Exercise- and strategy-based physiotherapy-delivered intervention for preventing repeat falls in people with Parkinson's: the PDSAFE RCT

Parkinson’s disease

Parkinson’s disease (hereafter referred to as Parkinson’s) is a progressive neurological condition characterised by impairments of movement and postural control (balance) and non-motor deficits. 1 There are estimated to be around 127,000 people with Parkinson’s in the UK. 2 The prevalence of Parkinson’s increases steadily with age. 2

Falls in Parkinson’s

Falls among people with Parkinson’s are both common and disabling. 3,4 Approximately 40–70% of people with Parkinson’s fall each year, and one-third fall repeatedly. 5 Overall, people with Parkinson’s are twice as likely to experience falls as a healthy elderly population. 6 There is evidence to show that falls have a major impact on the lives of people with Parkinson’s, including debilitating effects on confidence,7 decreased activities levels8,9 and reduced quality of life. 10 Such falls are financially costly for individuals and health-care systems. Repeat falls are a risk factor for further falls and carry devastating consequences, such as fractures, immobility and fear of falling leading to dependency and social isolation. 4 Although the incidence of falling increases with disease severity, falls are common even in the early stages of the condition. 11

Falling among those with Parkinson’s is associated with a host of risk factors, including disease severity, duration of disease,12,13 self-reported disability12 and impaired mobility,11 as confirmed by Canning et al. 6 The strongest predictor of falling, identified from meta-analysis, is having had a previous fall. 4,6 Loss of motor control [e.g. anticipatory and reactive postural control, reduced leg muscle strength, proprioception and gait speed, increased gait variability and freezing of gait (FoG)] have also been shown to be associated with, and predictors of, falls. 14,15 In addition to the motor symptoms, impaired cognition and orientation,14 as well as misjudgement and distraction,16 have also shown significant association with falling.

Falls prevention

Drugs are the main treatment used to manage the symptoms of Parkinson’s, while research into finding a cure continues. However, reduced postural control and falls do not respond to medication. 3 It has long been accepted that exercise is a fundamental component of treatment for people with Parkinson’s, alongside medical and surgical management, and its positive effects on symptoms are well supported in the literature. There is substantial evidence showing that regular exercise can be particularly beneficial for people with Parkinson’s to maintain postural control, mobility, daily living activities and general symptom management. 17,18 At the conception of this trial, there was no definitive evidence that an exercise programme would be as effective in the population of people with Parkinson’s as it is in the general population. Research has also shown that selecting and targeting each of the symptoms independently may not be enough to influence the falls rate. Instead, a multidimensional model that combines functional exercises with behavioural strategies is likely to provide a greater influence on reducing falls. 19

PDSAFE is an example of a multidimensional programme. 19 It is an individualised, home-based, physiotherapist-delivered, personalised treatment programme of exercises and strategies to help prevent falls in Parkinson’s. The novelty lies in both the content (i.e. disease-specific exercises and strategies for instability, use of motor relearning and cognitive awareness) and delivery [i.e. personalised feedback using a digital versatile disc (DVD) for adherence and self-management]. The programme comprises (1) exercises for postural control, gait and muscle weakness; (2) strategies for reducing freezing and encouraging stability and gait efficacy; and (3) a feedback model to promote learning and adherence. Frequency of intervention sessions is faded over time, 1 hour twice a week for 1 month, then once a week for a further 2 months, then once a month for another 3 months.

Aims and objectives

Prior to the main trial, a small pilot study was conducted as part of the protocol development work. The aims of the PDSAFE pilot study were to confirm the:

• content and delivery of the novel intervention

• implementation of procedures for recording and checking treatment fidelity

• battery of assessments to be administered at the baseline and follow-up visits.

The objectives of the PDSAFE pilot study were to:

• recruit a sample of up to 20 people with Parkinson’s: up to 10 in Southampton and up to 10 in Newcastle

• take informed consent from up to 20 people with Parkinson’s

• complete screening measures on up to 20 people with Parkinson’s

• test the collection of falls data through the use of diaries with up to 20 people with Parkinson’s

• complete baseline assessments with up to 20 people with Parkinson’s

• assess the time taken to complete assessments

• determine the feasibility of assessments

• determine the acceptability to participants of assessments

• deliver the PDSAFE intervention to up to 20 people with Parkinson’s.

Procedure

Ethics approval was given by the South Central – Hampshire B National Research Ethics Service (reference number 13/SC/0538). Sixteen participants were recruited [six from Newcastle and 10 from Southampton, including one patient and public involvement (PPI) representative]. All participants completed all screening measures [i.e. Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), Hoehn and Yahr (H&Y) scale and falls characteristics]. Six participants from each site completed baseline assessments, participated in the intervention for 1 month and completed the battery of assessments post intervention. The remaining four participants in Southampton provided falls diaries only, to test the diary completion required prior to randomisation in the main trial.

Two patterns of treatment delivery were tested to establish the most appropriate way of providing the intensity of treatment in the early stages. In Newcastle, participants had the intervention for two sessions a week for 3 weeks, then once a week for 1 week. In Southampton, participants had treatment for two sessions a week for 2 weeks and then once a week for 2 weeks.

Outcome

Overall, the participants enjoyed the treatment. Participants’ comments included ‘very pleased he participated – he has found it very beneficial and continues to use several strategies’, ‘Doing all home exercises and feels his balance is getting better’ and ‘have enjoyed the exercises and have learned a lot about keeping active’.

The therapists were enthusiastic about the programme and recognised the importance of mapping the house for problem mobility areas and integrating the exercises with video strategies appropriate for each individual. They commented that there was a lot to do in the initial stages around planning the strategies and fitting the weighted vests, so the pattern of treatment of two sessions a week for 3 weeks may be more valuable than two sessions a week for 2 weeks. However, it was agreed that this may not be appropriate for each person and that some flexibility in the way the sessions are delivered should remain. Training was recognised as very important. Some people found the weighted vests difficult to put on and required help from their partner; most people could tolerate only a small percentage of body weight in the vests, yet most people viewed them positively.

The outcomes of the pilot study were discussed at a face-to-face meeting of the grant holders, which was also attended by three PPI members; adjustments to the protocol were suggested and implemented. The pilot study confirmed the content and delivery of the novel intervention for the main trial (including printed versions of the exercises and strategies, video vignettes for play on a tablet or DVD and the use of weighted vests for strengthening exercises). Procedures for recording and checking treatment fidelity were also established, including weekly telephone and Skype™ (Microsoft Corporation, Redmond, WA, USA) consultations with the physiotherapists. As a result of the pilot, the assessment time was shortened and procedures tightened. For example, the handgrip strength measurement test20 was dropped from the main study, but retained as a substudy in the Portsmouth area; and the shorter (15- instead of 30-question) version of Geriatric Depression Scale (GDS) was used.

Primary outcome: risk of repeat falling between 0 and 6 months

The power and sample size calculations were based on the findings of a previous trial by Ashburn et al. 37 This was a rehabilitation trial similar to PDSAFE in design. In the EXSART trial,37 the risk of repeat falling in a 6-month period was 68% in the control group and 56% in the exercise group. However, falls risk was anticipated to be lower in PDSAFE because EXSART was restricted to people who had fallen twice or more often in the previous year. Therefore, it was assumed that the risk of repeat falling between 0 and 6 months would be 63% for the control group and 50% for the intervention group. This required 228 participants per group, with data for analysis of 456 participants in total. Allowing for a 5% drop-out rate between randomisation and 6 months meant that 480 participants needed to be randomised. Furthermore, allowing for a 10% drop-out rate between agreeing to the 3 months pre-randomisation falls collection and randomisation meant that 534 participants needed to be recruited to the pre-randomisation falls collection period. In total, 541 participants were recruited to the pre-randomisation falls collection period. Power calculation scenarios are summarised in Appendix 6, Tables 26 and 27.

Power calculations (see Appendix 6) for rates of falling were based on Tango38 and related to the number of falls during a fixed follow-up period, analysed using negative binomial regression conditioned on baseline counts: specifically, formula 23 in the paper38 was used, assuming equal rates in the baseline and follow-up periods in the control group and a follow-up period of twice the length of the baseline. Anticipating a falls rate ratio (FRR) of 0.8 between 0 and 6 months post randomisation, that is a 20% reduction in the rate of falling in the intervention group compared with the control group, and based on a rate of 2.5 falls in the 3-month baseline period, 197 participants per group were required at analysis, which led to 488 participants being recruited to the pre-randomisation falls collection period.

Other scenarios were considered for the differences in risk of repeat falling or for the FRR (see Appendix 6, Tables 26 and 27) and generally led to recruiting fewer than 600 participants to the pre-randomisation falls collection period. All the calculations aimed for 80% power in 5% two-sided tests between the intervention and control groups.

Therapists

Physiotherapists were recruited by each site. Trial requirements stated that each therapist should have experience in Parkinson’s or falls rehabilitation. It was initially designed that each site would have one treating therapist and a trained cover for periods of absence. However, owing to clinical workload and logistical delivery, a total of 18 therapists were trained over the study period. One lead therapist co-ordinated the team, delivered the training and development activities, and monitored the fidelity of intervention delivery.

Training, facilitated by the lead therapist, included attendance at one of the compulsory 2-day training events held on three separate occasions. In addition, therapists were asked to attend a virtual weekly meeting by telephone to discuss clinical cases and problem-solve within the boundaries of the intervention protocol. These sessions were chaired by the lead therapist, with a total of 122 telephone contacts made available over the intervention period. To maintain a high standard of clinical reasoning throughout the intervention period, therapists were also asked to attend (either physically or virtually) monthly ‘masterclasses’ on key clinical topics such as cognition and dual tasking, turning, FoG and balance. Alternating with masterclasses, therapists were asked to present case studies on key topics for team discussion. Twelve ‘masterclass’ topics (some were repeated for new therapists) and seven case study reviews were held over the intervention period.

Fidelity of the intervention was a priority to encourage uniformity of practice. The lead therapist observed each therapist in a treatment session with a participant, once a month for the first 3 months of delivery and then once every 3 months for the delivery period. Following the observation, a clinical reasoning discussion was completed and a report written. Therapists could also request additional joint sessions with the lead therapist if they had concerns or queries regarding a particular participant. This ensured that the PDSAFE intervention was uniformly delivered across all sites and by all therapists. A total of 75 fidelity sessions were held over the intervention period, with all therapists being assessed.

Intervention sessions

A total of 2587 sessions were delivered to the 291 intervention participants, with the majority of participants receiving the anticipated 12 sessions (mode = 12).

Figure 3 presents the total number of sessions received by participants allocated to the intervention arm of the trial. The majority of participants, 236 out of 238, received the exercise assessment and at least one supervised session. Two participants did not start because they had changed their mind and 19 received fewer than seven sessions; reasons for not fully engaging included admission to a nursing home, deteriorating health, commitment was too much and caring for others; in some cases, no reason was given.

FIGURE 3.

Total number of intervention sessions received by participants.

All interventions sessions included a brief review of falls; warm-up exercises; review, practice and progression of a participant’s individual exercise programme; and strategy training in functional scenarios as a basic structure. Each therapist tailored the strategies treated, exercises prescribed and functional tasks practised from a menu for each participant.

Selection of strategies

Evidence from the literature and from previous studies of falls among patients with Parkinson’s, as well as expert opinion, were used to determine the most frequent falls mechanisms in Parkinson’s. 19 Eight strategies were defined: avoiding tripping, dual tasking, freezing cues, moving in tight spaces, picking up an object, reaching, stepping backwards and turning. As described above, through the process of taking a detailed falls history, clinical assessment and advanced clinical reasoning, therapists determined the most likely ‘fall mechanism’ for each participant (levels 1 and 2 in Figure 2). For example, in the case of a participant who repeatedly reported catching their foot and falling, regardless of the task being undertaken or location, would be most likely to have a falls mechanism of tripping; thus, the strategy ‘avoiding tripping’ would be selected by the therapist.

Over the 291 participants who received the PDSAFE intervention, strategies were selected a total of 440 times, with a strategy being used in treatment a total of 3447 times over the period. This provides a large sample to consider the clinical reasoning process of the therapists selecting the strategies. Figure 4 demonstrates the number of times each strategy was selected as a potential falls mechanism corresponding to the number of times that strategy was used in a treatment session.

FIGURE 4.

Stratification of strategies used in delivery of PDSAFE intervention.

Importantly, this allows description of the strategies selected as primary falls mechanisms and thus treated for the majority of the intervention period versus strategies that may have been selected as a secondary or subsequent strategy and thus treated less frequently or for a shorter period of time.

It is apparent that ‘avoiding tripping’ was the most widely used strategy. It was selected following assessment a total of 116 times, used 1110 times during treatment sessions and accounted for 26% of all strategies selected. The figures for ‘turning’ are similar [selected 107 times and used 938 times during treatment (26% of the total)]. ‘Freezing cues’ was also frequently selected as a strategy [selected 79 times and used 365 times (24% of the total)]. It is clear from Figure 4 that all other strategies were selected and used in treatment with similar frequencies to each other.

Selection of exercise prescription

Once the strategy or strategies most appropriate for addressing a participant’s falls mechanism had been selected, therapists used advanced clinical reasoning and assessment skills to determine the physical impairments and deficits in physical falls risk factors that were most likely to contribute to the fall mechanism (see Figure 2). For example, the participant described above, who frequently caught their foot and subsequently fell, was allocated the ‘avoiding tripping’ strategy. The therapist must consider a number of reasons why the participant has a tendency to catch their foot, such as weakness of the muscles used to lift the toes, failure to transfer weight onto the supporting leg appropriately because of hip weakness or reduced limits of postural control stability, or failure to achieve enough clearance from the ground because of weakness in the hip flexors. Through assessment, the therapist determines the most likely impairment and designs a functional strength and postural control exercise programme from the available menu that treats this impairment.

Evidence from the literature and from previous studies of falls among patients with Parkinson’s, as well as expert opinion, were used to determine what exercises were available on the menu for therapists to select from. 19

Table 3 shows the menu of exercises available for the therapist to select from when putting together an individual participant’s PDSAFE intervention.

Exercises were available on six levels; working through the levels enabled progression (level 4 in Figure 2) and maintenance of intensity for each exercise. All programmes had to include at least one balance and one strengthening exercise. A total of 1693 exercises were selected by all therapists over the intervention period with an average of six (range one to eight) exercises prescribed per participant across the period. Figure 5 shows a spread of all exercises being used over the intervention period.

FIGURE 5.

Exercise use frequency across all of the intervention, in all participants.

Figure 5 demonstrates the predominance of the more dynamic postural control (compensatory step and lunge and heel/toe walking) and strengthening exercises over other exercises from the menu. The complex nature of the compensatory step and lunge exercise makes it suitable for the treatment of many of the falls risk factors associated with Parkinson’s; for example, high dynamic stepping actions help with motor control, compensatory stepping helps with the regain of an appropriate base of support from a loss of postural control or a trip by increasing stepping amplitude for those who freeze and expanding limits of stability for those who fall when reaching. The practice of stepping backwards with appropriate postural control and weight distribution will assist those who fall stepping backwards. A common symptom of the Parkinsonian gait is loss of foot clearance, heel strike and step length45 (hence the predominance of tripping as a falls mechanism); thus, the high frequency of the use of heel/toe walking to improve these impairments is also unsurprising. As each exercise programme had to include both strengthening and postural control exercises, the high frequency of selection of strengthening exercises can be attributed to the fact that the postural control menu contained fewer exercises that could be selected. This is less of a clinical reasoning observation and more related to the ratio of exercises in the menu.

Design

A longitudinal qualitative study was conducted alongside the main PDSAFE RCT. The qualitative study drew on the principles of grounded theory50 as follows:

• gathering rich, in-depth data through ‘intensive interviewing’

• conducting detailed analyses, which become increasingly theoretical

• using line-by-line coding to ensure a high level of familiarity with the data

• writing reflective memoranda to assist with data interpretation

• employing theoretical sampling.

Theoretical sampling refers to a strategy that ‘allows the researcher to generate theoretical insights by drawing on comparisons among samples of data’. 51 As per the theoretical sampling strategy, participants with a range of different characteristics were included, to confirm that the qualitative findings were broadly applicable to the variety of people with Parkinson’s included in the larger sample. Previous research has suggested that these key characteristics might affect participants’ experiences of the PDSAFE intervention.

Longitudinal qualitative studies are comparatively rare in health research, although they have been identified as particularly useful when exploring conditions of change such as might occur in those with long-term and progressive illnesses, as in this study. 52

Methods and data collection tools

Qualitative semistructured interviews were conducted to explore the experiences of participants in the intervention arm of the PDSAFE trial; they were chosen for their capacity to generate rich, in-depth data about the social realities of participants, in context. 53

Two qualitative interviews were carried out with people from a subset of those who participated in the PDSAFE intervention arm of the trial (see the following section). The initial interview took place after randomisation [time 1 (T1)], but prior to commencement of the intervention. It focused on the impact of Parkinson’s and the participants’ expectations of the intervention. The second interview was carried out 6 months later (T2), and elicited information about participation in the intervention arm of the trial (Table 12). Because the focus was on exploring participants’ experiences of the intervention, the control group were not included in the qualitative substudy.

Initial interview schedules were informed by scoping the relevant literature, which was carried out by referring to study protocols for primary questions of interest. Interview schedules were developed iteratively, in collaboration with the main trial team and stakeholders (including trained physiotherapists, rehabilitation experts and qualitative researchers), other clinical and academic experts (i.e. two qualitative researchers, two social networks experts) and two expert patients. Both interview guides were piloted with expert patients, to ensure that the questions made sense, and revised as necessary. The qualitative Research Fellow employed on the PDSAFE trial completed all the interviews at T1 and T2.

Sample and recruitment

Recruitment for the qualitative study took place from only the first four trial sites (Southampton, Portsmouth, Bournemouth and Poole, and Exeter), so that the qualitative Research Fellow had sufficient time to carry out both sets of interviews and complete analysis.

Participants who were eligible for the PDSAFE trial (see Chapter 2) were invited to ‘opt in’ to qualitative interviews at the time of recruitment to the main trial. Only those participants randomised to the intervention arm were approached to participate.

Theoretical sampling was used to represent the views of participants with a range of characteristics, and included:

• location (at least 10 people from each site)

• age (at least two people aged < 80 years and two aged > 80 years in each site)

• gender (a minimum of two men and two women per site)

• severity of disease (at least one participant per site with disease severity of H&Y scale stages 1, 2 or 3/4)

• number of falls [at least one person experiencing a single fall in the previous 12 months, one who had repeat falls (more than one fall and fewer than 10 falls in the last 12 months) and one who had multiple falls (more than 10 falls in the previous 12 months) per site]

• time since diagnosis [at least one person per site who had been diagnosed for (1) ≤ 5 years, (2) 6–10 years, (3) 11–15 years and (4) ≥ 16 years]

• living status (at least one person who lived alone per site).

Procedure

During the screening visit for the PDSAFE main trial, assessors asked participants if they would be interested in participating in the qualitative substudy, and left a qualitative study patient information sheet if participants expressed an interest. If a participant was randomised to the intervention arm, the qualitative Research Fellow telephoned them to answer any questions they had and to explore if they were still interested in participating in the qualitative substudy. Once their participation was confirmed, a convenient time for the first qualitative interview (before the treatment with the therapist commenced) was identified.

The Research Fellow then scheduled a second follow-up interview 6 months later. Just prior to the scheduled second interview, the researcher checked with the trial manager and administrator that the participant had not withdrawn from the trial for any reason and then telephoned the patient to confirm that the second scheduled interview was still convenient.

Ethics

Ethics approval and relevant NHS trust governance approvals to carry out this qualitative study were covered within the main trial approvals processes. All data were anonymised, and information about participants was kept separate from the data in locked filing cabinets. The qualitative Research Fellow did not contact potential participants until they had expressed an interest in taking part, and participants also had the opportunity to clarify any queries before confirming participation. Written informed consent was taken immediately before each interview.

Qualitative analysis

Interviews were transcribed verbatim and then checked for accuracy by a member of the research team who had carried out the data collection. Analysis of interview transcripts was carried out using inductive thematic analysis, in which dominant themes were identified through close examination of the data. 54,55 Interview transcripts were read and re-read to ensure a high level of familiarity with the data before line-by-line coding. A coding manual was then created to define emerging codes and themes for initial interviews at T1 and T2, and these codes were then applied to the remaining transcripts. The coding manual was developed iteratively and revised throughout the coding process to ensure that the codes adequately reflected the data. The coding manual was discussed and agreed by core members of the research team at various stages of the coding process.

All interviews were transcribed and analysed by the qualitative Research Fellow. In addition, two T1 and three T2 interview transcripts were independently coded by the senior investigator with qualitative expertise, and discussed with the Research Fellow to promote thorough, careful and reflexive practice in the development of the coding framework. 53

Participant characteristics

Forty-two participants were recruited at T1. Of these, 37 were interviewed at T2, although two of these participants had withdrawn from the main RCT. Consequently, 35 participants remained in the trial, and completed interviews at T1 and T2. Their demographic characteristics are shown in Table 13.

Participant characteristics in the qualitative sample were comparable to the overall trial sample, with the exception of time since diagnosis (the mean was slightly higher in the qualitative sample), number of falls in the previous 12 months (again, the mean was slightly higher in the qualitative sample) and living status (with slightly more participants in the qualitative sample living alone). The higher numbers of people within the last two categories reflect the theoretical sampling, in which we deliberately sampled for these characteristics. All transcripts were included in the thematic analysis.

Expectations of participants about the PDSAFE intervention

The first interview guide included a question enquiring about previous therapies, rehabilitation and experiences of exercise, in order to better understand participants’ expectations about the intervention. Almost all of the participants had experienced some form of physiotherapy in the past, although not necessarily in relation to their Parkinson’s. This previous experience of rehabilitation and exercises appeared to influence participants’ expectations about the intervention, with many anticipating that this would involve a programme of exercises:

Well, as I said, hopefully they will come up with a series of exercises that I can do, hopefully every day, or most days, which will keep me a bit mobile and stop me getting stiff. I dread the thought of being really really stiff and not being able to move and get around; I don’t want that to happen.

Participant with Parkinson’s (21/057)

(Throughout the report, the numbers at the end of the quotation sources indicate anonymised participant identifier/anonymised site identifier.) Some participants described in some detail what they expected, which included tailored exercises, stretching, freezing exercises and exercises for postural control, co-ordination, stiffness and muscle tone:

I’m hoping that they will give me some useful exercises that will deal with, well, the balance; now the question in my mind is whether they will go on from that and talk about also posture, which is something I desperately need and general muscle tone I suppose, so I’m sort of hoping that they will cover all three things.

Participant with Parkinson’s (43/017)

I anticipate that she will do things like check my balance and um, my mobility and coordination and give me exercises to help improve my balance.

Participant with Parkinson’s (54/040)

A few participants highlighted everyday tasks that they anticipated the programme might help with, such as getting dressed, and also wider benefits, such as improvement in confidence and helping to sustain social contacts:

Yes, so, that’s really why I put my name down [for participating in the trial], because I thought, well anything that can help me keep going as long as possible, because I have got one daughter in Dubai and one daughter in Germany, so, and no daughters in England [laughter] and they are not likely to be coming back, so ah, I want to keep going.

Participant with Parkinson’s (43/058)

Participants appeared to value the fact that the intervention was to be delivered at home, citing reasons such as convenience, problems travelling because of poor mobility and relief at not having to visit hospital or clinic sites, or park a car. One person highlighted that, although they were happy for the PDSAFE intervention to take place at home, they felt that appointments outside the home were important to fulfil the need to ‘broaden outlook’ on life.

Most participants talked about anticipated benefits of the PDSAFE intervention. These included general improved mobility, as well as enhanced functional and physical improvements, for example better postural control, stability, posture, co-ordination, muscle tone and pain control, help with FoG, maintenance or improvement of mobility and walking with more confidence:

Well, as I said, hopefully they will come up with a series of exercises that I can do, hopefully every day, or most days which will keep me a bit mobile and stop me getting stiff. I dread the thought of being really really stiff and not being able to move and get around; I don’t want that to happen.

Participant with Parkinson’s (21/057)

Is there anything that you are hoping that you might achieve from the therapy sessions?

Well, maybe that I can walk without, walk with more confidence possibly, and also I have tried using a stick before and I don’t like using a stick because I don’t think it helps your posture. My posture at the moment is pretty good and I’ve seen so many people with sticks, really bent over and I’d like to be taught how to use, if need be, a stick properly.

A few participants highlighted more general, functional tasks and a desire that partaking in the intervention would maintain or improve their general independence. The wish to help others by taking part in a trial was also another major motivator for participation in the study:

Anything that you are hoping it might help with?

Mainly co-ordination and manipulation of such things as my buttons. If I can remain independent so much the better.

OK, so is there anything you think it might help you achieve?

Helping others with Parkinson’s, if they can learn something about physio[therapy] and how it can help people, so that it helps others in the future.

OK, that’s important, anything you would like to achieve personally?

Cure Parkinson’s! I’m joking.

But it can’t do that, can it?! I hope it will help give me my independence back a bit, or some independence, some mobility, I miss my independence. Parkinson’s is so frustrating.

A small minority of people expressed feelings of anxiety or nervousness about the initial therapy sessions; some of this was related to perceived expectations about what it might involve and the intensity of the sessions. Others voiced that they had not previously had success with physiotherapy and so wondered how beneficial it might be:

I will be a bit nervous until I know quite the intensity of it, everybody’s a little bit aware you know that, I mean, what does intense mean?

Participant with Parkinson’s (21/049)

Perceived challenges or barriers to participation included concern about having a ‘bad day’ or an ‘off day’, including anxieties about medication:

I suppose one thing that might be worrying is if the physio[therapist] comes round to do some exercises and I am having an off spell ‘cause I have, you know, sometimes I’ve got a day that’s like that and it makes a difference; sometimes the drugs haven’t kicked in properly for whatever reason and it might be a down spell, an off spell, and I might sort of have to, you know, that might have to be taken into account.

Participant with Parkinson’s (21/039)

Some also had concerns about the time commitment, and expressed ambivalence about or dislike of physical activity:

I mean the thought of doing it every day for 6 months is a bit daunting and I said at the time, I am not absolutely sure I can promise that every day.

Participant with Parkinson’s (21/057)

And anything you’re not looking forward to at all?

Well, the, the further drain on my time and, you know, doing physical exercise is not my favourite thing.

At the time of the second interview, participants were asked if their expectations about the intervention had been fulfilled. The majority were not surprised by what they had been asked to do, and were pleased with the content of the PDSAFE intervention. Those few with unrealistic or inappropriate expectations acknowledged this at the second interview:

Did the physiotherapy meet your expectations, was it what you thought it would be?

Yes, I suppose generally there was not much else she could do really to make sure that everything I did was done correctly and that I understand what it was doing and why it helped.

I knew it would be personal to me but I didn’t expect it to be so personal, I know that sounds a bit ambiguous but she almost assessed how I was doing each time and geared it specifically to what I could and couldn’t do.

Participant with Parkinson’s (54/040)

A number of participants suggested that they had not fully appreciated the time and commitment the intervention programme would take, sometimes suggesting that the programme was onerous or a ‘bind’. This reflected anxieties and concerns from the first interviews about being able to find time in already busy schedules to complete the programme daily:

So it was not quite what you thought it was going to be . . .

No. It was alright in the winter when you, when I have got nothing to do, but when the vegetables started to come, it became a bit of a bind trying to find time to do it, and I wanted to pack it in.

The experiences and perceived impact of the PDSAFE intervention for participants

During follow-up interviews, participants were asked to describe the content of the PDSAFE intervention. All of the participants spoke about the exercises and strategies that they were encouraged to use. Most participants were able to describe these in detail, and described a range of exercises.

Most people found the level of the programme acceptable and usually recognised the progressive nature of the exercises over time:

They got harder [laughs] . . . The expectations were greater, um, for example I could just about put one foot in front of the other and walk the line, like you do if you are drunk, if I did it quickly; so then the next stage was to do it slowly and it is much more difficult and so it was graded.

Participant with Parkinson’s (43/058)

Participants also described a variety of strategies that they had been taught as part of the intervention, including swaying from side to side to promote motion, not cutting corners, walking with one foot in front of the other, using wider steps so as not to shuffle, having feet further apart to prevent falling and picking up feet.

Did the physio[therapist] give you any sort of strategies to use or hints?

Well, if I ground to a halt, because I have been freezing a little bit, just to psych meself up and to count from one to four and then I could go [i.e. move] on the four.

Most participants commented on the expectation that they would engage with the programme daily, or as close to this as possible, and tried to conform to this, although this appeared to be challenging for some, particularly if they had other commitments on the same day:

The idea is to do the exercises every day, she said, if you could do them three or four times a week that would be sufficient, not ideal, but it would be sufficient, so I did try.

Participant with Parkinson’s (21/057)

The PDSAFE intervention involved the use of a range of equipment, including a weighted vest or jacket, the step, the exercise folder and log, a metronome and/or a DVD. Some recognised that the intensity of the exercises increased with the use of equipment:

I would do the exercise, I’d be out the garden doing it out there during the summer and using the step for the legs . . . there in the sunshine, and then I got a weighted jacket, which was harder, more challenging.

Participant with Parkinson’s (21/039)

Experiences of using the equipment were mixed, with many participants commenting particularly on the weighted vest:

I was doing stepping on that [i.e. the step] and that was fine, um, and then she put little feet on it to make it harder, but every time she came, she made it a little bit more difficult and the real shock was when she came with the weighted vest. Have you seen the weighted vest? It’s a torture chamber, honestly, so I picked it up and I said ‘Am I supposed to be wearing this? . . . I can hardly lift it’ and it actually weighed 14 pounds and I had to really sort of struggle to get it on and then do it up and I had to do the stepping with this vest on. Well, I couldn’t believe it, I said ‘Are you sure this is doing me good?’ and she said ‘Yes, [laughter] it definitely is’ and then after I had done that for a few weeks she wanted to put some more weights in it. I said ‘Look I am sorry, but this is really as much as I can cope with’.

Participant with Parkinson’s (21/057)

Although views about use of the equipment were mixed, a small number of participants were disappointed that the vests, steps and folders had to be returned once the physiotherapy visits ceased.

Several people mentioned that the filming and resulting DVD helped them to understand issues with their gait or posture:

There was one of me, my posture, which was, I did not mean to be alarmed because I knew my posture was awful. I did try, as it is one of the things she has helped me with quite a lot actually is the posture and, erm, I suppose it has improved a bit. Well anyway, the DVD was helpful in seeing how awful it was.

Participant with Parkinson’s (21/026)

However, more commonly, participants disliked, or experienced problems with the technology. Some felt that the DVD/metronome was unnecessary or did not help them, some did not use it and some preferred to do exercises to counting or using the beat of music. Some participants described technical problems with the DVD or would have preferred to see the programme on a portable device such as a tablet.

Follow-up interviews also explored the perceived impact and benefits of the PDSAFE intervention for those taking part. The majority of participants felt that they had made some progress in terms of mobility (including improved posture, postural control, functional strength, walking better, falling less frequently and/or enhanced control of freezing symptoms). Some also described improvements in functional activities:

[Laughs] I thought ‘what have I let myself in for?’ initially, but once you start doing the exercises and you know you can do them, you know, like putting my socks on was very difficult and that’s better now, stretching, that helped a lot.

Participant with Parkinson’s (54/074)

In follow-up interviews, over one-third of participants reflected that either their falls had reduced or they were better able to recognise falls triggers since embarking on the PDSAFE intervention:

I think they [i.e. falls] are less frequent because I am thinking in my mind all the time, walk sensibly, open this door sensibly, do this sensibly, because otherwise you are going to fall.

Participant with Parkinson’s (32/014)

Other perceived benefits of the PDSAFE intervention included increased awareness of limitations, for example recognising the need to plan things carefully, increased awareness of safety and recognising poor posture. Linked to this was enhanced recognition of the need to ‘slow down’ and ‘not rush(ing) around’:

I think the most useful thing, probably, is that I have had to learn [pause], to slow down, . . . I mean before I would just dash everywhere, not run I don’t mean, but you know just sort of walk and then I would get my feet in a muddle and you know all that sort of stuff and I think that is probably the thing I have learned most, to control that so that, you know, if I am in a hurry I can still function without fear of tripping.

Participant with Parkinson’s (43/050)

Around half the participants interviewed at T2 also spoke of increased confidence, independence and general well-being:

It made you feel you could do it independent. It was a good help.

Participant with Parkinson’s (32/054)

I feel more confident.

Participant with Parkinson’s (43/027)

Although most participants interviewed at follow-up perceived some improvements, physically, functionally, in falls reduction or more generally (as described above), a small number felt that they had not experienced any benefit from participation in the PDSAFE intervention. Several of these participants simply questioned whether or not the intervention could make any difference to them in the face of a deteriorating condition, even if they felt that the programme was, in itself, good.

Barriers to participating in the PDSAFE intervention

During the follow-up interviews at T2, participants were asked for their views about the barriers to and facilitators of taking part in the PDSAFE intervention. The time commitment to complete the exercise programme was identified as one of the major barriers to continued participation, particularly towards the end of the programme and beyond. Participants often described the programme as being fairly time-consuming or requiring a big commitment. Although for some this was not an problem, others found this challenging, particularly if they already led busy lives, which seemed to be a particular issue for those whose Parkinson’s was not so advanced:

There was nothing difficult about it; the only difficulty about it was me staying on target with the exercises between the visits. I found, because I lead a fairly active life anyhow, I couldn’t always find the time to get to do the regular exercises.

Participant with Parkinson’s (43/055)

A lack of motivation was also commonly offered as the reason for giving up the exercise programme, particularly when visits from the treating physiotherapist became less frequent or stopped altogether:

It would be trying to do it every day [is the most difficult part]. But that day, I know that if I do it I would be better and more flexible and more co-ordinated but it is just motivating myself to do it sometimes. It’s the hardest, that’s . . . it’s the hardest thing to do.

Participant with Parkinson’s (54/040)

However, not all participants ceased the programme at the end of the visits from the physiotherapists (see the following section). Another common barrier alluded to previously, was use of the programme equipment or technology. The two items most often mentioned as problematic were the weighted vest and the DVD with metronome.

For some people, illness, injury and other life events, such as a family death, prevented them from participating fully or at all in the programme. Illnesses and injuries mentioned during the interviews as barriers included problems with knee and shoulder, heart attack, shingles, pain, cancer and high blood pressure. Difficult life events included bereavements and family illness:

I had one week off in March when I was away for the week and then in the beginning of June my husband died and that threw us all out completely.

Participant with Parkinson’s (32/014)

Participants also described their experience of ‘off’ or ‘bad’ days as another barrier. These days were characterised by difficulty in engaging with the programme, particularly at times when medication was not optimised, participants felt particularly ill or symptoms were at their worst:

And sometimes, you know, with Parkinson’s you have your off periods where your drugs haven’t kicked in and there are times when I am having an off evening and I try and do my exercises, but the off spell seemed to be longer than normal, so I might have to miss it then, so it’s just choosing the time of day.

Participant with Parkinson’s (21/039)

Finally, a couple of barriers mentioned by just a few people included fear and forgetting to engage with the programme.

Facilitators of continuing with the PDSAFE intervention

In the same way that a reduction in and eventual cessation of visits from the treating physiotherapist acted as a barrier to continued participation in the programme, the knowledge that the physiotherapist would be visiting again acted as a facilitator of participation for participants. Interviewees commented that they tried harder when they knew that the physiotherapist would be returning and expressed the view that they did not want to disappoint her. Some mentioned the importance of reinforcement for their efforts provided by the therapist, and looked forward to her visits. A few suggested that they engaged in the exercises only when the therapist was present. Even the occasional participant who did not feel that they gained a lot from the programme appeared to value the therapist’s presence:

Also, having the physiotherapist visiting regularly, that gives you something to aim for, like, when you know that she is coming next week and you want to make sure that everything is up to date and that you are doing the exercises properly, so having that sort of check and balance if you like helps to keep . . . and it ran very smoothly.

Carer (21/056)

Support and encouragement from one’s partner, spouse or carer was another important facilitator. This could take several forms, for example the partner or carer completing the exercises alongside the participant, or reminding the participant to complete them:

And it was nice as well because she included me, it was really nice because obviously I would do the exercises with him, but she didn’t make me feel that I shouldn’t be there and she included me, which I thought was really nice. I thought at first did she just want me to go out of the way and think I was interfering, but no she was really good. Obviously, I like to know what is happening, because when I do them with him I would like to know what she is doing. That is quite good.

Spouse (43/059)

Another facilitator mentioned by participants was the written information about the PDSAFE programme provided in a blue folder to those taking part in the intervention arm of the trial:

Well, on every session, whenever she came, we started straight away with the list of exercises which she gave me, which she showed me . . . I had a list, a chart that she gave me, so I could tick off each one as I did it.

Participant with Parkinson’s (43/050)

For some, particularly those with already busy lives, the intervention was easier to engage with if integrated into their daily lives. Participants found novel ways of building the exercises into their daily routines, for example doing them while completing chores, gardening or while in the kitchen. Others would integrate exercises, for example walking or stepping exercises, into leisure activities such as dog walking, playing golf or simply walking into town, or adapted exercises to their environment:

I did them in the morning. I did them while the kettle was boiling. Once you knew what to do it was quite simplistic so I would do my marching you know, and then I would go and put the kettle on and I would do this bit and that and by the time I went out in the morning it had all been done.

Participant with Parkinson’s (43/058)

Other facilitators of the PDSAFE intervention included seeing direct improvements or feeling a sense of progress.

Views on the PDSAFE concept and study involvement

On a conceptual basis, therapists responded positively towards PDSAFE; the intervention was described as being holistic, comprehensive and inherently relevant to an individual’s problems and environment. The approach focusing on strategies to enable functional activity was seen as differing from the prevalent approach used in usual practice/the NHS, which was described as being exercise driven:

I thought it was very good, it takes a slightly different tack than a traditional NHS assessment because it’s sort of strategy driven rather than exercise driven so it’s potentially looking at a person at a functional level, trying to kind of really drill down when they are falling and why they are falling but at a much more of a global level and trying to really focus in on that, not address that problem through the provision of exercises but more actually looking at practising functional activities that are linking in with the falls.

Therapist 6

Having PDSAFE in the individual’s environment, ‘where the falls happen’, was seen by all the therapists as the ideal. Being in the person’s home, setting their environment and situation was considered to give much better insight into how to tailor treatment:

It had relevance to them, [. . .] and their problems, they demonstrated their problems by when you look at them and you map the house and you have a really good look at them. So the relevance for them is excellent, so they know that this is their problem and you start working towards trying to solve that problem or a couple of problems that they do have with stuff that is functional, that they need to do for their everyday lives.

Therapist 1

All therapists reported enjoying the experience. One therapist commented that the PDSAFE approach was so different that working with it was initially difficult. Another therapist noted at times feeling constrained/unsupported when patients were unwell and the links back to clinical services seemed less apparent:

I suppose initially actually not approaching it from a training base but approaching it from a strategy training in a functional side was quite hard to embrace to start with but now I’ve got the hang of it, I absolutely love it and I think it’s a really ingenuitive different way of actually looking at how therapists approach the treatments to their patients. I’ve really, really enjoyed it.

Therapist 3

Intervention challenges

There was a consensus that working with patients who had cognitive issues and/or dyskinesia was the most challenging aspect of the delivery of the PDSAFE intervention. This group were noted as challenging to support in usual practice, and one therapist noted needing to seek support during the intervention from a colleague to help their patient progress. This difficulty was attributed to patients with cognitive issues being unable to retain an understanding of the PDSAFE structure, rationale and content covered. Some therapists felt that this group required different support to PDSAFE. Others consider PDSAFE to be capable of helping this group if they had support from a significant other or carer:

We had a fair few that did pass the cognitive screens to be part of the trial but then it is very different to pass a screen like that and then actually be able to participate in the trial, [. . .] It was fine if they had a partner or someone that they were living with to support them, but if they didn’t, that was really hard and we had to use so many different strategies and took lots of our time to try and get them on board and get them familiar with the programme and sometimes it still wouldn’t really work.

Therapist 5

Promoting exercises was seen as more challenging among patients who were unused to exercise/deconditioned, were less physically able and were unwilling to change their routines/habits. There was also comment that many patients struggled to grasp how to progress their exercises:

I think it’s very easy to follow the sort of structured exercises and things but that carry-over into function is the biggest, I found the biggest hurdle and was starting to say to patients at the start of my treatment. I said you know there’s gonna be exercises which support the use of a strategy and the crux of it for you I felt was to have that transferred into function and you know and that’s where the real hard work is for you is to try and change habits, things that you do on a daily basis without really thinking. So I try and sort of prep them, yeah, there was resistance to that I think.

Therapist 2

More individually raised challenges included working with patients who had overly protective spouses/partners and fitting the required number of sessions in around seasonable periods/patients’ busy schedules:

With the higher-level people, only that they had busy lives and getting a schedule to fit in. [. . .] and several of them had really active social lives and travelled and went away and went to rugby matches and so fitting in all the appointments at the appropriate times was difficult.

Therapist 1

Ingredients of intervention success

Most therapists associated having minimal/no cognitive difficulties as a key element for intervention success. Cognitive issues were explained as potentially reducing motivation (owing to reduced understanding of intervention intent) and preventing carryover from treatment:

I think the whole thing worked very well as long as they had the cognitive capacity to understand how it all knitted together, which aided their motivation.

Therapist 4

This issue then limited a patient’s ability to progress. However, the belief that cognitive acuity was important for intervention effectiveness was not universal; Therapist 1 did not feel that any group benefited more than any other and reported that all of her patients progressed, including those with cognitive issues:

Everybody progressed, everybody really enjoyed it, all the people I had, even the ones with cognitive impairment were very compliant [. . .] whoever benefits are the ones that are motivated to exercise and I have to say that every participant I saw, mine and (two other therapists) were all motivated. I can’t say that anyone benefited more than anybody else.

Therapist 1

Those described as ‘motivated’ to exercise and to integrate PDSAFE into their routines were thought more likely to derive benefit. Those with mild to moderate Parkinson’s, who were functionally independent, were seen by some as ideal candidates in terms of potential to benefit. Finally, being concerned over one’s personal risk of falls was discussed explicitly by one therapist as important for facilitating the required lifestyle changes:

I think people who have had a significant enough experience of a fall to see why they would want to work on it, so I think sometimes when patients maybe in the trial had had one fall a very long time ago or whose symptoms were so mild that they couldn’t really engage so much with changing their daily habits to prevent them falling, the problem wasn’t really significant enough I suppose.

Therapist 2

Factors associated with reducing intervention effectiveness

In line with therapist comments on the limitations of the PDSAFE intervention, cognitive impairment and/or dyskinesia were noted as limiting intervention success:

There were issues regarding patients with cognitive impairment which made it that much more difficult. [. . .] Patients with dyskinesia which made it again very difficult for anyone who wasn’t really used to cope with some of those issues but I think for the . . . mainly for the person who isn’t that severely affected, I think it had a better outcome.

Therapist 4

One therapist commented that PDSAFE could then become a burden to the patients who, on being given a programme to work on, could not obtain sufficient benefit. Support was noted as a key issue, particularly for those with cognitive problems; patients with no support reportedly struggled to complete all exercises and, as a result, could only partially engage in PDSAFE. The time since a patient’s last fall/the significance of the last fall to the patient was suggested as influencing intervention effectiveness. As PDSAFE was recognised as requiring changes to routines, some therapists commented that, if falling was not a big concern, patients may be less likely to alter these routines.

The extent to which a patient felt able to cope with new commitments was also identified as influential; those with multiple conditions and accompanying appointments, or busy lives generally, were noted as being more likely to withdraw from the treatment because of an inability to take on the substantial commitment posed by PDSAFE:

Where people haven’t progressed is where you’ve got some underlying cognition problems, whether they forget the appointments, where you turn up and they are out and where they just don’t get the intensity, you can’t build them up to that intensity and they have to commit like 6 days or up to 6 days, 5–6 days they have to be doing this for half an hour, so that is quite a commitment to them and so you need to be motivated, you need to be knowing what you’re doing, why you’re doing it and what benefit you are going to get from it and if you haven’t got that type of person, it’s a hard slog.

Therapist 6

Those patients that are so severe that they’ve got so much going on in their life and they’ve got different appointments to attend, they’ve got sort of other illnesses, they’re so tired, they are falling quite frequently and for want of a better phrase ‘in a pickle’ basically. The patients who have got so much on, the idea of doing these, you know, looking at this blue folder and using the blue folder to do anything extra on top of everything else that is going on is just too much for them really. It was quite a struggle and you would go and you would feel like it was just one thing too many at that time.

Therapist 2

Other factors reported as reducing intervention effectiveness included the following:

• patient attitude/motivation (patients who embraced illness and inactivity were seen as less likely to engage/benefit from PDSAFE)

• physical ability (e.g. those with walking aids participated at a lower level, which limited the benefit they could achieve)

• medication (side effects in some in cases hindered participation).

Strengths associated with the PDSAFE intervention

Several strengths were attributed to the PDSAFE intervention. The focus of PDSAFE on an individual’s functional problems and issues, in their own environment, was seen as a core strength of the programme. One therapist overtly related this patient-centred focus and approach to building patients’ confidence in their functional abilities:

It boosts a lot of people’s confidence, especially the fact that you can take them anywhere, take them outside, [. . .] I did get someone on the beach [. . .] she said to me that her friends had commented periodically, how much better her gait pattern was and how much more confident she was and she felt more confident.

Therapist 1

A number of therapists commented on how helpful the PDSAFE programme structure was not only for them, but also for their patients. This structure was described as being extremely clear and different from usual practice in that there was no ‘wait and see’ element:

The fact that you had very, very clear guidelines to work towards and I thought it was a clear and easy-to-follow intervention programme, [. . .] you had a clear progression and a clear way to lead but you also had a huge amount as a therapist of decision-making yourself and using your clinical reasoning to come to the decisions of what strategy you chose or exercises you chose, so again, it didn’t dumb you down.

Therapist 3

The opportunity afforded by PDSAFE to reinforce functional strategies was reportedly appreciated by therapists. The range of strategies available was also considered to be particularly good; this was seen as giving therapists the ability to respond flexibly and creatively to their patients’ individual needs:

It was a good variety and, from a strategy point of view, you could really open it up and could treat in many different sorts of ways out and about and in their own home and doing anything that they find particularly difficult. So it was quite open at that stage.

Therapist 6

Limitations associated with the PDSAFE intervention

Therapists described a number of limitations associated with the PDSAFE intervention. The three most commonly discussed related to the following:

1. PDSAFE was seen by a number of therapists as less suited to patients with cognitive issues and/or dyskinesia.

2. Some therapists reported that progressions posed ceiling and/or floor effects on exercises, namely exercises could not be made simpler than the lowest-level exercise (even though in usual practice it could be adapted) and, similarly, exercises could not be made more challenging than the uppermost exercise progression.

3. Therapists also reported that being unable to deviate from the set exercise positions inhibited their freedom to clinically reason and adapt the exercise to suit an individual, for example if a patient had a comorbidity, such as knee arthritis, that prevented them from adopting certain positions:

I probably would have treated them in a different position, so somebody for example, if they are, if they’ve got lots of arthritis or if they’ve got any other comorbidities, it’s difficult to then start loading joints if they are a bit arthritic in order to then strengthen. Previously, I would have then maybe looked at strengthening up in lying or on their tummy so I would have done it slightly differently [. . .] rather than on a protocol. I could only, for example, strengthen people up in, say, standing and obviously if you’ve got an arthritic knee or hip, it’s actually quite difficult. So there were some limitations with it but from an exercise point of view.

Therapist 6

Consequently, therapists reported that some patients remained ‘stuck’ at a certain point in the programme. Both of these issues led to therapists suggesting ways to improve the PDSAFE intervention for future implementation (see Theme 5: PDSAFE intervention usability and transferability). Other limitations concerned (1) exercises becoming repetitive for some patients and (2) some progression (specifically involving the metronome) promoted dual tasking at the expense of movement quality:

I definitely think the exercises, I want to be a bit more sort of creative with the patient and I think sometimes they potentially started to find things quite repetitive (Interviewer: yeah, yeah) so I probably would want to involve other exercise ideas, maybe to make them a bit more fun as well. That’s not necessarily a PDSAFE criticism, it’s a general, a lot of exercises we described are often quite boring.

Therapist 2

Yeah I think the variety of exercises and the fact that there is a lower limit and an upper limit to what you can do so you can’t progress further than a certain point. You can’t strip the exercise back past a certain point as well, I suppose.

Therapist 2

Perceptions of patient experience

Most therapists perceived that a majority of patients enjoyed the PDSAFE experience, reporting patients to be engaged and ‘on board with it’. However, one therapist noted occasions when patients had not seemed to enjoy the overall experience, for example when coexisting musculoskeletal conditions were exacerbated by PDSAFE activities or when patients were not open to changing their routines:

I think most people were on board with it [. . .] I don’t think I had any dropouts at all or not due to not wanting to do it. It was they were withdrawn because their diagnosis changed but most people were on board with the programme and wanted to do it and wanted to be part of the trial so that went well.

Therapist 5

There are a couple of patients who had musculoskeletal problems and had problems with pain who probably didn’t enjoy it as much just because they associated movement with discomfort.

Therapist 2

Therapists described a number of particular elements of PDSAFE that they believed patients to have most enjoyed. Three of the most commonly discussed of these were (1) the focus of the programme being on issues most relevant to the individual patient, (2) the clear structure attributed to the programme and (3) the intensity and level of input/contact with therapists to support activity. Other features reported as being positively received by patients related to the novel focus on strategy training, the chance to exercise, the opportunity to use equipment and the confidence regarding managing a fall and the resultant feeling of increased safety developed within the programme:

I think it’s having somebody supervise them, I think it’s somebody coming in and they are not being left with a piece of paper, doing a list of exercises, they are actually having somebody come in and teach them something that they can relate to and for us it was falling but it might have also been getting up out of a chair, rolling over in bed, getting their food out of the boot of the car from shopping, you know.

Therapist 6

The intensity of the programme, that it was centred around them as an individual and the strategy training practice which was different to what they had received before.

Therapist 3

Aspects of the programme that were least enjoyed included (1) completing paperwork, for example falls diaries (although it is worth noting that these were a product of the research trial rather than the intervention itself), (2) reduced contact in later stages of the programme, 3) changing habits/the home environment and (4) exercise. Regarding falls diaries, some therapists attribute this to the tedium of completing paperwork whereas others felt that embarrassment was a factor. Regarding exercise, for some this was due to the difficulty of the exercises whereas for others, more physically able, exercises might become a bore:

It’s always about the documentation, the tick sheets – it’s always a recurring issue so we got to the 6 months’ point and the majority, the ones that had done well without any cognitive impairment, didn’t need the tick sheets, they think ‘I do this every day’ or ‘Monday, Wednesday, Friday is my days’, I’ve got a reference for the exercises and they were glad to see the back of that. Although it’s just a tick sheet and doesn’t take long, it becomes a bit of a nuisance doing it so that’s the thing that I found people moaned about more than anything else.

Therapist 4

I think they probably found it difficult when it went down to monthly and they were used to seeing you more regularly than that and I think probably filling in their forms, none of them liked filling in their fall forms [laughter] and I can understand why because you are focusing in on your . . . falling is . . . whenever I fall over, which thank god is not very often, but if I fall over in the street or trip over the dogs fairly frequently, it’s embarrassing, isn’t it? I really don’t want to dwell on that and I cover up my grazes and I kick my dog out of the way and I carry on my way that nobody sees. If you ask somebody constantly to write down their falls and reminisce about them, I think people found that difficult.

Therapist 1

Patient engagement in the PDSAFE intervention

Therapists perceived most patients to be motivated during the most intensely delivered points of PDSAFE. Patient engagement following intervention tapering was described as variable; some therapists perceived patient engagement to reduce when contact changed to monthly visits, others felt that their patients continued to engage through to programme completion. All therapists raised uncertainty and/or concern regarding the likelihood of their patients continuing the exercise programme. There was a feeling that some would continue, many would ‘ease off’ in terms of intensity/frequency of exercise and some would not continue. Although it was recognised that therapy contact could not continue indefinitely, it was considered that there was a need for some form of follow-on programme/group. Therapist reports of patient feedback support this also:

I did find that a few people, quite a few, lost motivation towards the end. So they started off quite motivated and then towards the end they would lose it so when I was seeing them at a less intense level, they definitely lost motivation and a few of them, I would say to them at the end, will you continue they would say – probably not just doing it themselves and be more likely to go to a group, some of them do tai chi and other exercise classes that are Parkinson’s related or non-Parkinson’s related and they would continue with that.

Therapist 5

They all said ‘is that it?’ so I think my feeling is that with exercise-based programmes you’ve just got to keep seeing people or move them on to something else.

Therapist 1

Evaluations of therapist training

Training was highly praised, the knowledge of staff was commended and, for the most part, therapists reported feeling supported:

Absolutely fantastic, loved every second of it, comprehensive. I don’t think anything could train you for the completely different way, obviously [name] did the training and then we did the case studies to make us try and think strategy based and I think until you are actually out there and actually doing it, the training couldn’t have been any better and the early fidelity checks were paramount, I think.

Therapist 3

The one instance in which a therapist felt unsupported initially related to part-time working patterns and difficulty synchronising schedules to arrange a telephone conversation. The monthly group discussions and masterclasses were valued as very supportive, inspirational and good sources of continuing professional development. The peer support/double visits were particularly valued as learning and support opportunities. Some therapists noted that although the training in class was good, much learning occurred when implementing the programme, and peer support/double visits were vital to this. There were few negative comments; one therapist reported experiencing a large gap between receiving training and beginning the intervention, which resulted in the therapist struggling/feeling less confident when initially commencing treatment. Another comment was made regarding time spent working through paperwork (e.g. the treatment protocol, which detailed the treatment structure) during initial training, reducing time for practical learning:

The training was delivered and I then had a very long wait before I got my first participant. I can’t remember how long but I can’t remember if I’m honest, but it could have been over a month and I couldn’t . . . it was then difficult to really feel confident with delivering the protocol because you are new, you can’t totally remember, you read the paperwork and you read what’s given to you and you’ve got a reasonable idea but it’s all the conversations, it’s all the scenarios that was discussed so whether somebody could have come and did one . . . our first visit, a few of our first assessments, then that might have worked.

Therapist 6

Tool utility

Therapists commented on the utility of various tools, equipment and documentation used in the PDSAFE intervention. Some of these resources were praised by all participants as exceedingly useful. The use of one-to-one videos, whereby the therapist recorded a patient’s movement and then talked them through an assessment of this, was considered a very useful facility that was, for the most part, appreciated by a majority of patients. There were some therapists who noted some patients struggling to focus on the recorded movement, instead attending to their appearance:

The one thing I did find really useful was when you are doing a strategy with someone and you’ve got them doing a functional task and you video them and then give them immediate feedback, they can see themselves.

Therapist 1

The weighted jacket and the metronome received a more mixed review. The participants reportedly found it very difficult to put on the weighted jacket themselves; some patients found this a considerable struggle and required assistance. Concern was raised regarding being able to use the jacket if people did not have significant others/carers to help. The weighted jacket was also seen as (1) having potential to aggravate musculoskeletal issues and (2) being intimidating/off-putting, especially for those who were unused to exercise/weight training. It was not only patients, but sometimes also partners/spouses, who were intimidated by the jacket:

I would say there was an element of them trying to protect their partner a little bit and not want them to do too much, so things like the weighted vest, even if the participant themselves were quite keen on using it, their wives were not.

Therapist 5

However, some therapists reported that with practice donning the jacket became easier, and some patients responded well/enjoyed wearing it:

I think sometimes with the strengthening part of the programme they may have been quite daunted by the thought of wearing a vest, especially some of the females. The actual, the realisation that they could actually do far more than they ever thought that they were capable of that was really encouraging.

Therapist 4

The metronome was seen as being disliked by many patients because of the associated difficulty of dual tasking, although some therapists reported patients enjoying working with this. There was some suggestion that trying to maintain speed could reduce quality of movement (see Theme 2: benefits and limitations of the PDSAFE intervention).

The pre-recorded DVDs were unanimously reported to be used infrequently by patients. A number of therapists noted technical difficulties associated with this, and reasoned that the effort of setting up the DVD was not justified by the benefit patients derived from watching it. Another explanation was that some patients did not feel the need to remind themselves how to perform the exercises. Therapists noted being surprised by this finding.

Suggestions for improving the PDSAFE intervention

Increasing the number of exercises in the exercise library was a common suggestion/request, as was allowing more flexibility in modifying exercise progressions and postural sets to aid clinical reasoning and individual intervention tailoring. Other suggested improvements included the following:

• Increase the choice of functional activities and strategies.

• Allow a more flexible blue folder composition so that it can be simplified for people with cognitive issues. Remove old cards when new ones are issued to reduce confusion.

• Emphasise video feedback one-to-one sessions with patients and therapists over DVD usage.

• For exercises, consider having A4 laminated sheets with pictures and few words that can be put up on a patient’s wall.

• Consider spreading the first two sessions over four sessions, as there is a lot to assimilate.

• Make computer versions of paperwork more readily available for patients who struggle to complete forms in hardcopy.

• Add patient demo to PDSAFE training to help therapists understand the flexibility of the programme and how things may be adapted if initial choices fail.

• Potentially encourage more ‘double’/fidelity sessions and sooner in the programme as these are valuable learning opportunities.

Feasibility and desirability of integrating the PDSAFE in clinical practice

Therapists felt that PDSAFE could be delivered in the community. Although it was recognised that aspects of the programme would work in a hospital setting (e.g. exercises) and that group exercise particularly could be motivational, devoid of vital context, hospital delivery was still seen as a less desirable setting. The need for context in order for patients to learn strategies and attain carry-over was stressed by all. For some, delivering the programme in a hospital, divorced of this context, was possible but less optimal/effective, whereas others considered it infeasible.

There was recognition that PDSAFE, as delivered, amounted to double the therapy sessions usually delivered as part of routine care. This led to therapist uncertainty regarding whether or not PDSAFE would be funded adequately in the current climate; as a result, this issue was noted as being decisive when evaluating the feasibility of delivery of PDSAFE in clinical practice:

I guess it comes down to funding anyway, if the funding is there to deliver it as it is, then yes I imagine that it would be hard to get, just because of the time involved and it’s not just even the intensity of the programme, it’s all that extra stuff in terms of the admin[istration] and the [tele]phone calls to arrange the appointments and things and the yeah, whether it was put into a community service for people with Parkinson’s as a – you are seeing them anyway but we want you to deliver it like this rather than what you would normally do, then yes, I guess, but you would have to have that funding there.

Therapist 5

There was discussion regarding who should be involved in the delivery of PDSAFE; there was a feeling that a qualified physiotherapist would be needed to assess, develop the treatment programme and progress patients, but that therapy assistants could take on some sessions to check in and support patients when they were implementing their personal programmes.

Resource use

In line with NICE economic evaluation recommendations, this CEA adopted an NHS and Personal Social Services (PSS) perspective. 58 Data for CEA were collected at baseline and at 3, 6 and 12 months. Costs were evaluated in Great British pounds using a 2016 base year. All analyses were undertaken according to the principle of ITT and in Stata/SE version 12.0. Costs included intervention costs (apportioned per participant) and NHS and social care resource use costs. Intervention costs were collected directly from trial records. For resource use costs, unit costs were obtained from national sources: the Personal Social Services Research Unit’s (PSSRU’s) Unit Costs of Health and Social Care59,60 and NHS Reference Costs 2015 to 2016. 61 Quantities were collected from the trial participants. The resource use questionnaire collected information on (1) primary care services provided in the NHS system [e.g. general practitioner (GP) visits, physiotherapy, occupational therapy], (2) secondary care services provided in the NHS system [e.g. ambulance call outs, accident and emergency (A&E) attendances, hospital stay] and (3) social care services (e.g. home care visits, meals on wheels). Medication use was collected separately. All resource use data were originally collected from participants by interview with an assessor. This was revised after identifying a risk of unblinding of assessors during the interview by the question asking participants how many times they saw a physiotherapist. Therefore, the whole section containing the number of times a physiotherapist was seen was then collected via postal questionnaire, which was completed by participants. Questions on fall-related A&E attendances and hospitalisation admissions were also embedded into the falls diaries for validity checks. Unit costs in the financial year 2015/16 published by the PSSRU59 and NHS Reference Costs 2015 to 2016,61 when information was not available from the PSSRU, were attached to each item of resource use. Table 14 shows the unit costs applied to each resource use item and Table 15 shows details of assumptions varied in each sensitivity analysis.

Cost of personalised home-based physiotherapy

The cost of implementing the PDSAFE intervention was collected from the trial co-ordinator and lead therapist. The analysis followed an ITT principle; therefore, only the participants allocated to the intervention arm were assumed to incur the cost of treatment and any research-only cost was excluded. This ensures the relevance of the cost estimates if the programme were to be ‘rolled out’ to larger numbers of people.

Therapy session costs included the personnel costs (i.e. salary) of the therapist, travel costs, equipment costs and consumables. The salary cost of the therapist was calculated based on NHS band 6 full-time equivalent (£41,005.91) taken from the 2016 Health Service pay scale. 66 A 2-day training course was carried out with the involved therapists to train them in delivering the home-based personalised physiotherapy intervention. Cost of training included time spent by the trainer (lead therapist), room hire and training materials. During the delivery of the therapy sessions, all participants were provided with printed materials and compact discs (CDs) of the intervention demonstration. In addition, a proportion of participants used assisted equipment including weighted vests, balance pads and steps. This equipment was kept by the physiotherapists and could be used with different participants. Annuitisation was carried out to spread the costs over the anticipated lifespan of this equipment.

Analysis

A cost–utility analysis of the PDSAFE physiotherapy intervention compared with usual care was conducted after imputation. Differences in costs and QALYs between the two groups were estimated using generalised linear models (GLMs), which take into account the typically skewed nature of cost and QALY data. Histograms revealed that the cost data were right-skewed; therefore, gamma distribution was used. For QALYs, as suggested in a previous National Institute for Health Research Health Technology Assessment report,67 decrement of QALY was originally predicted in the regression using a gamma distribution. Decrement of QALY was calculated as the difference between the maximum QALYs that could possibly be accrued within the time frame and the actual QALY gained. However, this analysis cannot proceed with the imputed data; therefore, a Gaussian family was used instead given that the distribution of QALYs was close to a normal distribution. Covariates in the GLM were selected based on statistical significance using regressions with complete cases from the full list of demographics, medical history and screening measures. The chosen covariates included in the model were age, gender, H&Y scale stage, MoCA score, MMSE score, presence of diabetes, history of myocardial infarction, history of ischaemic heart disease, history of deep-brain stimulation and whether or not the participant had an informal carer. Baseline utility score and baseline cost were also included to adjust for any imbalance of utility between groups. 68

Mean costs and QALYs for each group were estimated using the method of recycled predictions. The ICER was estimated from the difference in cost and QALYs from the GLM regression. A 1000-iteration bootstrapping was conducted to investigate the uncertainty surrounding the ICER estimate and the probability that the intervention was cost-effective under a wide range of hypothetical thresholds (£0–200,000). These results were represented in a cost-effectiveness plane and cost-effectiveness acceptability curve (CEAC), respectively.

Utility and quality-adjusted life-year

Effectiveness was expressed as QALYs. QALYs were estimated by the area-under-the-curve approach using the utility index values at each data collection time point, which were generated using the EuroQol-5 Dimensions, three-level version (EQ-5D-3L), questionnaire, with preference weights from the UK general population. 69 EQ-5D-3L questionnaires were completed by participants during an interview conducted by assessors at baseline and at 3, 6 and 12 months. The EQ-5D-3L is a commonly used standardised generic preference-based quality of life (QoL) measure addressing five domains: (1) mobility, (2) self-care, (3) usual activities, (4) pain/discomfort and (5) anxiety/depression. 70 For each dimension, participants indicate the level of problems experienced on a three-level scale: (1) no problem, (2) some problems or (3) extreme problems. These responses were then converted to utility scores using the value set elicited from the UK general population. The EQ-5D-3L is the preferred measure of HRQoL in adults by NICE technology appraisal. 57 Participants who died had their utility score set to zero from the next assessment point after death. The area-under-the-curve method was then used to estimate the QALY score over the 12-month period, following the trapezium rule assuming a linear change in utility between each assessment time point.

Missing data and multiple imputation

Data were missing if participants did not return a questionnaire or returned an incomplete questionnaire. Multiple imputation was conducted at aggregate level for both cost and utility values to avoid convergence issues of the imputation model (the model becomes very statistically demanding when there are too many variables with missing data to predict, which usually leads to model convergence failure, meaning that the model cannot be executed). The EQ-5D-3L scores were imputed as utility values and cost items were imputed as categories, namely NHS cost excluding hospitalisation, hospitalisation cost and social care cost. To maximise the use of the completed data, a single item on resource use in both the interview and postal questionnaire was assumed to indicate that no resource was used for that item during that assessment period. 71 Hospitalisation data in falls diaries were checked and filled up the responses from the resource usage questionnaire if the response was missing. The aggregate-level missing data were imputed with multiple imputation with chained equations and predictive mean matching method for all imputed variables. The intervention and control groups were imputed separately. The multiple-imputed data sets were then used to estimate the difference in QALYs and costs, and the ICER.

Sensitivity analyses

A number of sensitivity analyses were undertaken to assess the robustness of the base-case results to alternative assumptions. These sensitivity analyses are summarised in Table 15.

The time horizon for the base-case analysis was 6 months, as all the participants completed their 6-month data collection and it was the primary end point. Nonetheless, for half of the participants, 12-month data were collected and the effect of intervention may last after the trial period; therefore, the time horizon was varied to 12 months. Sensitivity analyses were carried out on the number of physiotherapy sessions to gain an insight into the most cost-effective level of provision. Complete-case analysis was conducted assuming that data were missing completely at random to assess the impact of imputation strategies on the incremental cost and QALYs. Finally, the participants may have found it difficult to differentiate between the NHS routine physiotherapist and the trial physiotherapist; thus, the reported number of visits may not be accurate. Therefore, the use of an NHS physiotherapist between baseline and 6 months was excluded from the total cost in the sensitivity analysis.

Subgroup analysis

Following on from the statistical analysis plan outlined earlier, three subgroup analyses were undertaken to identify if the intervention was potentially more or less cost-effective than the overall group. The clinical effectiveness analysis found that the PDSAFE intervention reduced the number of falls in the participants who had scored in the middle level of the MDS-UPDRS (i.e. a score of 23–29), were not cognitively impaired (a score of ≥ 26 in the MoCA), did not have freezing symptoms at baseline and had two or three falls during the 12-month monitor period prior to recruitment. Therefore, the ICERs were estimated in the subgroups defined by the above criteria. Interaction variables of the treatment and the group indicator were added into the regression to obtain the incremental effects without splitting the sample. The incremental effects were generated using recycled prediction methods to balance the covariates.

Intervention cost

The PDSAFE intervention costs are detailed in Table 16. The cost per patient for each of the breakdown items is shown in Figure 8. The cost of the intervention was a main cost driver that made up a large portion of the difference between treatment arms. Within the cost of intervention, the therapist time and travel expenses account for > 95% of the total cost; thus, assumptions were varied around the number of sessions.

FIGURE 8.

Cost of PDSAFE home-based personalised physiotherapy intervention. CD, compact disc.

Missing data

The number and percentage of missing data for each collected cost item and EQ-5D-3L questions at the four assessment points are shown in Table 17 for the whole sample, and in Table 18 for the intervention and control (usual-care) groups, respectively. Baseline for the resource use refers to the 3-month monitor period prior to randomisation. Approximately half of the data were missing (primarily attributed to the termination of data collection) for the 12-month follow-up for both the resource use and the EQ-5D-3L data. For the 6-month follow-up, the proportion of data missing for variables collected via postal questionnaires (i.e. NHS resource use excluding hospitalisation) (≈28%) was almost twice that for variables collected through the face-to-face interview (≈15%). The proportions of data missing at each assessment were much higher (≈10% higher for all variables) in the PDSAFE intervention group than in the control group (see Table 18).

Resource use and costs

Table 19 reports the use of each resource item (mean number of visits, SD, median, minimum and maximum) accrued over the 6-month primary follow-up period in the intervention and the control groups. The total costs of NHS service use excluding hospitalisation and of hospitalisation and of social care services are presented in Table 20 (before and after the imputation) and Figure 9. Overall, there was no difference between the intervention and control groups in the cost of NHS and social care service use. The total 6-month cost of service use per patient was £3137 (95% CI £2602 to £3673) in the intervention group and £3069 (95% £2621 to £3518) in the control group. Figure 9 shows that resource use was slightly higher in the intervention group than in the control group (but the difference was not statistically significant); however, the total baseline cost (i.e. from 3 months pre baseline to baseline) was also higher in the intervention group. The most commonly used NHS services were GP, NHS physiotherapist, practice nurse, Parkinson’s nurse and outpatient clinic, with at least one visit per patient over the 6-month follow-up period. The intervention group was found to have fewer GP visits, practice nurse visits and home care help visits; none of the differences was statistically significant. The most frequently used social service was home care, of which, over the 6-month follow-up period, each participant in the intervention group had an average of 13 visits and each participant in the control group had an average of 30 visits. Medication was not included in the total costs because of the large number of prescriptions identified, the frequent low cost of medications, identified comparability across arms and poorly reported data.

FIGURE 9.

Mean cost of NHS and social care resources use per patient over the 6-month follow-up period.

Health-related quality of life

The EQ-5D-3L health utility values for each treatment group at baseline and at 3, 6 and 12 months are shown in Table 21 (complete case), Table 22 (after imputation) and visualised in Figure 10 (after imputation), assuming a linear change between each assessment point. The completeness of the EQ-5D-3L questionnaire at the 6-month follow-up was 79% (188/238) for the intervention group and 90% (213/236) for the control group. The utility values of the intervention group declined slightly less than the control group over the 6-month, as well as the 12-month, period. The differences in utility values between the arms were small (i.e. 0.031 at 6 months and 0.017 at 12 months) given that the full range of the EQ-5D-3L index score was –0.592 (worst) to 1 (full health), and not statistically significant.

FIGURE 10.

The EQ-5D-3L utility values at baseline and at 3, 6 and 12 months (after imputation).

Cost-effectiveness base-case analysis

The cost-effectiveness results for the PDSAFE home-based personalised physiotherapy intervention are presented in Table 23. Overall, the PDSAFE intervention is more costly within the 6-month time horizon but generates greater HRQoL than the control. The average cost per patient was £4020 (95% CI £3531 to £4510) in the intervention group and £3095 (95% CI £2694 to £3496) in the control group, an incremental cost of £925 (95% CI £428 to £1422). The average QALY gain was 0.34 (95% CI 0.326 to 0.345) QALYs in the intervention group and 0.328 (95% CI 0.319 to 0.337) QALYs in the control group, The intervention group had an incremental 0.008 (95% CI –0.006 to 0.021) QALY gain compared with the control group.

The cost-effectiveness plane for the base-case analysis is shown in Figure 11. The dyads come from the 1000 bootstrap iterations. The x-axis represents the bootstrapped incremental QALYs between the randomisation groups and the y-axis represents the incremental costs. All of the simulated cost–utility dyads were in the north quadrant, indicating that the intervention was always more expensive than the control. Similarly, the majority of the simulated cost–utility dyads were in the east quadrant, indicating that the intervention was very likely to improve health outcomes compared with the control. The flat oval shape of the cost-effectiveness plane indicates that there is a higher degree of uncertainty surrounding the estimates of the incremental QALYs than the incremental cost, leading to a wide 95% CI for the estimate of ICER.

FIGURE 11.

Cost-effectiveness plane representing 1000 bootstrapped cost difference and QALY difference pairs.

The CEAC for the base-case analysis is shown in Figure 12. The probability that the intervention was cost-effective at the £30,000 per QALY threshold was 0.5%. This corresponds to the very few dyads under the sand line in the cost-effectiveness plane in Figure 11.

FIGURE 12.

Cost-effectiveness acceptability curve.

Subgroup cost-effectiveness analysis

The results for the subgroups defined according to cognitive function, presence of freezing symptoms, MDS-UPDRS and retrospective falls at baseline are presented in Table 24. The results indicated that the ICER was lower in two subgroups (although still above NICE’s upper threshold of £30,000): (1) participants who were cognitively impaired (MoCA score of ≤ 25) and (2) participants whose Parkinson’s was of moderate severity (MDS-UPDRS score of between 22 and 39). Compared with the control group, those in the cognitively impaired subgroup of the intervention arm (n = 184) consumed less resource (–£462.80, excluding the intervention cost) and achieved a similar QALY gain to the control group; therefore, the ICER for the PDSAFE intervention versus control was £39,486. In the subgroup including those participants with Parkinson’s moderate severity of (n = 161), the PDSAFE intervention was associated with an incremental cost of £500 and an incremental QALY (0.016) that was double the incremental QALY gain of the overall group, leading to an ICER of £30,731.

Although still above the £30,000 per QALY threshold, compared with the base-case analysis the ICER is lower in the following subgroups: participants with moderate cognitive function (£84,925 per QALY gained), participants who had not experienced freezing symptoms (£61,687 per QALY gained) and participants who had had at least two falls in the 12-month period prior to recruitment (£53,860 per QALY gained for participants who had two or three falls and £82,532 per QALY gained for participants who had more than three falls). In addition, the increase in QALYs gained in the PDSAFE intervention arm for the participants who had not experienced freezing episodes was statistically significant [0.021 (95% CI 0.0007 to 0.041) QALYs]. In contrast, the PDSAFE intervention was found to be less effective than usual care and more costly in two subgroups, the subgroup with more severe Parkinson’s (i.e. those with a MDS-UPDRS score of ≥ 39) and those who had experienced freezing episodes; therefore, usual care dominated PDSAFE in both of these subgroups.

Sensitivity analysis

The results of the sensitivity analysis are presented in Table 25. Although the PDSAFE intervention was cost-ineffective compared with the control in all sensitivity analysis, the ICER decreased by more than half when estimated over a 12-month time horizon with around half of the data imputed based on the observed data. This was due to the linearly interpolated increase in QALYs over the 12-month period.

Administrators

Ms Brenda Colwell, Mr Matt Saxby, Mr Nick Parton, Mr Tim Worth and Mrs Sue McDonald.

Royal Devon and Exeter NHS Foundation Trust

Mrs Emily Rogers, Dr Ray Sheridan (Principal Investigator), Dr Kathleen Reilly, Mrs Jane Hall, Mr Robert James, Mrs Clare O’Reilly, Ms Kathy Polverino and Mrs Gayle Githens-Mazer.

Royal Bournemouth and Christchurch Hospitals NHS Foundation Trust

Dr Khaled Amar (Principal Investigator), Ms Carly Fricker, Ms Laura Tucker and Ms Chantel Cox.

Hampshire Hospitals NHS Foundation Trust

Mrs Debbie Dellafera, Mrs Trisha Norman, Dr Sam Arianayagam (Principal Investigator) and Mrs Angie Dempster.

Portsmouth Hospitals NHS Trust

Mrs Katie Castle, Mrs Liz Ashdown, Dr Sarah Williams and Professor Helen C Roberts (Principal Investigator).

University Hospital Southampton NHS Foundation Trust

Mrs Veena Agarwal and Professor Helen C Roberts (Principal Investigator).

Poole Hospital NHS Foundation Trust

Mrs Meryl Goddard, Mrs Pippa Collins, Dr Ralph Gregory (Principal Investigator), Mrs Anneke Fox, Mrs Sarah Chessell and Mrs Lynette Corney.

Royal Cornwall Hospitals NHS Trust and Cornwall Partnership NHS Foundation Trust

Dr Christine Schofield (Principal Investigator), Ms Vicky Farrell Mrs Alex Gorree-Wery, Mrs Alison James and Mrs C Stone.

Livewell Southwest

Mrs Bernadette Coles, Mrs Rena Truscott, Dr Jane Stribley (Principal Investigator), Mrs Louise Johnson, Mrs Nicola Thompson, Ms Kyra Higgins and Ms Maria Stevenson.

Southern Health NHS Foundation Trust

Mrs Trisha Norman, Dr Gill Turner (Principal Investigator) and Dr Peter Phiri.

Institute of Neuroscience, Newcastle University

Mr Philip Brown, Mrs Heather Hunter, Ms Rosie Morris and Professor Lynn Rochester (Principal Investigator).

University of Southampton

Mrs Meryl Rowlands, Mrs Tracey McElwaine, Mrs Megan Liddiard and Mrs Vanessa Pressly.