Phase 1
Recruitment
Out of 122 outpatient haemodialysis patients, 94 were assessed as eligible for participation. 70 participants declined to participate (74.5% of eligible participants). The most common reasons being a lack of interest in the arts (n=29), and anxiety related to a perceived lack of creative ability (n=11). 24 participants (25.5% of eligible participants) consented to participate in the study and were successfully randomised to control or intervention group. Therefore 80% of the target sample size was successfully recruited to the study. An overview of screening, recruitment and attrition can be seen in the Consort flow diagram in Figure 1. The demographics of recruited participants at baseline are presented in Table 1.
Retention
There was an overall attrition rate of 12.5% (n=3) during the study. Two participants withdrew from the intervention group and one from the control group. Reasons for withdrawal centred around physical health; one participant withdrew from the study following baseline data collection as they sustained injuries that made them physically unable to participate in the intervention, one participant withdrew at three-month follow-up due to a significant decline in their physical health, and one participant died shortly before the three-month data collection time point. No participants withdrew from the study whilst engaging with the intervention and no participants withdrew from the study because of participant burden.
Intervention adherence and fidelity
One participant withdrew from the study prior to starting the intervention. All 11 participants who started the intervention completed six art sessions over the course of three weeks. As there were only two arts sessions per participant each week, this provided flexibility to allow participants to rearrange their sessions if required.
While it was planned that sessions would be limited to an hour, during implementation it became apparent that this restriction was limiting engagement. As participants developed skills their engagement increased and they required more time and to complete their work. Consequently, the mean length of time increased for each participant over the course of the six sessions, with the initial session lasting hour (58.8 minutes, SD= .9) and the mean length of the final session lasting approximately 69.6 minutes (SD= .16.2).
Clinical outcome measures
The majority of participants completed the questionnaires with assistance of the researcher (CC), only two participants completed the questionnaires without assistance. Using this approach there was no missing data across all outcome measures at each time point, with the exception of participants who withdrew from the study.
Table 2 shows the mean, standard deviation and 95% confidence intervals of the clinical outcome measures at baseline for all participants. The mean HADS score for both anxiety and depression were lower in the intervention group (3.42 and 2.75 respectively) when compared to the control group (8.33 and 7.33). This suggests that the two groups were not similar at baseline in terms of their levels of anxiety and depression. The mean scores across the different KDQOL-SF 36 subscales were also consistently higher in the intervention group compared to the control group, suggesting the two groups were not similar at baseline in terms of their HRQoL.
Figure 2 shows the mean anxiety and depression scores for the intervention group over 3-months longitudinal follow-up. A similar reduction in anxiety and depression was also found in the control group over the three-month follow-up time period. The reduction in anxiety and depression is more pronounced in this group as they had notably higher mean anxiety and depression scores at baseline when compared to the intervention group. Figure 3 shows the mean KDQOL-SF 36 subscale scores for the intervention group over 3-months longitudinal follow-up, showing a general trend of mean increases at six weeks and three months follow-up, after the initial apparently inconsistent changes across subscales in the post-intervention period. The control group also demonstrated varied changes across different subscales of the KDQOL-SF 36 during the longitudinal follow-up period, but at three months all subscales had improved compared to the initial baseline mean scores.
Phase 2
An overview of interviewee characteristics are provided in Table 3. Nine participants were recruited from the intervention group, four from the control group, and nine healthcare professionals were recruited who had observed implementation of the intervention whilst working on the unit.
A total of four themes and 17 subthemes were identified through the semi-structured interviews. An overview of the themes, subthemes, and the participants who contributed to the identification of these themes are provided in Table 4. Data will be presented to outline each theme in turn.
Theme 1: Perception of art
This theme highlights the interviewee’s perception of art participation, both prior to the intervention, during and following implementation. Participants had mixed preconceptions and apprehensions of art, with patients saying they were ‘a bit nervous’ mostly due to a lack of experience, and were consistently critical of their artistic skills.
The first one was absolutely dreadful! Because it was painting, and I just ended up with this… blobs all over the place. It was supposed to be a scene of you know, eh, by the beach, a beach and water and clouds. Awful! Blob. Blob. Blob.
- EG06
However, participants’ reported a change in their perception of art participation following implementation of the intervention. The changes in perception were exclusively positive, with participants identifying an interest where they had none previously.
I was quite happy I done it and I was quite happy with the results… You know, and it gives you something – you achieve something,
- EG01
Theme 2: Effect of the arts-based intervention on patients and staff
This theme captured the effects the intervention had on both patients and HCPs. The art triggered positive emotional responses amongst participants, with improvements in self-esteem and the development of a sense of purpose being commonly reported as it helped participants ‘achieve something’.
‘I suppose satisfaction that you can add another string to your bow at 80 odd years of age.’
- EG03
There was an increase in social interaction both between patients and between patients and HCPs as ‘it gave them a bond’, and improved the dialysis experience for both patients and HCPs. Participants reported that the intervention required a degree of focus and concentration, and consequently they were distracted from both their thoughts and their surroundings, improving their intradialytic experience and altering patients’ perception of time in a beneficial way, by seemingly making the time pass quicker as they were focused on a task and no longer watching the clock.
Aye the time seemed to go in a bit quicker whenever … comes in, you watch TV but uh, you never notice the time so much. But whenever you’re painting time, that hour or whatever it was that she was in with, just seemed to go like that.
– EG05
Healthcare professionals reported that the intervention had an impact on the environment itself by transforming the atmosphere into one where the focus wasn’t exclusively clinical:
This was sort of taking it to the next level and saying ‘it’s not all about dialysis, it’s not all about kidneys here, we’re looking after you and we are seeing what can benefit you positively here’
– HCP04
Theme 3: Acceptability of the arts-based intervention
This theme explored the acceptability of the intervention within the haemodialysis setting for both patients and HCPs. There are six sub-themes related to the intervention’s acceptability, addressing the intervention, implementation strategies and context (Table 4). Participants identified potential barriers to implementation related to the haemodialysis setting, but highlighted how these barriers were easily overcome.
Not even patients with fistulas! Did that affect them? No! (Laughs) Not at all! So, if you can get over that barrier and they’re happy – yeah, there were no problems.
- HCP09
The presence of a facilitator was identified as beneficial, both for overcoming barriers and producing a positive experience, as they were perceived as ‘a good teacher’. Participants highlighted the importance of choice and variety for sustained engagement ‘because they weren’t Pidgeon-holed’, as well as the suitability of materials for the haemodialysis environment.
It was all… catered to the environment. It was fit for the environment. Even if you used water based, they were not like messy water based… what children would use, you know. So it was all very… very accurate. Very well done.
- HP06
Theme 4: Acceptability of research procedures
This theme highlighted the acceptability of trial design and procedures through four sub-themes. Participants identified curiosity and boredom as main motivating factors for participation.
Clustering was identified as essential for the acceptability of randomisation, as ‘it would have annoyed’ control participants to view others engaging in the intervention, whilst a waiting list was important to reduce attrition in the control group as participants reported that this as a core motivation for remaining in the study.
Just looking forward to it… to doing the art, just something different to amuse me here. Instead of lying sleeping.
- CG03
The outcome measures were not perceived as burdensome to participants however they weren’t seen as appropriate for capturing the benefits of the arts.
I answered the question, but I couldn’t see it’s got anything to do with me on dialysis and doing art.
- EG03