Assessing environmental barriers by means of the Swedish craig hospital inventory of environmental factors among people post-stroke

Abstract

Background

When occupational therapists address environmental barriers to occupational engagement, some barriers might not be possible to reduce for single individuals, because decisions have to be taken at community or societal level, for example changes in public transport. Investigating environmental barriers by means of the Swedish Craig Hospital Inventory of Environmental Factors (CHIEF-S) may increase our understanding of the environmental impact on occupation engagement and the methodological challenges to assess environmental barriers.

Aims

To investigate and describe the magnitude of encountered environmental barriers in a group of people post-stroke and to assess psychometric properties of the CHIEF-S.

Material and methods

In total, 34 participants, who had sustained a stroke in Sweden were recruited.

Results

The participants reported in average 2,7 barriers and the total frequency-magnitude score of barriers (CHIEF-S score) was 0.45. The Cronbach’s α for the total CHIEF-S was 0.80 and the analysis of test-retest reliability revealed ICC = 0.86. The entire instrument demonstrated better psychometric properties than the single sub-scales.

Conclusion

In this study, the frequency-magnitude of environmental barriers encountered by people post-stroke are reported at a group level and adds information to the cumulative knowledge generation on perceived environmental barriers in the society. However, to inform which interventions are needed at a more detailed level, other data collection methods have to be added.

Keywords:

• Assessment
• city planning
• community health planning
• occupational therapy
• participation
• total words

Introduction

Traditionally, occupational therapists mainly focus on assessments and interventions that aim to enhance each individual client’s engagement in occupations [1]. However, several of the environmental barriers and facilitators identified might be related to areas, which are societal affairs [2], e.g. changes in public transport and they are therefore difficult to change during an individual intervention. Occupational therapists have the knowledge and skills to influence environmental features at a societal level, e.g. policies and societal development [3]. Yet, they seldom intervene with a more systematic and societal approach, which have the potential to improve all humans’ equal opportunities to occupational engagement and well-being [1,4]. One explanation may be that occupational therapists often are part of the medical and health care system whereas a vast majority of physical and social planning take place in other organisations. To improve environments at community and societal level that support occupational engagement, systematic data collection of environmental facilitators and barriers are important not only at an individual level, but also at a group and a societal level. For example, the results from a study [5] of participation in public spaces demonstrated that participation and relevance of everyday technologies were significantly lower for people with dementia than for people with no known cognitive impairment. Another example is the results presented by Slaug and colleagues [6] showing how assessments of housing can be used to demonstrate accessibility problems that can be expected to emerge for people aging in typical dwellings in the ordinary housing stock. With this type of systematised data the evidence-base can increase and be used by policy-makers and managers as basis for their decisions [7].

However, assessing environmental features and its impact on occupational engagement is associated with major methodological challenges because of its complexity [7–9]. For instance, occupational engagement can be viewed from a taxonomic hierarchy where the level of complexity range from voluntary movement and mental processes, actions, tasks, activities to engagement in entire occupations [1]. Thus, when all relevant environmental features for each task, activity, etc. are addressed, assessments tend to be quite extensive and the results difficult to handle. Environmental features that influence e.g. occupational engagement might also be difficult to assess due to the dynamic of the physical environment [8], e.g. changing light and weather conditions or due to other dimensions such as attitudes and service [9].

One assessment instrument that focuses on perceived environmental demands on a complex level is the Craig Hospital Inventory of Environmental Factors (CHIEF) [10]. The tool measures the environmental demands in five areas; Attitudes/support, Service/Assistance, Physical/structural, Work/school and Policies [11]. Its psychometric properties have been investigated in population-based samples, where it was found to have high test-retest reliability (ICC=.93) and high internal consistency (Cronbach α=.93) [10], and it has been translated to different languages, for example Swedish. The Swedish version (CHIEF-S) has been used among people with late effect of polio [12]. Most of the identified barriers encountered by this sample were in the Physical/structural subscale, but the barriers were generally infrequent and the magnitude tended to be low. Yet, those that were encountered could prevent some people from participating in the desired occupations. Another Swedish study utilised the CHIEF-S in 156 people who used three types of assistive technology devices: assistive technology for cognitive support, myoelectric arm prosthesis or powered mobility device [13]. The highest magnitude of barriers for them were found within the Physical/structural subscale and the Service/Assistance subscale. Participants who used assistive technology for cognitive support experienced higher barriers compared to those using myoelectric arm prosthesis or powered mobility device. Thus the instrument has been used to outline environmental barriers in different groups of people [12,13], but data related to other groups in a Swedish context is lacking.

Stroke is a common cause of disability in adults, which can result in various types of impairments that influence their ability to participate in occupations they previously were engaged in [14,15]. In particular, these problems become evident when a person is discharged from hospital and try to reintegrate in society. Often environmental demands, such as attitudes from other people, infrastructure or different physical factors [16,17] are barriers during this process. Internationally, the CHIEF has previously been used among people with stroke [18,19]. For example, in a study of the Chinese version of the CHIEF, the highest magnitude of barriers was also found within the Physical/structural subscale [19]. In another study, among 400 Korean community-dwelling people with and without stroke [18], the most severe environmental barriers were identified within different subscales and items such as transportation (15,5%), natural environment (14,3%), personal equipment (13,5%) and surroundings (13,0%) were the most common barriers. To summarise, environmental features might differ among countries and client groups. In order to increase the understanding of methodological aspects when using the Swedish Craig Hospital Inventory of Environmental Factors (CHIEF-S) as a tool for assessing environmental barriers, the present study set out to describe environmental barriers among people with stroke in Sweden. The aims of this study were to investigate and describe the magnitude of encountered environmental barriers in a group of people post-stroke and to assess psychometric properties of the CHIEF-S.

Material and methods

Participants

The 34 participants (21 men and 13 women, mean age of 68.1 years) were recruited from a university hospital in the south of Sweden, for another project [20], between Oct 2012 and February 2014. All participants had to be 55 years or older, and be able to speak and understand Swedish in order to participate in the data collection. At the time of inclusion, they had sustained a stroke six to ten months prior to recruitment and reported being independent regarding in- and outdoor mobility at the three month follow-up past stroke. All were community-dwelling, living in an ordinary housing; either in a multifamily house (a building where families live in separate flats) or a single-family home (a residential house, used by a single dwelling-unit). Some of the participants reported cognitive impairments, depression, and some used mobility devices. One participant used a rollator and another participant used both cane/crutches and a rollator for indoor mobility. For outdoor mobility, four participants used only cane/crutches, three used a rollator, and two used both cane/crutches and rollator. All, except two, reported they were independent during outdoor mobility, and a majority were able to walk >400 metres. Further sample characteristics are found in Table 1.

Table 1. Sample characteristics, N = 34.

Variable	Descriptive measure
Sex, men, n (%)	21 (61.8)
Age, mean (min–max)	68.1 (58–86)
Months post stroke, mean (min–max)	8.0 (6–10)
First-ever stroke, n (%)	29 (85.3)
Stroke subtype
Infarction, n (%)	33 (97.1)
Haemorrhage, n (%)	1 (2.9)
Living condition
Living with partner, n (%)	22 (64.7)
Living alone, n (%)	12 (35.3)
Children living at home, n (%)	3 (8.8)
Housing condition
Multi-family house, n (%)	21 (68.8)
Single-family house, n (%)	13 (38.2)
Mobility device^a,b
Indoors, n (%)	2 (6.1)
Outdoors, n (%)	9 (27.3)
Walking distance^b
0–200 metres	7 (20.6)
201–400 metres	2 (5.9)
>400 metres	24 (70.6)
Independent mobility^b, n (%)	31 (91.1)
Self-rated cognitive impairment
Any cognitive impairment^c, n (%)	19 (55.9)
No of self-rated cognitive impairments^b, mean (min–max)	3.2 (0–20)
Geriatric depression scale (GDS)^d
Total score, mean (min–max)	5.3 (0–19)
Possible depression^e, n (%)	12 (35.3)

^aCane/crutches; rollator.

^bOne missing.

^cStudy-specific self-rated cognitive impairment questionnaire. According to Wendel et al. [21] this assessment has showed cognitive impairment to a greater extent compared to an objective assessment.

^dGDS [22].

^e>5p = possible depression.

Ethics

The Regional Ethical Review Board in Lund, Sweden, approved the study (2012/174). After having received verbal and written information about the study, participants gave their informed consent to participate, and were informed they could withdraw at any time without any consequences for their future care.

Chief-s

The CHIEF-S [8] comprises five sub-scales with 25 items in total. The sub-scales are Attitude/support (5 items), Service/Assistance (7 items), Physical/structural (6 items), Work/school (3 items) and Policy (4 items). For each item is assessed, how often the barrier is encountered on a scale of 0–4 (0 = never, 1 = less than monthly, 2 = monthly, 3 = weekly and 4 = daily). Further, the magnitude of each reported barrier is assessed on a scale of 0–2 indicating the size of the problem (0 = no problem since the barrier was never encountered, 1 = a little problem and 2 = a big problem). The product of the frequency score and the magnitude score (0–8) indicating the overall impact of the barrier. In the original publication by Whiteneck et al. [10], the CHIEF showed good test-retest (ICC = 0.93) and internal consistency reliability (Cronbach α = 0.93), with evidence of content, construct, and discriminant validity.

The instrument has thereafter been translated to other languages and psychometric properties have been investigated in different samples. Test-retest reliability and internal consistency reliability of the instrument are similar to what Whiteneck et al. [10] found. For example, in the study among Chinese people with stroke, ICC = 0.845 and Cronbach α = 0.93 were reported [19] and in a study among people with spinal cord injury, the Hindi version demonstrated ICC = 0.80 and Cronbach α = 0.92 [23]. The Brazilian version [24] used among children and adolescents with cerebral palsy also demonstrated ICC ≥0.92. In addition, Soni et al. [23] determined the content validity ratio to be 1.00 for all the items except item number 5, 11, and 12. The overall content validity index was 0.97 for the longer version.

Procedure

The data collection of sample characteristics as well as environmental factors by means of CHIEF-S was part of a larger data collection in another project (for more information see Carlstedt et al. [20]), at a neurology clinic in the south of Sweden.

Before the individual interview started, the researcher described the background and purpose of the study, and explained the content in each questionnaire, including CHIEF-S. She read the different CHIEF-S items loud to each participant, and recorded all answers in the form. All 34 participants completed the test at the first test occasion (T1) and then about three weeks later (T2) with no missing data. Each occasion lasted approximately 30–60 min, whereof the data collection regarding CHIEF-S lasted about 5–10 min.

Statistical and psychometric analysis

Descriptive statistics were used to report on the sample characteristics as well as the barriers encountered including the magnitude of the problems. Floor and ceiling effects, i.e. the percentage with highest and lowest score, were calculated as potential indicators of content validity [25]. In order to investigate construct validity, the correlation of each sub-scale with the total score was estimated by means of Spearman coefficient [26]. The hypothesis was that each subscale should positively correlate with the total score. Internal consistency reliability was calculated by means of Cronbach’s α on the basis of data from T1, striving towards reaching values above 0.7, and preferable 0.8 [27]. To investigate the test-retest reliability, the intra-class correlation (ICC) between T1 and T2 was calculated, aiming to reach standard reproducibility criteria stating values need to be >0.8 [28].

The mean age was 68 years and in accordance with the manual only participants who had recent experience of work/school answered questions related to this area, which means that only the sub-sample of 13 participants were used for these analysis.

Results

Environmental barriers

The 34 participants reported 93 barriers in relation to the 25 items assessed in CHIEF-S, i.e. in average 2,7 barriers per person (min: 0, max: 11). For 14 items, 31 or more participants reported that they never encountered any environmental barrier. Three of the items; ‘Education/training’ within the sub-scale ‘Service/Assistance’ and ‘Attitudes to Work/school’ and ‘Help at Work/school’ within the sub-scale ‘Work/school’ were not at all encountered as barriers (Table 2). Other barriers were more frequently encountered, for example ‘Transportation’ and ‘Information’ within the subscale ‘Service/Assistance’ and ‘Surroundings’, ‘Natural environment’ and ‘Design of the home’ within the subscale ‘Physical/structural barriers’. Among those who encountered these barriers, 13 were reported as large problems within the subscale ‘Service/Assistance’ and 17 within the subscale ‘Physical/structural barriers’.

Table 2. Environmental barriers as perceived by people who had sustained stroke, N = 34.

	Encountered at all^a	Perceived as large problem	CHIEF-S score^b
Environmental barrier	n (%)	n (%)	Mean (SD)
Attitude/support (5 items)			0.22 (0.48)
Support in the community	2 (6)	1 (3)	0.18 (0.76)
Attitudes in the community	3 (9)	1 (3)	0.21 (0.77)
Support in the home	2 (6)	1 (3)	0.21 (0.84)
Attitudes within the home	4 (12)	1 (3)	0.26 (0.83)
Discrimination	4 (12)	0 (0)	0.24 (0.70)
Service/Assistance (7 items)			0.47 (0.45)
Transportation	9 (26)	4 (12)	1.03 (2.22)
Healthcare	4 (12)	2 (6)	0.26 (0.83)
Help in the home	2 (6)	1 (3)	0.21 (0.84)
Help in the community	1 (3)	0 (0)	0.06 (0.34)
Information	9 (26)	4 (12)	0.97 (2.32)
Personal equipment	3 (9)	2 (6)	0.38 (1.52)
Education/training (n = 13)	0 (0)	0 (0)	0.00 (0.00)
Physical/structural (6 items)			0.87 (1.23)
Design of the home	7 (21)	5 (15)	1.09 (2.52)
Natural environment	10 (29)	4 (12)	1.03 (1.99)
Design of the work/school (n = 13)	3 (23)	2 (15)	1.31 (2.69)
Design of the community	6 (18)	1 (3)	0.65 (1.67)
Surroundings	13 (38)	5 (15)	1.32 (2.16)
Technology	1 (3)	0 (0)	0.09 (0.51)
Work/school (3 items)			0.08 (0.27)
Support at work/school (n = 13)	1 (8)	0 (0)	0.23 (0.83)
Attitudes at work/school (n = 13)	0 (0)	0 (0)	0.00 (0.00)
Help at work/school (n = 13)	0 (0)	0 (0)	0.00 (0.00)
Policies (4 items)			0.18 (0.53)
Business policies	2 (6)	1 (3)	0.12 (0.48)
Education/employment policies (n = 13)	2 (16)	0 (0)	0.23 (0.60)
Services in the community	3 (9)	2 (6)	0.32 (1.41)
Government policies	2 (6)	0 (0)	0.09 (0.38)
Total score			0.45 (0.64)

Note. questions regarding work/school only applicable for 13 individuals.

^aDaily, Weekly, Monthly and Less than each month have been summed due to low frequencies in each category.

^bScore calculated by multiplying frequency with magnitude. Theoretical min-max, 0–8 (higher scores mean more problems).

The total frequeny-magnitude score of barriers in the instrument was 0.45 (Table 2). The highest score was found for the ‘Physical/structural’ sub-scale (0.87), and the lowest for the ‘Work/school’ sub-scale (0.08). However, only 13 of the 34 participants were involved in activities related to ‘Work/school’. The second lowest score was found for ‘Policies’ (0.18). For single items the highest scores were found for the items ‘Surroundings’ (1.32), ‘Design of Work/school’ (1.31), ‘Design of home’ (1.09) and ‘Natural environment’ (1.03) in the sub-scale ‘Physical/structural’ and for the item ‘Transportation’ (1.03) in the sub-scale ‘Service/Assistance’.

Psychometric properties

The instrument demonstrated considerable floor effects at subscale level, but no ceiling effects (Table 3). The scores from each one of all sub-scales, beside work/school, were significantly correlated with the total score (p < 0.05). The Cronbach’s α indicator of internal consistency reliability for the total CHIEF-S was 0.80 (not including work/school items), whereas the Cronbach’s α were lower for the sub-scales (see Table 4 for details). The physical/structural sub-scale (Cronbach’s α = 0.64) and the attitude/support sub-scale (Cronbach’s α = 0.6) had the highest internal consistency.

Table 3. Indicators of content and construct validity of the CHIEF-S in a sample with people who had sustained stroke, N = 34.

	Floor and ceiling effects		Sub-scale correlation with the total score
	% lowest score	% highest score	Spearman coefficient	p-value
Subscale
Attitude/support	73.5	0.0	0.43	0.012
Service/Assistance	55.9	0.0	0.77	<0.001
Physical/structural	38.2	0.0	0.86	<0.001
Work/school (n = 13)	92.3	0.0	0.11	0.704
Policies	82.4	0.0	0.34	0.047
Total score	17.6	0.0

Note. questions regarding work/school only applicable for 13 individuals.

Table 4. Internal consistency and test-retest reliability of the CHIEF-S in a sample with people who had sustained stroke, N = 34.

	Cronbach’s α		Intraclass correlation
	Raw	Standardized	ICC coefficient (95% CI)	p-value
Subscale
Attitude/support	0.60	0.63	0.86 (0.72–0.93)	<0.001
Service/Assistance	0.40	0.48	0.80 (0.61–0.90)	<0.001
Physical/structural	0.64	0.56	0.83 (0.65–0.92)	<0.001
Work/school (n = 13)	0.00	–	0.00 (–2.71 to 0.73)	0.5
Policies	0.05	0.37	0.00 (–0.86 to 0.48)	0.5
Total score^a	0.80	0.79	0.86 (0.71–0.93)	<0.001

Note. questions regarding work/school only applicable for 13 individuals.

^aNot including items related to Work/School, i.e. based on 19 items.

The analysis of test-retest reliability for the total score of the CHIEF-S revealed ICC= 0.86 (0 < 0.001). However, for two of the subscales the internal consistency were zero, due to the low numbers of barriers identified (Table 4).

Discussion

In this study, the CHIEF-S was used to explore perceived environmental barriers among Swedish people with stroke. Our results show that the participants perceived in average 2.7 barriers and the total frequency-magnitude score for the instrument was 0.45, which is higher than what e.g. Liao et al. [19] found in a stroke sample (0.24). Still, our participants reported few barriers, which may be explained by the fact that overall, they reported few mobility problems, i.e. few used a mobility device for walking and the majority were able to walk more than 400 metres. At the same time, our participants reported cognitive limitations (55%), also known to influence the amount of perceived environmental barriers [17,29]. A paradox is that people choose occupations in environments that afford them the greatest opportunities and fewest barriers [7,9]. That is, people seem to avoid occupations that are connected to environmental barriers, and as a result, no barriers are reported. The average number of environmental barriers that our participants experienced can be viewed as low, but few barriers may influence a single individual’s occupational engagement [30,31]. For example, the attitude from a bus-driver, to drive before a passenger has been seated, might have serious consequences and impact on an individual’s occupational engagement [32].

The psychometric properties of the CHIEF-S had an internal consistency of 0.80 and the test-retest reliability had an ICC of 0.86, which is somewhat lower than in the original study [10]. The entire instrument demonstrated better psychometric properties than the single sub-scales. Even if some subscales of the instrument demonstrates considerable floor effects, our results reveal important data on environmental barriers. The highest frequency-magnitude score was found in the ‘Physical/structural’ sub-scale (0.87), which is higher than Widehammar et al. [13] reported (0.67) for the same sub-scale in a Swedish sample of people using different assistive devices. However, comparisons are difficult since frequency-magnitude scores are not calculated for the subscales, but only for single items in several studies [18,19]. Problems related to the ‘Physical/structural’ subscale was also frequently reported in a Swedish sample of people with late-effects of polio [12], even though they did not calculate any frequency-magnitude score. Turning to single items, our participants reported large problems with ‘surroundings’ and ‘natural environment’ within the ‘Physical/structural’ subscale, but also ‘transportation’ within the ‘Service/Assistance’ subscale. These barriers were also large problems in a Korean sample of community-dwelling people who had sustained a stroke [19]. On the other hand, the Korean study reported large problems with lack of personal equipment or special adapted devices whereas our participants reported problems with the design of the home, which made it difficult to pursue occupations they wanted or needed to do. Personal equipment such as assistive devices is regulated by different laws in Sweden, e.g. the Swedish Health and Medical Services Act [33] and has been accessible for many years. It is likely that our participants already received the assistive devices needed, whereas changes to optimize the home environment may be a longer process. Many people post-stroke struggle with re-integration into society [14,15], trying to overcome environmental barriers [34,35]. It is therefore surprising that few barriers were reported in some areas, for instance ‘technology’, which refer to computer technology. Even if technology includes mobile phones and computers, there is a variety of how much such technology is used [36]. Thus, we do not know how the participants interpreted this question. The floor effects detected could thus indicate that the items included do not sufficiently capture potential barriers in this area.

Individuals’ occupational performance involves everything from actions and tasks to entire occupations [1], i.e. the taxonomic hierarchy of engagement in occupations need to be considered in assessment of environmental barriers. For example, our participants reported problems with transport, which can be an occupation in itself, but people with disabilities have also identified transport as an important mediator for participation [7]. To improve transport, e.g. to travel by public transport, all aspects in a travel chain needs to be paid attention to [17,32]. The travel chain includes many tasks such as planning the trip, walking to the bus stop, buying the ticket, etc. If a person fails to perform only one task in the travel chain, for example to buy a ticket, the completion of the entire occupation may be compromised. However, barriers at this level in the hierarchy of occupational performance is not possible to identify by means of the CHIEF-S. At an individual level, when the occupational therapists meet their clients they can supplement a formal assessment with an interview and observation to identify the barriers and facilitators for occupational engagement the clients experience [32]. In order to get this information useful for policy-makers and managers the data needs to be systematised at different levels in the taxonomic hierarchy of engagement in occupations. For example by relating the environmental barriers and facilitators identified to an action such as opening the door or a task such as entering a building or buying a ticket.

The CHIEF is used to investigate environmental demands in different samples [10,12,13], and cover many environmental features. Most of the questions are posed at a general level, and the specific geographical area is not defined. The self-perceived environmental barriers reported are based on questions of how often, in the past 12 months, environmental features have been barriers to participation in activities that matter to each individual [10]. Heinemann et al. [11] describe other assessments that capture barriers in the environment, and report the most frequently used assessments are the Facilitators and Barriers Survey/Mobility (FABS/M) [35], the Home and Community Environment Instrument (HACE) [37]; and the Measure of the Quality of the Environment (MQE) [38]. Heinemann et al. [11] also report that the CHIEF, FABS/M, and MQE describe self-perceived barriers and/or facilitators in the environment, whereas the HACE focuses more on existing environmental barriers defined by others and their consequences. In some situations, barriers and facilitators have opposite counteracts, and in other situations the barriers completely hinder occupational engagement [7]. For example, if the staff at the travel centre are service minded, this might compensate for a lack of an automatic door-opener. However, a service minded staff do not compensate for stairs when a person using a wheelchair is trying to enter a building with only stairs and no elevators.

Being part of the decision-making processes can potentially empower the users of the environments, and hopefully reduce barriers. To some extent users are already involved in, for example design-processes [39], but there are a wide range of environmental barriers for occupational engagement and well-being that have not been addressed. To deal with the complex and multidimensional issues and open up for new interventions, Fransen and colleagues [40] suggest a shift from mainly focussing the individual who is engaging in occupations in a specific context, to instead focus on the broader perspective of citizenship, including a societal perspective. Whalley Hammel [4] highlight the right to occupational opportunities and that occupational therapy could advance the right of all people to engage in meaningful occupations. That is, occupational therapists need to find ways to translate the knowledge their clients generate on environmental barriers for occupational engagement, into data that will be usable also at group and societal level.

Instruments that are developed to capture complex issues that involve impacts of several different dimensions and how these dimensions are related to each other, implies the calculation of composite scores. The CHIEF-S measure of the frequency-magnitude of environmental barriers [9] is an example of such a composite score. The Housing Enabler that measures the amount of accessibility problems by matching the capacity of the individual with the demands of the environment [41] is another example. Such composite scores that combine several dimensions in a complex calculation mean that the range of possible values is exponentially increased, where most of the possible outcomes in the higher range are never encountered in reality. That is, already low scores have ‘clinical’ relevance and indicate distinct issues, while the maximum score is purely theoretical and not a realistic value. In some of the previous studies using the CHIEF, only frequencies and magnitudes are reported, but not the combination that produces the frequency-magnitude scores [e.g. 12]. The maximum frequency-magnitude score for CHIEF-S is 8.0, but that implies participants should experience all of the barriers as a large problem on a daily basis. Thus, there is no ceiling effect for this instrument. When the score is calculated, the value is often the same as the theoretical minimum, because if a barrier is not encountered the score is always zero. For the entire instrument, the floor effect was 17.6% and a common recommendation is that it should not exceed 15–20% [e.g. 27]. For the subscales, the floor effects were even more striking (up to 92.3%). However, as the construct is intended not only to capture different grades of a phenomenon but also the absence of the phenomenon itself, a substantial floor effect can be expected. In other studies of the CHIEF, the floor effect is also obvious, even if it is not calculated [e.g. 12,13]. That is, the standard recommendation for floor/ceiling effects may not be applicable for the CHIEF, but could be used to assess the plausibility of proportions perceiving or not perceiving the presence of certain barriers.

A limitation with this study is that a relatively small sample is included. The participants were recruited because they were active, and engaged in occupations outside home and thereby had experiences of environmental barriers in different arenas. Prior to recruitment, they reported being independent regarding in- and outdoor mobility three months after their stroke but at the same time, several of them reported having remaining cognitive impairments and some used mobility devices during walking (Table 1). That is, even though they are not representative for all people post-stroke, they seem to represent a variation in physical, cognitive and psychological functioning. Another limitation is unfortunately that no data was available in the project to assess the construct validity of the instrument.

Conclusion

The results of this study demonstrate that a sample of Swedish stroke participants, who reported being independent regarding in- and outdoor mobility at the three month follow-up past stroke, still experienced more or less environmental barriers in different areas and some of the problems were large. The highest frequency-magnitude score was found within the sub-scale physical/structural barriers. Potentially, many more barriers could have been reported, but for the single individual’s occupational engagement, only one or few barriers may be crucial. The psychometric properties of the CHIEF-S showed that the instrument is sufficiently valid and reliable for use in a Swedish context, even if the internal consistency (0.80) and the test-retest reliability (ICC: 0.86) were somewhat lower than the original study [10]. The information collected at an individual level is important for planning individual interventions, but in order to improve the possibilities for occupational engagement in the community, it is of uppermost importance to address the societal level as well. Occupational therapists face several challenges in this regard but by addressing the societal level in different ways, steps in the right direction can be taken and strengthen the evidence base in occupational therapy [4]. Even if data from a small sample was used and some subscales of the instrument have considerable floor effects, our results presents data on group level and thereby adds to the cumulative knowledge generation on perceived environmental barriers in the society.

Acknowledgements

The authors wish to thank all respondents who participated, Dr Emma Carlstedt who collected the data and employees at the Department of Neurology for assistance during the data collection, especially Dr Hélène Pessah-Rasmussen. This article was economically supported by the Ribbingska Foundation in Lund, Sweden and prepared within the Centre for Ageing and Supportive Environments (CASE) at Lund University, financed by the Swedish Council for Working Life and Social Research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

The present research was financially supported by The Ribbingska Foundation, Lund, Sweden.