Perception of autonomy among people who use wheeled mobility assistive devices: dependence on environment and contextual factors

Abstract Purpose To evaluate and compare the perceived autonomy of people using wheeled mobility assistive devices (WMADs) in five community-based environments. To evaluate how personal, environmental, and assistive device-related factors impact the perceived autonomy of WMAD users. Method A study-specific questionnaire was used to evaluate perceived satisfaction of WMAD users with their autonomy in five environments: the Home Environment, Buildings Outside of the Home Environment, Outdoor Built Environment, Outdoor Natural Environment, and Transportation. For each environment, participants rated their satisfaction with autonomy about 15 personal, environmental, and assistive device-related factors. Qualitative perceptions were also collected with open-ended questions. Results Participants included 123 full- and part-time community-dwelling WMAD users. Participants’ overall satisfaction with autonomy in the Outdoor Natural Environment was statistically significantly lower compared to the other four environments (p < 0.05). In all environments, the top factor respondents were most satisfied with was WMAD ease of use. Their least satisfaction was when negotiating stairs, curbs, or obstacles in the Home, Buildings Outside of the Home, and the Outdoor Built Environments. In the Outdoor Natural Environment, the most dissatisfaction was with manoeuvring on different terrains. Responses to open-ended questions supported the quantitative findings and highlighted the effects of various factors on autonomy (e.g., subject-environment familiarity). Conclusions WMAD users reported the greatest restriction to their autonomy in outdoor environments. Different context-specific factors were found to impact autonomy in different environments. Understanding how environment-specific contextual factors contribute to overall perception of autonomy may inform the development of future strategies to overcome identified limitations and challenges. Implications for Rehabilitation Wheeled mobility assistive device (WMAD) users experienced the highest autonomy in their home environments, specifically, when having access to home modification services. WMAD users had the lowest autonomy in the outdoor natural environment, with manoeuvrability on different terrains being the main predictor of the overall satisfaction with autonomy in this environment. Environment-specific contextual factors with significant impacts on perceived autonomy were identified that can inform the design and development of future WMADs (e.g., distance travelled, safety).


Introduction
Mobility disability is one of the most prevalent types of disability in the United States [1] and the third most common in Canada [2].Autonomy, which is defined as having decisional and executional control over one's life [3], is recognised as one of the most important subjective aspects of participation among people with disabilities [4].It has been identified as a major contributor to the quality of lives of people with mobility limitations [5,6].Perception of autonomy is "personally unique", meaning that it may vary across different individuals and is determined based on their personal values and preferences [7].
The significant place of autonomy in the lives of people with mobility conditions necessitates the development and use of autonomy-based evaluation schemes to characterise the experiences of these individuals [9].The results of research assessing autonomy of people with mobility disability has revealed that these individuals, and wheeled mobility assistive device (WMAD) users specifically, experience impeded autonomy in various life situations [10].In this regard, significant associations have been found between the perception of autonomy and physical/ structural barriers encountered in the environment [11] and WMAD users reported experiencing the least and most restrictions in the autonomy indoors and outdoors, respectively [12].Issues related to environmental design features such as reaching shelves or a mailbox have been reported as barriers hindering both the indoor and outdoor autonomy and problems associated with entering/exiting the house/buildings were found to impose restrictions on outdoor autonomy of WMAD users [13].
Existing autonomy assessment tools created for more generalised assistive technology use (e.g., Impact on Participation and Autonomy) may not accurately recognise the experiences of WMAD users and capture the nuances of WMAD use in specific environments [14,15].Previous studies have evaluated autonomy in generic environments (e.g., indoor/outdoor) and when performing certain activities only [16].For instance, although difficulties experienced by WMAD users in the context of transportation have been discussed before [17], impacts of these issues on the autonomy of WMAD users have not been thoroughly examined.Moreover, little is known about the differences in perceived autonomy in more diverse environments (e.g., outdoor built vs. outdoor natural environments).
To address the above-mentioned limitations, we used an autonomy-based framework to qualitatively and quantitatively evaluate subjective concerns of WMAD users in various environments.In this regard, the main objectives of this research were to (1) evaluate and compare the perceived autonomy of WMAD users in five environments; and (2) evaluate how personal, environmental, and assistive device-related factors impact the perceived autonomy of WMAD users in each environment.In contrast to the previous approaches, we aimed to conduct an environment-based autonomy assessment rather than an activitybased assessment (e.g., when fulfilling personal/social roles), thereby obtaining a broader view of WMAD users' experiences in various environments.

Methods
An online survey including both quantitative and qualitative open-ended questions was developed by the research team to achieve the study objectives.We used the Checklist for Reporting Results of the Internet E-Surveys to report our findings [18].This study was approved by the local university and health authority ethics board.

Survey development and content
A cross-sectional semi-structured online survey with closed-and open-ended questions was developed to evaluate satisfaction with autonomy of WMAD users.The development of this survey was an iterative process in which the survey's content was analysed and reviewed by all members of the research team, including 1 Occupational Therapist (with academic and clinical experience in WMAD use), 1 Rehabilitation Engineer (with 30þ years of personal experience using WMADs), 1 Ph.D. student (Biomedical Engineer), and 2 Masters of Occupational Therapy students.The survey content was developed based on the existing frameworks (e.g., the human activity assistive technology [19]), findings of published literature, and our previous research [20].The research team reviewed published literature to identify main environments in which WMAD users have expressed or experienced challenges regarding their participation/autonomy [12,[21][22][23][24][25][26][27][28][29][30][31][32].Five environments were chosen for the present survey to cover a range of common places where people use their WMAD.These environments were (1) Home Environment; (2) Buildings Outside of the Home Environment; (3) Outdoor Built Environment; (4) Outdoor Natural Environment; and (5) Transportation (i.e., transitioning between locations).Moreover, we sought to identify interactions between personal, environmental, and WMAD-related factors that contribute to WMAD users' autonomy in each environment.The graduate students created a list of relevant factors that emerged from their literature review and shortlisted the ones that were quoted or cited most frequently in previous publications.The selection of factors contributing to autonomy was informed by the feedback from the advisory team including a skilled manual wheelchair user.Ultimately, we selected 12 factors associated with WMAD use in the context of Transportation and 15 factors associated with the other four environments (see Supplementary Appendix for a copy of the questionnaire).The survey's internal consistency was assessed using the Cronbach's alpha, which was calculated for the abovementioned factors in each environment.
We first piloted the survey with four non-experienced WMAD users to check for potential ambiguities in the content and logical or technical issues.Feedback from this group was used to refine the usability, technical functionality, and readability of the questionnaire.Next, we piloted the survey with eight individuals who were WMAD users or worked in the assistive technology field.The pilot participants' responses were used to refine question descriptions.This included clarifying the importance of using the keyword "autonomy" in each environment, and adding specific types of transportation (i.e., bus, trains, cars) to increase the specificity of responses.
The final version of the survey was implemented on Qualtrics [30], which is a secure online platform.The survey comprises three sections: background, environment-and contextual-factorspecific autonomy-related questions, and demographic questions.The first part of the survey included information regarding the purpose of the survey, estimated time to complete the survey, contact information for investigators, and the option to accept or reject participation in the survey.For each environment, participants were asked to (1) determine their choice of WMAD; (2) rate their satisfaction with autonomy when using their WMAD; (3) rate their autonomy concerning listed personal, environmental, and WMAD-related factors; and (4) use text boxes to provide further comments regarding their autonomy.A five-point Likert scale ranging from 1-Very Dissatisfied to 5-Very Satisfied was used to rate satisfaction with autonomy.We did not include or exclude any specific activities in the questionnaire.However, to guide participants in recalling relevant activities, we provided example tasks in the introductory description of each environment.Once the survey was started, participants had the option to change their responses or stop at any point and return to complete the survey at a later time.IP addresses were collected and examined to identify potential duplicate responses.

Data collection and sample characteristics
To be eligible to participate in the survey, respondents were required to be 19 years of age or older, fluent in English, community-dwelling, and have experience using a WMAD independently.We excluded WMAD users who resided in institutions.Eligible WMADs included manual wheelchairs (MWCs), manual wheelchair with add-ons (MWC þ AOs), power wheelchairs (PWCs), and scooters.We distributed the survey's URL address through flyers posted in the [Name of institution withheld] and the [Name of institution withheld], both located in [Name of city and country withheld], as well as their respective online recruitment boards.In addition to this, we sent mass emails to databases of participants from previous research who had expressed interest in participating in future studies and contacted several local and international community groups.The survey was open to anyone who had the URL to the survey's online webpage, and data were collected between February and September 2019.

Data analysis
Descriptive statistics, including mean, standard deviation, and frequencies, were used to report demographics and Likert scale data.Since assumptions of a parametric statistical test were violated (e.g., non-normal distribution), for objective 1, we used the non-parametric Friedman test to compare the difference in satisfaction with autonomy in the five pre-defined environments.The significance level was set at p<. 05.If significant main effects were found, a Wilcoxon signed-rank test, with Bonferroni adjustment (p<.005), was used to identify pairwise differences.For objective 2, we performed a linear multiple regression analysis to predict the overall satisfaction with autonomy in each environment using the identified personal, environmental, and WMAD-related factors as dependent variables.Cases with standardised residual greater than 3 or less than À 3 were identified as outliers.All analyses were performed using SPSS [31].Responses to open-ended questions were analysed using inductive content analysis [32] to determine environment-specific (i.e., environmental) factors promoting or hindering the autonomy of WMAD users.Two researchers [33] separately coded the qualitative responses and identified categories related to barriers and facilitators of autonomy in each environment.These categories were then reviewed and finalised amongst the larger research team.

Participants
One hundred and seventy individuals consented to participate in the survey; however, only 123 respondents completed the survey.Demographic information about these respondents is presented in Table 1.The study sample includes 72 females and 51 males with an age range of 24 to 77 years (mean age ± SD, 49.1 ± 13.3 years old).Participants were from seven countries with the majority living in North America.Various types of mobility disorders that require using WMADs were reported, with spinal cord injury being the most prevalent diagnosis (61.8%).Most survey participants had over 16 years of experience with WMAD use (51.2%).Although participants indicated using various modes of transportation (e.g., bus, subway), they were mainly relying on personal vehicles when using their WMADs.The frequency of WMAD use and the choice of WMAD varied in different environments.Detailed information about the types of WMADs that participants used in different environments and user satisfaction with those devices can be found in our previous work [33].

Overall satisfaction with autonomy
As noted in Figure 1, WMAD users' satisfaction with their autonomy was high in all environments with the majority of participants being "Satisfied' or "Very Satisfied" (Figure 1).We found a statistically significant difference in overall satisfaction with autonomy in the five environments, v 2 (4) ¼36.00, p < 0.05.Ranking of the five environments from the highest to lowest satisfaction with autonomy was the Home Environment, Buildings Outside of the Home Environment and Transportation, Outdoor Built Environment, and Outdoor Natural Environment.A Wilcoxon signed-rank test for paired comparisons showed that the overall satisfaction with autonomy when performing activities in the Outdoor Natural Environment was significantly lower than the Home Environment (p < 0.005, Z=-1.37),Buildings Outside of Home Environment (p < 0.005, Z=-1.26),Outdoor Built Environment (p < 0.005, Z=-1.25), and Transportation (p < 0.005, Z=-1.31).However, there was no significant difference in the overall satisfaction with autonomy in other environments.

Satisfaction with autonomy and associated factors
Satisfaction rates regarding environment/factor-specific responses are presented in Figure 2. The internal consistency (Cronbach's alpha) of the perceived satisfaction with the environment-specific factors (i.e., 12 items for Transportation and 15 items for the rest of the environments) was between 0.91 and 0.94 across the five environments.Participants' satisfaction with some factors was generally consistent across the five environments, examples of which include WMAD users' health conditions and reach as well as manoeuvrability through narrow spaces when using a WMAD.However, we observed environment-specific variability in respondents' satisfaction with other factors such as negotiating stairs, curbs, or obstacles and manoeuvrability on different terrains.In all environments, respondents were the most satisfied with the usability of their WMAD (i.e., ease of use).In contrast, respondents reported the least satisfaction when negotiating stairs, curbs, or obstacles in the Home, Buildings Outside of the Home Environment and Outdoor Built Environments.When moving around in the Outdoor Natural Environment, participants were the most dissatisfied with manoeuvring on different terrains.In the context of Transportation, respondents were the most dissatisfied with their reach and health conditions as well as the portability and weight of their device.
We developed linear regression models to examine the extent of the chosen contextual factors' contribution to overall satisfaction with autonomy in different environments.The results of these models are presented in Table 2.We removed the outliers from our dataset including 2 cases in the Home Environment, 8 cases in the Buildings Outside of the Home Environment, and 1 case in the Outdoor Built and Transportation.Our analyses indicate that some contextual factors predicted the overall satisfaction with autonomy in all environments, while explaining the least and most variance in the overall satisfaction with autonomy in the Home and Buildings Outside of the Home Environment, respectively.However, further analysis revealed that only a few factors (indicated in Table 2) added significantly to the overall satisfaction with autonomy prediction in each environment.In the Home Environment this factors was "Manoeuvrability through Narrow Spaces"; in Buildings Outside of the Home Environment these factors included "Manoeuvrability through Narrow Spaces", "Reach", and "Safety"; in the Outdoor Built Environment these factors included "Manoeuvrability through Narrow Spaces" and "Distance"; in the Outdoor Natural Environment these factors included "Manoeuvrability on Different Terrains" and "Health Conditions"; and in the context of Transportation these factors included "Durability and Reliability", "Dimensions", "Portability", "Ease of Use", and "Safety".

Content analysis
Analysis of the open-ended questions showed that various environmental characteristics contributed to the perception of autonomy across different environments.In each environment, we identified the top categories that emerged from the analysis of the open-ended questions.In addition to these categories, we selected some relevant but singular comments discussing unique autonomy-related topics.Representative examples of the participants' quotes from each category are labelled (e.g., Q1) and presented in Table 3.
In the Home environment, several quotes were related to home modifications and their impact on autonomy (Q1-Q3).The most commonly mentioned problems encountered in the Home Environment were related to doorways (Q4), level changes and stairs (Q5), and WMAD users' reach (Q6).Several participants commented on the accessibility issues they face when entering/exiting and manoeuvring inside Buildings Outside of the Home Environment.These issues were mainly associated with passing through doorways (Q7), manoeuvrability in narrow spaces (Q8),  and navigating soft surfaces (Q9).The number of concerns reported regarding navigating the Outdoor built Environment was greater than the previous two environments (i.e., Home and Buildings Outside of the Home).These problems were generally related to navigating terrains (Q10), navigating level changes, curbs, stairs (Q11), and weather conditions (Q12).In the Outdoor natural environments, participants' comments were dominated by issues concerning the accessibilities of natural environments (Q13, Q14).Difficulties accessing and navigating natural environments and terrains such as soft sand, grass, and gravel were frequently mentioned in the comments (Q15-Q17).Many participants reported how using accessible public transportation [33] and personal vehicles positively impacted autonomy in this environment (Q18).WMAD users' safety was one of the major concerns when using transportation (Q19, Q20).Participants referred to navigating in confined and crowded spaces as another main issue when using public transportation (Q21).Some participants commented on unique and broader topics, such as (un)familiarity with the environment (Q22) or expectations from their WMAD (Q23), determining their autonomy in different contexts.

Discussion
This study aimed to evaluate satisfaction with autonomy of WMAD users in various environments and regarding different contextual factors.In this research, we expanded the conventional dichotomous indoor-outdoor classification of environments to explore some of the less examined contexts and to differentiate the experiences of WMAD users in different environments.On average, WMAD users perceived their highest satisfaction with autonomy in their home.This could be associated with different factors promoting autonomy in the home environment (e.g., the importance of home adaptations was frequently commented upon) and specific capabilities of WMADs (e.g., elevation features).Similarly, others have observed strong associations between (lack of) home modifications, wheelchair usability (manual and power), and autonomy/participation [12,13].The significance of home modifications is also reflected in the outcome of the regression analysis, where manoeuvrability through narrow spaces (e.g., accessibility of doorways, confined spaces) was the main predictor of the overall satisfaction with autonomy in the Home Environment, and frequently commented upon.Our findings also provide further evidence regarding WMAD users' concerns about their reach and manoeuvrability through narrow spaces when moving around Buildings Outside of the Home Environment [13].
Survey respondents were the most dissatisfied with their autonomy when performing activities in more taxing environments (e.g., outdoor natural).This is aligned with the outcomes of previous research in which WMAD users indicated experiencing more restrictions with their autonomy outdoors compared to autonomy indoors [34].Based on participants' comments we can infer that the autonomy restrictions experienced in the outdoor natural environments were associated with limitations imposed by physical barriers [11], weather conditions [17,35], or inherent accessibility issues or unforeseen contingencies in the outdoor settings.This is further supported by the outcome of the regression model, in which manoeuvrability on different terrains was found to be the main predictor of the overall satisfaction with autonomy in that environment.Environmental factors may become even more restrictive in places that are not familiar to WMAD users, as one of our participants stated.However, it is not surprising that WMAD users were less satisfied with their autonomy outdoor versus indoor.Previous research showed the impacts of the primary criteria for WMAD prescription and funding is specific device use indoors.Thus, devices better equipped for outdoor use are difficult to get funded [36].In the context of Transportation, we found associations between availability of transportation and outdoor autonomy.The high satisfaction trends observed among the majority of participants could be associated with most of them residing in places with relatively accessible transportation facilities.Moreover, survey responses showed that in many cases, lack/availability of accessible public or personal transportation determined WMAD users' outdoor mobility.In other words, WMAD users' autonomy in the context of transportation had a broad influence on many life experiences, such as accessing outdoor natural environments [37], performing recreational activities [38], and participation as a whole [39], especially in the winter [38].
In addition to providing insights regarding the overall satisfaction with autonomy across various environments, the outcome of this study provides environment-specific contextual factors' rankings that could be used to identify the needs and challenges of WMAD users.The outcomes of our analyses suggest that although WMAD users had very low satisfaction with autonomy concerning certain factors, those factors did not significantly contribute to their overall sense of autonomy (e.g., negotiating stairs, curbs, or obstacles).This could be mainly related to WMAD users' expectations of their device and their capabilities which was discussed in some of participants' responses.In contrast, our analyses showed that certain factors such as manoeuvrability through narrow spaces, manoeuvrability on different terrains, safety, and distance travelled had significant impacts on autonomy, to the extent that satisfaction with autonomy in each environment could be predicted by satisfaction with one or a combination of these factors.
The outcome of this research can inform policy-makers, assistive technology developers, or clinicians regarding the development of future strategies to overcome identified limitations and challenges.Considering the main predictors of autonomy across all environments, we propose the following topics for future research and developments (1) designing wheelchairs to improve manoeuvrability through narrow spaces and portability of the device; (2) developing advanced assistive technologies enabling WMAD users to seamlessly traverse uneven or soft terrains, as well as handle level changes and other obstacles, [40]; and (3) improving the capability of WMADs to go further distances.Considering the low satisfaction rates in outdoor autonomy and due to the strong associations between autonomy outdoors and life satisfaction [6], particular attention should be paid to improving WMAD users' autonomy outdoors.
Supplementary investigations should be done to address the limitations of the current research.For example, the average high satisfaction with autonomy that was observed across survey participants, which were mostly North Americans presumably having access to high-quality devices and health care coverage, might not be generalisable to all WMAD users.Further information should be gathered from a more diverse group of WMAD users to improve the generalisability of our findings.Moreover, future research could further validate our survey and examine the feasibility of using these questions as part of a standard autonomy assessment measure for WMAD users.
Although we did not have a large enough sample size to analyse the potential impacts of underlying factors such as e.g., age, gender, type of disability, time since injury, region, etc., on individuals' autonomy, future studies with larger sample sizes should investigate effects of these factors on autonomy.As discussed previously [41], even though a satisfactory level of autonomy was reported in most environments, specific activities within these dimensions may still be significantly restricted among WMAD users.Therefore, it is necessary to perform more in-depth studies (e.g., interviews, focus groups) to gather a more accurate representation of autonomy among WMAD users.

Conclusion
In this study, we examined the subjective impressions of WMAD users in some of the less explored life contexts.The outcome of this research provides an in-depth view of contributors to the autonomy of WMAD users in various environments.These findings build on existing evidence regarding WMAD users' perception of their autonomy concerning a combination of personal, environmental, and WMAD-related factors and expands this knowledge to more comprehensive and diverse contexts.The findings of this research highlight the need for transforming the accessibility of certain environments such as outdoor natural environments.Moreover, the knowledge regarding contextual factors' ranking and their contribution to overall autonomy can provide a strong case for prioritising future research with the focus on addressing identified difficulties experienced by WMAD users.

Previous presentations
Parts of this work were presented orally at the Canadian Association of Occupational Therapists Conference (Niagara Falls,

Figure 1 .
Figure 1.Overall satisfaction with autonomy in five environments.Bar graphs illustrating the cumulative percentage regarding the level of satisfaction with autonomy in each environment.

Figure 2 .
Figure 2. Satisfaction with autonomy regarding personal, environmental, and WMAD-related factors in five environments.Bar graphs illustrating the cumulative environment-/factor-specific satisfaction percentage.To achieve higher clarity, we presented the Very (Dis)satisfied and (Dis)satisfied categories with similar colours.

Table 1 .
Demographic Characteristics (N ¼ 123).These countries include Australia, the United Kingdom of Great Britain and Northern Ireland, and India.b Only 107 participants indicated using their WMAD in the context of Transportation. a

Table 2 .
Association between overall satisfaction with autonomy and contextual factors.Unstandardised Coefficients, b Beta: Standardised Coefficients, c p < 0.05.d These factors were not part of the questionnaire in the context of Transportation.

Table 3 .
Exemplar participants' comments on their autonomy.My biggest issue is with bus drivers who don't want to safely secure me."Q20."I am particularly concerned and dissatisfied with the safety of using my power wheelchair in tie-down spaces in taxis, vans etc." Q21."When getting into a bus it is very difficult to manoeuvre into a space without running over toes."General Topics Q22."You get used to going places where you live.When you are in a different town there is a level of trepidation that comes with the unknown."Q23."It [my scooter] does not do stairs.It does not move on a sidewalk without curb cuts/sidewalk ramps.It does what it is designed to do.No more, no less." "[my home] has been modified to fit me" Q2. "I built my home to my needs" Q3. "My home was not designed to accommodate my chair."Q4. "I wish there were push to open door buttons everywhere.That's where I struggle the most."Q5. "Stairs are the issue I need help with."Q6. "I'd so wish to be able to reach things up high and not worry … about falling out."Buildings Q7. "Most of the barriers to perfect autonomy are when I need to go through a closed door."Q8. "The only challenges I have are going down narrow isles in shops."Q9. "Carpets are terrible, they slow me down so much and make me work SO HARD."