What influenced people with chronic breathlessness and advanced disease to take part and remain in a drug trial? A qualitative study.

Recruitment and retention in clinical trials remains an important challenge, particularly in the context of advanced disease. It is important to understand what affects retention to improve trial quality, minimise attrition and reduce missing data. We conducted a qualitative study embedded within a randomised feasibility trial and explored what influenced people to take part and remain in the trial. Qualitative study embedded within a double blind randomised trial (BETTER-B(Feasibility): BETter TreatmEnts for Refractory Breathlessness) designed using a person-centred approach. Participants with cancer, Chronic Obstructive Pulmonary Disease (COPD), Interstitial Lung Disease (ILD), or Chronic Heart Failure (CHF), with a Modified Medical Research Council Dyspnoea Scale grade 3/4 were recruited from three UK sites. A convenience subsample completed qualitative interviews after the trial. Interviews were analysed using thematic analysis. Results were considered in relation to the core elements of person-centred care and our model of the person-centred trial. It was important that participants felt actively involved and as though they were contributing to the trial. Knowing that the trial may benefit patients in the future, as well as providing an opportunity for individuals to give back were motivating factors for completing the trial. Several participants described how they found the trial process rewarding on an individual level.


success (3).
There is an increasing literature around person centeredness in trials with growing evidence that involving patients at the research design stage can direct recruitment and retention strategies and improve enrolment (23)(24)(25). Patient and public involvement (PPI) is one method of applying person centeredness to trials, and can help to ensure that the research process is participant friendly and trial information is relevant, readable and understandable (26,27). While studies evaluating personcentred care in trials remain limited, Chhatre et al applied a conceptual model of patient-centred recruitment and retention to a RCT of patients with newly diagnosed prostate cancer (28). The study identified strategies which may aid recruitment and retention, although acknowledged limitations due to time and resource constraints, with attrition of 26% at one of three sites (28).
As more people approach the end of their lives with chronic and complex conditions, the need for robust research and evidence has never been greater. However, clinical trials in palliative care remain sparse often limited by poor funding and methodological weaknesses (2,21,22). It is therefore important to understand what affects retention so that we can minimise attrition and ensure high quality clinical trials of palliative care interventions in the future. We conducted a qualitative study embedded within a randomised feasibility designed using a person-centred approach. The study aimed to explore what influenced participants to take part and remain in the trial.

Design
We conducted a qualitative study embedded within a randomised trial of mirtazapine for chronic breathlessness (BETTER-B(Feasibility): BETter TreatmEnts for Refractory Breathlessness). The trial design aimed to optimise recruitment and retention through the use of a person-centred approach, which has been shown to enable engagement and improve patient outcomes in advanced disease (29)(30)(31).
Based on the core concepts of person-centred care identified by Kitson (32) we developed the model of a person-centred trial (figure 1). Our design aimed to put the patient at the centre of the trial, and minimise study burden, therefore enabling participants to be actively involved and able to participate. The design focused on developing a genuine relationship between the researcher and participant, with emphasis on continuity. Burden from the trial was minimised by offering home visits and helping participants to complete trial related questionnaires to ensure a supportive system. PPI contributed to all stages of the trial, from design to analysis with representatives on the Trial Management Group (TMG) and the Trial Steering Committee (TSC). Trial burden was highlighted as important, and changes were made to the patient information sheet to ensure a clearer explanation of trial processes including the concept of randomisation.

Study participants and sampling
Those eligible for the feasibility trial were adults with cancer, Chronic Obstructive Pulmonary Disease (COPD), Interstitial Lung Disease (ILD), or Chronic Heart Failure (CHF), with a Modified Medical Research Council (mMRC) Dyspnoea Scale grade 3 ("I stop for breath after walking about 100 yards or after a few minutes on the level") or 4 ("I am too breathless to leave the house" or "I am breathless when dressing"), with no current diagnosis of severe depression, not currently prescribed an antidepressant medication. For full eligibility criteria see appendix one. A sampling frame was agreed which included characteristics considered to be important including; gender, diagnosis, trial completion/ non completion, and age (<65 years / >65 years). However, due to the limited pool of participants we decided to take a pragmatic approach and used convenience sampling, offering each trial participant the opportunity to participate in a qualitative interview. Participants were approached by telephone or in-person to arrange an interview. All participants provided written informed consent prior to their interview.

Trial schedule
Patients and carers were approached by their usual clinician and provided with some initial information about the trial. If they were in agreement they were then contacted by a researcher who was able to provide more detailed information including; the rationale for doing the study, the trial design, and what it would mean if they agreed to take part in terms of the intervention and study assessments. All members of the research team had training and experience of working with people living with advanced disease. Patients were given a minimum of 24 hours to consider the trial and discuss with friends and family. Participants then provided written informed consent, and a more detailed eligibility assessment followed. After randomisation the medication was provided along with a diary to complete, details of who to contact with any questions or concerns, and emergency contact details for out of hours. Participants received 28 days of trial treatment (either oral mirtazapine or placebo capsules). They were assessed face to face on day 0, day 14 and day 28, and via telephone on day 7, day 21, and day 35. was Assessments were organised at a time which was convenient for the participant with some flexibility (+/-1 day). Participants were offered to be visited at home and assistance was provided with completing the trial-based questionnaires. Continuity of the researcher was prioritised where possible.

Data collection
Qualitative interviews were conducted at the end of the trial. Interviews were conducted in a place of the participants choosing. This was usually their own home, but some interviews were conducted in hospital. The topic guide (appendix two) was developed using existing literature and refined following feedback from PPI representatives and the Trial Management Group (1-6). The interview schedule included questions about experience of recruitment to the trial, why they had decided to take part, expectations of the trial, and experience of trial processes (taking the trial medication, experience of trial visits, and experience of completing the trial questionnaires). Interviews were digitally audio recorded and transcribed verbatim. A distress protocol was used to minimise the risk of potential harm. All interviews were conducted by one female researcher (NL) with a medical background, who had completed training in in-depth interviewing. Interviews took place between January 2017 and December 2017.

Analysis
The qualitative interviews were analysed using Braun and Clarke's framework for thematic analysis [29] using NVIVO version 10 (QSR International (UK) Ltd.).
Transcripts were read and re-read and then coded inductively for themes relating to; reasons to participate in the trial, reasons not to participate in the trial, reasons to remain in the trial and reasons to discontinue the trial. Results were considered in relation to the core elements of person-centred care and our model of the personcentred trial (figure 1) (32). Three transcripts were double coded by another researcher (SE) who produced their own coding frame. Areas of agreement and disagreement were then discussed until consensus was achieved.

Results
The were male. 20 had completed the trial, whilst 2 withdrew due to reported adverse effects of the trial medication. The mean interview duration was 33 minutes (range . Despite the use of convenience sampling, variation was achieved, and we interviewed participants from all 3 research sites, all disease groups, both age and gender categories, with 2 non-completers also participating in interviews. No trial decliners agreed to complete a qualitative interview. The relationship between patient and professional, potential for benefit, trial processes and the intervention all influenced the decision to participate in the trial.
The relationship and continuity with the research team, perceived benefit, and aspects relating to trial processes and the intervention influenced the decision to remain in the trial.
What influenced people to take part in the trial

Approach
The way in which potential participants were approached was important when considering whether to take part in the trial. What influenced people to remain in the trial Importance of the relationship and continuity of care The importance of the relationship between the participant and the researcher was identified across all interviews and was substantial when considering the reasons why people remained in the trial. Attempts by the researcher to minimise burden and ensure a calm environment were recognised and appreciated by participants.
The personal attributes of the researcher were also central to remaining in the trial.
Participants described the importance of effective communication, being treated with respect, and not feeling rushed during trial visits.

Discussion
This study identifies important considerations which may influence recruitment and retention in clinical trials. We found that the relationship between patient and professional, potential for benefit, trial processes and the intervention all influenced the decision to participate in the trial. The relationship with the research team and continuity, perceived benefit, and aspects relating to trial processes and the intervention influenced the decision to remain in the trial. In this trial recruitment targets were met and attrition levels were low, suggesting that a person-centred approach can support successful recruitment and retention.
What influences potential participants to take part in a clinical trial (or not) is recognised to be a complex multifactorial process (34)(35)(36)(37)(38)(39). In this study we found that the initial approach by both clinician and researcher was key in developing a genuine relationship built on trust, a concept which has been identified as important when deciding whether to participate in a clinical trial (34,35,40). In this study participants described the potential benefit to self and others as a motivating factor, comparable to the findings of previous qualitative research conducted in the palliative care setting (37). While concerns about randomisation and the potential for side effects can be deterrents to participating in a clinical trial (36), this was not a major influencing factor for the participants we interviewed. The trial design was important and attempts to minimise burden were viewed favourably by participants.
This is an important consideration as missing data in trials has been shown to increase with the number of questionnaires/ tests (17).
In this study the relationship between the patient and professional was crucial, and improve retention within trials we need to ensure that funding allows adequate resource allocation to spend time supporting participants with trial processes (28).
While our study suggests a benefit to having the same researchers working across all stages of a trial, current funding models in the UK focus specifically on recruitment and not on retention and therefore the funding for follow-up often needs to be pooled from other budgets (42). In practice, continuity of research staff is not a commonly reported outcome and so it is difficult to know the impact of this across different specialties, and for larger trials. To ensure that the same researchers are able to work across trials funding models need to be revised to rebalance of emphasis of recruitment and retention (43).
It is important to acknowledge that the researchers in our trial all had training and experience working with people living with advanced disease. Participants valued the personal attributes of the professional, a quality which has been identified as critical in person-centred care (32). Characteristics which have previously been identified as important for Palliative Care Professionals include: interpersonal skills, a willingness to listen, being someone the patient feels able to talk to, demonstrating an interest in knowing patients' as people, and recognising that patients may need to feel in control (32,44). Therefore the attributes of professionals delivering PCC and Palliative Care are closely aligned (45). Increased opportunities for the training of research staff has been highlighted as important if we are to improve retention in clinical trials in the future (20,43).

So how can PCC be applied to clinical trials in practice?
To improve retention clinical trials need to be individualised, with processes in line with individual capabilities, and considered alongside the other challenges in life (19). We propose that implementing a person-centred approach can support  The study was limited by one female researcher (NL) with a medical background conducting all of the interviews. In addition some of the interviewees had met this researcher during the feasibility trial, therefore increasing the risk of social desirability bias, and participants may have been reluctant to offer criticisms about the trial intervention and/ or processes. The time period between the trial ending and a qualitative interview being conducted varied, and this may have increased the risk of recall bias in the qualitative interviews. Some interviews were conducted with a carer present which may have impacted on the answers given. Although we achieved a varied sample of trial participants, we only interviewed two participants who did not complete the trial, and were not able to interview anyone who declined to participate in the trial.
The trial itself was of short duration with an arguably simple intervention and may therefore be perceived as easier in terms of recruitment and retention when compared to a longer trial or one of a complex intervention. However, challenges with recruitment (in part due to eligibility) and high attrition levels have previously been demonstrated in short duration drug trials conducted in people with advanced disease (5). 16

Availability of data and material
Requests for data should be made to the corresponding author.

Competing interests
The authors declare that they have no competing interests. The