Qualitative substudy

Background

A secondary aim of the PEDAL trial was to explore and document the views of participants and members of the renal care teams in relation to their experiences of usual care and usual care augmented with intradialytic exercise. The qualitative substudy used a constructivist phenomenological approach35 to provide insights into the progress and conduct of the trial in all five study regions. Views and experiences of service providers and study participants were explored, including control group participants and those who did not complete the intervention. This report describes the positive and negative impacts of participating in intradialytic exercise (where a person chose to and was able to participate). It explains individual and contextual influences on decision-making and experiences as well as influences of the RCT context.

Qualitative substudy aims and objectives

The aim was to explore participants’ and renal care teams’ expectations and experiences of exercise training, with the particular aims of understanding:

• the beliefs and attitudes of participants and of the renal care teams towards exercise training as a treatment and how this might affect QoL
• the acceptability and experience of the intradialytic exercise training intervention as well as how being in the usual-care group affected behaviour.

Methods

Study design and methodology

The qualitative substudy of the multiregion PEDAL RCT used a constructivist phenomenological approach.35 This aimed for ‘empathic neutrality’ through researcher reflexivity36 and was led by a researcher (CB) who was not involved with the rest of the trial. Face-to-face interviews were conducted using a semistructured topic guide. The substudy was included in the initial RCT ethics submission, and major amendments were submitted to the relevant NHS Research Ethics Committee to include relevant documentation.

Participants and study context

Anyone who met the criteria for and participated in the PEDAL trial in the intervention group or usual-care group, or who had dropped out of the intervention but not the study, was invited to participate as long as they could communicate sufficiently to be understood in English. To ensure representation of diverse views, people from each of these ‘categories’ were invited to participate at each site. Potential interviewees were provided with an invitation letter and response form.

‘Providers’ were also recruited, with the aim of including a range of people involved in providing the study intervention and whose working roles might be impacted by the trial. Representation was sought from people in various medical, nursing and support roles in the renal unit and from the PEDAL trial team.

Selected and willing individuals were interviewed once during the life of the study. Trial participants had consented to the substudy as part of the PEDAL RCT; however, it was considered appropriate to issue an invitation with more specific information at the point of qualitative recruitment at each site. It was also important to give renal care providers information. At each site, potential participants were given the invitation letter by a member of the PEDAL trial team based at that site at least 1 week before the day of data collection. The interviewer was introduced to a potential participant by the same person and started by providing an opportunity for the potential participant to ask any further questions and accept or decline interview. All PEDAL trial participants who were approached were happy to be interviewed.

Data collection procedure

Face-to-face interviews involved primarily one researcher and one participant (patient or provider). On one occasion, two providers wished to be interviewed together for pragmatic reasons, and on one occasion two patients in a side room wished to be interviewed together. Patients were interviewed when they came in for HD and providers were interviewed opportunistically to minimise disruption to their days. The majority of participants were interviewed by the same person (CB), who has expertise in health-related qualitative research, with a minority of interviews conducted by a supervised research assistant (who was trained by CB).

Topic guides were designed to focus on the study aims and were informed by review of relevant literature. Questions of interest were designed and sequenced using guidelines from Kreuger37 to move from more general to more specific information and to enable an ‘easing’ into the interview process. Finally, participants are asked for the most important aspect of the discussion in their view. Questions were reframed as needed to ensure that it was possible to elicit both the required topics or information as well as unexpected views and experiences, without implying values or demonstrating expectations in relation to responses. Draft topic guides were reviewed and approved by the PEDAL trial project team prior to ethics review. The final topic guides are provided on the NIHR project page (URL: www.journalslibrary.nihr.ac.uk/programmes/hta/122309#; accessed 14 April 2021). Interviews were audio-recorded. General patient characteristics emerged through interviews, with more specific details extracted after unblinding of the PEDAL trial. Profession and role were noted for providers.

Data processing and analysis

Data protection guidelines were carefully adhered to for all transportation of data through download of audio-recordings to an encrypted laptop. Study rigour was enhanced through purposive sampling, multiple coding of a sample of data, ensuring an audit trail during analysis, and triangulation of data from different sites and regions, as well as through inclusion of people with a diversity of roles in relation to the PEDAL trial and intervention.35 Data processing and subsequent framework analysis aimed at increasing abstraction while reflecting nuances, with development of theory to explain views and experiences in this study38 (Table 10). Analysis of all data collected in two different regions was carried out by the lead researcher (CB), who created the initial thematic framework. Subsequently, a research assistant (Jane Scullion) analysed all remaining data and coded this in NVivo version 10 (QSR International, Warrington, UK), applying and modifying the thematic framework through regular discussion with the lead researcher (CB). Mapping and interpretation, with triangulation and development of overarching theory, were carried out by the lead researcher (CB).

Results

Summary of data and participant characteristics

Interviews took place in all five regions: two sites in London, two in Central Scotland and one in each of the other regions. Owing to the staggered start to recruitment at different sites, and recruitment over time within sites, people were interviewed at different points in their trial journey. Interviews took place over 16 months, between May 2016 and September 2017.

Tables 11 and 12 summarise participant recruitment numbers in each region (patients and providers, respectively) for the qualitative substudy. According to baseline numbers, 17% of participants in the trial were interviewed, which increased to 22% when considering numbers at the 6-month follow-up. When looking at the 6-month follow-up time point, 16% of all intervention participants who completed participation, 13% of control group participants and 40% of those who did not or could not continue participating in the intervention were interviewed.

Trial participants included in the qualitative substudy included 26 women (46%) and 31 men (54%). There was substantial diversity in terms of age, months since starting dialysis and ethnicity. There was a mean age of 60 years, with a mean HD vintage of 43 months (data unavailable for 14 participants). Most people were receiving HD for 4 hours (range 3.5–5.0 hours) three times per week. Ethnic diversity among participants was greatest in London, the West Midlands, and North Wales and North West; it was lowest in Central Scotland and at the site of interviewing in the East Midlands. Where ethnicity was noted, there was representation from Asian British (Pakistani and Indian), black Caribbean and white British people.

People were interviewed at different points in their trial journey, at a mean of 11 months from consenting to participate in the trial and 8 months from their baseline assessment.

Explanatory theory and supporting thematic analysis

Analysis of the substantial quantity of data resulted in an explanatory model, depicted in Figure 3. This will be both explained and justified, supported by illustrative data. Explanation will move from the centre of the figure outwards, focusing first on the outcomes when people sustained participation in the intervention over time. The individual influences on decision-making about participation over time will then be explored, followed by contextual influences. Finally, an explanation of how the RCT context of participation affected decision-making will be addressed. Table 13 provides theme names and definitions.

FIGURE 3 Explanatory theory relating to individual-decision-making about starting and continuing participation in exercise while on HD. (Figure not reproduced here; see the original article.)

Impacts of pedalling while on dialysis

The final explanatory model (see Figure 3) provides the key outcomes in the centre; these were the aim of the intervention and were described very well by multiple participants who were able to sustain participation in both the trial and the intervention. Impacts included positive physical effects, such as more energy; increased fitness, stamina and strength; weight reduction; improved circulation; and reduced shortness of breath, symptoms from dialysis and restless leg:

Well, a bit more energy about me, certainly, than what I had a year ago. I’m getting less of the . . . problems you might get after you dialyse. When I was doing it [dialysis] first I was still getting . . . quite sick . . . but as this has gone on it’s improved. And I’m sure that the exercise must be part of that, you know, it must be helping.

This resulted in functional improvements for some, such as increased mobility and balance and better sleep patterns. This was described both by people who completed the intervention and by people who did not or could not continue for the whole period for different reasons such as comorbidities:

Noticing the change in me . . . walking, going up the stairs, coming down the stairs. It’d become so difficult for me. But after exercise it became easier and easier.

A variety of positive psychological impacts were described, including feeling happier and more hopeful, having feelings of pride and having increased confidence and self-esteem. Participants felt that time passed faster, less time was spent with negative thoughts and they enjoyed the routine, and they felt more ‘normal’ and independent again. One participant explained:

It’s been a really, really good journey. Should I say, I’ve had many ups and downs in that study. But, I mean, now I’m feeling myself, so I’m much more confident and much quicker . . . because I know if I can go on to the bike for so long and on a certain resistance . . . I’m getting a bit stronger. So that has basically boosted my confidence.

Participants and providers described increased participation in physical activity when outside the renal unit and greater participation in life more generally, which they attributed to intervention participation:

We’ve got a treadmill at home, hardly used it. And now I’m on it, ‘cause I asked [a PEDAL Trial employee] what’s the best way to . . . so he did explain and that’s helping me a lot as well.

It was also evident that there was a feedback loop between the experience of these positive impacts, enjoyment of the experience overall and the motivation to continue participating in exercise during dialysis. Some people who did not continue the intervention felt this and were disappointed to have to stop:

Yes, I felt it was doing me good. I was actually enjoying it.

Impacts of cycling were not all positive, with some describing feeling uncomfortable or unsafe on the bike, that there was a lack of interaction with others and that exercise exacerbated fatigue and depression and other co-existing conditions such as low blood pressure and musculoskeletal pain. For some people this was a barrier to continuing intervention participation, whereas others completed the intervention.

People in the control group did not describe these positive or negative feelings or experiences purely from taking part in the trial. Some had participated to give back and experienced related satisfaction but did not discuss the kinds of physical, functional, psychological and social impacts of participating in the trial that were discussed by those experiencing the intervention.

Individual influences on decision-making: physical activity culture, ability to cycle and social environment

In the explanatory theory (see Figure 1) there is an individual layer of influence on decision-making and a further contextual layer in the overall background of the RCT scenario. In the individual and contextual layers there are interacting aspects of the beliefs and culture relating to physical activity, ability to cycle and social influences.

The patient’s physical activity culture – or personal views, goals and social support relating to physical activity and their health status – affected their decisions to continue cycling. Motivators included a drive to participate in physical activity or exercise due to previous and/or current participation and interest in improving personal health outcomes or goals:

Well, I used to do a lot, I used to go to the gym quite frequently, but because I’ve got polycystic kidneys and pressure on my lungs it makes it harder for me to do . . . I try to find alternative ways to exercise, so cycling and walking’s helping quite a bit.

Ability to cycle mattered as well, and most frequently manifested as a barrier to participation (i.e. as a lack of ability to cycle). The highly variable nature of renal symptoms, influenced by dialysis days, affected whether or not people were having ‘good days’ or ‘bad days’. This was more likely to affect the degree of exercise participation over time during dialysis sessions. By contrast, comorbidities and changes in renal health status, such as fluid overload, blood pressure changes and musculoskeletal pain, were more likely to affect a person’s ability to continue the intervention:

When it got to the time to do it I discovered I had the problems with my hips, which is why I couldn’t do the cycling.

Relationships with other patients, renal unit staff and PEDAL trial staff also affected decisions to continue cycling. The motivational role of the person directly supporting cycling was very important on a day-to-day basis:

I talk to them about the benefits of cycling and I remind them of how they felt last time, because I find generally patients after they’ve cycled do feel good and they enjoyed it and they’re feeling happier with themselves – and then it’s kind of negotiating . . . I do try to be as encouraging as I can but not pushy . . . showing a bit of empathy.

Camaraderie, competition and encouragement between patients in the renal unit influenced some people, more often where there were multiple people situated together in a bay or in a longer ‘ward’ layout. This was described as a gap where people were in side-rooms:

They make jokes about it and we’ll say, ‘oh, are you on your bike today?’ You know, ‘where have you been today?’ and things like that . . . The rest of the group were quite encouraging . . . people were encouraging her.

This camaraderie linked strongly with the contextual layer of the theory: a supportive, proactive culture in the renal unit.

Contextual influences on decision-making: social environment, physical activity culture and ability to cycle

A supportive social environment in the renal unit was described at some sites, with positive relationships within the unit and culture of empowerment:

We’ve done lots of things down here that we’ve tried to motivate the patients with. I mean, we’ve done art in hospital, we’ve done lots of things, staff here are quite motivated.

A patient explained how important relationships are in the renal unit:

We sort of all really get to know each other, don’t we? And the staff are really supportive, really good. You’re trusting people with your life, because your fistula’s your life, isn’t it?

Another provider explained that, from her experience, she thought renal unit staff could empower or disempower patients unintentionally, through beliefs about ensuring safety, efficiency and patients’ capabilities:

Of course you can come in and do it yourself, and the biggest resistance I had was not the patients but the nurses and a lot of it came from . . . that maternal or paternalistic attitude of ‘I’ll do it and I know it’s done well, or it’s done properly, or I can get done quicker’ . . . you know, you just want to do it right for them, so I do think that creates an awful lot of it, because they’re with us for a long time, and if we don’t instil that confidence in them that you can do it, then the patients aren’t going to believe that they can do as well.

Some renal units described having a proactive attitude to change:

We like embracing new things on the unit . . . It’s new at first but you adjust to it, no trouble – we seem to do that here.

The social culture of the renal unit also linked with attitudes towards physical activity participation and engagement with the intervention. Staff enthusiasm about physical activity and exercise as necessary for renal patients was important and related to their awareness of the rationale for it, prior experiences of exercise for renal patients, personal beliefs about physical activity and exercise as beneficial for all, and noticing positive changes in participants during the trial. One provider explained:

I would love to see a lot more exercise on the unit. I want to see my patients feel better and have better outcomes. I certainly see it as one of those things that could be good for that.

Some providers described having initial scepticism about whether or not renal patients would engage with the intervention:

Prior to starting I’d say that I had . . . if anything, sceptical as to what the impact would be . . . They’re not the most active bunch, renal patients.

This may also have interacted with participants’ ability to cycle, owing to structural barriers affecting whether or not the exercise bike was available at a suitable time:

Sometimes it does get forgotten, and part of that is being busy as well, and short staffed.

One participant explained:

Well, everybody just encouraged me . . . patients and staff. Oh aye. Most of the staff managed to put me on. But there was a few just didn’t know how to and they kept saying I need to learn how to do this.

Influences of the randomised controlled trial context on decision-making

It is important to remember that the context of participating in a RCT underlies the decision-making about both starting and continuing to participate in the intervention group and control group.

It was reassuring from a trial perspective that the evidence suggested that participants in the control group did not increase or decrease their physical activity or exercise during the trial period. However, allocation to either the control group or the intervention group had an impact on people in other ways that affected their willingness to continue participating. People gave different reasons for consenting to participate in the trial. Some people wished to ‘give back’ because they felt that they had received so much from the health system and wished to improve knowledge and help people receiving dialysis in the future. Others wished to participate to pass the time, as they found being in the renal unit boring. Both of these motivators were valuable if a person was allocated to the control group, because they might be less invested than someone in the intervention group of the study. For example, as one control group participant explained:

It’s good to be doing something rather than just sitting here doing nothing. The daytime telly’s not very good, you know.

If allocated to the intervention group, however, a person may have had less motivation to exercise than others in the intervention group. Some people consented to the trial because of social influences, for example seeing others in the unit benefitting from participation. This again was positive if the person was allocated to the intervention, but less so if allocated to the control group:

So one or two of my co-patients, they volunteered, and they went ahead. So I said, why should I be left out?

A strong motivation to consent involved interest in physical activity and exercise due to having been active in the past, and currently being invested in achieving health-related benefits, including those relating to renal condition and broader health and mobility needs. Some people wished to lose weight specifically to increase their likelihood of transplant. These were emotionally loaded reasons for participation and meant that a person was very affected by their randomisation:

It’s unfortunate the randomisation of the exercise to control or exercise, and it disappoints, that’s another thing that we have to take into consideration is somebody’s just been turned down and you’re exercising somebody in the next bed to them. How do you keep them interested in the trial for 9 months now?

A participant reinforced this:

. . . if I’m honest I’ve been a bit disappointed because I wasn’t selected to do the cycling and a lot of people who were selected don’t seem to want to do it. I’m one person who actually did and I was keen to do it, but I just wasn’t picked, so I’ve been really disappointed about that.

In addition, it is important to remember that the trial context formed a barrier for some people who were interested in exercising but did not wish to attend assessments, particularly on non-dialysis days. This was complex in relation to conflicting priorities, because people valued their ‘good days’ when not in the renal unit and were sometimes living at a substantial distance from the location for trial assessments:

The difficulty with the unit here is that we have a fair number of patients who have extended travel times to get here because they don’t live on the mainland and they have to catch the ferry. And what the study protocol dictates is that they have to come on a non-dialysis day to have their assessment. And during the study it was particularly challenging trying to explain this to the patients. Particularly those who were, you know, there were a fair number who were unhappy about having to attend for assessment if they didn’t get the intervention.

One provider stated:

A lot of patients are asking can they not be on the study and still use the bikes, so it’s like a gym on dialysis . . . There are a few that have said they don’t want to be on the study, but can they use the bike?

Another issue relating to the trial context is that in all but one site the intervention was supported by a dedicated PEDAL trial employee. In the site at which the renal unit supported intervention provision, a provider described difficulties in prioritising intervention provision at busier times. This could affect the ability of the person to cycle, owing to access. It is important to note, however, that multiple providers at other sites explained that, although often enthusiastic about the intervention, they would not be able to provide it without dedicated staff after completion of the trial:

My ideal scenario would be to have a physio[therapist] that would come in and work with patients, not just doing PEDAL but doing maybe other stretches and things for renal, that would benefit renal patients while they’re on dialysis and that would take the nurses out of it.

Another provider stated:

We have got quite a lot of work coming in . . . and sometimes there’s a staff shortage and . . . I don’t think we’ll be able to support them . . . and if there is nobody to initiate that programme then I don’t think it’s going to go very far.

Interlinkages between influences and impacts: explanatory theory

Data clearly support positive impacts from cycling during HD where a person is fit to do so and able to continue, and these impacts motivated people to continue participation. The trial context affected whether or not some people decided to take part because, although interested in exercise, they did not wish to attend assessments on non-dialysis days. The trial context also influenced whether or not people continued participation in either the control group or intervention group depending on their reason for consenting to participate.

Where people participated in exercise, their willingness and ability to continue was affected by their own individual capability to do so, their attitudes, beliefs and family support and their relationships with those around them in the renal unit. The motivational roles of other patients and PEDAL trial and renal unit staff cannot be underestimated, and availability of this support was influenced by the culture and beliefs in the unit. A positive culture of involvement and empowerment was important. This was particularly well harnessed by one PEDAL trial employee who was able to enlist the support of all staff in the unit, including medical, nursing and support staff. The impact of this was a unit where all were encouraging those cycling, resulting in a sense of enthusiasm and camaraderie. This person was also highly strategic in motivating others to consent to the trial, for example by building relationships with patients and asking the person who volunteered first to move bays and influence others into participating through role modelling.

Discussion

Results of the qualitative substudy are largely consistent, with the few existing qualitative studies exploring health-care provider and patient experiences of intradialytic exercise. The current study findings go further in relation to theory generation; therefore, it is possible to make some more in-depth suggestions for further development. The literature search conducted early in the trial was repeated in June 2020 [searching PubMed, MEDLINE, Cumulative Index to Nursing and Allied Health Literature (CINAHL), PsycInfo^® (American Psychological Association, Washington, DC, USA) and SPORTDiscus (EBSCO) using the search terms ‘exercise’ and ‘haemodialysis’ and ‘qualitative’, with no limiters] and no further studies were found.

Young et al.39 conducted a study in the UK of patient and staff perceptions of intradialytic exercise. They carried out focus groups with 24 patients and 9 staff before implementation and semistructured interviews with 11 exercising patients and 8 staff members 6 months later. Thompson et al.40 interviewed 25 patients and 11 staff in Canada to explore perceptions of intradialytic exercise. Heiwe and Tollin41 conducted individual interviews in Sweden with six men and four women who were experiencing implementation of intradialytic cycling. There was a focus in the studies of Young et al.39 and Heiwe and Tollin41 on identifying barriers and motivators in the data; both looked at the stage of deciding whether or not to take part, and the former also looked at experiences 6 months later.

There were multiple common themes between the PEDAL trial themes and those in previous studies. In particular, Young et al.39 found similar functional, psychological and physiological benefits, and viewed these as providing positive feedback that enhanced ongoing participation. Thompson et al.40 found social benefits but did not focus on analysis of the impacts of exercise. Heiwe and Tollin41 conducted their interviews in the early stages of implementation and found that participants were motivated partly by a desire to pass the time and felt that this was a good opportunity for exercise. In addition, they noted that the positive experiences of well-being benefits fed into ongoing participation, which was evident in our study.

There were common barriers that revolved around staffing and workload, limited resources and concerns over roles and appropriate training or expertise.39,40 Interestingly, these three studies delivered the intervention with existing staffing: kinesiologist and nursing staff,40 nursing staff41 and physiotherapist staff.41 By contrast, in all but one of the renal units represented in the current study, a dedicated PEDAL trial employee delivered the intervention. The renal unit in the PEDAL trial that provided its own support for the intervention shared the workload of study implementation between staff and discussed similar barriers to the previously published studies. Staff in the other renal units commented on the necessity of dedicated staff time. This may relate also to a theme that was common to all studies: the importance of motivation, encouragement, reinforcement, feedback and support from staff to facilitate ongoing participation.

In the context of qualitative data collection and analysis, evidence of support from similar, previous studies provides further evidence of credibility and trustworthiness of the results, so the similarities in themes from different countries and contexts are encouraging. Our study was, to our knowledge, the largest of its kind by far, and had quite diverse participants, which provides valuable triangulation.

The results of this qualitative substudy clearly support the value of exercising while on dialysis for those who can. We also have useful insights about how to enable people to continue participation over time, both from feed forward of the positive impacts and through complex social and cultural dynamics in the renal unit. It is important to consider the insights of analysis for ongoing implementation of the intervention. The results suggest that a renal unit is an integrated system, with complex interdependencies and mutual influences on expectations, beliefs and attitudes. Staff and patients are intertwined as a community and, when considering changing complex behaviours, it is important to remember this. It would be useful to analyse the findings of this study in relation to current behaviour change theory as a secondary analysis, for example using the capability, opportunity, motivation, behaviour (COM-B) behaviour change model, which emerged from the behaviour change wheel of Mitchie et al.42 Interestingly, Young et al.39 considered the importance of influencing behaviours of staff as well as of patients; they recommended staff training and patient education to increase engagement. We suggest that the need for change goes beyond knowledge to culture; renal units with proactive and relationship-focused cultures appeared to engage people in the intervention in different ways, which may be more sustainable in the long term. In the future it would be valuable to explore ideas around ways in which providers empower and enable, how this is influenced by leadership and context and how this affects changes in the behaviours of people engaging in and sustaining interventions such as exercising while on dialysis.

Conclusions

The explanatory theory developed through the qualitative substudy of the PEDAL trial has provided in-depth insights into the influences on people’s decision-making relating to participation in both the trial and the intervention over time. Themes are supported by the few previous qualitative studies; however, our data are more extensive, and the resulting theoretical explanation goes further to enable suggestions for future implementation and intervention sustainability. A novel finding relates to the importance of culture, and leadership to enable this, in the renal unit. Providers interviewed felt that the trial did not impinge much on their jobs, which supported trial implementation; however, this may not be conducive to sustainability of the intervention beyond the trial. Moving forward beyond the PEDAL trial, it will be most important to influence the attitudes, beliefs and behaviours of providers. Where providers are engaged and empowered, they are much more likely to engage and empower patients in participating. Once a patient is participating (where capable), the numerous physical, functional, psychological and social benefits are likely to support their ongoing participation. This in itself may then feed forward into a positive, proactive culture in the renal unit. Further research is needed to explore the complexities of intervention implementation in the longer term, with a view to aspects of behaviour change and insights from implementation science.