1 INTRODUCTION

Globally, an estimated 2,500 limb amputations are undertaken daily (Moxey et al., 2011). Most of these amputations are of the lower limb and related to peripheral vascular disease, neuropathy and soft tissue injury secondary to diabetes mellitus (Moxey et al., 2011). In the next 10 years, there is a proportionate rise expected in lower limb amputations amongst people with diabetes and an associated need for quality rehabilitation (Moxey et al., 2011).

The rehabilitation process varies depending on the cause of amputation (traumatic or vascular), the mental and physical capabilities of the individual and the potential of the amputated leg to be fitted with a prosthesis (Fiedler et al., 2014). Regardless of whether they will be a prosthetic user, most Australian people experiencing lower limb amputations are prescribed a manual wheelchair, in the anticipation that it will support independence with mobility (Fiedler et al., 2014; Gupta & Kumar, 2019). Simply providing a wheelchair does not equate to its safe and functional use. Ability to use a wheelchair is dependent on age, confidence, pain, strength and endurance, as well as environmental factors (Fiedler et al., 2014). Education about using a wheelchair is pivotal for independence and safety (Sakakibara et al., 2013).

Many wheelchair users and their carers receive insufficient wheelchair skills training (Best et al., 2015; Kirby et al., 2020). Concerningly, more than half of experienced community dwelling wheelchair users have reported one wheelchair related accident over a 3-year period and 17% experienced two or more (Chen et al., 2011). Several factors contribute to inadequate wheelchair skills training, including limited time and resources of clinicians, limited length of stay, lack of training resources and low clinician confidence to demonstrate and teach (Kirby et al., 2020). Quantitative research suggests clinical confidence to teach wheelchair skills is improved with the use of a structured curriculum for wheelchair skills training and attendance at conferences, workshops, or in-service training. However, clinician perspectives about these strategies are largely unknown (Giesbrecht, Wilson, et al., 2015b).

Competent use of a wheelchair supports independence in activities of daily living and return to work, reduces reliance on carers/families, avoids admission to long term care facilities, and positively impacts stress, social interaction and economic engagement (Best et al., 2015; Kirby et al., 2020). It is expected that implementation of a wheelchair skills training programme for wheelchair users may support development of their confidence and decrease tips and falls (Keeler et al., 2019; Kirby et al., 2020; MacPhee et al., 2004). One Canadian programme that has a growing body of evidence, is the wheelchair skills training program (WSTP). This includes the assessment and teaching of 32 different skills broken down into indoor, community and advanced (Kirby et al., 2018b).

According to two systematic reviews, WSTP has a clinically meaningful effect on wheelchair users' skill performance in the short term (Keeler et al., 2019; Tu et al., 2017). However, many of the included studies involved powered and experienced wheelchair users (Best et al., 2005; MacPhee et al., 2004; Routhier et al., 2012). Of the 13 studies outlined in the most recent systematic review (Keeler et al., 2019), there were only three studies including people with amputations amongst a mixed participant group. Two studies that included people with amputations demonstrated statistically significant improvements in skill from pre to post training (Best et al., 2005; MacPhee et al., 2004), and the other study suggested an increase in skill particularly with community wheelchair skills (Routhier et al., 2012). Outcomes were not analysed depending on diagnostic group, therefore more research exploring the effectiveness of the WSTP specifically for people with amputations is warranted. The studies that have investigated wheelchair skill training have focused primarily on wheelchair performance and confidence outcomes, but have not considered qualitative perspectives, including users' experiences of training and perceptions on individual or group training formats.

While the WSTP supports individual and group training formats, it is less clear whether one mechanism is more efficacious than the other. Two papers have reported on the facilitation of skills in a group setting, but these occurred with experienced community wheelchair users, largely with spinal cord injuries (Best et al., 2016; Worobey et al., 2016). Given the success of group programs to support people pre and post amputation (Marzen-Groller & Bartman, 2005), clinician led group wheelchair skills training, that encourages support between group members was piloted in an inpatient rehabilitation setting in Australia.

The following research questions were explored through this pilot study: What are the outcomes for people with lower limb amputation participating in group WSTP? What are the perspectives of participants and facilitators about the process of the WSTP within the inpatient rehabilitation setting? 2 METHOD 2.1 Study design/ethics

A nested mixed method design (Creswell & Plano Clark, 2007) was used to collect pilot data across two phases (see Figure 1). This design is suitable for qualitatively evaluating the process of an intervention as well as quantitatively evaluating the outcomes. With this study design, the two types of data do not require integration during analysis and interpretation (Creswell & Plano Clark, 2007). A reporting guide for mixed methods was used to ensure the research occurred with rigour (Leech & Onwuegbuzie, 2010). The first phase involved the collection of quantitative baseline data for people with newly acquired lower limb amputations who were completing their rehabilitation at Hampstead Rehabilitation Centre. The second phase involved collection of quantitative post intervention data, and nested within this were individual semi-structured interviews with a sub-set of this sample. Within this phase the perspectives of group facilitators were also gathered through a focus group. The qualitative data were collected to augment the quantitative findings (Leech & Onwuegbuzie, 2010). Written consent was obtained from all WSTP participants and facilitators. The research was approved by Central Adelaide Local Health Network (CALHN) Human Research Ethics Committee and University of South Australia Human Research Ethics Committee Ref no. R20190129.

Participant recruitment strategy and overview of phases of data collection

2.2 Participant recruitment and sampling 2.2.1 Phase 1

Convenience sampling (Patton, 2002) was used to invite participation from people with newly acquired lower limb amputations who arrived on the rehabilitation ward between May 2019 and February 2020. To participate, they needed to be over 18 years old, have no aggressive behaviours, and use a self-propelled manual wheelchair. Participants were excluded if they were acutely unwell, had a cognitive impairment that required direct 1:1 supervision or based on functional observation did not have sufficient upper limb coordination/strength to propel a manual wheelchair. Based on historical numbers of suitable people passing through the ward, a sample size of 24 was anticipated.

2.2.2 Phase 2

After completion of the WSTP, a sub-set of participants were purposively invited to an interview (Patton, 2002). To seek maximum variation in this sample, participants were selected that reflected diversity in confidence scores and skill performance (Patton, 2002). Using convenience sampling, all WSTP facilitators were invited to a focus group.

2.3 Intervention

The intervention was based on the WSTP- Version 5.0. (Kirby et al., 2018b). The WSTP is validated and has structured teaching of skills for wheelchair use. For this pilot, only indoor and community skills were taught thus excluding advanced skills and ascending and descending a low kerb. This decision was based on the highly variable admission length and the short time frames available for building up strength and endurance of wheelchair users, many of whom had chronic health comorbidities. This reduction left 21 of the 32 skills in the WSTP (Table 1).

TABLE 1. Manual wheelchair skills in WSTP (Kirby et al., ) No. WSTP skill names included in this research Skill level No. WSTP skill names excluded in this research Skill level 1 Rolls forward short distance and stops Indoor 22 Ascends low curb Community 2 Rolls longer distance Community 23 Descends low curb Community 3 Rolls backward short distance and stops Indoor 24 Ascends high curb Advanced 4 Turns in place Indoor 25 Descends high curb Advanced 5 Turns while moving forward Indoor 26 Performs stationary wheelie Advanced 6 Turns while moving backward Indoor 27 Turns in place in wheelie position Advanced 7 Manoeuvres sideways Indoor 28 Rolls forward and backward in wheelie position Advanced 8 Picks objects from floor Indoor 29 Descends high curb in wheelie position Advanced 9 Relieves weight from buttocks Indoor 30 Descends steep incline in wheelie position and stops Advanced 10 Performs level transfers Indoor 31 Ascends stairs Advanced 11 Folds and unfolds wheelchair Community 32 Descends stairs Advanced 12 Performs ground transfers Community 13 Gets through hinged door Indoor 14 Ascends slight incline Community 15 Descends slight incline and stops Community 16 Ascends steep incline Community 17 Descends steep incline and stops Community 18 Rolls across side-slope Community 19 Rolls on soft surface Community 20 Gets over obstacle Community 21 Gets over gap Community 2.4 Study procedure

The WSTP was delivered for 45 minutes for three consecutive weeks facilitated by one or two occupational therapists or allied health assistants. Circuits for skill acquisition around the grounds at the inpatient rehabilitation centre were determined prior to the pilot. Each facilitator shadowed the first author (KC) for two sessions before running sessions on their own. The first author has more than 5 years' experience in teaching wheelchair skills but all other facilitators had no prior experience. Facilitators were provided with written copies and links to the WSTP webpage. The day and the time of the WSTP remained flexible to accommodate participant availability. Over the three sessions, indoor skills were consolidated before progressing to community skills. In circumstances where only two sessions were completed, all indoor skills were covered, but sometimes there was insufficient time to fully address community skills. While the intent was to deliver WSTP in groups, this was not always possible with some one-to-one sessions occurring based on the availability of programme participants on the ward at the time of running the sessions. Participants who received individual training were able to have higher frequency of skill practice due to the facilitator only having one person to spot and all were encouraged to practice their skill acquisition in their own time if the occupational therapists agreed they were safe. The WSTP was completed using a manual wheelchair provided by the rehabilitation centre until participants sourced their own wheelchair.

2.5 Data collection tools 2.5.1 Phase 1: Quantitative data collection

Following provision of consent, electronic health records were used to collect demographic data regarding: sex, age, weight, diagnosis, other comorbidities, living/social situation and time since amputation.

Wheelchair skills test—Questionnaire

The first 21 of the 32 skills outlined in the Wheelchair Skills Test—Questionnaire (WST-Q) (Version 5.0) (Kirby et al., 2018a) were completed by participants pre and post intervention, with the treating occupational therapist providing support if required (Table 1). The remaining 11 skills were removed from the scoring. Pre-testing was completed less than a week before the first WSTP session and post-testing was completed less than a week post completion of the WSTP. The WST-Q, is highly correlated with practical skill performance and is a self-reported tool focusing on performance (the ability to do an activity in the everyday setting), confidence (to complete skills safely and consistently in own environment) and frequency (how often the skills were completed in their environment) (Kirby et al., 2018a). The WST-Q provides a percentage score across three domains of capacity, confidence and performance. This tool was chosen over a practical test as it was less resource intensive and safe for participants prior to receiving their WTSP. The WST-Q has been previously used in the early stages of rehabilitation to prioritise which wheelchair skills to address (Kirby et al., 2018a).

Goal attainment scale

The Goal Attainment Scale (GAS) was used to measure achievement of occupation specific wheelchair goals through quantifying the meaning of goal achievement in specific and measurable terms. GAS is being increasingly used as an outcome measure in rehabilitative research (Turner-Stokes, 2009). Given the individual nature of goals, the GAS' psychometric qualities cannot validly be compared; however, it is suggested that the GAS has good content validity when used by experienced facilitators (Krasny-Pacini et al., 2013). In this case, all goal setting was facilitated by the first author (KC). Participants were asked to focus on one major goal which was measured pre and post WSTP.

Functional Independence measure

Functional Independence Measure (FIM) scores relating to mobility were captured to track wheelchair mobility changes pre and post WSTP (Linacre et al., 1994). The FIM is widely used within rehabilitation settings and offers excellent content validity across 18 items on a seven-point ordinal scale (Dodds et al., 1993).

2.5.2 Phase 2: Qualitative data collection People with amputations

Participants' perspectives about the WSTP were collected through semi structured interviews (Liamputtong, 2013) either face to face or by telephone within 2 weeks of completion of the WSTP. Due to limited availability of the independent research assistant, she only conducted one interview with the remaining seven interviews conducted by the first author (KC). A flexible topic guide was developed with seven open ended questions to gain participants overall perspectives of the WSTP, including areas of strength and improvement. The guide was developed by the second author (CM) and refined by the research team after piloting with four allied health professionals working with people with amputations (see supporting information).

Allied health professionals

All allied health professionals who had facilitated the WSTP participated in a focus group at the end of the 10 months of data collection. To reduce the introduction of bias, the first author (KC), was excluded from the focus group. The focus group was completed on the ward by the second author (CM), an experienced researcher external to the programme. A flexible focus group guide, encompassing seven questions was developed by the second author (CM) and refined by the research team. Facilitators were asked about their experience of the WSTP, and any perceived benefits, challenges and recommendations (see supporting information).

2.6 Data analysis

Descriptive statistics were used to analyse and report data from each domain of the WST-Q (capacity, confidence and performance), FIM and GAS into frequency distributions. Mean calculations with standard deviation were used to report data distribution and range (DePoy & Gitlin, 2016). Comparative analysis was not completed due to insufficient sample size.

A qualitative descriptive approach was used to explore the perceptions of participants and facilitators (Stanley & Nayar, 2014). All interview and focus group data were digitally recorded and professionally transcribed. Data were analysed using reflexive thematic analysis (Braun & Clarke, 2012). Initially, the interview and the focus group transcripts were analysed separately but were then synthesised into themes that triangulated the data from both participant groups. Analysis involved independent line by line coding of all transcripts by the first author and duplicate coding of some transcripts by the second author. Following this, the research team met to discuss the coding before the first author clustered codes into categories. Seventeen categories were formed from initial coding of the interviews and a further 22 from the focus group. Further engagement with data, including returning to the original transcripts, enabled the inductive synthesis of findings into four themes (Liamputtong, 2013). Any differences in interpretation amongst the research team during the analysis process were resolved through discussion. A reflective journal detailing the first authors' preconceptions and assumptions was kept and reviewed during data analysis to minimise bias in interpretation of the data. An audit trail was recorded to track analytical decisions (Liamputtong, 2013). Pseudonyms were allocated for data reporting.

3 RESULTS 3.1 Participants

Eighteen people were invited to participate in the WSTP, five declined and two dropped out (see Figure 1). Of the 11 attendees of the WSTP, not all attended over three consecutive weeks or completed sessions in a group format. Eight of these attendees participated in qualitative interviews. Interviews lasted an average of 7 min (range 3.34–17.31).

The 11 participants had newly acquired amputations (see Table 2), 10 due to vascular related complications and one through traumatic events, and all were new wheelchair users. They commenced the WSTP an average of 19.6 ± 8.3 days post lower limb amputation. There were two females, nine males with an average age of 58.7 ± 15.9 (range 22–85 years).

TABLE 2. Results of WST-Q Pseudonym Sex Diagnosis traumatic/non-traumatic No. sessions attended Mode of sessions Method of interview Percentage change in WST-Q capacity Percentage change in WST-Q confidence Percentage change in WST Q performance Adam M Vascular-below knee amputation 3

2 individual

1 group

Phone 26 2 4 Brad M Vascular- above knee amputation 3

2 individual

1 group

Face to face 19 14 18 Chris M Vascular- below knee amputation 3 3 individual sessions Face to face 43 36 16 Dave M Vascular below knee amputation 3

1 individual

2 group

Face to face 44 47 33 Edward M Vascular- above knee amputation 3 3 individual Face to face 69 69 71 Freddie M Traumatic amputation- below knee 3 3 individual Face to face 36 36 90 Gerda F Vascular- above knee amputation 3 3 individual Face to face 57 54 46 Henry M Vascular- above knee amputation 3 3 individual Face to face 41 33 0 Ian M Vascular- below knee amputation 2 2 group NA 43 32 29 Jack M Vascular- bilateral below knee amputation 2

1 individual

1 group

NA 46 30 42 Kate F Vascular- below knee amputation. 2

1 individual

1 group

NA 41 1 24

Two occupational therapists, two allied health assistants and one fourth year occupational therapy student provided the WSTP intervention. Between the WSTP and the focus group, the allied health assistants qualified as occupational therapists and attended the focus group along with one of the other occupational therapists who had less than 3 years' experience. Participants in the focus group had not been educated in wheelchair skills training beyond university and this programme.

3.2 Outcome measures 3.2.1 Wheelchair skill test-questionnaire

Participants reported perceived improvements in performance across indoor and community skills with a mean percentage increase at the completion of the WSTP of 42.3 ± 13.4. Similarly, confidence in completing the wheelchair skills was higher (mean percentage increase of 33.9 ± 20.7). After the programme, they more often completed wheelchair skills when they needed or wanted to (mean percentage increase of 33.9 ± 27.3). Table 2 details percentage of change for individuals and Figure 2 compares pre and post scores.

Average outcomes from the wheelchair skills test—Questionnaire (WST-Q)—Performance, confidence, frequency

3.2.2 Goal attainment scale

Due to the varied functional abilities and comorbidities experienced by participants, goals identified by participants varied considerably in task demand (i.e. complex community access to simple self-propelling at home). Ten participants experienced improvement in their GAS, achieving or exceeding their identified goal, with an average GAS change of two points ±1. An example of goal achievement included one participant moving from being pushed by a therapist within his home environment to being able to independently navigate his home with increased time and effort. One participant did not achieve their goal but did not decline in GAS score.

3.2.3 Functional Independence measure

Participants had an average mobility FIM change of 3.9 ± .5, with 82% (n = 9) recording a mobility FIM of six, meaning they could operate their wheelchair independently for a minimum of 50 m, turn around, manoeuvre the wheelchair to a table, bed and toilet, manoeuvre over rugs and over door sills and can negotiate a 3% graded ramp. The remaining two participants recorded a mobility FIM of five, indicating they could operate a wheelchair independently for short distances only, a minimum of 17 metres and there may be safety considerations or more time may be required to complete the task.

3.2.4 Qualitative outcomes

The collective perspectives of programme participants (n = 8) and facilitators (n = 3) are described using four themes including: motivators driving learning; delivery methods; structure and profile of the WSTP; managing risk and safety and confidence in wheelchair use.

Theme 1: Motivators driving learning There were conflicting opinions between participants and facilitators about the best way to learn wheelchair skills with participants not always being motivated.

“We had a few issues from patients who did not want to come” (Evie-facilitator).

“Some people do not want to learn, want to sleep” (Chris-participant)

According to facilitators, participants would engage sporadically and, on their terms, depending on who was in the group or their fatigue. Conflicting appointments and explanations of the programme influenced willingness to participate.

“you sort of pitch it that this is essential for your rehab and this is your main way of getting around and this is your mobility” (Maxine-facilitator)

Communication and rapport building were essential in maintaining engagement in the programme as well as knowing the participants' capacity well enough to set small realistic challenges and provide skill specific positive reinforcement.

“I feel like rapport is a big one. … as I developed my relationship with him, I could see that he realised that all the tips I am giving him actually makes his life easier.” (Maxine-facilitator)

Participants not knowing whether they were a candidate for a prosthesis meant they were uncertain about what they wanted to achieve, leading to facilitators being over-involved in their goal setting.

“You're kind of like helping them create goals, so I guess it is probably a little bit skewed in regards to being really (emphasis) patient driven.” (Robin-facilitator)

Theme 2: Delivery methods, structure and profile of the WSTP Prior to the introduction of the WSTP, the training provided in wheelchair use was adhoc. Facilitators liked the structure of the WSTP and the manual to ensure that all key skills were covered.

“I found it really easy to just come in and facilitate the session with the handout” (Robin-facilitator)

A large proportion of participants were ambivalent about the structure of sessions, including intensity and length. Freddie, Gerda and Edward found that the training grading and intensity were adequate and covered skills for a diverse age group, but Henry wanted more than one session a week to consolidate learning and Brad suggested sessions be longer to allow for the preservation of energy. One facilitator, Maxine, reflected that having three sessions provided flexibility to repeat and consolidate skills that participants found difficult. Maxine reported that session structure depended on the abilities of participants and their goals.

Participants additionally requested inclusion of wheelchair use during everyday tasks such as carrying objects when self-propelling and suggested incorporating family members into training and provision of education around wheelchair maintenance.

“I think family members must know about wheelchair. I know family help to make lunch or something else, but they do not know how to help transfer from wheelchair to other chair.” (Chris-participant)

Visual learning tools were highly valued as participants had difficulty following verbal instructions, particularly with some of the advanced skills that were difficult to explain.

“Seeing it done is way different to having someone say you have got to push here … it is a lot more instructional.” (Brad-participant)

Facilitators developed a strategy of asking participants to visualise their wheelchair wheel as a clock to assist with hand positioning and propulsion and they suggested learning could be further facilitated through watching videos of skills being performed (i.e. iPad) either prior to or during WSTP.

All facilitators and four participants agreed that there was value in the WSTP occurring in a group setting as well as one-to-one due to the peer support that was offered.

“I love it when patients encourage other patients. It is more effective than when the therapists encourage them, just because they have that bond … in it together.” (Robin-facilitator)

“you can bounce off each other and sort of share- do it with someone else and yeah help each other through.” (Freddie-participant).

Brad,