Understanding recruitment: recruiter perspectives
All recruiters readily described the barriers that they were convinced hindered recruitment. The clearest and most common barriers were organisational difficulties, fewer eligible patients than anticipated, and strong preferences expressed by patients for particular treatments or against being randomised:
T2–D1: “The actual difficulty was in the logistics and it’s usually because the nurse can’t attend the clinic or the nurse can’t attend the meeting or the multidisciplinary team doesn’t actually go through the list of patients that are eligible at that time.”
T1–D2: “So, so there’s a local nuance there that we do get a significant number of patients who will plump for [intervention 1] if they live in [district a]. They don’t want to travel. They live next to the [intervention 1] centre which is I guess, what, ten miles difference or something like that, eight, ten miles? And conversely patients will plump for [intervention 2] for the fact that it’s going to happen in the hospital down the road.”
T1–N2–FG: “All the patients that have said ‘no’ , have said ‘no’ for personal reasons, not because of anything that we could change. One of them didn’t want a computer to decide what was happening about his treatment, he had to have control over his own destiny, and he just didn’t like the idea of being randomised.”
These barriers were consistently reported and recruiters often seemed overwhelmed by the difficulties, leading to low morale and very poor recruitment levels:
T7–D5: “I’m a genuine believer in the principle that [intervention 1] might be as good as [intervention 2] and therefore I’m a strong supporter of the basis of this trial … But it’s been a nightmare, frankly … I love this trial. I think it’s beautifully designed, I think it is extremely clever, which fits [name of CI] who is extremely clever … But, I honestly have never had as many problems with recruiting to a trial in all my born days.”
As would be expected, the leaders of the RCTs tried hard to find solutions to these difficulties, by changing organisational arrangements of clinics, or making sure that all potentially eligible patients could be included:
T1–N2–FG: “The patients get the diagnosis and they see the surgeon and the oncologist on the same day. They mention [RCT] to the patients and then if they want to go in it, they refer the patient to me then they come in to me sometime in the next week or so … They’ve had both treatments explained to them a couple of days ago by the time they get to me, so it’s quite good.”
T7–D4: “[We are going to] make certain that all potentially eligible patients have a discussion about the trial and know that it’s available for them … have a research nurse attending the MDT [multi-disciplinary team meeting, where treatments are decided]. In terms of the organisational things, it’s challenging, but it’s not impossible.”
Some organisational changes slightly improved recruitment in some RCTs, but it remained problematic. Recruiters had been encouraged, in the interviews, to recount their actual experiences of recruitment in detail. As they did so, it became apparent that there were issues associated with their role as a recruiter that they found difficult, particularly in relation to their other roles as practising doctors or nurses. As they described their roles and the challenges and conflicts within them, they began to express emotion and discomfort, sometimes defensiveness, and a number of previously hidden challenges to recruitment emerged; these are detailed below.
1. Doctors’ roles
Doctors were involved in recruitment in all these RCTs, most in assessments of eligibility and sometimes undertaking the process of informed consent on their own or with the assistance of nurses or other researchers. When asked directly, these doctors described themselves as scientists or practising clinicians, or some combination of the two. Some were clear that these roles were complementary, and that it was really important to them to be able to combine the often urgent need to collect robust evidence through RCTs with a clinical role.
T3–D1: “I see my role as a researcher and as a clinician as engaging my patients and the community out there in a long-term relationship, not just for today, not just for this child, not just for this illness, but a long-term relationship for us together, to improve clinical practice and reduce uncertainty. And that’s about saying to people ‘okay that’s fine this time you don’t want to do it perhaps another time you will, perhaps another time the condition will be the right one for you or the timing will be better’.”
T6–D11: “I’m convinced fully about what I’m doing … Unless we take this attitude of impartiality we will not find out and we will continue to over-treat patients or to under-treat patients.”
When asked directly, the doctors were positive about their involvement in the RCT and their commitment to it, but then went on to describe a number of underlying conflicts with clinical practice that they experienced to varying degrees. These related particularly to the balance between carrying out research and ensuring that they safeguarded the best interests of the individual patient, and their willingness or ability to suspend their clinical judgments and recruit an individual patient to the RCT. In talking about the RCT, it was easy for them to agree that there was a lack of evidence about the most suitable treatment in general, but many admitted that when faced with a particular patient, or those with specific clinical characteristics, they became more uncomfortable:
T7–D2: “There’s always a slight conflict between the patient sitting in front of you and their wishes but also taking part in the trial and wanting to support the trial, wanting to increase recruitment. Yes and it’s an element of frustration when someone says, ‘no’. So there can be a conflict between your management responsibilities for an individual compared to your wish to make something else a success.”
T6–D2–F: “Sometimes you get the feeling that we are counselling them to the extent that you are doing a disservice to them. Are you helping or hindering their management when you yourself know there are so many uncertainties and at the moment, or as far as you know, there are controversies, but all treatments can be similar with minute variations … So, it’s inevitable at that times you feel uncomfortable.”
Some went on to describe particular patients or groups of patients who were technically eligible for the RCT, but who they felt were not really suitable and were either uncomfortable about recruiting, or definitely would not recruit:
T6–D12: “I wouldn’t be involved with the trial because it wouldn’t be ethical if I had a strong view that this wasn’t right … But, there are a subset of younger men with more aggressive tumours where I find that really quite difficult.”
T7–D6: “My bias is that in a younger person, [intervention 1] probably is a better treatment … Rather than putting them into trial, I think what I’d like to do is give them [intervention 1] up-front.”
While doctors continued to identify recruitment barriers as being organisational difficulties, the lack of patients fitting the RCT eligibility criteria, and patients expressing strong treatment preferences or rejecting randomisation, a careful analysis of the interview data indicated that these obstacles were not quite as straight-forward as they seemed. In the interviews, doctors readily expressed considerable discomfort and difficulty in relation to their roles as clinicians and recruiters, particularly around the eligibility of particular patients or groups, and their own treatment preferences, clinical instincts, and ability to feel uncertain about the most appropriate treatment. Many of these barriers seemed to originate from the doctors’ views or at least to be reinforced by their views. Their lack of equipoise for groups or individual patients contributed to the lack of eligible participants as they did not recruit those outside their comfort zone. If patients expressed the same preference for a treatment as the clinician, it was easy to allow them to have that treatment rather than offer recruitment to the RCT:
T6–D9: “If you believe [intervention 1] is the best for them, you should say so … There’s somebody I saw recently … I actually took him off the trial. I said, ‘Well you’d be better off [with intervention 1 rather than intervention 2]’. So it’s a small number of people, but I hope I would offer them what I thought was the best treatment and take them out of the trial.”
T6–D13: “I believe that patients come through the door with preconceived ideas and in a lot of cases with their mind made up, and they’re checking that prejudice out, if you like, with the doctor when they come in.”
T2–D2: “I think there is a problem with equipoise … There are patients whom [specialists] regard as really high risk and they would be uncomfortable about putting them in, so they don’t actually.”
Doctors were thus able to express commitment to the RCT because of the need to gather evidence but also considerable discomfort about particular aspects of recruitment, only rarely linking their views to recruitment problems.
2. Nurses’ roles
Nurses were employed as the primary recruiters in T3 and in the main recruitment phase of T6, and were closely involved in recruitment in the other RCTs as network nurses (T1, T2, T7) or care co-ordinators (T5). They discussed their roles in great detail, identifying three major aspects: caring clinical nurse, patient advocate, and recruiter/scientist. All the nurses perceived potential conflicts between these roles, particularly on the boundary between recruiter and caring nurse. Only a very small number thought their caring and recruiting roles could be complementary, but even they struggled with how to accommodate advocacy:
T7–N2: “I hope that one complements the other really, rather than be conflicting. If it’s a study that I’m involved in then I think my background as a nurse would be – I would be there in a supportive sort of role, and also to be able to put something across to a patient where I felt that they didn’t feel intimidated, that they didn’t feel that they had to go into something, that it was, you know, really their choice to do that. I mean if you’re thinking about being an advocate, I don’t think that as a research nurse I can be.”
The most common and powerful view that emerged was how concerned nurses were about the conflicts they perceived between caring for patients and recruiting them to RCTs. The boundary between research and clinical activity was a particular source of anxiety, and for those relatively new to research, adapting to a combined role was seen to be very challenging because of the differences they perceived in their status, employment, and the purpose of their activity:
T3–N1: “I had to learn to adjust to quite a different role where actually I very much had to be invited to go into their family rather than being the person who’s going to go there and help them because although there is an element of that to my job, it’s not my main job. My main job is the research and to recruit. We’re not in uniform, because our first role is not as one of a nurse, it’s a researcher and I’m not employed as a nurse, I’m employed as a research associate by [University] … How will I know when I should be a nurse and when I should be a researcher? It’s really hard and there’ve been a couple of times when I’ve got a bit confused about that and thankfully it hasn’t led to anything drastic.”
T2–N3: “I haven’t had a real conflict yet, but I can imagine that sometimes it would be … You know, how many times do you pursue them, how many times do you say? And there is the bottom line that they want us to recruit people to these trials, don’t they? Obviously, in order for them to be a success, so you do feel, yes I do sometimes feel like a salesman.”
When asked directly, most of the nurses stated that they enjoyed their roles as recruiters and their involvement in the RCTs, but as they discussed these issues in more detail in the interviews, it was clear that many of them wanted to define their position as caring clinical nurses, and they defended that role tenaciously. The primacy of their clinical/caring role was often asserted strongly, sometimes defensively, and many ensured their clinical role took precedence over the RCT:
T6–N3–G: “I just tell them that they’re in control whatever they decide to do. If they decide to randomise, fine; if they reject the randomised treatment, fine; if they decide that they want the randomised treatment – whatever they choose to do as long as it’s an informed choice and I am satisfied it’s an informed choice, it’s entirely up to them. They’re in control of the whole process … We are nurses, we are here for them … We will support them through it because they are our priority. They come first, last and always. And that doesn’t always fit well with sort of, academics, but that is what we’re here for.”
T2–N2: “I always have the patient’s best interests at heart at the expense of the research. I’m not here to recruit patients onto trials and I always say to them, ‘I don’t mind what you do, the decision is yours, and if you’re not happy to do the trial, don’t do it.’ I would never recruit anyone that I didn’t feel was happy. I’d never talk them into doing it because that’s not – I am a nurse first and foremost.”
The role as a caring clinical nurse was contrasted with the (lesser, rejected) role of researcher:
T6–N6–G: “I don’t think we ever think of ourselves as being a researcher.”
Interviewer: “So – you always think of yourselves as nurses?”
All: “Yes.”
T6–N1–G: “Nurses carrying out research, but not researchers.”
Nurses had considerable influence in these RCTs over which patients were informed about the RCT and how it was presented to them. They strongly defended their professional right to make decisions about these things, but as they recounted decisions they had made, they sometimes expressed strong personal views about particular types of patients, the RCT design, and its interventions. Most assumed they could use their clinical judgment to decide who to approach about the RCT and that they would be able to do this, and recruit patients to the RCT, without transmitting their views:
T2–N2: “[Patients] need to have great trust in their [doctor], you know, that’s what I say to them. If you’re happy with your [doctor], OK. If you’re not, it’s probably not the trial for them to go into … I know I must not have opinions on the studies and mustn’t influence them but for their own peace of mind I think they do need to have trust in their [doctor] just in case they’re randomised to [intervention 1].”
T7–N3: “You can tell when you’re speaking to patients, if they absolutely want the [organ] removed, you’re not going to be able to sway them, and I don’t think it would be ethical to try to be honest … I think if it was me in the situation or my dad in the situation I think I’d … rather [have intervention 1] … I keep that inside my head … If someone asked me, what would I do, if I can get out of answering the question, I would, but I’d never lie to somebody.”
Some nurses were much more open with the interviewer about how they could and did make decisions that affected recruitment. In terms of their ability to decide who would or would not be informed about the RCT, two themes emerged.
Using discretion over whom to approach
Some nurses defended their right to use their discretion, based on what they believed to be their clinical/caring or advocacy role, and did so assertively. There may have been good clinical reasons for decisions, but in others, it appeared to be their own discomfort with the process of recruitment, or aspects of the RCT, and even prejudiced views about particular groups or individuals:
T2–N3: “He was a very indecisive man generally, very indecisive, and in the end he literally just couldn’t make up his mind what he wanted to do … That’s a nursing decision if you like. You get to know what people are like and whether some – as I say like that indecisive man, he wouldn’t have coped with [having intervention 1]. You just know he wouldn’t because it’s his nature.”
Interviewer: “Does that mean that you actually make the final eligibility decision when you see the patient? You might decide not to approach somebody, at that point?”
T2–N3: “Yes I think that would be fair to say. Yes.”
T7–N3: “There was a patient who was young and he was clinically eligible, and basically he wasn’t very intelligent, and he would’ve been very easy to coerce to get on the trial. He would’ve gone on it, but I didn’t feel he was able to make a rational fully-informed decision, so I made the decision that he wasn’t going on the trial.”
T3–N2: “Sometimes you have to use your discretion about who you approach and not. You can see that a [potential participant] is there and she’s obviously very stressed, she’s got a screaming child and you kind of think ‘I don’t think she’s going to take part in the study anyway ‘cause her stress levels are just too high.’ So you don’t.”
Apologising for ‘bothering’ potential participants and avoiding recruitment
Some nurses found the decisions about who to approach about the RCT so uncomfortable that they became very hesitant. They became awkward and apologetic about approaching potentially eligible patients, or found reasons to avoid this discomfort:
T3–N1: “Recruiting [participants] to a study when they’re ill, that’s the most difficult time to ask a [participant] to take part in something. And that’s quite difficult and at times I almost feel apologetic, I think is the word, for approaching them … At the beginning, there were times where I felt I don’t know whether to approach them. I don’t know why, maybe it’s stereotyping, maybe it’s the way they look, the way they dress. I don’t know what else might be coming in and affecting me but I’m thinking ‘I’m not sure whether to approach them because I’m worried about being rejected, being told to go away’ … It was quite difficult to approach families that you thought didn’t speak English and that took quite a lot of courage on my part because you’re in the middle of a health centre, where everyone’s listening to you, and there’s nothing worse than actually not being understood or feeling that they’re not understanding you or that you’re causing them concern as well so I found that quite difficult … There was actually a [potential participant] I didn’t approach last week in [location]. Her attitude at the desk was very aggressive … and I thought I’m just not going there. No, so I didn’t.”
Nurses had considerable opportunities to inform patients about ways to overcome organisational issues, but they tended to just reinforce the difficulties, or transmit their own views:
T2–N1: “I can’t imagine having been already told by my doctor that this is what they are going to do for me and then be told by somebody else that hang on a sec, we are going to cancel all that and you could then potentially have [another intervention] than had already been scheduled. I mean I think that is a massive stumbling block.”
T1–N1–G: “I haven’t recruited anyone to the trial. Of the fifteen patients which have been registered, six of them came to us from outside [city] … It wasn’t in my conversation to ask these patients to be randomised because they were coming for [intervention 1].”
T1-N2-G: “My patients are right smack bang in the middle of a socially deprived area, very socially deprived, not amazingly well educated – not to say that everyone’s stupid because you get some who are quite well educated – but the majority of them aren’t and they’re just, so you know, they’ve just got these preconceived ideas and they don’t want to think about it. Cancer means they’re dead. That’s it … They don’t want to go to [name of] hospital because they think once they go … there, they’ll die because that’s where you go to die. So you’re asking them to go there to have [intervention 1], but they don’t want to go there because that means they’re going to die.”
There were particular difficulties in T5 where the nurses and researcher struggled to understand some of the basic principles of the RCT design, and where they also had strong personal treatment preferences:
T5–N3: “What is an RCT, sorry? … [Intervention 1] is better and the reason why I believe that is that it provides the extra support that the [patient] requires.”
T5–N4–G: “A couple of our [patients] have gone along and then they’ve got allocated to the control arm … I know that it’s a randomised trial, I’ve done research but from a clinician’s point of view – and this is what we’ve always quite openly sort of communicated to the [T5] project and this is why sometimes we have had difficulties – because we’d like to sort of say ‘right okay can we refer to the [intervention arm]?’”
T5–N1: “I feel a bit uncomfortable … The fact that I want them to get the intervention group, I’m sure that’s coming through. I was sure that I was sometimes trying to downplay that but I’m sure that my voice changes, there’s a lilt [laughs].”
Nurses identified recruitment barriers as being organisational difficulties, the lack of patients fitting the RCT eligibility criteria, and patients expressing strong treatment preferences or rejecting randomisation. Analysis of the interview data revealed, however, the nurses’ considerable discomfort and difficulty in relation to their roles, particularly in terms of which patients they approached to discuss recruitment to the RCT, a lack of knowledge for some about the RCT design, and their own views about logistical issues, treatment preferences, and patient characteristics. These views were expressed similarly by those employed by the wider networks and in the specific RCTs. It was also clear that many found their role as a recruiter conflicted with their perception of their clinical, caring, and advocacy responsibilities, leading to considerable discomfort and some clear instances of avoided recruitment. As with the doctors, above, some of the recruitment barriers could thus be seen to be originating from, or being reinforced by, the nurses’ views. Their unwillingness to approach individual or groups of patients contributed to the lack of eligible participants. If patients expressed a preference for treatment that accorded with theirs or fitted their view of the logistical difficulties, it was easy to allow them to have that treatment rather than offer recruitment to the RCT.