546769
research-article2014
CRE0010.1177/0269215514546769Clinical RehabilitationBarclay et al.
CLINICAL REHABILITATION
Article
Factors describing community ambulation after stroke: a mixedmethods study
Clinical Rehabilitation 2015, Vol. 29(5) 509–521 © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0269215514546769 cre.sagepub.com
Ruth Barclay1, Jacquie Ripat2 and Nancy Mayo3,4
Abstract Objective: To develop a model of community ambulation after stroke based on: Canadian data from community-dwelling individuals post-stroke; the experiences and opportunities for community ambulation expressed by individuals with stroke; and current literature. The model presents a visual depiction of the relationships between the different factors of community ambulation after stroke. Design: A quantitative/qualitative explanatory sequential mixed-methods design was utilized. Secondary data analysis with structural equation modeling resulted in a community ambulation model. Two focus groups of individuals with stroke were conducted to verify and explain the model. Setting: Community. Subjects: Quantitative data from 227 participants: 142 (63%) male; 63.4 (12.0) years of age and 2.6 (2.5) years post stroke. Eleven individuals participated in the focus groups: 6 (55%) male; 61.4 (6.9) years of age and 5.8 (3.3) years since stroke. Main measures: Model variables: items from the EuroQol, Preference Based Stroke Index, gait speed, Reintegration to Normal Living Index, the Community Health Activities Model Program for Seniors, and the Geriatric Depression Scale. Results: The model had reasonable fit with three latent variables: ambulation, gait speed, and health perceptions (normed χ2 = 1.8, root mean square error of approximation = 0.060 (0.043; 0.075)). Depression was also a component of community ambulation. Participants verified the model and added endurance and the environment as additional components. Participants used self-awareness and knowledge of the environment to engage in cognitive strategies related to community ambulation. Conclusions A model of community ambulation after stroke was developed and verified. Recognizing important components of community ambulation may assist physiotherapists in determining community ambulation goals, needs, and opportunities in partnership with clients.
1Department
of Physical Therapy, University of Manitoba, Winnipeg, Manitoba, Canada 2Department of Occupational Therapy, University of Winnipeg, Manitoba, Manitoba, Canada 3Division of Clinical Epidemiology, Department of Medicine, McGill University, Montreal, Quebec, Canada
4Research
Scientist, McGill University Health Centre Research Institute, Montreal, Quebec, Canada
Corresponding author: Ruth Barclay, Department of Physical Therapy, University of Manitoba, R106-771 McDermot Ave., Winnipeg, Manitoba R3E 0T6, Canada. Email: [email protected]
510
Clinical Rehabilitation 29(5)
Keywords Community ambulation, participation, stroke Received: 9 December 2013; accepted: 6 July 2014
Introduction “…walking around the living room isn’t going to cut it”. Focus group participant
Walking in the community is a key activity for many individuals after stroke: in a survey of 130 individuals surveyed with stroke, approximately 75% specified that community ambulation was considered essential or very important.1 Community ambulation incorporates indoor and outdoor walking, as well as a social aspect, and has been defined as “independent mobility outside the home, which includes the ability to confidently negotiate uneven terrain, private venues, shopping centers and other public venues”.1 In one study, approximately onethird of community-dwelling individuals with stroke reported not being able to walk outdoors independently.1 Of those who were independent in their community ambulation, most were hesitant to use public transportation and relied on others for transportation, which could in turn further limit community ambulation.1 Benefits of walking outside include “transportation, health benefits, and recreation”,2 increased physical activity,3 and improved balance, endurance, and walking ability.4,5 Community ambulation enables participation in activities that are meaningful to the individual,6 and is associated with self-rated health.7 A number of factors have been associated with the ability to be a community ambulator after stroke, these include: faster gait speed,1,8,9 increased endurance,10 falls self-efficacy,9,11 and a supportive environment (e.g. walking location).12 While not specific to stroke, in older adults gait speed, health perception, and balance perception have all been shown to be associated with the distance walked outside.13 Gait speed and stepping height represented mobility in a study of adults age 75–80 in Finland,14 while health worry (anxiety from a
concern for health, or health perception) predicted physical activity, which in turn predicted walking difficulty in older American adults.15 Walking speed and endurance are considered key elements of community ambulation as they directly relate to crossing the street in the time of a walk signal and walking sufficient distances to complete essential activities such as shopping.16,17 A visual depiction, or model, of the relationships between the different components of community ambulation after stroke (impairments, activity limitations, participation restrictions, and contextual factors of the person and environment) would be helpful for both clinicians and researchers. This model could be used in a clinical context to help formulate client goals and to direct assessment and treatment strategies with the client wishing to improve community ambulation after stroke. The model could also be helpful in forming research questions about improving community ambulation after stroke. The purpose of this research was to develop a model of community ambulation after stroke based on: Canadian data from community dwelling individuals post-stroke, the experiences and opportunities for community ambulation expressed by individuals with stroke, and current literature.
Methods Overview We used an explanatory sequential mixed-methods design18 to guide this study. The first (quantitative) phase of the study was the development of a model of community ambulation. The second, qualitative phase, followed up on the quantitative phase for the purpose of confirmation and revision of the model of community ambulation from the
511
Barclay et al. Table 1. Health indices and items in proposed model. Health index
Constructs (latent variables) Indoor mobility
Outdoor mobility
Health perceptions
Gait speed
VAS health state Health today
Comfortable Maximal
(combined for final model) EQ5D PBSI
Mobility Walking Stairs Physical activities
CHAMPS
RNL 5-m gait speed
Move in home
Jog or run Walk uphill Walk fast for exercise Walk to do errands Walk leisurely for exercise or pleasure Move in community
EQ5D: EuroQol; CHAMPS: Community Health Activities Model Program for Seniors; PBSI: Preference-Based Stroke Index; RNL: Reintegration to Normal Living Index; VAS: visual analogue scale. Italicized variables were removed from model due to their poor fit.
perspective of individuals with stroke. The benefit of using a mixed methods approach in this situation was that the qualitative experiences of stroke survivors could be used to explain and add to the quantitative theory-based and statistically derived model of community ambulation. The authors have used this method of developing a quantitative model and verifying it with qualitative data previously in developing a model of participation after stroke.19
Quantitative phase – developing the initial model
(RNL),23 the Community Health Activities Model Program for Seniors (CHAMPS),24 and the Rasch version of the Geriatric Depression Scale for individuals with stroke.25 The variables used in developing the model were based on current literature, as described above, with the qualification that the variables had also been measured in the trial. The latent variables of indoor mobility, health perceptions, gait speed, and outdoor mobility were chosen, as they summarized the available ambulation-related variables. Table 1 provides a summary of how each item or subscale was associated with each latent variable.
Participants. The quantitative phase involved secondary data analysis of a multi-center randomized controlled trial of participation of 227 Canadians with stroke living in the community.
Procedure and analysis. Structural equation modeling (SEM) was used to develop the model. Statistical analysis was completed with SPSS and AMOS, with the following steps, using SEM.
Variables. Several variables related to ambulation had been collected during the study. Measured variables that were available in the data and relevant to this study include those from the EuroQol (EQ5D),20 Preference-Based Stroke Index (PBSI),21 5-meter gait speed,22 Reintegration to Normal Living Index
1. A measurement model was developed to demonstrate associations between the latent variables of indoor mobility, health perceptions, gait speed, and outdoor mobility and the related measured variables. This was completed using the baseline evaluation time.
512 2. A structural model (associations and paths showing directionality, but not cause) was developed from the measurement model. Variables of age, sex, and depression were added at this point. As is common in studies with multiple outcomes, not all participants answered all items for all measures. For this reason, Full Information Maximum Likelihood (FIML) estimation was used, as this method accounts for missing variables.26 Missing data is not imputed with FIML, as parameter estimates are estimated from the raw data.26 Fit indices are used to determine the fit of a model. The model chi-square (χ2) is commonly reported in the SEM literature. As the χ2 increases, model fit worsens.27 The χ2 tests the difference between a model that has a perfect population fit and the observed model.27 A χ2 that is non-significant means that there is little difference between the models; this suggests a good fit. Models with large sample sizes can often be rejected since the model χ2 is affected by sample size.28 A variety of fit indices are therefore used. Some researchers use a normed χ2 (χ2/degrees of freedom) to decrease the effect of the sample size on the χ2. Normed χ2 values up to 5.0 have been used to suggest a reasonable model fit.27 The root mean square error of approximation (RMSEA) measures the lack of fit in a model compared with the population. A close fit is suggested with an RMSEA of ⩽0.05; a reasonable fit from 0.05 to 0.08; and ⩾0.10 a poor fit.27 The comparative fit index (CFI) is an incremental fit index in which a reasonably good fit is suggested when the CFI is greater than 0.90.27
Qualitative phase – Exploring community ambulation and verifying the model Two focus groups were conducted to explore community ambulation from the perspective of individuals with stroke who live in the community. Each focus group ran for approximately two hours. Participants. Individuals were recruited through a local Stroke Recovery association, and by contacting participants from previous studies who had
Clinical Rehabilitation 29(5) indicated their willingness to be re-contacted for research purposes. Inclusion criteria were: adults with stroke who were living in the community and were community ambulators by self-report. We conducted purposive sampling with the intent to “bring together people of similar backgrounds and experiences to participate in a group interview about major issues that affect them” (Patton, page 236).29 While participants varied in terms of sex, age, and length of time since stroke, all indicated that they were community ambulators, and all had gone through their stroke rehabilitation within the same regional health authority. Procedure and analysis. A directed content approach30 was used to guide development of the focus group questions and an initial coding scheme; this approach is useful when theory exists that suggests relationships among variables (in this case, the community ambulation model). The community ambulation model was not shared with participants prior to, or during, the focus group in an attempt to elicit a range of divergent and confirmatory opinions. Focus group questions were openended, with additional probes used to address aspects of community ambulation established in the model, as well as other factors in the literature but not included in the developed model.30 Questions addressed participant-perceived barriers and facilitators to community ambulation, reasons why participants did or did not walk outside, and situations that encouraged or hindered their ability to walk outdoors. Focus groups were facilitated by the second author (JR) and detailed field notes were taken by a research assistant. Focus groups were digitally audio-recorded and all data was transcribed verbatim. The transcripts were reviewed and coded, first individually and then collaboratively by RB and JR, using a set of categories based on the model while staying open to the inclusion of additional emergent categories. Field notes were used to corroborate category assignment. To verify the model of community ambulation post-stroke, the final stage of mixed methods analysis was theory driven, focusing on linking the emergent categories to the model of community ambulation that was developed with SEM. Categories
513
Barclay et al. Table 2. Description of the study participants – quantitative and qualitative phases. Gender freq (%) Quantitative phase n = 227 Male 142 (63) Qualitative phase n = 11 Male 6 (55)
Age in years Mean (SD) [range]
Years post-stroke Mean (SD) [range]
Comfortable gait speed meters/second Mean (SD) [range]
63.4 (12.0) [24–93]
2.6 (2.5) [0.1–15.7]
0.78 (0.40)* [0–2.0]
61.4 (6.9)
5.8 (3.3)
Not measured – all participants were community ambulators
[48–71]
[1–11]
*Normal gait speed for women aged 60–69 = 1.16 m/s, for men = 1.28 m/s.32
that differed from the developed model were considered for inclusion into the model. In particular, a substantial portion of the focus group data referred to the importance of environmental factors (not captured in the original SEM). The International Classification of Functioning (ICF) environment chapters31 were first considered as a framework for coding the environmental factors data, and once we confirmed this to be a reasonable coding frame, the chapters were subsequently adopted. Trustworthiness of the findings were addressed through triangulation (the use of two primary researchers involved in data analysis, and the use of two focus groups of stroke survivors), by considering various theories in the interpretation of the data, by clearly describing the focus group participants, and by using the participants’ direct quotes in supporting the development of the categories.
Ethics Ethics approval was received from the Health Research Ethics Board at the University of Manitoba and participants in the focus groups provided informed consent prior to proceeding.
Results Quantitative phase Participants. Data from 227 participants were included in the model, using the baseline assessment
time. See Table 2 for characteristics of the participants. A total of 92% of the participants had received rehabilitation after stroke. Variables. The initial measurement model included four latent variables: indoor mobility, health perceptions, gait speed, and outdoor mobility. Two measured variables were removed from the model as the paths were not significant (Table 1). The latent variables indoor mobility and outdoor mobility were correlated at 0.91, indicating they are measuring the same thing, and thus were combined and renamed ‘ambulation’. Therefore the three latent variables in the final measurement model were ambulation, gait speed, and health perceptions. Analysis. The measurement model had a reasonable fit: χ2 = 111.90, df = 62, p = 0.000, normed χ2 = 1.81, RMSEA = 0.060 (0.042; 0.077), CFI = 0.94. The p values for the χ2 do not suggest a good fit; a significant χ2 frequently occurs with large sample sizes.28 It is therefore important to use normed χ2 and other fit indices. For the structural model, variables of age and gender did not demonstrate significant paths with the latent variable ‘ambulation’. Only the addition of depression demonstrated a significant path with the latent variable ‘health perceptions’. The structural model had a reasonable fit: χ2 = 135.04, df = 75, p = 0.000, normed χ2 = 1.80, RMSEA = 0.060 (0.043; 0.075), CFI = 0.94. See Figure 1 for the structural model.
514
Clinical Rehabilitation 29(5) Move in home
0.40
Walking
0.81
Stairs
0.62
Physical ac vity Move in community
Maximal
0.70 0.94
Comfortable
Ambula on
0.60
0.29
Walk uphill
0.30
Walk for errands
0.99
0.64
Walk fast
Walk leisurely
Gait speed
χ2 = 135.04, df=75, p=0.000, normed χ2 = 1.8, RMSEA=0.060(0.043; 0.075), CFI=0.94
0.46
0.21 0.42 0.91
Health percep ons
Health state Health today
-0.46
Depression
0.28
Figure 1. Community ambulation model (standardized model).
Qualitative phase Participants. A total of 11 participants were involved in the focus groups (Table 2). Of these 11 participants, five had been participants in the trial that was used to develop the model. After completing analysis on the second focus group, the researchers agreed that data saturation had been achieved. Analysis. All variables included in the final model (Figure 1) were verified as relevant and important by participants in both focus groups. Aspects that were added through the analysis process included discussion of the endurance aspect of ambulation, the importance of the environment in promoting or preventing community ambulation, and two sets of cognitive strategies used to make decisions about community ambulation – preplanning and goal-setting. Participants described the myriad of locations that they walked, including places where the activity they engaged in required walking (e.g. malls, grocery stores, cutting the grass), locations that they chose to walk for leisure or exercise (e.g. walking trails, around the block, indoor tracks),
and locations they had to walk to get to (e.g. to the bus stop, to the car, to appointments from the car). They identified specific community locations where, by the nature of the set-up, the demand for walking was high: Yeah, the main walking is the big stores. I mean they’re so enormous that you get a mile and a half out of those anyway just to get from one department to the next.
Health perception, ambulation, and depression Health perception related to participant’s opinion of their current state of health. They supported how having had a stroke was linked to mood and affect for many people: Speaking of anxiety and depression, I believe it goes hand in hand with strokes and I think if you have anxiety and depression it does make you stay home a lot.
However, participants also clearly linked perception of their health with ambulation following stroke:
515
Barclay et al. I had stiffening of muscles after my stroke and it took, I had to be on T3s for about the first few months … within the first year of my stroke and I found the more that I walked and I stayed mobile, the less my muscles would stiffen and the less they would ache.
having adequate endurance to ambulate in their communities. They often viewed improved endurance as an important marker of their improvement and recovery post-stroke:
For most participants, the positive physical, emotional, and social benefits of ambulation were reinforced by engaging in walking, i.e. if they walked, they felt better:
… at first I was making it 3 houses and then back and then I’d have to lie down. And then it was 5 houses and then to the corner and back and then to the other corner and back, and as I did this, it was more and more time I was able to start having conversations instead of just thinking, don’t fall down, don’t fall down.
I’m not sure how much of it is that I’m not in the house anymore, that I’m out doing something with people or how much is the actual movement that’s making me start to feel better, you know, maybe the actual act of doing is part of me cheering up a little bit.
However, one participant was clearly an exception to this positive reinforcement. This participant experienced significant pain when walking, which created a cycle of negative reinforcement: I’ve been trying to figure out, and it’s gotten worse, rather than getting better. I mean I could walk better two months after my stroke than I can now and every time I try to extend my range, it got worse.
Gait speed Gait speed was affected for almost all of the participants, and was highlighted as an important factor related to safely and successfully crossing streets:
Environment Discussion of environmental factors as constraining or facilitating participants’ abilities to engage in community ambulation featured prominently in the focus group discussion. All five chapters of the Environment in the ICF were supported (products and technology, natural environment and human made changes to environment, supports and relationships, attitudes, and service systems policies).31 Ambulation aids and devices (canes, ice picks, devices to increase shoe traction), accessibility of environments, existence of benches and stairs, and available public washrooms were all considered as products and technologies relevant to community ambulation. Selection of ambulation routes was highly dependent on the existence of these types of features:
Having time to cross a major intersection is a problem. I often find that, like even I’m doing quite well now, and to get across in the amount of time the light is, can be kind of dangerous.
I found it very helpful when I was at home in the first year of my stroke to walk along a route that had a lot of benches because I had to stop and sit and rest frequently. So if there was a route to a grocery store I needed to go to, it had a lot of benches.
In the focus group discussions, gait speed also related to the ability to “keep up” with others they hoped to walk with:
Participants identified how weather conditions impacted their ability to walk, how they adapted their community ambulation in demanding conditions.
Now I walk around the neighbourhood all the time with my wife [but] I couldn’t after I had my stroke because I was just too slow.
Walking indoors was better because of the weather conditions but it’s also more populated, but mind you on the walkways, they have railings and you can hold on to those railings, so that’s handy, so it’s better all around.
Endurance While not present in the model constructed through SEM, participants highlighted the importance of
Crowds were a particular challenge that participants faced, and often avoided. They described
516 how they alternatively bumped into others or others bumped into them, creating instability and increasing the chance of falls. The social aspects of community ambulation were highlighted throughout the focus groups. For example, one participant described the benefits she gained when ambulating in the community as follows: I’d go there once a day, twice a day, for my normal walk route, and I really enjoyed it because people would often be out walking their dogs and they would, I was obviously struggling to walk but they would see me every day do this and they would always give me a smile of encouragement and that doesn’t only feel better about my achievements and about myself, but it also, just emotionally felt better period.
The importance of peer support in taking risks in the community was stressed by some participants. One man described, with humor, how he and a peer had gone golfing together after they had both had a stroke, and the challenges they faced: So here we are laughing at each other because we both fell down, and we’re both on the ground, so who’s helping who? So anyway, we got up. But it was funny. We were laughing. It was a hoot. Here we are both lying in the bush.
Public attitudes and perceptions were prevalent in the community ambulation experiences conveyed by participants; however, often these situations were poignantly described as events that caused embarrassment or distress. In some cases, the public was described as overly helpful: I’ve fallen seven times in two years and it was always out in the open, just walking down the street, klunk. You draw a crowd real quick of people trying to get you up on your feet. Nobody really checks to see if you’re hurt, they pick you up, get you back on your feet.
In other situations, ignorance and lack of awareness seemed to be the greater issue: I’ve been knocked down more than once. Simply knocked to the ground by people pushing past me and I can’t catch my balance … Somebody will come,
Clinical Rehabilitation 29(5) shove me down, they’ll keep going and say, get out of my way, you drunk.
Available public services, such as snow-clearing, well-maintained sidewalks, and affordable public facilities, were important facilitators to community ambulation. Oftentimes, participants reported that facilities where they might be able to walk indoors were too expensive, or programs were not set up for people with disabilities to easily partake in. The city programs, city spaces, cause I wound up walking around some of the pool tracks, they’re pretty good, but again you know, you’re paying five bucks [accessible public transportation] to get back and forth or your gas money, and then you’re paying for, I mean to go to the swimming pool is five dollars a shot if you buy a pack of tickets. So we’re not talking a minor expense.
Self-awareness and environment With an average time since stroke of six years, participants had a strong sense of self-awareness of their ambulation abilities. They could clearly identify ambulation activities that challenged their individual situation: Sometimes though when I get overly tired, to this day, I, walking is not a problem but all of a sudden going up the stairs, and it’s only when I’m exhausted, you know, when I’m overtired.
Strategies: preplanning and goal-setting Coupling their self-awareness of their own abilities/ limitations with their knowledge of the community environments in which they planned to ambulate, participants engaged in two sets of cognitive strategies, goal-setting and preplanning. These strategies were evident in participants’ approach to, and decisions about, community ambulation. Goal-setting was a strategy used by participants to regain lost function or ability. When participants engaged in goal-setting, they sought to challenge their ambulation abilities by selecting environments that created an increased demand on their current abilities, and then graded their efforts towards a goal:
Barclay et al. Female: Well my husband dropped me off at (shopping mall) first thing in the morning before they open up, so there’s nobody there, only the people that’s walking. And slowly, step by step, I learned to walk. Before I can’t go around, I used to sit down all the time. JR: On the benches along the way? Female: Yeah, ‘til I was able to walk around. JR: So that was kind of your goal, to be able to walk around the mall? Female: Yeah, and not have to sit down.
In contrast, in preplanning, participants matched their ambulation abilities to the known demands of the environment in order to successfully participate (rather than challenge their abilities) in a community activity, for instance, one man described how he made decisions about where to sit at the baseball game: All season, I sat in the 20th row cause I didn’t have to go down the stairs. If you want to sit by the home plate, you have to go down to the third or fourth row, and that’s pretty tough.
Discussion Based on the study findings, a unique model of community ambulation after stroke has been developed. Figure 2 depicts the revised model, verified and expanded upon based on the experiences of the focus group members who discussed challenges and facilitators regarding their own community ambulation after stroke. Community ambulation after stroke appears to be represented by associations between ambulation, gait speed, and health perceptions. As identified in the Figure 2, all of the existing relationships between variables depicted in Figure 1 were verified. However, several notable additions were made. Through focus group discussion, experience of depression was thought to relate to ambulation in a bidirectional manner (shown by the dotted arrows). The heavier dotted line, leading from ambulation to depression depicts the emphasis focus group participants placed on the impact increased ambulation has on positive affect or mood. Endurance was added by
517 focus group participants as a variable that interacts in a bidirectional manner with a person’s community ambulation (i.e. increasing ambulation increases endurance; decreased endurance limits ambulation). This model differs from a widely used model of patient outcomes, the Wilson–Cleary model, primarily with respect to the role of health perception in ambulation.33 In the Wilson–Cleary model, health perception is theorized to be a consequence of biological factors, symptoms (of depression for example), and function (here gait speed and ambulation). In our model, ambulation was depicted as a consequence of health perception. However, in this context, members of the focus groups expressed health perception as the person’s opinion of their health status, a construct that would reflect biological or physiological health, and likely influenced by the fact that they had a stroke. Self-perceived health is known to have a biological basis34 and hence its placement in this model as contributing to ambulation, rather than as a consequence. Depression as measured here represents symptoms and has a biological basis in the context of stroke, and hence its position on the model is compatible with the Wilson–Cleary model.35 All aspects of the environment, as described in the ICF,31 were discussed in the focus groups as important factors affecting community ambulation. These environmental aspects were not included in the initial SEM model (as there were no initial data available), but have been added to Figure 2 as a dotted line around the perimeter, indicating the pervasive impact of environmental factors on community ambulation. This is the advantage of combining a theory-based, statistically derived model with the experience of individuals. Patla and Shumway-Cook described a conceptual framework for mobility environmental challenges, which included the dimensions of: minimum walking distance, time constraints, ambient conditions, terrain characteristics, external physical load, attentional demands, postural transitions, and traffic level.36 All of these aspects were confirmed as contributing to community ambulation by focus group participants, based on
518
Clinical Rehabilitation 29(5)
Figure 2. Revised community ambulation model.
All parts of the model that are associated with dotted lines have been added based on focus group input.
their experiences. Neighborhoods with high walkability have been shown to be associated with increased physical activity,37 while increased precipitation, outdoor temperatures below 17 °C, and decreased day lengths are modestly associated with low levels of physical activity.38,39 This information further supports our revised model, with the inclusion of the environmental factors. It was evident from focus group participants that stroke survivors considered their own abilities within the context of their knowledge of the community environments (Figure 2). The strategies of goal-setting and preplanning in order to engage in community ambulation were situated within this knowledge (Figure 2). Similar to the current study findings, Kubina and colleagues proposed that, over time, stroke survivors use the concept of being in charge of the performance and timing of meaningful activities to enable testing of their abilities and
subsequent change in expectations and adaptation of an activity.40 The theoretical significance of this project is the development of a unique model of community ambulation post-stroke, supported and modified by the contributions from individuals with stroke on community ambulation. Learning from stroke survivors about their perceptions as to what facilitated or hindered their ability to walk in the community, in combination with the statistically developed community ambulation model and other current research in the field, will assist in developing a strategic proposal for a treatment intervention to improve community ambulation and participation. The experiences of the focus group members also reinforce the value of the model in clinical and community use. For example, to assess community ambulation, a clinician may consider evaluating
519
Barclay et al. components of the community ambulation model, such as ambulation, gait speed, and health perceptions. For treatment intervention, a clinician may consider focusing on components of the model to improve overall community ambulation, for example, practicing stairs or walking with increased speed. Practical applications also include the importance of evaluating the environment, considering the importance of community ambulation and social connectivity for our clients, and assisting our clients with goal-setting and developing preplanning strategies.
Limitations The model was developed using secondary data analysis. We utilized a unique and relevant source of data; however, only the variables collected could be used, and no other aspects of community ambulation were able to be added. Although this is a limitation, the use of focus groups with stroke survivors to verify and modify the model gives strength to the approach we used. It is important to note that with any structural model, equivalent models are possible.27 The model could be described with the paths between the latent variables and the depression variable in different directions, and the fit would be the same. The direction of the paths chosen was based on the literature and clinical experience. In summary, the model of community ambulation developed using SEM consists of the constructs of ambulation, gait speed, health perceptions, and depression. The model was augmented by incorporating the perceptions of stroke survivors, and the revised model included aspects of the environment which have an overall effect on community ambulation, endurance, and an individual’s self-awareness of their ambulatory abilities, which lead to individual goal-setting and preplanning related to community ambulation. Using the community ambulation model to identify important aspects of community ambulation may assist occupational therapists and physiotherapists in determining community ambulation goals, needs, and opportunities in collaboration with clients.
Clinical messages •• Community ambulation after stroke is represented by associations between ambulation, gait speed, health perceptions, and depression. Understanding one’s own ambulatory abilities may lead to goal-setting and preplanning for the individual, while endurance is an important aspect of community ambulation, and factors in the environment have a pervasive impact of an individual’s ambulatory ability in the community. •• Clinical application of the model includes: the importance of evaluating the environment, considering the importance of community ambulation and social connectivity, and assisting our clients with goal-setting and developing preplanning strategies. Acknowledgements Thanks to Brenden Dufault from the Biostatistical Consulting Unit, University of Manitoba for statistical assistance.
Conflict of interest The authors declare that there is no conflict of interest.
Funding Funding support for this study was from an operating grant from the Manitoba Health Research Council. Funding support for Ruth Barclay was from an establishment grant from the Manitoba Health Research Council.
References 1. Lord S, McPherson K, McNaughton H, Rochester L and Weatherall M. Community ambulation after stroke: how important and obtainable is it and what measures appear predictive? Arch Phys Med Rehabil 2004; 85: 234–239. 2. Ripat J, Redmond J and Grabowecky B. The winter walkability project: Occupational therapists’ role in promoting citizen engagement. Can J Occupat Ther 2010; 77(1): 7–14. 3. Canadian Society of Exercise Physiologists. Canadian Physical Activity Guidelines for older adults – 65 years and
520
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Clinical Rehabilitation 29(5) older. 2012. wwwcsepca/guidelines (accessed December 5, 2013). Pohl P, Duncan P, Perera S, Liu W, Lai S and Studenski S. Influence of stroke-related impairments on performance in 6-minute walk test. J Rehabil Res Dev 2002; 39: 439–444. Pohl P, Perera S, Duncan P, Maletsky R, Whitman R and Studenski S. Gains in distance walking in a 3-month follow-up poststroke: what changes? Neurorehabil Neural Repair 2004; 18: 30–36. Rantakokko M, Manty M, Iwarsson S, et al. Fear of moving outdoors and development of outdoor walking difficulty in older people. J Am Geriatr Soc 2009; 57: 634–640. Kerr J, Sallis J, Salelens B, et al. Outdoor physical activity and self rated health in older adults living in two regions of the US. Int J Behav Nutr Phys Act 2012; 9(89). Perry J, Garrett M, Gronley J and Mulroy S. Classification of walking handicap in the stroke population. Stroke 1995; 26: 982–989. Rosa M, Marques A, Demain S and Metcalf C. Fast gait speed and self-perceived balance as valid predictors and discriminators of independent community walking at 6 months poststroke – a preliminary study. Disabil Rehabil. Epub ahead of print 23 April 2014. DOI: 10.3109/09638288.2014.911969. Fulk G, Reynolds C, Mondal S and Deutsch J. Predicting home and community walking activity in people with stroke. Arch Phys Med Rehabil 2010; 91: 1582–1586. Robinson C, A S-C, Ciol M and Kartin D. Participation in community walking following stroke: Subjective versus objective measures and the impact of personal factors. Phys Ther 2011; 91(12): 1865–1876. Lord S, Rochester L, Weatherall M, McPherson K and McNaughton H. The effect of evironment and task on gait parameters after stroke: a randomized comparison of measurement conditions. Arch Phys Med Rehabil 2006; 87: 967–973. Talkowski J, Brach J, Studenski S and Newman A. Impact of health perception, balance perception, fall history, balance performance, and gait speed on walking activity in older adults. Phys Ther 2008; 88: 1474–1481. Sakari R, Era P, Rantanen T, Leskinen E and Laukkanen P. Mobility performance and its sensory, psychomotor and musculoskeletal determinants from age 75 to age 80. Aging Clin Experi Res 2010; 22: 47–53. Li K, Cardinal B and Vuchinich S. Health worry, physical activity participation, and walking difficulty among older adults: A mediation analysis. Res Quart Exerc Sport 2009; 80(1): 12–21. Andrews A, Chinworth S, Bourassa M, Garvin M, Benton D and Tanner S. Update on distance and velocity requirements for community ambulation. J Geriatr Phys Ther 2010; 33: 128–134. Dennett A, Taylor N and Mulrain K. Community ambulation after hip fracture: completing tasks to enable access
to common community venues. Disabil Rehabil 2012; 34: 707–717. 18. Creswell J and VL PC. Designing and Conducting Mixed Methods Research. Thousand Oaks, CA: SAGE Publications, 2007. 19. Barclay-Goddard R, Ripat J and Mayo N. Developing a model of participation post-stroke: a mixed-methods approach. Qual Life Res 2012; 21(3): 417–426. 20. Dorman P, Slattery J, Farrell B, Dennis M and Sandercock P. Qualitative comparison of the reliability of health status assessments with the EuroQol and the SF-36 questionnaires after stroke. Stroke 1998; 29: 63–68. 21. Poissant L, Mayo N, Wood-Dauphinee S and Clarke A. The development and preliminary validation of a Preference-Based Stroke Index (PBSI). Health Qual Life Outcomes 2003; 1(43). 22. Steffen T, Hacker T and Mollinger L. Age- and genderrelated test performance in community-dwelling elderly people: Six-minute Walk Test, Berg Balance Scale, Timed Up & Go Test, and gait speeds. Phys Ther 2002; 82: 128–137. 23. Wood-Dauphinee S, Opzoomer M, Williams J, Marchand B and Spitzer W. Assessment of global function: The reintegration to normal living index. Arch Phys Med Rehabil 1988; 69(8): 583–590. 24. Stewart A, Mills K, KIng A, Haskell W, Gillis D and Ritter P. CHAMPS Physical Activity Questionnaire for older adults: Outcomes for interventions. Med Sci Sports Exerc 2001; 33(7): 1126–41. 25. Cinamon J, Finch L, Miller S, Higgins J and Mayo N. Preliminary evidence for the development of a stroke specific geriatric depression scale. Int J Geriatr Psychiatry 2011; 26: 188–198. 26. Enders C. Analyzing structural equation models with missing data. In: Hancock G and Mueller R (eds) Structural Equation Modeling – A Second Course. Quantitative Methods in Education and the Behavioural Sciences: Issues, Research, and Teaching. Greenwich: Information Age Publishing, 2006. pp.313–342. 27. Kline R. Principles and Practice of Structural Equation Modeling. 2nd ed. New York: The Guilford Press, 2005. 28. Hays R, Revicki D and Coyne K. Application of structural equation modeling to health outcomes research. Eval Health Prof 2005; 28(3): 295–309. 29. Patton M. Qualitative Research and Evaluation Methods. 3rd ed. Thousand Oaks, CA: SAGE Publications, 2002. 30. Hsieh H and Shannon S. Three approaches to qualitative content analysis. Qual Health Res 2005; 15(9): 1277–1288. 31. World Health Organization. International Classification of Functioning, Disability and Health. Geneva: World Health Organization, 2001. 32. Öberg T, Karsznia A and Öberg K. Basic gait parameters: Reference data for normal subjects. J Rehabil Res Dev 1993; 30(2): 210–223.
Barclay et al. 33. Wilson I and Cleary P. Linking clinical variables with health-related quality of life. A conceptual model of patient outcomes. JAMA 1995; 273: 59–65. 34. Undėn AL, Andrėasson A, Elofsson S, et al. Inflammatory cytokines, behaviour and age as determinants of self-rated health in women. Clin Sci 2007; 112: 363–373. 35. Yu L, Liu CK, Chen JW, Wang SY, Wu YH and SH Y. Relationship between post-stroke depression and lesion location: a meta-analysis. Kaohsiung J Med Sci 2004; 20(8): 372–380. 36. Patla A and Shumway-Cook A. Dimensions of mobility: Defining the complexity and difficulty associated with community walking. J Aging Phys Act 1999; 7: 7–19.
521 37. King A, Sallis J, LD Frank, et al. Aging in neighborhoods differing in walkability and income: Associations with physical activity and obesity in older adults. Social Sci Med 2011; 73: 1525–1533. 38. Aoyagi Y and Shaphard R. Steps per day – The road to senior health? Sports Med 2009; 39(6): 423–438. 39. Sumukadas D, Witham M, Struthers A and McMurdo M. Day length and weather conditions profoundly affect physical activity levels in older functionally impaired people. J Epidemiol Comm Health 2009; 63: 305–309. 40. Kubina L, Dubouloz C, Davis C, Kessler D and Egan M. The process of re-engagement in pesonally valued activities during the two years following stroke. Disabil Rehabil 2013; 35(3): 236–243.
Copyright of Clinical Rehabilitation is the property of Sage Publications, Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
research-article2014
CRE0010.1177/0269215514546769Clinical RehabilitationBarclay et al.
CLINICAL REHABILITATION
Article
Factors describing community ambulation after stroke: a mixedmethods study
Clinical Rehabilitation 2015, Vol. 29(5) 509–521 © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0269215514546769 cre.sagepub.com
Ruth Barclay1, Jacquie Ripat2 and Nancy Mayo3,4
Abstract Objective: To develop a model of community ambulation after stroke based on: Canadian data from community-dwelling individuals post-stroke; the experiences and opportunities for community ambulation expressed by individuals with stroke; and current literature. The model presents a visual depiction of the relationships between the different factors of community ambulation after stroke. Design: A quantitative/qualitative explanatory sequential mixed-methods design was utilized. Secondary data analysis with structural equation modeling resulted in a community ambulation model. Two focus groups of individuals with stroke were conducted to verify and explain the model. Setting: Community. Subjects: Quantitative data from 227 participants: 142 (63%) male; 63.4 (12.0) years of age and 2.6 (2.5) years post stroke. Eleven individuals participated in the focus groups: 6 (55%) male; 61.4 (6.9) years of age and 5.8 (3.3) years since stroke. Main measures: Model variables: items from the EuroQol, Preference Based Stroke Index, gait speed, Reintegration to Normal Living Index, the Community Health Activities Model Program for Seniors, and the Geriatric Depression Scale. Results: The model had reasonable fit with three latent variables: ambulation, gait speed, and health perceptions (normed χ2 = 1.8, root mean square error of approximation = 0.060 (0.043; 0.075)). Depression was also a component of community ambulation. Participants verified the model and added endurance and the environment as additional components. Participants used self-awareness and knowledge of the environment to engage in cognitive strategies related to community ambulation. Conclusions A model of community ambulation after stroke was developed and verified. Recognizing important components of community ambulation may assist physiotherapists in determining community ambulation goals, needs, and opportunities in partnership with clients.
1Department
of Physical Therapy, University of Manitoba, Winnipeg, Manitoba, Canada 2Department of Occupational Therapy, University of Winnipeg, Manitoba, Manitoba, Canada 3Division of Clinical Epidemiology, Department of Medicine, McGill University, Montreal, Quebec, Canada
4Research
Scientist, McGill University Health Centre Research Institute, Montreal, Quebec, Canada
Corresponding author: Ruth Barclay, Department of Physical Therapy, University of Manitoba, R106-771 McDermot Ave., Winnipeg, Manitoba R3E 0T6, Canada. Email: [email protected]
510
Clinical Rehabilitation 29(5)
Keywords Community ambulation, participation, stroke Received: 9 December 2013; accepted: 6 July 2014
Introduction “…walking around the living room isn’t going to cut it”. Focus group participant
Walking in the community is a key activity for many individuals after stroke: in a survey of 130 individuals surveyed with stroke, approximately 75% specified that community ambulation was considered essential or very important.1 Community ambulation incorporates indoor and outdoor walking, as well as a social aspect, and has been defined as “independent mobility outside the home, which includes the ability to confidently negotiate uneven terrain, private venues, shopping centers and other public venues”.1 In one study, approximately onethird of community-dwelling individuals with stroke reported not being able to walk outdoors independently.1 Of those who were independent in their community ambulation, most were hesitant to use public transportation and relied on others for transportation, which could in turn further limit community ambulation.1 Benefits of walking outside include “transportation, health benefits, and recreation”,2 increased physical activity,3 and improved balance, endurance, and walking ability.4,5 Community ambulation enables participation in activities that are meaningful to the individual,6 and is associated with self-rated health.7 A number of factors have been associated with the ability to be a community ambulator after stroke, these include: faster gait speed,1,8,9 increased endurance,10 falls self-efficacy,9,11 and a supportive environment (e.g. walking location).12 While not specific to stroke, in older adults gait speed, health perception, and balance perception have all been shown to be associated with the distance walked outside.13 Gait speed and stepping height represented mobility in a study of adults age 75–80 in Finland,14 while health worry (anxiety from a
concern for health, or health perception) predicted physical activity, which in turn predicted walking difficulty in older American adults.15 Walking speed and endurance are considered key elements of community ambulation as they directly relate to crossing the street in the time of a walk signal and walking sufficient distances to complete essential activities such as shopping.16,17 A visual depiction, or model, of the relationships between the different components of community ambulation after stroke (impairments, activity limitations, participation restrictions, and contextual factors of the person and environment) would be helpful for both clinicians and researchers. This model could be used in a clinical context to help formulate client goals and to direct assessment and treatment strategies with the client wishing to improve community ambulation after stroke. The model could also be helpful in forming research questions about improving community ambulation after stroke. The purpose of this research was to develop a model of community ambulation after stroke based on: Canadian data from community dwelling individuals post-stroke, the experiences and opportunities for community ambulation expressed by individuals with stroke, and current literature.
Methods Overview We used an explanatory sequential mixed-methods design18 to guide this study. The first (quantitative) phase of the study was the development of a model of community ambulation. The second, qualitative phase, followed up on the quantitative phase for the purpose of confirmation and revision of the model of community ambulation from the
511
Barclay et al. Table 1. Health indices and items in proposed model. Health index
Constructs (latent variables) Indoor mobility
Outdoor mobility
Health perceptions
Gait speed
VAS health state Health today
Comfortable Maximal
(combined for final model) EQ5D PBSI
Mobility Walking Stairs Physical activities
CHAMPS
RNL 5-m gait speed
Move in home
Jog or run Walk uphill Walk fast for exercise Walk to do errands Walk leisurely for exercise or pleasure Move in community
EQ5D: EuroQol; CHAMPS: Community Health Activities Model Program for Seniors; PBSI: Preference-Based Stroke Index; RNL: Reintegration to Normal Living Index; VAS: visual analogue scale. Italicized variables were removed from model due to their poor fit.
perspective of individuals with stroke. The benefit of using a mixed methods approach in this situation was that the qualitative experiences of stroke survivors could be used to explain and add to the quantitative theory-based and statistically derived model of community ambulation. The authors have used this method of developing a quantitative model and verifying it with qualitative data previously in developing a model of participation after stroke.19
Quantitative phase – developing the initial model
(RNL),23 the Community Health Activities Model Program for Seniors (CHAMPS),24 and the Rasch version of the Geriatric Depression Scale for individuals with stroke.25 The variables used in developing the model were based on current literature, as described above, with the qualification that the variables had also been measured in the trial. The latent variables of indoor mobility, health perceptions, gait speed, and outdoor mobility were chosen, as they summarized the available ambulation-related variables. Table 1 provides a summary of how each item or subscale was associated with each latent variable.
Participants. The quantitative phase involved secondary data analysis of a multi-center randomized controlled trial of participation of 227 Canadians with stroke living in the community.
Procedure and analysis. Structural equation modeling (SEM) was used to develop the model. Statistical analysis was completed with SPSS and AMOS, with the following steps, using SEM.
Variables. Several variables related to ambulation had been collected during the study. Measured variables that were available in the data and relevant to this study include those from the EuroQol (EQ5D),20 Preference-Based Stroke Index (PBSI),21 5-meter gait speed,22 Reintegration to Normal Living Index
1. A measurement model was developed to demonstrate associations between the latent variables of indoor mobility, health perceptions, gait speed, and outdoor mobility and the related measured variables. This was completed using the baseline evaluation time.
512 2. A structural model (associations and paths showing directionality, but not cause) was developed from the measurement model. Variables of age, sex, and depression were added at this point. As is common in studies with multiple outcomes, not all participants answered all items for all measures. For this reason, Full Information Maximum Likelihood (FIML) estimation was used, as this method accounts for missing variables.26 Missing data is not imputed with FIML, as parameter estimates are estimated from the raw data.26 Fit indices are used to determine the fit of a model. The model chi-square (χ2) is commonly reported in the SEM literature. As the χ2 increases, model fit worsens.27 The χ2 tests the difference between a model that has a perfect population fit and the observed model.27 A χ2 that is non-significant means that there is little difference between the models; this suggests a good fit. Models with large sample sizes can often be rejected since the model χ2 is affected by sample size.28 A variety of fit indices are therefore used. Some researchers use a normed χ2 (χ2/degrees of freedom) to decrease the effect of the sample size on the χ2. Normed χ2 values up to 5.0 have been used to suggest a reasonable model fit.27 The root mean square error of approximation (RMSEA) measures the lack of fit in a model compared with the population. A close fit is suggested with an RMSEA of ⩽0.05; a reasonable fit from 0.05 to 0.08; and ⩾0.10 a poor fit.27 The comparative fit index (CFI) is an incremental fit index in which a reasonably good fit is suggested when the CFI is greater than 0.90.27
Qualitative phase – Exploring community ambulation and verifying the model Two focus groups were conducted to explore community ambulation from the perspective of individuals with stroke who live in the community. Each focus group ran for approximately two hours. Participants. Individuals were recruited through a local Stroke Recovery association, and by contacting participants from previous studies who had
Clinical Rehabilitation 29(5) indicated their willingness to be re-contacted for research purposes. Inclusion criteria were: adults with stroke who were living in the community and were community ambulators by self-report. We conducted purposive sampling with the intent to “bring together people of similar backgrounds and experiences to participate in a group interview about major issues that affect them” (Patton, page 236).29 While participants varied in terms of sex, age, and length of time since stroke, all indicated that they were community ambulators, and all had gone through their stroke rehabilitation within the same regional health authority. Procedure and analysis. A directed content approach30 was used to guide development of the focus group questions and an initial coding scheme; this approach is useful when theory exists that suggests relationships among variables (in this case, the community ambulation model). The community ambulation model was not shared with participants prior to, or during, the focus group in an attempt to elicit a range of divergent and confirmatory opinions. Focus group questions were openended, with additional probes used to address aspects of community ambulation established in the model, as well as other factors in the literature but not included in the developed model.30 Questions addressed participant-perceived barriers and facilitators to community ambulation, reasons why participants did or did not walk outside, and situations that encouraged or hindered their ability to walk outdoors. Focus groups were facilitated by the second author (JR) and detailed field notes were taken by a research assistant. Focus groups were digitally audio-recorded and all data was transcribed verbatim. The transcripts were reviewed and coded, first individually and then collaboratively by RB and JR, using a set of categories based on the model while staying open to the inclusion of additional emergent categories. Field notes were used to corroborate category assignment. To verify the model of community ambulation post-stroke, the final stage of mixed methods analysis was theory driven, focusing on linking the emergent categories to the model of community ambulation that was developed with SEM. Categories
513
Barclay et al. Table 2. Description of the study participants – quantitative and qualitative phases. Gender freq (%) Quantitative phase n = 227 Male 142 (63) Qualitative phase n = 11 Male 6 (55)
Age in years Mean (SD) [range]
Years post-stroke Mean (SD) [range]
Comfortable gait speed meters/second Mean (SD) [range]
63.4 (12.0) [24–93]
2.6 (2.5) [0.1–15.7]
0.78 (0.40)* [0–2.0]
61.4 (6.9)
5.8 (3.3)
Not measured – all participants were community ambulators
[48–71]
[1–11]
*Normal gait speed for women aged 60–69 = 1.16 m/s, for men = 1.28 m/s.32
that differed from the developed model were considered for inclusion into the model. In particular, a substantial portion of the focus group data referred to the importance of environmental factors (not captured in the original SEM). The International Classification of Functioning (ICF) environment chapters31 were first considered as a framework for coding the environmental factors data, and once we confirmed this to be a reasonable coding frame, the chapters were subsequently adopted. Trustworthiness of the findings were addressed through triangulation (the use of two primary researchers involved in data analysis, and the use of two focus groups of stroke survivors), by considering various theories in the interpretation of the data, by clearly describing the focus group participants, and by using the participants’ direct quotes in supporting the development of the categories.
Ethics Ethics approval was received from the Health Research Ethics Board at the University of Manitoba and participants in the focus groups provided informed consent prior to proceeding.
Results Quantitative phase Participants. Data from 227 participants were included in the model, using the baseline assessment
time. See Table 2 for characteristics of the participants. A total of 92% of the participants had received rehabilitation after stroke. Variables. The initial measurement model included four latent variables: indoor mobility, health perceptions, gait speed, and outdoor mobility. Two measured variables were removed from the model as the paths were not significant (Table 1). The latent variables indoor mobility and outdoor mobility were correlated at 0.91, indicating they are measuring the same thing, and thus were combined and renamed ‘ambulation’. Therefore the three latent variables in the final measurement model were ambulation, gait speed, and health perceptions. Analysis. The measurement model had a reasonable fit: χ2 = 111.90, df = 62, p = 0.000, normed χ2 = 1.81, RMSEA = 0.060 (0.042; 0.077), CFI = 0.94. The p values for the χ2 do not suggest a good fit; a significant χ2 frequently occurs with large sample sizes.28 It is therefore important to use normed χ2 and other fit indices. For the structural model, variables of age and gender did not demonstrate significant paths with the latent variable ‘ambulation’. Only the addition of depression demonstrated a significant path with the latent variable ‘health perceptions’. The structural model had a reasonable fit: χ2 = 135.04, df = 75, p = 0.000, normed χ2 = 1.80, RMSEA = 0.060 (0.043; 0.075), CFI = 0.94. See Figure 1 for the structural model.
514
Clinical Rehabilitation 29(5) Move in home
0.40
Walking
0.81
Stairs
0.62
Physical ac vity Move in community
Maximal
0.70 0.94
Comfortable
Ambula on
0.60
0.29
Walk uphill
0.30
Walk for errands
0.99
0.64
Walk fast
Walk leisurely
Gait speed
χ2 = 135.04, df=75, p=0.000, normed χ2 = 1.8, RMSEA=0.060(0.043; 0.075), CFI=0.94
0.46
0.21 0.42 0.91
Health percep ons
Health state Health today
-0.46
Depression
0.28
Figure 1. Community ambulation model (standardized model).
Qualitative phase Participants. A total of 11 participants were involved in the focus groups (Table 2). Of these 11 participants, five had been participants in the trial that was used to develop the model. After completing analysis on the second focus group, the researchers agreed that data saturation had been achieved. Analysis. All variables included in the final model (Figure 1) were verified as relevant and important by participants in both focus groups. Aspects that were added through the analysis process included discussion of the endurance aspect of ambulation, the importance of the environment in promoting or preventing community ambulation, and two sets of cognitive strategies used to make decisions about community ambulation – preplanning and goal-setting. Participants described the myriad of locations that they walked, including places where the activity they engaged in required walking (e.g. malls, grocery stores, cutting the grass), locations that they chose to walk for leisure or exercise (e.g. walking trails, around the block, indoor tracks),
and locations they had to walk to get to (e.g. to the bus stop, to the car, to appointments from the car). They identified specific community locations where, by the nature of the set-up, the demand for walking was high: Yeah, the main walking is the big stores. I mean they’re so enormous that you get a mile and a half out of those anyway just to get from one department to the next.
Health perception, ambulation, and depression Health perception related to participant’s opinion of their current state of health. They supported how having had a stroke was linked to mood and affect for many people: Speaking of anxiety and depression, I believe it goes hand in hand with strokes and I think if you have anxiety and depression it does make you stay home a lot.
However, participants also clearly linked perception of their health with ambulation following stroke:
515
Barclay et al. I had stiffening of muscles after my stroke and it took, I had to be on T3s for about the first few months … within the first year of my stroke and I found the more that I walked and I stayed mobile, the less my muscles would stiffen and the less they would ache.
having adequate endurance to ambulate in their communities. They often viewed improved endurance as an important marker of their improvement and recovery post-stroke:
For most participants, the positive physical, emotional, and social benefits of ambulation were reinforced by engaging in walking, i.e. if they walked, they felt better:
… at first I was making it 3 houses and then back and then I’d have to lie down. And then it was 5 houses and then to the corner and back and then to the other corner and back, and as I did this, it was more and more time I was able to start having conversations instead of just thinking, don’t fall down, don’t fall down.
I’m not sure how much of it is that I’m not in the house anymore, that I’m out doing something with people or how much is the actual movement that’s making me start to feel better, you know, maybe the actual act of doing is part of me cheering up a little bit.
However, one participant was clearly an exception to this positive reinforcement. This participant experienced significant pain when walking, which created a cycle of negative reinforcement: I’ve been trying to figure out, and it’s gotten worse, rather than getting better. I mean I could walk better two months after my stroke than I can now and every time I try to extend my range, it got worse.
Gait speed Gait speed was affected for almost all of the participants, and was highlighted as an important factor related to safely and successfully crossing streets:
Environment Discussion of environmental factors as constraining or facilitating participants’ abilities to engage in community ambulation featured prominently in the focus group discussion. All five chapters of the Environment in the ICF were supported (products and technology, natural environment and human made changes to environment, supports and relationships, attitudes, and service systems policies).31 Ambulation aids and devices (canes, ice picks, devices to increase shoe traction), accessibility of environments, existence of benches and stairs, and available public washrooms were all considered as products and technologies relevant to community ambulation. Selection of ambulation routes was highly dependent on the existence of these types of features:
Having time to cross a major intersection is a problem. I often find that, like even I’m doing quite well now, and to get across in the amount of time the light is, can be kind of dangerous.
I found it very helpful when I was at home in the first year of my stroke to walk along a route that had a lot of benches because I had to stop and sit and rest frequently. So if there was a route to a grocery store I needed to go to, it had a lot of benches.
In the focus group discussions, gait speed also related to the ability to “keep up” with others they hoped to walk with:
Participants identified how weather conditions impacted their ability to walk, how they adapted their community ambulation in demanding conditions.
Now I walk around the neighbourhood all the time with my wife [but] I couldn’t after I had my stroke because I was just too slow.
Walking indoors was better because of the weather conditions but it’s also more populated, but mind you on the walkways, they have railings and you can hold on to those railings, so that’s handy, so it’s better all around.
Endurance While not present in the model constructed through SEM, participants highlighted the importance of
Crowds were a particular challenge that participants faced, and often avoided. They described
516 how they alternatively bumped into others or others bumped into them, creating instability and increasing the chance of falls. The social aspects of community ambulation were highlighted throughout the focus groups. For example, one participant described the benefits she gained when ambulating in the community as follows: I’d go there once a day, twice a day, for my normal walk route, and I really enjoyed it because people would often be out walking their dogs and they would, I was obviously struggling to walk but they would see me every day do this and they would always give me a smile of encouragement and that doesn’t only feel better about my achievements and about myself, but it also, just emotionally felt better period.
The importance of peer support in taking risks in the community was stressed by some participants. One man described, with humor, how he and a peer had gone golfing together after they had both had a stroke, and the challenges they faced: So here we are laughing at each other because we both fell down, and we’re both on the ground, so who’s helping who? So anyway, we got up. But it was funny. We were laughing. It was a hoot. Here we are both lying in the bush.
Public attitudes and perceptions were prevalent in the community ambulation experiences conveyed by participants; however, often these situations were poignantly described as events that caused embarrassment or distress. In some cases, the public was described as overly helpful: I’ve fallen seven times in two years and it was always out in the open, just walking down the street, klunk. You draw a crowd real quick of people trying to get you up on your feet. Nobody really checks to see if you’re hurt, they pick you up, get you back on your feet.
In other situations, ignorance and lack of awareness seemed to be the greater issue: I’ve been knocked down more than once. Simply knocked to the ground by people pushing past me and I can’t catch my balance … Somebody will come,
Clinical Rehabilitation 29(5) shove me down, they’ll keep going and say, get out of my way, you drunk.
Available public services, such as snow-clearing, well-maintained sidewalks, and affordable public facilities, were important facilitators to community ambulation. Oftentimes, participants reported that facilities where they might be able to walk indoors were too expensive, or programs were not set up for people with disabilities to easily partake in. The city programs, city spaces, cause I wound up walking around some of the pool tracks, they’re pretty good, but again you know, you’re paying five bucks [accessible public transportation] to get back and forth or your gas money, and then you’re paying for, I mean to go to the swimming pool is five dollars a shot if you buy a pack of tickets. So we’re not talking a minor expense.
Self-awareness and environment With an average time since stroke of six years, participants had a strong sense of self-awareness of their ambulation abilities. They could clearly identify ambulation activities that challenged their individual situation: Sometimes though when I get overly tired, to this day, I, walking is not a problem but all of a sudden going up the stairs, and it’s only when I’m exhausted, you know, when I’m overtired.
Strategies: preplanning and goal-setting Coupling their self-awareness of their own abilities/ limitations with their knowledge of the community environments in which they planned to ambulate, participants engaged in two sets of cognitive strategies, goal-setting and preplanning. These strategies were evident in participants’ approach to, and decisions about, community ambulation. Goal-setting was a strategy used by participants to regain lost function or ability. When participants engaged in goal-setting, they sought to challenge their ambulation abilities by selecting environments that created an increased demand on their current abilities, and then graded their efforts towards a goal:
Barclay et al. Female: Well my husband dropped me off at (shopping mall) first thing in the morning before they open up, so there’s nobody there, only the people that’s walking. And slowly, step by step, I learned to walk. Before I can’t go around, I used to sit down all the time. JR: On the benches along the way? Female: Yeah, ‘til I was able to walk around. JR: So that was kind of your goal, to be able to walk around the mall? Female: Yeah, and not have to sit down.
In contrast, in preplanning, participants matched their ambulation abilities to the known demands of the environment in order to successfully participate (rather than challenge their abilities) in a community activity, for instance, one man described how he made decisions about where to sit at the baseball game: All season, I sat in the 20th row cause I didn’t have to go down the stairs. If you want to sit by the home plate, you have to go down to the third or fourth row, and that’s pretty tough.
Discussion Based on the study findings, a unique model of community ambulation after stroke has been developed. Figure 2 depicts the revised model, verified and expanded upon based on the experiences of the focus group members who discussed challenges and facilitators regarding their own community ambulation after stroke. Community ambulation after stroke appears to be represented by associations between ambulation, gait speed, and health perceptions. As identified in the Figure 2, all of the existing relationships between variables depicted in Figure 1 were verified. However, several notable additions were made. Through focus group discussion, experience of depression was thought to relate to ambulation in a bidirectional manner (shown by the dotted arrows). The heavier dotted line, leading from ambulation to depression depicts the emphasis focus group participants placed on the impact increased ambulation has on positive affect or mood. Endurance was added by
517 focus group participants as a variable that interacts in a bidirectional manner with a person’s community ambulation (i.e. increasing ambulation increases endurance; decreased endurance limits ambulation). This model differs from a widely used model of patient outcomes, the Wilson–Cleary model, primarily with respect to the role of health perception in ambulation.33 In the Wilson–Cleary model, health perception is theorized to be a consequence of biological factors, symptoms (of depression for example), and function (here gait speed and ambulation). In our model, ambulation was depicted as a consequence of health perception. However, in this context, members of the focus groups expressed health perception as the person’s opinion of their health status, a construct that would reflect biological or physiological health, and likely influenced by the fact that they had a stroke. Self-perceived health is known to have a biological basis34 and hence its placement in this model as contributing to ambulation, rather than as a consequence. Depression as measured here represents symptoms and has a biological basis in the context of stroke, and hence its position on the model is compatible with the Wilson–Cleary model.35 All aspects of the environment, as described in the ICF,31 were discussed in the focus groups as important factors affecting community ambulation. These environmental aspects were not included in the initial SEM model (as there were no initial data available), but have been added to Figure 2 as a dotted line around the perimeter, indicating the pervasive impact of environmental factors on community ambulation. This is the advantage of combining a theory-based, statistically derived model with the experience of individuals. Patla and Shumway-Cook described a conceptual framework for mobility environmental challenges, which included the dimensions of: minimum walking distance, time constraints, ambient conditions, terrain characteristics, external physical load, attentional demands, postural transitions, and traffic level.36 All of these aspects were confirmed as contributing to community ambulation by focus group participants, based on
518
Clinical Rehabilitation 29(5)
Figure 2. Revised community ambulation model.
All parts of the model that are associated with dotted lines have been added based on focus group input.
their experiences. Neighborhoods with high walkability have been shown to be associated with increased physical activity,37 while increased precipitation, outdoor temperatures below 17 °C, and decreased day lengths are modestly associated with low levels of physical activity.38,39 This information further supports our revised model, with the inclusion of the environmental factors. It was evident from focus group participants that stroke survivors considered their own abilities within the context of their knowledge of the community environments (Figure 2). The strategies of goal-setting and preplanning in order to engage in community ambulation were situated within this knowledge (Figure 2). Similar to the current study findings, Kubina and colleagues proposed that, over time, stroke survivors use the concept of being in charge of the performance and timing of meaningful activities to enable testing of their abilities and
subsequent change in expectations and adaptation of an activity.40 The theoretical significance of this project is the development of a unique model of community ambulation post-stroke, supported and modified by the contributions from individuals with stroke on community ambulation. Learning from stroke survivors about their perceptions as to what facilitated or hindered their ability to walk in the community, in combination with the statistically developed community ambulation model and other current research in the field, will assist in developing a strategic proposal for a treatment intervention to improve community ambulation and participation. The experiences of the focus group members also reinforce the value of the model in clinical and community use. For example, to assess community ambulation, a clinician may consider evaluating
519
Barclay et al. components of the community ambulation model, such as ambulation, gait speed, and health perceptions. For treatment intervention, a clinician may consider focusing on components of the model to improve overall community ambulation, for example, practicing stairs or walking with increased speed. Practical applications also include the importance of evaluating the environment, considering the importance of community ambulation and social connectivity for our clients, and assisting our clients with goal-setting and developing preplanning strategies.
Limitations The model was developed using secondary data analysis. We utilized a unique and relevant source of data; however, only the variables collected could be used, and no other aspects of community ambulation were able to be added. Although this is a limitation, the use of focus groups with stroke survivors to verify and modify the model gives strength to the approach we used. It is important to note that with any structural model, equivalent models are possible.27 The model could be described with the paths between the latent variables and the depression variable in different directions, and the fit would be the same. The direction of the paths chosen was based on the literature and clinical experience. In summary, the model of community ambulation developed using SEM consists of the constructs of ambulation, gait speed, health perceptions, and depression. The model was augmented by incorporating the perceptions of stroke survivors, and the revised model included aspects of the environment which have an overall effect on community ambulation, endurance, and an individual’s self-awareness of their ambulatory abilities, which lead to individual goal-setting and preplanning related to community ambulation. Using the community ambulation model to identify important aspects of community ambulation may assist occupational therapists and physiotherapists in determining community ambulation goals, needs, and opportunities in collaboration with clients.
Clinical messages •• Community ambulation after stroke is represented by associations between ambulation, gait speed, health perceptions, and depression. Understanding one’s own ambulatory abilities may lead to goal-setting and preplanning for the individual, while endurance is an important aspect of community ambulation, and factors in the environment have a pervasive impact of an individual’s ambulatory ability in the community. •• Clinical application of the model includes: the importance of evaluating the environment, considering the importance of community ambulation and social connectivity, and assisting our clients with goal-setting and developing preplanning strategies. Acknowledgements Thanks to Brenden Dufault from the Biostatistical Consulting Unit, University of Manitoba for statistical assistance.
Conflict of interest The authors declare that there is no conflict of interest.
Funding Funding support for this study was from an operating grant from the Manitoba Health Research Council. Funding support for Ruth Barclay was from an establishment grant from the Manitoba Health Research Council.
References 1. Lord S, McPherson K, McNaughton H, Rochester L and Weatherall M. Community ambulation after stroke: how important and obtainable is it and what measures appear predictive? Arch Phys Med Rehabil 2004; 85: 234–239. 2. Ripat J, Redmond J and Grabowecky B. The winter walkability project: Occupational therapists’ role in promoting citizen engagement. Can J Occupat Ther 2010; 77(1): 7–14. 3. Canadian Society of Exercise Physiologists. Canadian Physical Activity Guidelines for older adults – 65 years and
520
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Clinical Rehabilitation 29(5) older. 2012. wwwcsepca/guidelines (accessed December 5, 2013). Pohl P, Duncan P, Perera S, Liu W, Lai S and Studenski S. Influence of stroke-related impairments on performance in 6-minute walk test. J Rehabil Res Dev 2002; 39: 439–444. Pohl P, Perera S, Duncan P, Maletsky R, Whitman R and Studenski S. Gains in distance walking in a 3-month follow-up poststroke: what changes? Neurorehabil Neural Repair 2004; 18: 30–36. Rantakokko M, Manty M, Iwarsson S, et al. Fear of moving outdoors and development of outdoor walking difficulty in older people. J Am Geriatr Soc 2009; 57: 634–640. Kerr J, Sallis J, Salelens B, et al. Outdoor physical activity and self rated health in older adults living in two regions of the US. Int J Behav Nutr Phys Act 2012; 9(89). Perry J, Garrett M, Gronley J and Mulroy S. Classification of walking handicap in the stroke population. Stroke 1995; 26: 982–989. Rosa M, Marques A, Demain S and Metcalf C. Fast gait speed and self-perceived balance as valid predictors and discriminators of independent community walking at 6 months poststroke – a preliminary study. Disabil Rehabil. Epub ahead of print 23 April 2014. DOI: 10.3109/09638288.2014.911969. Fulk G, Reynolds C, Mondal S and Deutsch J. Predicting home and community walking activity in people with stroke. Arch Phys Med Rehabil 2010; 91: 1582–1586. Robinson C, A S-C, Ciol M and Kartin D. Participation in community walking following stroke: Subjective versus objective measures and the impact of personal factors. Phys Ther 2011; 91(12): 1865–1876. Lord S, Rochester L, Weatherall M, McPherson K and McNaughton H. The effect of evironment and task on gait parameters after stroke: a randomized comparison of measurement conditions. Arch Phys Med Rehabil 2006; 87: 967–973. Talkowski J, Brach J, Studenski S and Newman A. Impact of health perception, balance perception, fall history, balance performance, and gait speed on walking activity in older adults. Phys Ther 2008; 88: 1474–1481. Sakari R, Era P, Rantanen T, Leskinen E and Laukkanen P. Mobility performance and its sensory, psychomotor and musculoskeletal determinants from age 75 to age 80. Aging Clin Experi Res 2010; 22: 47–53. Li K, Cardinal B and Vuchinich S. Health worry, physical activity participation, and walking difficulty among older adults: A mediation analysis. Res Quart Exerc Sport 2009; 80(1): 12–21. Andrews A, Chinworth S, Bourassa M, Garvin M, Benton D and Tanner S. Update on distance and velocity requirements for community ambulation. J Geriatr Phys Ther 2010; 33: 128–134. Dennett A, Taylor N and Mulrain K. Community ambulation after hip fracture: completing tasks to enable access
to common community venues. Disabil Rehabil 2012; 34: 707–717. 18. Creswell J and VL PC. Designing and Conducting Mixed Methods Research. Thousand Oaks, CA: SAGE Publications, 2007. 19. Barclay-Goddard R, Ripat J and Mayo N. Developing a model of participation post-stroke: a mixed-methods approach. Qual Life Res 2012; 21(3): 417–426. 20. Dorman P, Slattery J, Farrell B, Dennis M and Sandercock P. Qualitative comparison of the reliability of health status assessments with the EuroQol and the SF-36 questionnaires after stroke. Stroke 1998; 29: 63–68. 21. Poissant L, Mayo N, Wood-Dauphinee S and Clarke A. The development and preliminary validation of a Preference-Based Stroke Index (PBSI). Health Qual Life Outcomes 2003; 1(43). 22. Steffen T, Hacker T and Mollinger L. Age- and genderrelated test performance in community-dwelling elderly people: Six-minute Walk Test, Berg Balance Scale, Timed Up & Go Test, and gait speeds. Phys Ther 2002; 82: 128–137. 23. Wood-Dauphinee S, Opzoomer M, Williams J, Marchand B and Spitzer W. Assessment of global function: The reintegration to normal living index. Arch Phys Med Rehabil 1988; 69(8): 583–590. 24. Stewart A, Mills K, KIng A, Haskell W, Gillis D and Ritter P. CHAMPS Physical Activity Questionnaire for older adults: Outcomes for interventions. Med Sci Sports Exerc 2001; 33(7): 1126–41. 25. Cinamon J, Finch L, Miller S, Higgins J and Mayo N. Preliminary evidence for the development of a stroke specific geriatric depression scale. Int J Geriatr Psychiatry 2011; 26: 188–198. 26. Enders C. Analyzing structural equation models with missing data. In: Hancock G and Mueller R (eds) Structural Equation Modeling – A Second Course. Quantitative Methods in Education and the Behavioural Sciences: Issues, Research, and Teaching. Greenwich: Information Age Publishing, 2006. pp.313–342. 27. Kline R. Principles and Practice of Structural Equation Modeling. 2nd ed. New York: The Guilford Press, 2005. 28. Hays R, Revicki D and Coyne K. Application of structural equation modeling to health outcomes research. Eval Health Prof 2005; 28(3): 295–309. 29. Patton M. Qualitative Research and Evaluation Methods. 3rd ed. Thousand Oaks, CA: SAGE Publications, 2002. 30. Hsieh H and Shannon S. Three approaches to qualitative content analysis. Qual Health Res 2005; 15(9): 1277–1288. 31. World Health Organization. International Classification of Functioning, Disability and Health. Geneva: World Health Organization, 2001. 32. Öberg T, Karsznia A and Öberg K. Basic gait parameters: Reference data for normal subjects. J Rehabil Res Dev 1993; 30(2): 210–223.
Barclay et al. 33. Wilson I and Cleary P. Linking clinical variables with health-related quality of life. A conceptual model of patient outcomes. JAMA 1995; 273: 59–65. 34. Undėn AL, Andrėasson A, Elofsson S, et al. Inflammatory cytokines, behaviour and age as determinants of self-rated health in women. Clin Sci 2007; 112: 363–373. 35. Yu L, Liu CK, Chen JW, Wang SY, Wu YH and SH Y. Relationship between post-stroke depression and lesion location: a meta-analysis. Kaohsiung J Med Sci 2004; 20(8): 372–380. 36. Patla A and Shumway-Cook A. Dimensions of mobility: Defining the complexity and difficulty associated with community walking. J Aging Phys Act 1999; 7: 7–19.
521 37. King A, Sallis J, LD Frank, et al. Aging in neighborhoods differing in walkability and income: Associations with physical activity and obesity in older adults. Social Sci Med 2011; 73: 1525–1533. 38. Aoyagi Y and Shaphard R. Steps per day – The road to senior health? Sports Med 2009; 39(6): 423–438. 39. Sumukadas D, Witham M, Struthers A and McMurdo M. Day length and weather conditions profoundly affect physical activity levels in older functionally impaired people. J Epidemiol Comm Health 2009; 63: 305–309. 40. Kubina L, Dubouloz C, Davis C, Kessler D and Egan M. The process of re-engagement in pesonally valued activities during the two years following stroke. Disabil Rehabil 2013; 35(3): 236–243.
Copyright of Clinical Rehabilitation is the property of Sage Publications, Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
