Qualitative analysis of feasibility of recruitment and retention in a planned randomised controlled trial of a psychosocial cancer intervention within the NHS

Abstract

Background

The randomised control trial (RCT) is the most rigorous method of evaluating interventions. Recruitment is often slower and more challenging than expected. The aim of the current paper is to understand the feasibility of recruitment within the NHS and the barriers and motivators to recruitment from the perspective of patients and healthcare professionals (HCPs).

Methods

NHS HCPs were surveyed to establish their willingness to participate. Twenty HCPs were interviewed to establish barriers and motivators to recruitment. Eleven patients were interviewed to understand their willingness to participate. Interviews were analysed using thematic analysis.

Results

HCP interviews identified key barriers to recruitment: practical barriers included workload and time; clinical barriers included terminology and concern that the trial implied criticism of their current practice; and patient barriers included gender and cultural factors. Motivators to recruitment included: regular communication between research and clinical teams; feedback on findings; and patient and individual benefits for clinicians. Patient interviews suggested that participation in a trial of a psychosocial intervention would strengthen existing coping skills and develop mechanisms for those who were struggling.

Conclusions

Survey results demonstrated that recruitment to an RCT of a psychosocial intervention for people living with and beyond cancer would be feasible within the NHS if specific barriers are addressed. From a clinician point of view, barriers should be addressed to improve recruitment, particularly training and education of clinicians and clear communication. From a patient perspective, interventions and RCT should be tailored to target those not routinely represented in RCTs.

Background

In randomised control trials (RCTs), participants are randomly assigned to two or more clinical interventions to reduce bias. The RCT is seen as the most scientifically rigorous method of evaluating the effectiveness of interventions and is widely regarded as the gold standard in experimental research design [1]. RCTs are important to establish the evidence base for clinical practice in the NHS [2]. Recruitment targets are set to avoid study findings being statistically underpowered, which may lead to a Type I or Type II error with researchers erroneously detecting an effect where none exists or failing to detect an effect that is, in fact, present. Such errors can result in interventions being provided that lack benefit or, conversely, patients being denied interventions that could significantly improve their health and quality of life (QoL) [3].

Recruitment to trials is frequently much slower and more challenging than expected, with up to 66% of trials not achieving their target sample size [4] and 53% of trials being required to extend the recruitment period, which increases costs and delays the application of research findings to clinical practice [5]. Reasons for unmet targets and delays included issues with local clinical arrangements, with some trials reporting pre-identified centres not participating as planned [5]. Losses of eligible participants were attributable to refusals by patients in 25% of trials and refusals by physicians to recruit in 29% of trials [4]. Achieving set recruitment targets is of paramount importance for the success of a RCT. An understanding of the factors that influence recruitment is therefore vital when planning a RCT.

When considering what may increase motivation to recruit, a monetary incentive is considered less important than factors such as interest in the research, patient benefit and good communication [6]. A systematic review of effective mechanisms to increase the recruitment activity of clinicians involved in RCTs suggested the use of qualitative research to identify and overcome barriers to recruitment, clinicians having a reduction in expected clinical workload when working on RCTs, and training being provided [7]. Good communication with the research team, a clear beneficial effect on patients and clinical practice, an understanding of the purpose of the research, and the individual benefit for clinicians aided recruitment, while HCPs’ perception that barriers are patient-related and out of their control was a barrier [7].

The ProtecT (Prostate testing for cancer and Treatment) RCT used qualitative methods to analyse recruitment discussions and data from interviews with potential participants. Findings suggested that within recruitment to the trial, recruiters had difficulty discussing equipoise and presenting treatments equally, often inadvertently using terminology difficult for patients to interpret [8]. The ProtecT trial successfully optimised recruiters’ communication strategies to increase recruitment rates and demonstrated the benefits of qualitative methods being included either at the feasibility stage of a RCT, or fully integrated into the RCT [9]. The Quartet (Qualitative research to improve recruitment to randomised controlled trials) study collaborated with trial management groups to improve recruitment in trials experiencing poor recruitment by focusing on the way that trials were presented to potential participants [9].

Recruitment to trials with patients with mental health problems face similar issues to trials in general, but there are additional considerations which are specific to this population [10]. The five main reasons for recruitment difficulties in such trials include: misconceptions about trials; a perception of lack of equipoise; HCPs misunderstanding the trial arms and eligibility; and staff paternalism regarding their role as a ‘carer’ [10]. These findings have relevance for planning a RCT of a psychosocial intervention for people living with and beyond cancer.

The aim of this research was therefore to establish the feasibility of recruiting participants to a RCT of a psychosocial intervention designed to improve the QoL of those living with and beyond cancer. In particular, it examined:

• 1. the willingness of clinics across the NHS to participate in such a RCT;

• 2. barriers and motivators to recruitment, with a particular focus on cultural barriers;

• 3. patients’ perceptions of potential recruitment to such a trial.

Methods

Survey of HCPs

The approval for the survey as a service evaluation was gained via Bart’s Health NHS Trust (Reg. No. 6131). The four-item scale used was drawn from a larger 22-item standardised survey [11]. The questionnaire items can be found in Table 1. Questions were asked to all participants regardless of their response to item 1. The short survey was sent to a range of professional bodies to try to capture all relevant HCPs involved in the management of cancer patients working in a range of clinical settings. These included the Association of Cancer Physicians, the UK Oncology Nurses Society, the Royal College of Radiologists, the UK Breast Intergroup, British Psychological Society and the Association of Coloproctology of Great Britain and Ireland. The professional bodies distributed the survey to their members via an email link that was open from August to December 2015. The participants completed the survey through the online tool SurveyMonkey (www.surveymonkey.com). Descriptive analysis was used to create proportions with the survey data.

Table 1 Survey response to participating in a randomised control trial

Interviews with HCPs

The approval for the interviews with HCPs as a service evaluation was gained via Barts Health NHS Trust (Reg. No. 7070). HCPs who were potential recruiters working within Bart’s Health NHS Trust and University College Hospital were first approached by email to participate in a 10–15-min semi-structured in-person or telephone interview. Consent was sought to record the interview. HCPs were asked about the feasibility of recruiting patients who had finished active treatment; their confidence in (i) explaining a trial, (ii) identifying those with low QoL, (iii) using screening measures; and barriers and motivators to recruiting to a trial within clinics. Interviews were anonymised and transcribed verbatim. Transcripts were analysed using thematic analysis. MD immersed herself in the data, highlighting key sections of text to develop an initial list of codes using the qualitative data analysis software NVivo. The list was then debated with GE to sort into key themes. A table of the full list of themes and codes can be found in the Appendix.

Interviews with patients

Ethical approval was obtained for this part of the study (Health Research Authority NRES Committee East of England, REC reference: 15/EE/0198 IRAS project ID 168246). Patients who had completed active treatment for cancer were approached in Bart’s Health NHS Trust outpatient clinics and asked whether they would participate in a single semi-structured interview to explore their potential willingness to participate in a trial of a psychosocial intervention aimed at improving QoL. Patients were recruited consecutively but with the aim of including people diagnosed with different types of cancer. To be eligible to participate, patients must have finished active treatment and be aged > 18 years. Patients were excluded if they were not able to speak English. The interviews lasted around 40–90 min and were recorded and transcribed. Analysis proceeded as for the interviews with HCPs.

Results

Survey of HCPs

There were a total of 278 responses, of which 166 had completed all four questions resulting in a 60% completion rate. One hundred and eight of the 154 (70%) acute NHS trusts were represented in the survey. We received responses from the Association of Cancer Physicians, the UK Oncology Nurses Society, the Royal College of Radiologists and the UK Breast Intergroup. No response was obtained from The British Psychological Society and the Association of Coloproctology of Great Britain and Ireland. We were unable to calculate an absolute response rate as the respective professional bodies acted as intermediaries in the process in order to protect the privacy of their members and handled dissemination of the survey link. The survey responses to questions asking about participating in a trial can be found in Table 1.

Interviews with HCPs

In total, 20 HCPs participated in the interview: six Clinical Nurse Specialists and 14 Consultants. HCPs were working in a variety of areas of cancer: nine colorectal, six breast, three head and neck, one haematology and one prostate cancer clinician. Thematic analysis revealed four key themes: practical barriers; clinical barriers; patient barriers; and motivators to improve recruitment.

Practical barriers

HCPs identified barriers to recruitment that were out of their control, such as under-resourced staff, limited time within clinic, workload concerns in a busy clinic and the recruitment to other ongoing trials.

One of the things we are facing at the minute are just how busy clinics are and there’s lots of the new technology coming in like new electronic systems, which means clinic time is squeezed even more. So, what I am certainly finding is I am spending less time with patients and more time at a computer screen, and I am sure that would have a knock-on effect potentially with trials.

Clinical skills barriers

Clinical skills barriers included factors relating to the clinicians perceived ability to promote a RCT. The terminology used and equipoise of clinicians was raised as an issue to recruitment:

I think the wording would probably need to be changed – I think those words are all very understandable to us as clinicians and scientists but if I said to my average patient, how is your quality of life, they would not necessary know what that means so I think there would need to be some description.

Other clinical barriers that were highlighted included the risk of clinicians viewing a RCT, especially in psychology, as criticism of their current practices:

The main barrier was that because we think we have quite a good set up already one of the nurses was a bit sort of suspicious of the study that it was here to sort of investigate what we do.

The final barrier from clinicians was remembering to mention a RCT to patients within a busy clinic, especially with multiple RCTs to recruit from:

The difficult bit is remembering and identifying the people who and remember to ask.

Patient barriers

It was a general consensus among clinicians that there would be a certain group of patients that would simply not be interested in a RCT or psychological support:

There are some patients you will find will want to access everything and there are some patients that will just want to go away and forget it all.

However, some HCPs identified specific potential barriers relating to the patients included demographic factors such as gender and cultural factors:

Now I do not want to be sexist but it’s just how I find it, with head and neck cancer it’s very much still such a male-dominated cancer but often they do not like these sorts of things that touch on the psychological aspects, because they are a man and they have got to be strong.

It’s not language, it’s culture because people have a lot of people of non-European origin, non-western European origin, intrinsically they do clinical studies, they enjoy it. So, our recruitment is always lower than it should be, just because of that cosmopolitan effect.

Furthermore, HCPs suggested that RCTs should focus on educational requirements and mental health problems that may prevent a patient from completing extensive outcome measures:

Quite a lot of our patients have mental health problems, which might mean they do not want to engage in this sort of thing.

A significant amount of our patients are from a low socio-economic background and do not always have great educational skills. But in an interview, they would be fine but I know sometimes they struggle with the long questionnaires.

Motivators to improve recruitment

After identifying the barriers to recruitment, HCPs highlighted factors that may improve recruitment. These included regular contact with the research team and/or research staff within clinics:

We certainly know the best recruiting studies are the ones that have either the research assistant or whoever actually in clinic reminding clinicians of the study, making sure we are asking the questions and giving the patients information.

The HCPs were more specific about the contact with the research team and felt updates on patient benefits would improve motivation:

Feedback from patients about how it helped them fed to the recruiting clinicians so we can see how it’s being helpful.

HCPs felt that communication was also key throughout the whole clinical team surrounding the patients. The clinicians should be aware of the RCT and actively promote it within clinics:

Making sure all members of the team whether it be me, a CNS, a ward nurse or a radiographer are saying the same thing. Because it’s almost that attrition and you say it enough times to a patient and they sort of give in. So I think that’s an important thing to make sure we are not saying different things and it’s not just one of us continuously saying it but it’s the team.

Important factors to promote HCPs recruiting patients included both personal benefit such as ‘being acknowledged in papers’ and wider patient benefits:

Just the fact that trials help in the future, so it’s just the fact that every trial is successful and will help many more patients.

HCPs felt that more in depth preparation for recruitment to a RCT trial may help, such as ‘more written information, and probably short teaching sessions’.

Interviews with patients

Eleven patients completed the interview: four head and neck cancer patients; four haematology patients; and three breast cancer patients. Demographic information can be found in Table 2.

Table 2 Patient demographics

Patients’ opinion was divided on whether they would personally participate in the intervention, but they believed it should be provided to those that need the support. Thematic analysis resulted in four main themes for those that would take up the intervention—positives of the intervention, timing of the intervention—and for those that would not—current coping mechanisms and positives of the intervention for others.

The patients that said they would personally be interested in taking part felt that the intervention would improve aspects such as fatigue, returning to work and depression:

If people are feeling really tired, I think for them that would be good because it’s something positive because I think it might be quite easy to look at the negatives and hang on to them.

I wanted to get back to work, but I really did not want to face anybody, because I am on reception so I did not want the constant, ‘How are you feeling? Duh-duh-duh-duh-duh.’ It’s like, yes, I needed to just get back to normal. Yes, it’s the initial first going back, the first week was hard.

If they have a couple of follow-up things then it might be you won’t… you will be depressed but you won’t be depressed as intense[ly].

These patients further suggested that the timing of the intervention being offered was important:

If it was offered to me when first discharged I may have considered it at the time, but as that wasn’t offered at the time then….

The feeling of elation is unbelievable, to be told that you are free of cancer… I wish I had not have had to have that experience but to have had it you cannot describe that feeling and that may override everything else that’s going on up here at that moment in time. So it might be that this needs to settle down a bit, so a month or so later say ‘Look, just, if you feel that you want it, here’s our contact number, get in touch.’ These people might cope with it perfectly well anyway.

The patients that said that they would not be personally interested tended to report current coping mechanisms such as acceptance and exercise:

I probably would not but only because of the fact I have someone that I talk to and I think I am working on that generally, accepting that you cannot control everything in life.

I think I am fine because I think I have accepted I cannot chew salad with people, cannot do that.

I think I am quite lucky that I feel into gym and that is my let up, but I think for other people, yeah, if people are feeling really tired, I think for them that would be good because it’s something positive because I think it might be quite easy to look at the negatives and hang on to them.

However, they felt that the intervention should be available to those that may need it:

In my particular case I don’t think so, but in general I think it would be a good idea. But those who don’t have that therapist and counselling would be very good, but those are… because I’m very outward, I do go out and talk to people and make myself busy and everything, but some people are not like that. So for those kinds of people it would be nice for people go and have a chat because instead of being a lonely person they can go and chat with somebody and tell their problems and somebody is there to listen to them and you get advice on that. I think that would be very helpful for patients. Not for me.

I think if you, I think it should be offered. Certainly to say, give people the opportunity, explain to them that they might not want it straight away, but say, you know, leave the door open for them and say to them ‘Look, you might not think about it now, you’ve probably got so many other things on your mind’.

That’s not me. That is not me. Sometimes group meetings are not me because I want positivity, I do not want negativity.

Discussion

The current findings combined survey data from NHS cancer clinics across the UK, HCP interviews and patient interviews regarding the feasibility of recruitment to a RCT of a psychosocial intervention to improve QoL of those living beyond cancer. The survey found that conducting a RCT within the NHS across the UK may be feasible, with most services agreeing to be involved, to allow staff training and the running of an intervention within their service.

The interviews with the HCPs were generally positive, with all participants welcoming RCTs within clinics. The results replicated previous findings of the practical barriers to recruitment including workload concerns, time and under-resourced staff [7]. Similarly, the current findings support previous research demonstrating barriers to recruitment that involved clinical skills such as difficulty discussing equipoise and the use of terminology that is difficult for patients [8]. However, the current findings advanced previous research by suggesting that HCPs were often protective of their current service provision and sometimes saw RCTs as a criticism of their practice. This is an interesting finding as it suggests that to improve recruitment and clinical arrangements an important focus is to acknowledge current practice and explain the purpose of RCTs in providing an evidence base for current practice in specific clinics to provide a case for providing equal care across the UK.

A further important finding was that patient barriers included demographic factors, suggesting that future RCTs should do more to incorporate those that are perhaps underrepresented in cancer QoL research, such as men and those from ethnic minorities. Furthermore, language, educational levels and mental health concerns may be considered a barrier when faced with extensive written outcome measures – which should be addressed in RCT designs to ensure generalisability of findings. Researchers both within and outside of the NHS would benefit from keeping in mind patient barriers when designing their research to improve generalisability of findings and benefit patients that are often underrepresented in research and psychological care.

Participants identified factors that may increase motivation that are simple and easy to incorporate into RCT research, such as regular contact and communication from the research team and throughout the clinical team, regular feedback on the findings and success of the trial, and in line with previous research [7] the importance of the effect of the trial on patients and the individual benefit for clinicians. This finding is high in external validity as all RCTs would benefit from introducing these ideas to increase motivation and recruitment and retention in trials.

The findings from the patient interviews further suggest that a RCT would be welcomed, for those that would benefit – suggesting interventions should be tailored to a patient’s individual needs. The findings demonstrate that those that cope well already use coping mechanisms that would be developed throughout a psychosocial intervention; and those that do not would benefit from support in these key areas of work, exercise and acceptance.

Limitations

All HCPs who participated in the interviews were working within clinics which were aware of the planned trial and within clinics that provided strong psychological support following active treatment – which is not always the case [11]. Therefore, the current findings based on HCPs already within strong research trusts may overestimate the positive response to recruitment of centres throughout the NHS more generally. Similarly, the patients interviewed were those that volunteered to take part in research and reported a positive experience of care, which may provide a biased sample of patients that perhaps represent those less likely to struggle post treatment and those less likely to be recruited.

Future research

The SURECAN team have successfully received NIHR funding to develop the intervention further and then conduct a multicentre RCT of a tailored psychological intervention for people who have undergone successful treatment for cancer, but who have a low QoL. The current findings suggested that a RCT was feasible within the NHS and informed the research team how best to improve recruitment and retention within the trial. The team will use the findings of the current study in the new research programme, in particular in acknowledging current practice, providing training and information to HCPs about the purpose of the RCT and will be feeding back findings to the clinical team. The success of these strategies will be revealed in our recruitment rates.

Conclusions

Survey results demonstrated that recruitment to a RCT of a psychosocial intervention for people living with and beyond cancer would be feasible within the NHS if specific barriers are addressed. From a clinician point of view, barriers should be addressed to improve recruitment, particularly training and education of clinicians and clear communication. From a patient perspective, interventions and RCTs should be tailored to target those not routinely represented in RCTs.

Appendix

Table 3 Initial list of codes and themes from the HCP interviews