Summary of findings
Although it is often stated that there is a need to provide training to those recruiting patients in trials [
4,
6,
9‐
12,
34], only a small number of training programmes were identified in this systematic review. Most training programmes were uncontrolled observational studies of moderate or weak quality. Training was most commonly found in the context of cancer trials and tended to consist of workshops with a mix of health professionals over one or two consecutive days covering both generic and trial specific issues. The effectiveness of training was assessed by various measures including recruitment rates, numbers of patients approached for recruitment to host RCTs, recruiter self-confidence, patient understanding of trial information and perceptions of recruiter communication, and observation of recruiter-patient trial consultations using pre-determined criteria. Findings suggest that RCT recruiter training programmes are acceptable to recruiters and may increase their self-confidence and communication of key RCT concepts to patients. Studies with less robust study designs also suggested that training has the potential to improve recruitment rates and aspects of patient satisfaction and understanding of RCTs. However, the review found limited high quality evidence of interventions aimed at recruiters and therefore demonstrates the need to develop more robust designs to develop an evidence base on how best to target this group for training in trial recruitment. More comparative studies, especially randomised or clustered randomised trials, are the ideal method to assess the effectiveness of such training programmes.
Feedback from recruiters in several studies suggested that they had found the training intervention useful and evidence from a number of non-randomised and qualitative studies showed recruiter confidence in communicating trial information to patients had improved following training. Such findings are encouraging as previous research has highlighted that many find this challenging [
9,
10]. Despite this, it appeared that recruiters still struggled with the amount, clarity and completeness of information to provide and had difficulty with explaining key RCT concepts such as randomisation. It is therefore perhaps not surprising that some studies reported no differences in patient satisfaction and understanding of trial information, or in host RCT recruitment rates, so more work is needed to address these identified training needs.
Due to the variation in content, delivery and format of training sessions it is difficult to determine precisely what is and is not effective at improving the process of recruitment to trials. Most of the training included presentations, videos, group discussions and role play. Feedback suggested that role play appeared to be particularly useful for recruiters as it was felt to provide a learning environment to apply new strategies and an opportunity to receive constructive feedback. Conversely, previous research shows that didactic based learning may not necessarily be most effective for learning [
35]. In line with this, a randomised controlled study which used only didactic methods of delivery found no difference in trial recruitment rates between intervention and control groups [
14].
The period of time required for training is likely to differ according to the experience of the recruiters and the complexity of the trial [
31]. In the studies identified, the majority of training sessions ran over one or two days. Two studies, which utilised a randomised controlled and an uncontrolled pre-test/post-test design, found limited effects of a one day intervention and suggested that recruiters may have benefited from more training [
23,
27]. There is some suggestion of a dose–response effect based on studies of general communication skills [
35], although minimal training duration for a sustainable effect is unclear [
27]. In the current review, one study which held multiple training sessions over a period of time found that recruitment rates increased from 65 to 81 % [
15]. However, another study had training which lasted only 4 h in total and reported that recruitment increased from 43 to 58 % [
21]. Bearing in mind the limitations of these observational uncontrolled designs, this suggests that content and delivery may be more important than length.
The majority of training workshops provided general information about the key principles of RCTs and also how to discuss trial concepts with patients. The interventions which appeared to have limited impact on outcomes did not address communicating with patients about concepts such as equipoise, randomisation, uncertainty, or exploring patient preferences [
14,
23,
27]. As research has demonstrated that these are key issues that recruiters can find particularly challenging [
4,
9,
10], it seems likely that training interventions which address these topics would be most beneficial to the experience of recruitment for recruiters and patients.
Research has suggested that recruitment training should address both generic and study specific skills [
31]. According to deSalis and colleagues, ‘each RCT has a unique–and uniquely complex–recruitment pathway and its own set of issues that need to be resolved’ ([
36] p.95). It therefore appears important to address issues specific to a particular trial, such as treatment arms, pathways or side effects, so that recruiters can discuss them confidently with patients, in addition to more generic trial recruitment issues. Fletcher et al. reported that the most successful interventions were studies which used qualitative research to identify the key training topics and then develop interventions based on this to improve recruitment [
10].
Most of the studies in this review provided training for a mix of health professionals. Donovan and colleagues suggested that nurses and doctors who recruit to randomised trials experience different issues within their roles [
4,
9]. It was recommended that doctors could benefit from support in relation to assessments of eligibility and equipoise [
4], whereas nurses could benefit from support pertaining to perceived conflicts in their roles as recruiter, patient advocate and clinician, and helping them to be comfortable with approaching all eligible patients [
9]. This suggests that training programmes may need to be targeted to the needs of different health professionals separately.
Methodological considerations
This review was written in accordance with PRISMA guidelines to ensure that it was reported fully and transparently [
37]. To minimise the risk of bias and errors, data extraction and the quality assessment of each study was performed independently by two researchers and disagreements were resolved by reaching a consensus through discussion with the third author if necessary. A wide range of sources were searched to identify training interventions on recruitment to RCTs for health professionals and trial staff recruiting to clinical trials, and a range of study designs were included. However, there is always the possibility that studies may have been missed as the search strategy was heavily reliant on text word searching, thus limiting searches to the terms used by authors in the title and abstract fields of each reference [
38]. As non-English papers were not included due to lack of resources, there may have been studies in other languages that were missed.
The studies reviewed were of varying quality, limiting the conclusions that can be drawn. Two studies excluded participants from the training session as they did not recruit sufficient numbers [
23,
27], although it could be argued that these individuals may have benefited most from training and support. Furthermore, six studies used simulated role play with actor ‘patients’ [
20,
22‐
25,
33], and one study used participating research nurses to act as patients [
28]. Whilst this may represent a more convenient option, it is possible that the simulated patients were not representative of ‘real’ patients. However, a study comparing role plays of trial recruitment discussions with real patients and simulated patients found that the latter were better informed about the purpose of a consultation and provided more specific feedback [
39].
Relatively few studies investigated the effect of training on recruitment rates, making it difficult to draw conclusions as to how training could translate into practice. General communication training has been found to influence communication style in a clinical setting if both competence and self-confidence are improved [
40]. Future research should use a range of measures to accurately understand effects of recruiter training, including the impact on recruitment rates, but ensuring also the capacity to assess levels of informed consent and to capture improvements in the ‘quality’ of the recruitment to trial consultation for both patients and recruiters. The interventions reviewed in this study focused mostly on training for the consultation in which trial recruitment is discussed. There are other issues outside of this consultation where training in trial recruitment might be beneficial, such as the assessment of eligibility, screening logs and recruitment pathways [
9], which would benefit from further research.