Phase 3: paediatric early rehabilitation and mobilisation during intensive care feasibility study – part 2 process evaluation

Introduction

Alongside the PERMIT non-randomised implementation feasibility study, we undertook a process evaluation focused on the feasibility and acceptability of ERM and study processes. The implementation evaluation was informed by NPT. NPT identifies factors that promote and inhibit the routine incorporation of complex interventions into everyday practice. It also explains how these interventions work, looking not only at early implementation, but beyond this to the point where an intervention becomes so embedded into routine practice that it ‘disappears’ from view (i.e. it is normalised). (Text adapted from ‘Normalisation process theory: a framework for developing, evaluating and implementing complex interventions’ by Murray et al., under CC BY 4.0 licence http://creativecommons.org/licenses/by/4.0/).87

Study management

The work package was co-led by BRS, JMen and FK. The study management group provided input into protocol, ethics application design and data interpretation. Sati Sahota was trial co-ordinator. JMen led the process evaluation, with methodological expertise by TR. Qualitative analysis was performed by JMen, TR and Natalie Read.

Methods

We undertook a mixed-methods process evaluation across the three PICU sites. Evaluation of the implementation process was conducted through five key components:

1. The weekly debriefs with PERMIT champions from each of the three PICUs.
2. Survey of wider PICU healthcare professionals conducted at three time points at each of the three sites throughout the 5-month study period [normalisation measure development (NoMAD) survey – implementation evaluation tool104].
3. Qualitative feedback from interviews with healthcare professionals from all three PICUs.
4. Qualitative feedback from interviews with parents from all three PICUs.
5. Completion rates of the parent-completed questionnaires (to review feasibility and acceptability).

Sampling and recruitment

We invited two groups of health professionals to take part in feasibility and acceptability work: firstly, ERM champions, those people identified, as part of the PERMIT intervention, as leading the implementation of ERM in the unit and secondly PICU multidisciplinary health professionals involved in any aspect of preparing for or delivering ERM across the unit. Each site aimed to recruit 2–5 ERM champions to take part in weekly debrief conversations (20 debriefs per site; n = 60 total). PICU health professionals were invited to take part in a survey at three time points, aiming for 30 in total per site (n = 90) and consenting staff could also volunteer for a qualitative interview aiming for four or five participants per site (n = 12–15). Parents/legal guardians of CYP recruited to receive the PERMIT intervention were also asked about completing optional outcome measurement tools and an interview. This included those who consented for their child to participate in the PERMIT study (aiming for a sample size of 12–15) and those who declined their child’s participation (aiming for one participant per site, n = 3). The study received Health Research Authority approval (ref 21/SC/0127).

Data collection

Weekly debriefs were conducted with ERM champions from each site, either in person or remotely (telephone/video conference). They were recorded in contemporaneous fieldnotes. Qualitative interviews with PICU health professionals and parents were conducted in person or remotely (telephone/video conference/face to face) and were audio-recorded, intelligently transcribed.105 and edited to ensure respondents’ anonymity. Finally, PICU health professionals also completed a brief online survey at three time points (beginning, middle and end of study period).

Debrief conversations, interviews and survey with health professionals focused on exploring implementation progress, determinants and key learning; this happened throughout the study period. Interviews with parents focused on exploring acceptability and feasibility of PERMIT intervention and study processes; this happened towards the end of PICU admission or following PICU discharge (up to 30 days post admission to PICU).

Initial topic guides for debrief and health professionals and parent interviews were designed, drawing on prior literature, experience, NPT71,72 and TFA.73 They evolved during the course of data collection, allowing for tailoring and gradual integration of a variety of follow-up issues and topics. The survey was based on the NoMAD Questionnaire, a 20-item self-report instrument104 (see Appendix 3, Table 41 for NoMAD survey). The survey was distributed by the local PERMIT champion team through internal e-mail lists, and survey participants had the option to consent to be contacted about an interview with a member of the central PERMIT study team. To ensure both the surveys and interviews captured diverse views across the three units, local teams were encouraged to promote the survey widely to promote different professional groups to engage and participate (criterion-based recruitment).106 Although data saturation is aspired to within qualitative research, that is when no new themes are emerging,107 we recognised this would not be achievable within the time scales of this feasibility study.

Data analysis

Notes from the debriefs were imported into NVivo 12 and deductively coded and analysed using a qualitative content analysis approach of systematic coding and categorising.58 Two researchers (JM, NR) independently familiarised themselves with the data and conducted open-coding, utilising NVIVO™ software for data management. Codes were then discussed and organised into higher-order subcategories and categories.59,60 In relevant sections of the paper free-text quotes from respondents are reported to add clarity, staying close to the original meanings and context.61

The interviews were analysed using a thematic analysis approach. This approach was taken because thematic analysis allows for the identification of common threads that extend across an entire interview or set of interviews.108 It also values the detailed and nuanced account of data, particularly emphasising the context.60 Analysis was informed by the work of Braun and Clarke,109 with NVivo™ software used to assist in the organisation and coding of data. The Framework approach was then used to facilitate systematic, rigorous and transparent data management.110,111

The survey data were analysed descriptively. Three questions explored: (1) how familiar ERM felt, (2) if ERM was a ‘normal’ part of work and (3) if it will become a normal part of work – responses were on a scale of 0–10 (e.g. 0 very new up to 10 completely familiar). The remaining 20 questions required a Likert-scale response (strongly disagree, disagree, neutral, agree, strongly disagree). Responses were allocated a score −2 (strongly disagree) to +2 (strongly agree) for interpretation; median scores for each question are presented.

Results

Study participants

We undertook 48 debrief sessions with ERM champions (Table 23). Despite recruiting well to the champion role at each site, the debrief calls were attended consistently by the same people. Debrief attendees included nurses (n = 2), medical representatives (n = 3), physiotherapists (n = 2), OT (n = 1) and clinical research nurses (n = 3).

We received survey responses from 118 health professionals across the MDT. The largest response group (66/118; 56%) were registered nurses and overall most had been in PICU 3 years or longer (88/118; 75%) (Table 24).

We interviewed 13 health professionals, including medical representatives (n = 4), PICU physiotherapists (n = 3) and one physiotherapist from a different team (provided on-call cover), advanced nurse practitioners (ANP) (n = 2), one OT, one bedside registered nurse and one team leader registered nurse.

In total, 21 parents consented to participate with the optional questionnaires and the interview, with 15 parents interviewed. Of the six who did not participate, two subsequently declined the interview when contacted, one was only in the study for a few hours so did not feel they could offer any insight and the second child had been re-admitted to hospital so parents did not feel able to participate at the time. One child died before discharge from hospital and three parents did not respond when contacted, after a median of three attempts to contact them. One parent was interviewed but then withdrew therefore interviews from 14 parents were included in the analysis. Interviews took place between November and December 2021 and were conducted in a variety of ways, subject to parental preference. Two were conducted face to face at BCH, five took place on Zoom and seven took place over the telephone. The final analysis included mothers (n = 12) and fathers (n = 2). All of the parents had experience of their child having been intubated and mechanically ventilated during their admission to PICU, and 12 were parents of a child admitted as an unplanned admission.

Pre-paediatric early rehabilitation and mobilisation during intensive care context – questions of capacity

The PERMIT study was introduced into three PICU sites. Across all the sites, potential issues around capacity, especially in relation to levels of MDT staffing, as well as priorities, in relation to place of ERM in ongoing care, were present (Table 25). All sites reflected that although ERM was part of care, they ‘didn’t do it actively or in a structured way’ (Site Two: Medic interview). Patients with defined pathways – neurology patients, those classified as ‘long stay’, infants and patients who were elective and had a more predictable recovery – were more likely to receive ERM. Delivery did not often occur ‘early’ and was undertaken inconsistently:

quite limited in terms of what was being done on the unit. There were elements of, I guess, what you could classify as ERM starting to emerge, but I think the timing was definitely at a later point of the child’s stay.

In addition, it was more likely to occur on weekdays, in office hours, and interventions were likely to be ‘low level’ and often movement-focused. In Site 1, an ERM programme had been launched in 2019, with good leadership and good staff engagement. However, the programme had floundered – with maternity leave, staff turnover and service pressures – and in 2021 there was little awareness of the programme. The champions felt that the feasibility study would be an opportunity to reinvigorate an ERM programme and secure staff buy-in. Neither of the other two sites had a pre-existing ERM programme, although they both reported being very enthusiastic about the idea of commencing one.

Introducing PERMIT – (re)organising the work

Throughout the whole study period all three units experienced huge service demand, running significantly over capacity with reduced staffing. In addition, alongside the impact of COVID-19, there were unanticipated challenges or competing demands. At Site 1 this included the move of the whole PICU to a temporary location. The sites progressed through the first four phases at different paces (see Figure 15). Due to the prior history of an ERM programme, Site 1 progressed through initial phases quickly, with more time spent on Phase 3 – supporting staff engagement and training – and Phase 4 – adapting the manual and planning how to effectively deliver the ERM intervention and the trial at their site. Site 2, in contrast, spent more time on Phases 1 and 2. There was less focus on education and training and the specifics of where and how discussions about ERM took place, and more on the logistical factors associated with running the research study and capturing the data. Site 3 experienced a significant delay (9 weeks) in commencing the study, due to local site approval issues. This delay was not only frustrating for a site enthusiastic to start staff training and education, but also added pressure about the feasibility of recruiting 10 patients within an 11-week period. As a result, the site spent relatively little time in each of the phases, with the site starting to screen and recruit by week 16, only 1 week after Site 2 and only 3 weeks after Site 1.

To embed ERM, all sites realised – often more strongly in retrospect – the fundamental importance of ‘education, education, education! … I think that’s the key thing I think to try and make it work’ (Site One: Physio interview). Staff training and education were central to conveying information to all staff on the PICU, supporting them to understand the key principles of the ERM programme and build experience and confidence in delivering ERM. Ongoing education and teaching within formal contexts (e.g. organised study days), alongside informal, ad hoc moments (e.g. at bedside) and being ‘more opportunistic with teaching when clinical’ (Site Two: Debrief) were needed to support delivery of ERM. A refined, considered, educational strategy was introduced in Site 1:

Clear education plan, trainers are being trained and there is a training log planned to keep track of who has had what training.

This strategy focused on identifying what key information staff needed to understand, how many people needed training and a plan for how to train them.

There were challenges to providing the education and training sites that were felt to be required to achieve adequate staff commitment, confidence and skills. A core factor was the short trial period – 20 weeks from start to finish, less for Site 3, due to delays. At Site 2, someone commented that ‘the speed at which we had to put it in place and the speed at which we had to then recruit I don’t think gave us time to grow the roots enough’ (Site Two: ANP interview). At this site an additional challenge emerged, with educational resources having to be directed towards getting staff trained as quickly as possible on a new ventilator. Sites felt that the absence of any pre-prepared education and training materials accompanying the manual had been a challenge throughout the 5-month study period. As a result, Site 1 shared training material (which was then localised and rebranded) as well as access to a core trainer to support training and education at the other sites.

Delivering PERMIT – integrating early rehabilitation and mobilisation

Despite all the challenges associated with education and the speed with which the study was rolled out, PERMIT was seen by health professionals and parents as worthwhile (Table 26). All three sites reported positively about how the ERM programme cultivated multidisciplinary working, helped with clarification of roles and helped with elements about leadership and staff buy-in around ERM. MDT work was recognised as being central to delivering the desired level of ERM. This meant embracing multidisciplinary working to optimise service provision:

There are not enough hours in the day for a therapist to deliver the amount of ERM they’d like to be seeing going on. … It needs to become an MDT-based focus … everybody’s role. Therapists can’t be the only ones delivering this.

In particular, the role of the bedside nurse was recognised as vital to support delivery. The challenge was about raising the profile of ERM as a priority in nurses’ daily work so ERM was no longer seen as just a desirable component, but an essential aspect of care provision. At moments, people struggled with whether to prioritise supporting colleagues:

Nursing staff report feeling guilty that they’re doing nice things with patients – play, music etc. when the rest of the staff are very busy and struggling with unit demands.

However, exposure to ERM over time, especially seeing ERM in action on a unit, could help in the process of it becoming embedded as ‘standard care’.

Exposure was, in part, challenging, however, because there were differences across sites to the extent that PERMIT was rolled out. At Site 1, ERM following the PERMIT manual – that is assessment of acuity, activity planning, risk assessments and family involvement – was regarded as the standard for all patients. At the other sites, only patients recruited to PERMIT received ERM as per the manual. Although there was variety in the use of the bedside bundle, there was recognition of the value to support ERM delivery and a shift in the traditional mind-set of a patient needing to be stable to be eligible to receive ERM. In practice, exposure emerged in a variety of ways, be that through strategies in place to promote staff buy-in and enhance involvement over time:

there’s posters going up, conversation in the coffee room taking place. Physio colleagues are more active in reminding clinicians at the bed space, or at least reminding the nurses to ask doctors doing the ward rounds … lots of general awareness that early rehab and mobilisation is actually important for the patients.

In and through taking part in PERMIT, staff exposure to ERM increased over the study period. Ward rounds also offered a space in which to undertake ERM discussions, at the very least in relation to whether a patient may be eligible for the study, as well identifying activities of rehabilitation for those recruited, if not for others outside PERMIT. Additionally, positive family feedback about involvement with ERM, when shared with staff, increased exposure and understanding of potential impact.

Over time, health professionals understood that ERM was important for the physical and psychological recovery of the CYP, as well as the psychological well-being of parents/carers supporting their involvement in their child’s care. For example, one site noted that:

Parents are engaged and involved, really keen on the ERM process. Parents are filling out the activity booklet and adding lots of detail. Seems to give them a focus and a process to be involved in.

All the sites reported that PERMIT has supported parental involvement. Feedback from parents was reported as ‘overwhelmingly positive and everyone’s saying how important it was within their journey’ (Site Three: Physio interview). All parents commented that there had been benefits to participating. This seemed to span from the emphasis the programme placed on parental involvement to the purpose that the programme provided. However, parental involvement in activity selection varied, with some referring to paperwork to help guide choices, while for others it appeared to be more led by the health-care professionals, whereas other parents reported not seeing a list or feeling actively involved in activity selection. Engagement in process was clearly enabling for parents:

It was quite nice to know, “Oh we can move onto the next stage” or you know, “Oh look there’s a new activity on the list, oh I hadn’t thought about doing that activity with him”. Or realising that sitting in the bed is a milestone almost. You know, it’s a new activity. As a parent you can see that as a milestone of their recovery.

Overall, most of the activities were undertaken while their child was ‘in bed’. The most commonly reported activities were cuddles, reading, singing, talking, touch or massage and stretches, which matched recorded activities completed by the bedside staff. The impression was that most parents felt the activities were tailored and appropriate for the stage the child was at.

Many parents reported feeling more confident in becoming involved in supporting ERM, that for them ‘I think it was so empowering’ (Site One: Parent interview). Parents reflected on how sick their child had seemed to them initially, but once they were familiar with what ERM involved and how they could be involved it gave them a sense of purpose:

But it also gives us something to do to feel useful because you’re in a situation where it’s one to one care. They know exactly what they’re doing, we’ve got no idea. We can’t help do anything in ICU, but you’re just looking at your child and you can’t do anything. But being able to have a structure as to what we can do to help, all of the activities that we could do … it meant that we could do something to actively help his recovery.

Parents perceived that being recruited to the study provided guidance to staff about activity assessment and selection and also encouraged them to consider the child holistically. They also felt more confident leaving their child to have a much-needed break because they trusted that staff would also follow the programme. Some saw a direct benefit to their child. Others were less sure or queried whether it was fully relevant to their child because they were an infant, but still saw the study as worthwhile. For many the main positive impact for their child was the contact the programme encouraged with parents. This spanned all aspects of care, particularly cuddles. Parents recognised how central they were to their child and their presence was therefore important to help support their child. However, outside the context of PICU delivery, support and guidance for delivering ERM were lacking. Several families reported that leaving PICU is only the first stage in recovery and more needed to be done to support them with ERM beyond PICU discharge.

Delivering PERMIT – integrating trial processes

Having access to research delivery support was central to supporting recruitment, data collection and data entry. This reduced demand on bedside staff as well as freeing up the site team for education and ERM delivery. Sites 1 and 2 worked with research nurses – with Site 1 having access to a clinical research nurse from Stage 4 – whereas at Site 3 did not. At Site 3 the lead champion took on all the workload associated with implementing the study and evaluating ERM delivery and they noted that:

The research side is very time consuming. The data collection and research collection are the biggest barrier to doing the ERM, the data collection, the consent process, the questionnaires – it’s all so time-consuming.

All three sites commented on the large volume of work associated with conducting the PERMIT study. A research team presence, ideally to cover all days and times of the week, was seen as essential. All parents also provided positive feedback about the role and presence of the research team during recruitment: their professionalism, the sensitivity with the timing of discussions and their ability to respond to questions.

Overall, the families were positive about the study recruitment process. Two parents referred to the challenge of written leaflets for parents with additional needs. As ERM was rendered as the ‘usual’ standard of care within the units – by the PIS and research team – many parents had low expectations about the potential risk of participation:

My main question was about if there was anything extra that was going to be done, anything invasive, and if it wasn’t then I didn’t see any reason why I wouldn’t consent to it.

For the majority the perception of ‘minimal’ to ‘no’ risk meant that they found this a relatively simple decision. The choice of a staged consent model was significantly influenced by other studies within the PICU environment66,102 and the PPIE work conducted within the PERMIT study (see Chapter 8). Of the 30 children recruited to the trial, 21 of the parents/legal guardians consented to participate in the ‘additional’ aspects of the PERMIT study (70% consent rate). Seven families (33%) consented in a staged approach, with 14 (67%) happy to consent to all parts in one go. Some parents felt signing up in stages had been really helpful for them, not overloading them at a difficult time. There was also a sense from some families that because ERM was standard practice on a PICU that an opt-out consent approach where families did not have to actively consent or could be asked at a later stage could be acceptable.

Introducing and embedding the study paperwork was challenging at times. This was only produced as physical documents and Site 2 did not have time or capacity to adapt it to their electronic patient management system, so

[h]aving the documentation all on paper created resistance from staff, they like everything to be done on the computer. This was perceived as a lot more work to have to do.

Easy access to documents at the bedside helped support and sustain study work. Sites also had some concerns about the outcome measurement tools provided to families, in part tied to the logistics of managing the process – with uncertainty about the time points required and the challenge of getting these completed. They also were uncertain about the volume of questionnaires to be provided to families. However, on the whole the questionnaires were viewed as acceptable by parents. Negative comments reflected that some questions were inappropriate for their child in terms of their ability to do the developmental outcomes (mostly for the parents of neonates), length and the slightly repetitive nature of some questions.

Embedding ERM post-PERMIT and PERMIT trial – (continuing) questions of capacity

As part of Phase 4 work, a debrief was held 2 months after the study had closed to recruitment. Despite huge enthusiasm about the potential for ERM to continue after PERMIT closed, all ERM activity related to the programme had stopped at Sites 2 and 3. However, at Site 1, even in the contexts of increased PICU workloads and COVID creating challenges for staffing, they still felt that ERM had shifted to being seen as part of routine care. They noted how ‘the research has increased frequency and awareness of ERM’ (Site One: Debrief), with most patients now having their acuity assessed daily and staff considering what ERM could be conducted, even if it was not always possible to undertake the planned activities. Bedside nurses appeared to have increased confidence in initiating and conducting ERM, with increased awareness of what each MDT member could contribute to ERM.

Core issues to embedding ERM after PERMIT remained, with Site 3 reflecting that

the challenges have taught us so much! … Need increased presence and visibility of an ICU specific therapy team to fully integrate into ICU. We learned also that therapy doesn’t need to be so structured and planned, we can do it in so many ways.

They felt they would ‘review, revamp and re-launch!’ with more time factored in for the programme to embed, and that the ‘minimum time to make this really embed would be 12 months’ (Site Three: Debrief). All three sites felt that additional resources were required to support (ongoing) staff education and training; that visibility, experience and (positive) feedback around multidisciplinary ERM working was key, as well nursing staff taking ownership being central to embedding.

The NoMAD survey results align with the way that sites, over time, began to realise the complexity of the work of introducing and embedding an ERM programme. Between time point one and two there was an increase in ERM familiarity through initial engagement, with a realisation that ERM was currently part of normal practice and would become part of normal practice (see Appendix 3, Figure 20). However, by the final and third survey the level decreased back to baseline.

A similar temporal pattern was seen in four NPT items: around questions of the legitimacy of ERM being a part of their role (Q15); being open to working with colleagues in new ways to use ERM (Q16); and notably, continuing to support ERM (Q17) and that sufficient resources are available to support ERM (Q23) (see Appendix 3, Figure 21). However, across all time points there was strong (and stable) agreement about seeing the potential value of ERM in their work (Q13) and strong (and stable) disagreement across all time points that ERM disrupts working relationship (Q19).

Summary of findings to inform PERMIT study

Although there were many challenges to implementation of the PERMIT intervention within the context of busy PICUs during a global pandemic, there was significant multidisciplinary input and support from parents/carers for ERM to be standardised within clinical practice. A future trial was therefore viewed as acceptable and we identified a number of recommendations for a future trial (Tables 27, 28 and 29). Key messages include:

• Within a future trial, ERM needs to be embedded as standard routine practice for all CYP within the PICU. This would also enable an opt-out informed consent to be utilised and potentially reduce some of the challenges that parents outlined in relation to consent processes.
• Research delivery support is vital to successfully complete the trial processes, but the research team cannot lead the implementation of a programme or it will not fully embed.
• Bedside nurses are ideally placed to help deliver ERM but this workforce is already stretched. Currently ERM is not in their priorities and is seen as desirable, rather than essential. Adequately resourced PICU therapists will be required to support implementation and delivery.
• There is no spare capacity to deliver ‘extra’ ERM within current resources.
• ERM is seen as important for the physical and psychological recovery of the CYP, but is also important for the psychological well-being of parents/carers supporting their involvement in their child’s care.
• Being involved with ERM is seen as hugely important for parents/carers. Further consideration of how we provide training and support for parents to facilitate this involvement is needed.