Article Text

Original research
We Walk: a person-centred, dyadic behaviour change intervention to promote physical activity through outdoor walking after stroke—an intervention development study
  1. Jacqui H Morris1,
  2. Linda A Irvine1,
  3. Stephan U Dombrowski2,
  4. Brendan McCormack3,4,
  5. Frederike Van Wijck5,
  6. Maggie Lawrence5
  1. 1School of Health Sciences, University of Dundee, Dundee, UK
  2. 2Kinesiology, University of New Brunswick Fredericton, Fredericton, New Brunswick, Canada
  3. 3Nursing, Queen Margaret University Edinburgh, Musselburgh, UK
  4. 4Nursing and Midwifery, University of Sydney, Sydney, New South Wales, Australia
  5. 5School of Health and Life Sciences, Glasgow Caledonian University, Glasgow, UK
  1. Correspondence to Dr Jacqui H Morris; j.y.morris{at}


Objectives To develop We Walk, a theoretically informed, 12-week person-centred dyadic behaviour change intervention to increase physical activity (PA) in community-dwelling people with stroke (PWS) through outdoor walking.

Design Three-phase intervention development study. Phase 1: we reviewed literature on barriers and facilitators to PA after stroke and mapped them to the Behaviour Change Wheel and Theoretical Domains Framework to define intervention components. The Health Action Process Approach determined intervention structure underpinned by person-centred principles. Phase 2: stakeholder focus groups involving PWS, their companions and health professionals reviewed the draft intervention, and experts in behaviour change were consulted. Phase 3: informed by phases 1 and 2, the intervention and form of delivery were refined, with final review through patient and public involvement.

Setting Three Scottish community rehabilitation stroke services.

Participants Twenty-three ambulatory community-dwelling PWS and their companions, thirty-seven health and exercise professionals, seven behaviour change experts.

Results Phase 1 determined key intervention components: information about benefits of walking; developing motivation and confidence to walk; facilitating dyadic goal setting and making plans together; monitoring walking, overcoming challenges; and maintaining walking behaviour. Phase 2 review by stakeholder focus groups and behaviour change experts endorsed intervention components and structure, emphasising dyadic relational aspects as central to potential success. In phase 3, intervention content and handbooks for PWS and buddies were finalised. Healthcare professionals proposed third-sector delivery as most appropriate for intervention delivery. A detailed delivery manual was developed. Participants preferred facilitated face-to-face and telephone delivery.

Conclusions Our multilens intervention development approach ensured this novel intervention was evidence-informed, person-centred, theoretically coherent provided appropriate social support, and addressed issues of concern to PWS. This study established intervention components and structure and identified operational issues critical to future success. Future research will pilot and refine We Walk and evaluate acceptability, feasibility, effectiveness and cost-effectiveness.

Trial registration number ISRCTN34488928.

  • Stroke

Data availability statement

Data are available upon reasonable request. Data are available on reasonable request. Access to data can be arranged by contacting the study chief investigator: Dr Jacqui Morris ( to discuss data sharing, data requirements and conflicts of interest, in line with UK and other regulations, including ethics approvals.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Strengths and limitations of this study

  • Our previous qualitative research and an evidence review of current qualitative studies examining barriers and facilitators to physical activity after stroke informed a theory-based behavioural diagnosis that guided the development of intervention content and form of delivery.

  • Stakeholder and expert consultation informed intervention refinement, guiding change to content and form of delivery, with emphasis on the relational aspects of dyadic working as central to potential success.

  • Person-centred principles ensured relevance, meaningfulness and acceptability of the intervention to people with stroke, whatever their context.

  • Most participants were male and already physically active; so their views may not apply to all of the stroke population.

  • Third sector agencies, which may be important in future intervention delivery, were not part of the stakeholder consultation; their role should now be explored.


Globally, 26 million people live with post-stroke disability.1 Improving physical function and fitness is crucial in reducing stroke burden for individuals, families and society.2 Regular physical activity (PA), including walking outside, improves walking speed, capacity and cardiovascular fitness in people with stroke (PWS)3 and reduces cardiovascular risk factors.2 Walking is inexpensive, valued, familiar and accessible for 80%–95% PWS who recover ambulation.4 5 Furthermore, 20–30 min daily moderate-intensity walking may reduce all-cause mortality in PWS by up to 41%.6 7 Therefore, outdoor walking has potential as a target behaviour for an intervention to increase PA and improve mobility outcomes and cardiovascular health after stroke.8

Average daily step counts PA after stroke are half those of age-matched healthy people.9 10 PWS are sedentary for >80% of the day,11 even those with ambulatory capacity, suggesting factors other than physical disability influence PA behaviour. Qualitative studies show PWS experience many psychological, physical, social and environmental barriers to PA participation,12–14 suggesting individually tailored behavioural approaches are necessary to promote long-term participation in regular walking. Systematic reviews show incorporating theoretically informed behaviour change strategies into post-stroke PA promotion programmes is effective, but to date, studies are small and behavioural strategies are poorly reported.15–17

Social support provided by family or friends predicts long-term engagement in PA in older adults18 19 and may enhance uptake and maintenance of post-stroke walking.20 Interventions involving pursuit of PA goals by two individuals with a socially significant relationship (eg, spouse, friend, family member, buddy) are known as dyadic interventions.21 One line of evidence supporting a dyadic approach to PA promotion after stroke comes from our previous qualitative study with 38 PWS and their carers.22 The study examined barriers to PA after stroke, showing many carers adopted a facilitatory role in supporting PWS to be active, so they could resume enjoyable shared activities together. Carers reported using informal approaches to support goal setting and goal achievement with the PWS and monitoring progress to support their partners’ confidence and motivation. These informal, common-sense behavioural approaches led us to conclude that a dyadic approach to PA promotion could effectively support PWS to be more active.

A second line of evidence to support dyadic interventions comes from a meta-analysis of studies with non-clinical populations, indicating dyadic interventions have moderate effects on PA outcomes when dyads agree shared goals.21 Systematic reviews have evaluated dyadic stroke survivor/carer interventions,23 but their focus was stroke recovery and caregiving, not PA promotion. No dyadic interventions were identified in a review of interventions to promote PA after stroke.16 However, a systematic review of dyadic interventions for PA promotion, including other clinical and disabled populations,21 showed psychosocial and physical health benefits for both caregiver and care recipient, suggesting a dyadic behaviour change intervention could effectively promote post-stroke walking.

Person-centred principles emphasise autonomy and empathic understanding24 to accommodate participants’ psychosocial context, roles, experiences, concerns and aspirations.25 Co-producing a dyadic intervention with users and key stakeholders that combines person-centred principles with theoretically based, individually tailored behaviour change strategies should increase enjoyment, enhance engagement and support PA maintenance.26 Such an intervention has potential to promote PA after stroke through walking, as PWS complete rehabilitation and move to independent community living.

This paper describes initial development of We Walk, a multistage project to develop and pilot a theory-based, person-centred, dyadic behaviour change intervention to promote PA through outdoor walking in community-dwelling PWS.

Study design

Intervention development involved three phases (figure 1). Phase 1. Step 1a: evaluating existing evidence. Step 1b: evidence mapping to the Capability, Opportunity, Motivation, Behaviour (COMB) framework of the Behaviour Change Wheel (BCW; Michie)27 and the Theoretical Domains Framework (TDF)28 for behavioural diagnosis. Defining intervention components and delivery in practice, underpinned by the Health Action Process Approach (HAPA).29 Phase 2: Step 2a: conducting focus groups (FGs) and reviewing the draft intervention with PWS, their companions, and health professionals. Step 2b reviewing for theoretical coherence with experts in behaviour change intervention development. Phase 3. Step 3a: integrating Phase 2 findings for intervention refinement. Step 3b: patient and public involvement (PPI) representatives reviewing intervention materials for clarity and relevance. Phases are reported below.

Figure 1

Intervention development and refinement stages.


Phase 1: intervention development methods

Step 1a: evaluating existing evidence

This project was conceptualised from earlier qualitative research22 and systematic reviews13 conducted by our team in which barriers and facilitators to PA after stroke were explored. To inform the present study, we updated our earlier systematic review of barriers and facilitators to PA after stroke.13 Systematic searches of CINAHL, Medline and PsycInfo from July 2011 to July 2018 identified 11 additional studies. We described and summarised themes within those studies and integrated them with our previous qualitative research and earlier review as the basis for intervention development.

Step 1b: initial intervention development

The BCW is a theory-based system for behaviour change intervention development.27 28 Following BCW methods, we mapped barriers and facilitators to post-stroke PA identified in phase 1a against theoretical constructs described in the COM-B model and related TDF. We identified intervention functions using behavioural diagnosis to specify target behaviours that, if changed, could bring about increased outdoor walking and identified the behaviour change techniques (BCTs) that would form the basis of the intervention.

The HAPA29 is a sociocognitive framework addressing adoption, initiation and maintenance of behaviour change. It describes motivational (intention formation) and volitional phases (planning, action, maintenance) of behaviour change, supported by self-efficacy development across each phase. We used the HAPA to inform the intervention structure for delivering BCTs.

Dyadic working

Dyadic planning involves creating plans with a partner about when, where and how PA goals are achieved.30 Transactive Goal Dynamics Framework31 conceptualises seven dyadic roles and goal pursuit permutations, illustrating how a dyad’s goals are linked and coordinated. The framework enabled us to initially specify dyads would pursue shared target-orientated goals; that is, walking goals for PWS would be the intervention target, with PWS and buddies working together to pursue them.

Person-centred principles

We adapted person-centred principles from a framework for person-centred practice25 to guide intervention development and delivery. Throughout the development process, we examined decisions made, language and tone used in written materials, interactions with stakeholders, and data collection and analysis, to ensure we followed these principles, which also informed planned delivery. We intended that those delivering the intervention would take person-centred approaches to guide the dyad in implementing the intervention. For example, facilitators would encourage goal setting by sharing information, listening, ensuring conversations focused on what mattered to the dyad, identifying meaningful goals aligned with their values and lives and respecting their preferences, needs, values, interests and contexts. Agreed activities to achieve goals would be tailored accordingly.

Patient and public involvement statement

PPI informed all stages of this project. Members of two Scottish Chest Heart and Stroke Scotland support groups and local exercise groups supported the initial grant application. They subsequently advised on initial intervention content, selecting walking buddy as an appropriate term and providing feedback on study and intervention materials. Four group members provided ongoing study oversight through steering group attendance as the project progressed reviewed intervention materials at the end of the study.

Phase 2: stakeholder consultation and intervention refinement

We next sought stakeholder views on intervention content and delivery format, enabling us to adapt and refine We Walk to meet their needs.

Step 2a Consultation with PWS, their companions and health professionals

Stroke rehabilitation clinical leads in three Scottish health boards, selected for urban and rural populations, identified community-dwelling, ambulatory PWS who could provide informed consent. PWS invited and attended with their chosen companions. The clinical leads also invited health and exercise professionals involved in stroke services within their local areas. All participants received study information before giving written informed consent for participation.

Participants attended one of six stakeholder meetings held at local universities in three health board areas. Consultations aimed to collect data to inform intervention content and delivery format; examine intervention acceptability and ensure person-centredness. Findings would inform intervention refinement before pilot testing (to be reported elsewhere).

Data collection

Intervention structure and draft materials were introduced to all participants, and initial discussions and questions were addressed in small groups. Discussions were recorded in contemporaneous field notes. To obtain all perspectives and ensure no one group dominated, we next conducted FGs separately with PWS and companions; and with professionals (12 FGs in total), respectively led by LI and JM. FGs captured data relevant to each stakeholder group, exploring participants’ views of intervention usefulness, acceptability and likely feasibility, dyadic working, form of delivery and person centredness. Healthcare professionals also considered intervention relevance, fit and timing concerning rehabilitation context and pathways. Groups were audio recorded and transcribed (topic guides in online supplemental material 1).

Data analysis

JM and LI read transcripts and agreed on coding. Codes were reviewed and grouped into categories relevant to intervention components. Using the six stages of thematic analysis,32 the research team interpreted findings relevant to intervention development, developed descriptive themes, and incorporated findings into the intervention in Step 2b. Data were managed in Nvivo V.12.

Step 2b Theoretical coherence of the intervention: review by experts in behaviour change intervention development

We invited an international group of experts in health psychology and PA research (n=14) to inform intervention refinement further by commenting on the draft intervention. Seven responded; five health psychologists, one expert in PA promotion, and one physiotherapist. Respondents received documentation detailing intervention functions, BCTs, and a summary intervention delivery plan. We invited them to provide comments on (1) theoretical coherence, (2) relevance and sufficiency of incorporated BCTs, (3) person centredness, with participants valued as equal partners to ensure the intervention meets their needs, (4) whether the intervention would engage stroke survivors and walking buddies.

Phase 3: intervention refinement

Step 3a Refining the intervention

Researchers summarised and collated findings and feedback from preceding phases to decide how to refine the intervention. Feedback and decisions were tabulated to illustrate the steps and decisions in deriving the intervention.

Step 3b Review by PPI group

The PPI group members reviewed the intervention structure and all intervention materials for relevance, appropriateness and ease of reading. Responses informed final intervention refinement prior to piloting.


Phase 1: intervention development methods

Step 1a Evaluating existing evidence

Collated findings of our updated review are presented in (table 1). Barriers to PA after stroke included negative emotional responses to PA, such as embarrassment and fear, believing PA was not helpful, limited knowledge about appropriate PA, perceptions of limited capability due to stroke, cognitive and communication impairments, low mood and confidence, and low recovery expectations. Facilitators included enjoyment, previous success in PA, and social support.

Table 1

Description of summarised barriers and facilitators to physical activity after stroke

Step 1b Evidence mapping/behavioural diagnosis and defining intervention components and delivery in practice

We mapped the qualitative review findings to the COM-B and TDF (table 2). Of nine intervention functions, five were appropriate for We Walk: education, persuasion, training, modelling, and enablement (table 2). We next identified BCTs relevant to intervention functions from Michie et al’s taxonomy of BCTs28 (table 2).

Table 2

Intervention functions and behaviour change techniques

Using the HAPA, we developed a structure for intervention delivery that included discussion with the dyad of benefits of walking, exploration of how the dyad would work together, identifying and setting meaningful walking goals, monitoring progress, and undertaking coping planning, reviewing goals and problem solving. Person-centred principles appropriate for We Walk underpinned the intervention structure (box 1).

Box 1

Person-centred principles applied to we walk

  1. Account for what the person values about his/her life and how he/she makes sense of what is happening from their perspective, psychosocial context and social role.

  2. Facilitate involvement in decision-making by the person and their buddy by considering values, experiences, concerns and future aspirations.

  3. Be underpinned by knowledge of the person, their beliefs and values and their experience

  4. Recognise the uniqueness of the individual, and maximise personally relevant resources through recognition of important agendas in their life

  5. Pay attention to the whole person

Deciding form of delivery

To avoid burdening rehabilitation services, we initially intended We Walk be delivered via a self-directed workbook. Rehabilitation professionals would provide initial instruction, followed by one follow-up telephone contact. However, in development, we recognised participants would need in-person support and time to engage with complex intervention components. Additionally, maintaining positive, person-centred written language within workbooks while explaining barriers and problem-solving strategies, conveying dyadic planning, and buddy support was challenging and undermined person-centred tailoring. We determined the intervention should be delivered through face-to-face (n=3 sessions) and telephone facilitation (n=3 sessions) with short handbooks to explain intervention components and dyad members’ roles. Incorporating all components would require at least 12 weeks. We developed a draft facilitator manual explaining intervention component delivery and BCTs to be used at each contact.

Phase 2: stakeholder consultation and intervention refinement

Stakeholder views on intervention content, delivery format are reported below. The findings enabled us to adapt and refine We Walk in response to their views.


Participants were 23 community-dwelling, ambulatory PWS (6 female, 17 male, aged between 50 and 83 years, and between 3 months and 24 years post stroke). PWS were from a range of sociodemographic backgrounds, with Scottish Index of Multiple Deprivation Index scores ranging between 1 (most deprived) and 10 (least deprived),33 and across urban and rural settings. All were ambulatory, and 15 participated in regular exercise or walking as PA. PWS brought 14 companions (13 female, 1 male). Thirty-seven rehabilitation, health and exercise professionals participated (22 physiotherapists, 8 occupational therapists, 5 stroke nurses, and 2 local exercise services coordinators). The 12 FGs each involved between 6 and 8 participants.

FG findings

FG findings relating to each intervention component are presented below, supported by illustrative quotes.


Most PWS emphasised enjoyment, social and health benefits of walking after stroke, but acknowledged developing confidence and skills could be slow. Finding personally relevant solutions to challenges and developing determination, resilience and motivation to persist, when fatigued or unmotivated, in inclement weather and difficult terrain was challenging but essential because persistence in walking more led to increased confidence and well-being.

the biggest thing is trying to tell people to have motivation to do, to get them self better… it’s self-motivation, I think that’s the secret. (FG4 Male PWS1)

Before the stroke I didn’t believe I would enjoy walking, but I enjoy walking now and get cabin fever if I hadn’t had a walk… (FG4 Male PWS 2)

Walking goals

In line with person-centredness principles, PWS agreed with our premise that walking goals should be personally meaningful but achievable while challenging enough to push them to progress.

There’s my goal there in front of me, that’s what I’m going to try an achieve rather than saying, ‘Oh, wait a minute, I’ve got to do ten thousand steps today.’ Well, if you don’t do your ten thousand steps you can get disheartened. But then, if there is a goal in front of you if you choose your goal each day… Aye, it becomes a whole different entity then… (FG1 Male PWS 1)

Form of delivery

PWS considered written material impersonal, reflecting our concerns about We Walk delivery using a self-directed handbook. Many PWS were overwhelmed by post-stroke paperwork, preferring tailored face-to-face delivery that addressed their heterogeneity (eg, mobility, balance, ability of the buddy to provide support, and the type of relationship between the two).

I think it should be face to face. After a stroke they give you piles of paperwork…You don’t read it. Everybody that has had a stroke, their needs are different. Anything you put in black and white isn’t gonnae suit everyone… (FG 1 Male PWS 2)

The dyad: buddy selection

The term ‘Walking Buddy’ was considered appropriate for the walking companion. Participants preferred to select buddies, but developing a relationship with a new buddy was also considered possible. Shared interests and time to develop trust were critical. Buddies would require knowledge of stroke and its impact and be knowledgeable and sensitive to PWS’s experiences of stroke and physical capabilities. Other PWS could be ideal buddies because of their shared experience.

I’ve got a skill set now I didn’t have before that people could tap into. … if somebody has shared interests then you can do things together quite easily and talk when you’re doing it. (FG6 Male PWS 1)

Dyadic goal setting and planning

Many participants thought walking buddies could motivate PWS through external accountability; however, adapting to fluctuations in fatigue and how PWS felt was critical. Relational aspects of dyads were considered easier when the buddies were known to PWS, but only if their autonomy in decision-making and goal-setting was respected. Aligned to the goal-setting framework we had identified, participants were clear that walking goals for PWS would be the intervention target, with PWS and buddies working together to pursue them

And then if you’ve got something to motivate you to move, once you get out of bed daily, that helps. Because if you get into a routine of doing stuff then that makes it much easier whereas if you do it weekly, I’d say if you’ve got a helper who comes… male stroke survivor in (FG 5, Male stroke survivor)

Every time we went out a walk, my wife grabbed my arm and it’s stifling me, stopping me moving…the person you’re going out with walking, has got to realise, you’ve got to let them do their own thing (FG 1 PWS Male Stroke survivor 2)

Health professionals, PWS, and companions suggested buddies needed tailored guidance about the role. Overburdening buddies was a concern expressed by healthcare staff in particular, where buddies were the primary carer.

… whatever sense of partnership they’re going to commit to you know helping this person on their journey towards recovery. So it’s important the facilitator has that conversation with both of them (FG4 PWS Female Companion 2).

I think a lot of the patients will use the word burden, they don’t want to put more onto the families and things, they’re maybe already doing part of the caring role so, it does get really complicated. (FG2 HCP Female Physiotherapist 1)


Monitoring walking was seen as important for motivation. Many PWS already monitored walking, including places walked to, steps per day, time spent walking, and distance covered. Some used wearable fitness trackers and mobile phones, but physical capability to apply them, experience and attitudes to technology and cognitive capability to use data to influence behaviour determined preference use and utility. PWS and buddies viewed pedometers as potentially useful.

‘Well if this suits you, use this, if you prefer pedometers you could use that, if you prefer Fitbit you can use that’. So you give flexibility, people can pick and choose what they think works best for them. (FG3 PWS Female companion 1)

Most PWS viewed recording and reflecting on progress in achieving walking goals as motivational, and would complete diaries. However, for some, physical and mental effort involved seemed overwhelming. Diary format had to be short and simple, with some preferring electronic diaries.

Just by writing it down, it would be visible to see for those that like doing a diary… Write it in the diary ’cause you don’t always see progress. (FG4 Female Companion 2)

So, I don’t need to mark that down because I know I’m capable of doing that and I’m not lying in the chair done in. But it’s not for everybody.(FG1 Male PWS 3)

Intervention timing

We intended to deliver We Walk as community rehabilitation was ending. However, most PWS said adjusting to post-stroke life took time, and they may not be ready for We Walk immediately. Instead, they wanted information about We Walk during rehabilitation and to decide when participation was appropriate for them.

We’ve seen people at the self-management course, ‘Oh, I had my stroke nine weeks ago.’ And it’s really too early for them to take it in… the stroke nurse or the OT people, they’re the ones that should say, ‘Why don’t you try that, have you ever thought about that?’ (FG2,PWS Male Stroke Survivor 4)

Facilitator role

We intended, and pre-project discussions had suggested, that rehabilitation professionals would facilitate intervention delivery, given their role in facilitating transitions between rehabilitation and active lifestyles in the community. While they viewed PA promotion as providing continuity with rehabilitation, many considered intervention delivery over 12 weeks unfeasible within community rehabilitation, which was typically shorter. Although supportive in principle, time-limited service delivery protocols left little flexibility for intervention delivery.

The amount of time you can give them when someone’s achieving outdoor mobility in the community, you almost have to discharge at that stage. You know you can’t follow them on weeks after weeks as we’re not really covered. (FG5 HCP, Female Physiotherapist)

Delivery through professional or volunteer facilitators within stroke charities or other community organisations was viewed as more feasible, given that charities already provide some supportive services. Healthcare professionals proposed that they could be involved in training and supporting those facilitators and in facilitating links between rehabilitation and third sector organisations delivering We Walk.

So I think there’s erm, we have to look at third sector and, and council exercise providers…You know a charity based person, you know all the professionals could feed into a training session that’s run by somebody that has been upskilled by whoever. You know from the charity. (FG5 HCP, Male Physiotherapist)

The person-centred approach

Throughout intervention development, we sought consistency with person-centred principles. Congruently, PWS and their companions viewed themselves as people not defined by their stroke. Data showed they considered good communication, shared decision-making, respect for their values, capabilities, and desired achievements would be crucial to the intervention’s success and dyadic working.

I think it’s about the relationship between you and your buddy and it’s that listening thing so you both know what each other are thinking… If you don’t listen to somebody, truly listen to them, then you’ll never get anywhere. (FG2 PWS, Female stroke survivor)

Overall usefulness of we walk

PWS were positive about the concept of We Walk because they saw outdoor walking as a fundamental skill that met their desire to prevent further stroke and improve health. Walking was an enjoyable activity that could be sociable because it enabled the community participation valued by PWS. Confidence for walking could be improved through small successes that incrementally enhanced confidence and improved walking and recovery over time, and We Walk could support that process.

What I found helpful was identifying and believing in the benefits of walking. When I had the fall two years ago, I had to start walking again outside. Walking on the road, the bumps on the road, getting the confidence to get up there and back doon again… (FG2, male stroke survivor)

You can do it once, but you have to keep it going to get your brain working and things like that…the more you do it, the easier it gets. (FG2, male stroke survivor)

In summary, participants agreed We Walk could support PWS to meet their desire to participate in regular outdoor walking. An individually tailored, person-centred approach was vital to engagement and making the intervention meaningful irrespective of physical capabilities. Personal selection of walking buddies sensitive to needs of PWS was another key finding. Face-to-face facilitation at the right time for PWS; clarity of the buddy’s role and the dyadic relationship, person-centred tailoring, and delivery outside health service care pathways were considerations for intervention refinement.

Step 2b Theoretical coherence of the intervention: review by experts in behaviour change intervention development

Feedback from experts is described in detail in table 3. Key suggestions were: a reduction of content in session 1 by reorganising BCTs across sessions 1 and 2, ensuring walking buddies understood their role through explanation within the buddy handbook and in sessions 1 and 2; discussing habit formation throughout the intervention, and clarifying how the facilitator would provide feedback through review of diary entries. Findings were incorporated into the intervention, as described in phase 3a, below.

Table 3

Summary of feedback and decisions made to refine the intervention

Phase 3: Intervention refinement

Step 3a Refining the intervention

Details of feedback obtained in phase 2, decisions made in response to findings and their implications are described in table 3.

Overall stakeholders viewed We Walk as relevant and coherent, but some key changes were required. Change in delivery form from workbooks to two facilitated face-to-face and three telephone contacts over 12 weeks, guided by a detailed delivery manual, with short handbooks for PWS and buddies would facilitate personally tailored intervention delivery. The timescale would enable dyads to engage with We Walk and develop personally relevant walking goals and plans, give time to change behaviour and establish PA patterns that could be maintained in the long term.

PWS viewed the relationship with a compatible buddy as crucial. Ensuring dyadic compatibility and clarity about roles and expectations was vital, but must be without obligation if PWS felt the arrangement would not work. The buddy role required clear explanation by the facilitator. Buddy handbooks and intervention delivery manual would emphasise respect for goals and desire for autonomy of PWS, while demonstrating sensitivity to fluctuating capabilities. PWS emphasised the need to tailor the intervention to account for fatigue, which is a major problem for many PWS.11 BCTs for problem-solving and coping planning were included to support the PWS in pacing to manage fatigue and maintain motivation. Weekly review of progress would enable participants to reflect on progress and to adjust goals to account for fluctuations in how they felt. Buddies needed to support the PWS to adapt to fluctuating fatigue and emotional states, and their role will be emphasised in the buddy handbook and throughout face-to-face intervention sessions with the facilitator.

Self-monitoring and diary completion needed to be pragmatic, reflecting our person-centred philosophy. Participants, including buddies, would receive pedometers; but could use preferred monitoring methods (phone, wearable fitness tracker, distance, time). Completing simple written or electronic diaries would facilitate reflection on progress, informing dyadic planning and weekly goal-setting. Diary entries would be reviewed in face-to-face sessions with facilitators and returned by post or electronically for discussion in telephone consultations, prompting reflection.

Phase 1 and phase 2 data showed that a facilitator should deliver the intervention, explaining and discussing dyadic working, the role of the buddy and delivering BCTs according to the delivery manual. In subsequent piloting, this would have to be a researcher facilitator, who would collect detailed fieldwork data to better understand the facilitator role, dyadic dynamics and determine how to meet the needs of PWS, enabling refinement of the delivery manual for later implementation. For testing in a trial and future implementation, our data showed healthcare professionals would not have resources to act as facilitators to deliver We Walk, suggesting external facilitators, such as those from third sector organisations, could support delivery. Future piloting will therefore establish training required for third sector facilitators; how they can support dyadic planning, and resource implications for the third sector (time, travel, training). Undertaking these steps will pave the way for delivery by third sector organisations.

The review of included BCTs indicated not all BCTs were core; therefore, we determined core BCTs with remaining BCTs used selectively to tailor the intervention. Finally, we developed a logic model to illustrate pathways between proposed intervention inputs, core BCTs, processes and outcomes for behaviour change (figure 2).

Figure 2

We Walk logic model. PWS, people with stroke; PA, physical activity.

Step 3b Review by PPI group

Intervention materials were reviewed by our PPI group, comprising four male PWS and their female companions. They suggested simplifying text, foregrounding benefits of walking early in the intervention, and including relevant images of PWS, including those who use walking sticks, into written materials. (table 3). The materials will be used in the subsequent piloting phase of the study, to be reported elsewhere.


We integrated review evidence, theoretical constructs, and person-centred principles with qualitative data from FGs to develop a theory-informed dyadic behaviour change intervention to promote outdoor walking after stroke. Stakeholder views guided refinement of content, form of delivery and dyadic working and highlighted the importance of person-centredness to PWS and their companions.

To our knowledge, this is the first dyadic intervention for PA promotion after stroke. We combined evidence, theory, person-centred principles and stakeholder perspectives to ensure the intervention was adaptable to the varied contexts, experiences and values of PWS. This integrated approach was, to our knowledge, novel in intervention development for stroke. Our approach aligns with other studies of people with disabilities, in recognising individualised tailoring as vital for promoting PA participation.17 34

The BCW, specifically COM-B and TDF28 35 36 guided behavioural diagnosis, specifying intervention functions and BCTs. Using HAPA29 to structure We Walk enabled development of the intervention structure to support shifts from motivation to enact and maintain walking. It supported careful selection of relevant BCTs identified by behavioural diagnosis, a challenge noted by other researchers.37 We combined selected BCTs in a logical, staged sequence to enhance walking behaviour.

FG participants were mainly male with female companions therefore findings may not apply to all stroke survivors. Many PWS were also physically active; therefore, their views may not reflect those of less active survivors. Nonetheless, their challenges and successes in maintaining motivation and developing self-efficacy shaped the intervention. Despite functional and ambulatory capability, many PWS remain inactive for up to 81% of the day.11 To address the capacity/behaviour gap, we developed a person-centred approach to intervention development. The dyadic person-centred intervention approach intends to support PWS in developing motivation and increasing outdoor walking by identifying personally meaningful goals, person-centred action planning, problem-solving, monitoring, and reflection on their progress. The person-centred approach means We Walk has potential to provide support for behaviour change to ambulatory PWS irrespective of their levels of disability. Nonetheless, further research, including empirical piloting and feasibility studies, will be necessary to enable us to identify stroke sub-populations with whom the intervention is most likely to be effective before undertaking a definitive randomised controlled trial.

We concluded in section 3 a) the likely scenario for delivery by third sector organisations. A study limitation was not including a range of third-sector stakeholders to explore how they could support delivery. Our initial intention was that healthcare professionals would deliver We Walk. They indicated the 12-week duration of the intervention would not fit well with timescales in community rehabilitation and that third-sector delivery would be more appropriate. Development of the facilitator role and exploration of how third sector organisations can deliver the intervention will be the subject of future piloting work.

PWS supported the partner-oriented goal-pursuit we identified through the Transactive Goals Dynamics Framework.31 A meta-analysis of randomised controlled trials of dyadic interventions for PA supported this approach,21 showing dyadic interventions targeting goals of one dyad member had larger effect sizes for PA compared with those targeting shared goals. Findings reflect our data indicating PWS would like the dyad to focus on their walking goals. PWS also emphasised relational aspects of the dyad, specifying sensitivity and adaptability as desired walking buddy characteristics. Reflecting our findings, the meta-analysis of dyadic interventions found type of dyad (spousal, friend/peer) can influence PA.21 Peer-friend dyads exerted larger effects than other dyad types, with authors suggesting responsiveness to each others’ goals within close relationships may best facilitate goal achievement. Our qualitative data suggested spousal interactions could positively or negatively influence dyadic working. Perceptions of pressure or control by spouses may not always support PA in a dyadic intervention. Other PWS suggested family involvement was central to the dyad because family supported their management of recovery. These positive and less positive relational nuances on dyadic working align with a qualitative study of dyadic support for PA people with osteoarthritis.38 The studies highlight relational aspects of dyadic interventions, indicating careful training of facilitators in supporting dyads is vital.

We found no other studies of dyadic interventions promoting community walking after stroke. However, a qualitative evaluation of a dyadic intervention to promote participation in Tai-Chi with people with dementia39 also highlighted the necessity of clarifying the buddy’s role and the important role of the instructor/facilitator. Congruent with our intervention, the study emphasised action and coping planning for developing self-efficacy in both dyad members.

Development and feasibility studies of a community-based peer-led walking intervention to increase PA in inactive older people40 reflected similar intervention development processes to We Walk. Authors used social-cognitive theory and a socioecological model to develop and include BCTs; however, the final BCTs and intervention structure aligned with We Walk. Congruently, participants in both studies emphasised tailoring, selecting personal goals, using self-monitoring to increase motivation, planning, enjoyment and desired buddy attributes. The community delivery model with peer supporters as buddies used in that study could provide a template for implementing We Walk through third sector organisations.


The theoretically informed person-centred approach we took to developing a dyadic intervention to promote walking behaviour is novel in stroke and seeks specifically to address known barriers and facilitators to PA experienced by PWS. Careful stakeholder consultation throughout informed development ensured the intervention meets needs of PWS and is theoretically and pragmatically coherent. Uncertainties remain about how stroke dyads will work in practice; future research will pilot and optimise We Walk, and evaluate dyadic working, acceptability, feasibility and effectiveness.

Data availability statement

Data are available upon reasonable request. Data are available on reasonable request. Access to data can be arranged by contacting the study chief investigator: Dr Jacqui Morris ( to discuss data sharing, data requirements and conflicts of interest, in line with UK and other regulations, including ethics approvals.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by ID number 19/NS/0077 North of Scotland Ethics Committee. Participants gave informed consent to participate in the study before taking part.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Twitter @jacquimorris11, @sdombrowski, @Profbrendan

  • Collaborators None.

  • Contributors JHM led the design and overall conduct of the study as chief investigator and led the writing of the manuscript. JHM is responsible for overall content as the guarantor. LAI, with JHM and SUD undertook intervention mapping and initial intervention development, study setup, recruitment and data collection, coordinated PPI consultations, and contributed to data analysis and interpretation and manuscript writing. SUD advised on behavioural theory and supported intervention development and manuscript development. BM was a co-investigator who contributed to study design, interpretation of findings, and led the development and integration of person-centred principles across the study. FVW was a co-investigator who contributed to study design, recruitment of participants and interpretation of findings, advising on physical activity matters. ML was a co-investigator who contributed to the design and interpretation of qualitative data.

  • Funding This work was supported by the Chief Scientist Office, Scottish Government, grant number HIPS-17-03.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.