the recruiting phase. However, further research into the extent to which hard-to-reach people are considered for FG may be warranted, not least because our approach only looked at published studies.

Participatory approaches for recruiting

Focusing on the perspective of study participants relates to a rather peripheral issue explored in this paper, namely fostering the active involvement of the target group in study. There is now a wealth of work that has not only provided arguments for how patient and public involvement (PPI) can benefit research, and which methods can be used at which stages of the process and for which purposes [50,51,52], but also ongoing discussion about the extent to which PPI is now a regular part of research, how to overcome the various practical challenges such as avoiding tokenistic involvement, and how to generate evidence of its precise benefits to the research process (see for example [53]). In our case, there was some evidence of PPI activities alongside the planning and running of FGs, with about 1 in 5 studies referring to this approach. However, these rare references to PPI may also confirm the difficulties and questions associated with implementing a more thorough, regular PPI practice, not least because the few reports do not help to generate more evidence about how it contributes to ‘better’ research. In fact, recruiting for FGs is a good opportunity to make research more participatory, as approaching study participants through PPI contributors can be considered a realistic, manageable task in the sense that peers often have better access to certain target groups than researchers. The benefits of participatory approaches to FG recruitment also seem likely to apply to other aspects addressed in this study, such as developing compelling arguments for why participation is worthwhile, and identifying additional recruitment channels that in turn contribute to sample diversity.

Limitations

Although we used a detailed approach to identify and analyse published FG studies, our research has a number of limitations. Firstly, to keep the work feasible, we did not include all eligible publications identified during the search, which would have contributed to an even larger sample. This is also true for the language restriction, but differences in the methodological approaches to recruitment are unlikely to occur according to where the study has been conducted. Since we randomly selected the studies, it is unlikely that more or different studies would have led to very different findings. Further research may still be relevant to investigate recruitment for online focus groups, e.g. to understand whether the method of data collection also impacts the method of recruiting participants. Also, our results are based on the information reported in the publications, and this information may not always be complete. To be certain, it would have been necessary to interview or survey corresponding authors about their recruitment practices. Nevertheless, there are basic, often mandatory, requirements for methodological details to be provided in a publication, and researchers are generally aware of what information they should provide as part of qualitative research. Therefore, it cannot be assumed that authors ‘hid’ major parts of how they actually recruited FG participants. By assessing not only whether a particular aspect of recruitment was reported or not, but also what was mentioned, we were able to make substantive findings.