Psychometric properties of the client-centred rehabilitation questionnaire-is derived from a large and varied group of rehabilitation users

Abstract

Purpose

The purpose of this study was to inspect and establish the factor structure of the Icelandic Client-Centred Rehabilitation Questionnaire [CCRQ-is] and investigate the service experiences of a large and varied sample of rehabilitation users.

Materials and methods

Altogether 499 rehabilitation users completed the questionnaire. Confirmatory and exploratory factor analysis was conducted for analysis of items. Mann-Whitney’s U and Kruskal-Wallis test was used to compare subscale responses based on participants’ characteristics.

Results

Four factors explained 53,2% of the variance: Respect and attentiveness, Interaction with significant others, Responsiveness to needs and preferences and Education and enablement. Subscales showed internal consistency from 0.72–0.91 and 0.92 for the whole instrument. The subscale Respect and attentiveness represented user centred rehabilitation the most and Interaction with significant others the least. Significant differences in relation to health conditions and age were obtained on all four subscales and differences by gender on one subscale.

Conclusions

Our results suggest the CCRQ-is is a reliable tool that can be used with rehabilitation users with a broad range of characteristics within the Icelandic context. The extent to which the intersection of age, gender and health issues influence users’ perception of services needs to be further scrutinized.

Implications for Rehabilitation

• Increasingly users of rehabilitation have complex and composite health issues as physical and mental health conditions commonly co-exist.

• The four subscale Client-Centred Rehabilitation Questionnaire [CCRQ-is] demonstrated strong reliability for assessing the client-centredness of rehabilitation services for people with different health issues in Iceland.

• People receiving rehabilitation for mental health reasons seem to find the service to be less client-centred than other rehabilitation service users.

• The lowest scores on the CCRQ-is were consistently found on the dimension Interaction with significant others.

• The key role often played by families during and after the rehabilitation process should be recognized.

Keywords:

• Client-centredness
• assessment
• psychometric
• reliability
• holistic approach
• cross-cultural

Introduction

In rehabilitation services, person-, client- or user-centred approaches are central to understanding and accommodating the personal needs and preferences of people seeking services [1–3]. While there is no agreed upon definition of these approaches, overall, the social, psychological, cultural and ethical sensitivity of encounters between professionals and people seeking services are emphasized [4–8]¹. In this context the uniqueness of rehabilitation has been pointed out [6,9] as user’s active engagement in specific activities as part of the rehabilitation process is typically required, rather than just participation in decision-making or adherence to medication prescriptions.

Much of the literature on client-centredness within rehabilitation has represented professional understandings [10,11] and the need for psychometrically sound instruments that measure users’ experiences across a wide range of rehabilitative services has been pointed out [8,12]. The Client Centred Rehabilitation Questionnaire (CCRQ) is a self-report measure designed to evaluate user’s experience of client-centred care in rehabilitation [13,14]. The CCRQ was designed in Canada for adult rehabilitation users with chronic physically disabling conditions. The 33 questions of the CCRQ are organized into seven subscales that represent different aspects of client-centredness.

While the CCRQ has demonstrated good reliability and validity in Canadian studies [13,14] there is a need to examine the psychometric properties and applicability in other cultures [15]. For example, a factorial study by Körner et al. [16] which was undertaken with 496 inpatients at 10 German rehabilitation centres did not verify the original seven CCRQ subscales. Instead, a 15-question short version instrument containing three subscales was confirmed. In contrast, Fisher and colleagues [12] who gathered data from 408 rehabilitation inpatients at 20 facilities across Australia found that a modified version of the CCRQ demonstrated internal consistency reliability and factorial validity of each of the seven subscales in the Australian context. Similarly, research undertaken with 305 rehabilitation inpatients in Catalonia, Spain, confirmed the seven subscales after eliminating four questions from the original instrument [17]. The findings of a qualitative study conducted in Denmark revealed that except for physical comfort all CCRQ domains reflected client-centred practice from the view of twenty-five hand-rehabilitation service users [18].

In Iceland there has been a notable lack of information about the experiences of rehabilitation users regarding the degree to which services reflect their needs and preferences [19,20]. Our earlier research focused on the face-validity of the CCRQ within the Icelandic context and consequently concerns regarding the conceptual domains and applicability of the instrument were raised. This led to the development of revised instrument, the Icelandic Client-Centred Rehabilitation Questionnaire (CCRQ-is).

As aforementioned the CCRQ was designed for people with physical health issues. However, users of rehabilitation often have health issues spanning different body systems i.e., physical, mental and emotional. Many of these issues are complex and composite and consequently call for holistic and integrated rehabilitative services [21]. Furthermore, users with different characteristics and health issues may experience the rehabilitation process differently [17].

Thus, the purpose of this study was twofold: Firstly, to test psychometric properties of the CCRQ-is, especially to inspect and establish the factor structure of the instrument. And secondly, to scrutinize the experience of a large group of rehabilitation users with a broad range of personal characteristics and health issues of the service they receive. Three research questions were posed:

1. What is the factor structure and reliability of the CCRQ-is subscales?

2. To what extent do rehabilitation services in Iceland reflect the needs and preferences of persons of different age and gender, as measured by the CCRQ-is?

3. What are the similarities and differences in responses of rehabilitation users with different health issues as measured by the CCRQ-is?

Materials and methods

This study employed a cross sectional survey design. Participants were recruited from the largest rehabilitation centre in Iceland and three rehabilitation wards at two of the country’s largest hospitals. The rehabilitation centre serves people with various health issues, i.e., physical, mental and emotional. The focus of two of the wards is mainly on physical rehabilitation and the third specializes in mental health rehabilitation. Interdisciplinary teamwork is emphasized at all institutions, and all serve both in- and outpatients, except for the specialized mental health ward which serves outpatients only. The study was approved by the Icelandic National Ethics Committee (VSN nr. 14-047).

Measure

Data were collected using the Icelandic Client Centred Rehabilitation Questionnaire (CCRQ-is). The translation and adaptation of the CCRQ-is into Icelandic has occurred over a number of years [19,20] and began with (1) primary translation of the CCRQ to Icelandic; and (2) cognitive interviews with eleven mental health service users to explore the relevance of the underlying domains, acceptability of questions, and cultural appropriateness of the language used for this particular group of users [22,23]. Consequently, language modifications were made to improve clarity and two questions were removed, one was a repetition of another question, and the other was appropriate only for people with physical impairments [20]. In accordance with recommendations from the primary author of the CCRQ [14], questions with negative orientation were changed to positive and the option ‘does not apply’ was removed. In addition, the orientation of the Likert scale was reverted so the highest score (i.e., five) would represent most client-centred practice instead of the opposite. This adapted Icelandic version for mental health clients (Client-Centred Practice Questionnaire-mental health [CCPQ-mh]) – or the first version of the CCRQ-is – was then used in a mixed method study with users of mental health services [19]. Cognitive interviews with six physical rehabilitation users were then conducted, which led to further semantic refinements of a few questions. The CCRQ-is questionnaire used in this study contained 31 questions (as opposed to 33 questions in the CCRQ) about to what extent the service is client-centred, answered on a 5-point Likert scale.

Data collection

The same data gathering procedures were implemented at each institution. Each of the three institutions assigned contact persons who introduced the survey briefly to participants, supervised distribution of the questionnaires at the end of participants’ rehabilitation period, collected them back and handed to the first author. Everyone who had completed their rehabilitation, were able to read and write Icelandic, and could complete the survey independently were offered to participate. Prospective participants received an information letter explaining the research prior to their participation. Responding to the survey anonymously and returning the questionnaire in a closed envelope was interpreted as informed consent. No health records were consulted.

Altogether 558 rehabilitation users responded to the questionnaire. Response rate was 89.7%−97.7% for different institutions and 93.7% for the whole sample. Fifty-nine participants were excluded, because the number of missing items exceeded our predetermined threshold of 3 missing items. The resulting sample included 499 participants. Participants’ age ranged from 17 to 84 years, with a mean age of 51 year. Women were in majority (59%). The two most common self-reported health-issues were Musculoskeletal (56.3%) and Cardiac and/or Pulmonary (38.6%), altogether 250 or 50.1% of participants reported more than one health issue. Table 1 provides an overview of the characteristics of the participants.

Table 1. Characteristics of participants (N = 499).

	Number	(%)
Age
<39	111	22.2
40–54	151	30.3
55–64	119	23.8
65>	107	21.4
Missing	11	2.2
Gender
Women	294	58.9
Men	198	39.7
Missing	7	1.4
Residence
Capital area	305	61.1
Outside of capital	187	37.5
Missing	7	1.4
Weeks in rehabilitation
≤4	181	36.3
5–6	181	36.3
≥7	123	24.6
Missing	14	2.8
Health issue – self reported^a
Musculoskeletal	276	56.3
Cardiac and/or pulmonary	189	38.6
Mental	133	27.1
Obesity	130	26.5
Central nervous system	57	11.6
Cancer or immune system	28	5.7
Perception	18	3.7
Other	56	11.4
Missing	9	1.8

^aParticipants reported more than one issue.

Data analysis

In order to establish the factor structure and internal consistency reliability of the CCRQ-is, we conducted an exploratory factor analysis (EFA), i.e., principal axis factor analysis (PAF) with Varimax rotation [24]. PAF is an appropriate method for this analysis because subscales represented different aspects of user-centred practice. Eigenvalues for each factor were obtained and the variance explained by each factor. Furthermore, internal consistency (Cronbach’s coefficient alpha) was calculated using item-to total correlations. Recognizing that there has been some criticism concerning Cronbach’s alpha we also calculated Mc Donald’s Omega [25]. A confirmatory factor analysis (CFA) was carried out to test amendment of the structure, using a maximum likelihood estimation and the following fit indices: χ² and its differences of degrees of freedom (χ²/df ≥ 0.90), the comparative fit index (CFI ≥0.90); the standardized root mean squared residual (SRMR <0.06) and the root mean squared error approximation (RMSEA < 0.08) and Pclose >0.05. The calculation of the rest of the statistical test was carried out based on results and dimensions obtained in the PAF, for this reason items with cross-loading (1,2,4,8,10 and 16) were removed and 31 because of lack of loading [26]. In addition, we calculated sub-scale intercorrelations using Spearman’s rho coefficient correlation. Correlation coefficients that are > 0.30 are considered to be strong [24].

Since the data were not normally distributed, the non-parametric Mann-Whitney’s U and Kruskal-Wallis test were used to analyse client responses, through descriptive and inferential statistics. Descriptive statistics for background variables of all participants included mean (M) and standard deviation (SD) for continuous variables and counts and percentiles for categorial variables. Crosstabs were used to investigate the relationship between two categorial variables. When calculating subscale means we used the ‘individual mean’ method which has shown to produce valid imputed values [27].

Mann-Whitney’s U (95% significance level) to compare means of different subscales (dependent variables) and potentially influencing factors (independent variables). Kruskal-Wallis test was used to relate subscale means to characteristics of participants and their environments. For all statistical analysis the IMB SPSS software package with Amos plugin, v 28 was used [28].

Results

Factor structure and reliability of the CCRQ-is subscales

The exploratory factor analysis calculation revealed a KMO value of 0.95, supporting the sampling suitability for the analysis [24]. Bartlett’s test of sphericity (χ² = 7510.5, df = 465, p < 0.001) showed that correlations in the correlation matrix did not occur by chance, and that correlations between questions were adequate for the analysis. Altogether four factors explained 53.2% of the variance: Respect and attentiveness, eigenvalue 11.4 and 36.7% variance; Education and enablement eigenvalue 2.5 and 8% variance; Responsiveness to needs and preferences 1.4 and 4.5% variance, and Interaction with significant others 1.2 and 4.0% variance. Question nr 31 The professionals only gave me information that I was ready to receive, only had a loading of 0.175 and was therefore removed. Twenty-seven questions had factor loadings that exceeded 0.40 and for the three remaining questions a factor loading greater than 0.3 was considered sufficient due to the large sample [24]. Three questions in the Education and enablement (nr. 2,10 and 16), and three in the Responsiveness to needs and preferences (nr. 1,4 and 8) had high cross-loadings with other factors and where therefore removed. After eliminating the seven questions described above, the subscales showed internal consistency from 0.69–0.91 and 0.92 for the whole questionnaire, which is considered to be strong [29]. The psychometric properties of the four factor, 24 items were also assessed using a CFA with a maximum likelihood estimation method, to determine how well items represented the proposed construct. The results showed that all the fit indices were situated within the recommended values, i.e.: χ²/df = 2.12; CFI = 0.95; SRMR = 0.043; RMSEA = 0.047; Pclose = 0.768.

Confirmatory analysis was used to compare Model fit for a one-factor model, the original seven-factors model and the four factors model. The four-factor model outperformed the two others, χ²/df = for the 7-factor model and 7,59 for the one-factor model. Similar for the CFI = 0.865 and 0.661. SRMR = 0.055 and 0.169. RMSA = 0.069 and 0.115 and Pclose <0.001 and <0.001. Hence, the CCRQ-is contains 24 questions that reflect four dimensions of client-centred rehabilitation. The subscales Interaction with significant others and Education and enablement were normally distributed but the other two had slight negative skewness., see Table 2. Internal consistency (α) Omega values (Ω), names of factors/subscales and questions’ loading after rotation are shown in Table 2. In Table 3 subscale intercorrelations are shown. In all cases the correlation coefficient showed strong relations, i.e., > 0.30.

Table 2. Factor loading reliability of the CCRQ-is after varimax rotation.

Questions	Component				α	Ω	Subscales
	1	2	3	4
28. The professionals tried to make me feel good.	.709	.287	.167	.106
30. The professionals in my team worked well together.	.692	.248	.224	.096
20. I received the information I needed when I asked for it.	.673	.24	0.23	.117
26. The professionals showed manner that made it easier for me to express my feelings and state of mind.	.656	.252	.229	.15
12. I was met with respect and consideration	.653	.404	.104	.05
17. I knew whom to talk to if I had a problem or had any questions during the rehabilitation.	.594	.327	.204	.093
29. The professionals acknowledged my emotional needs (such as worries. fears and anxieties) and took them seriously.	.575	254	.282	.098	0.91	0.91	Respect and attentiveness
21. I was taught what I needed to know to cope with my illness at home.	.521	.125	.323	.159
14. The professionals considered my goals and priorities to be very important.	.52	.253	.2	.096
19. I was encouraged to participate in setting goals.	.498	0	.219	-.001
15. My rehabilitation plan was explained to me in a way that I understood.	.492	.201	.344	.167
18. I was given a fair amount of time to rest.	.456	.244	.161	.248
7. The professionals treated my pain as much as they possibly could.	.341	.607	.076	.159
13. The professionals took note of my descriptions of pain.	.428	.537	.063	.098
5. The professionals consulted with me and took my needs into account when creating a special program (schedule) for me.	.11	.529	.394	.119
8. All the services I received on site were organized according to my personal needs.	.425	.514	.26	.151	0.69	0.70	Responsiveness to needs and preferences
4. The professionals treated me as a person and not as every other case.	.486	.504	.041	.069
1. The professionals and I decided together what could help me.	.199	.409	.375	.045
6. I did not have to repeat the same information for different professionals.	.272	.394	.126	.184
24. The different aspects of rehabilitation were well explained to me.	.402	.075	.619	.196
27. I was told what to expect when I got home.	.392	.07	.584	.225
2. I was informed on a regular basis about the progress of my affairs.	.208	.436	.474	.155
16. The professionals and I talked about how I was doing in the rehabilitation and made changes if needed.	.438	.276	.457	.182	0.74	0.75	Eduction and enablement
23. I was told whom I can contact if a problem arises after I am discharged.	.295	.126	.445	.309
9. The professionals informed me about services and resources in my local community.	.191	.203	.374	.204
10. I achieved the outcome I had expected in the rehabilitation.	.299	.184	.360	.258
31. The professionals only gave me information that I was ready to receive.	.058	.046	.175	.114
11. My significant others received instruction and information from the professionals (with my consent).	.106	.127	.177	.808
3. The professionals gave my significant others the support they needed.	.084	.183	.183	.807
22. The attitude of professionals towards my significant others reflected respect for them.	.138	.073	.089	.708	0.85	0.85	Interaction with significant others
25. My significant others were invited to participate in the rehabilitation as much as I wanted.	.037	.056	.304	.591

α = Cronbach’s alpha.

Ω = Mc Donalds’s Omega.

Loadings in bold represent questions that loaded on each component.

Questions nr. 8,4,1,2, 16 and 10 were removed because of cross-loading to other factors. Question 31 was removed because of lack of loading.

Please note that the questions were back-translated from Icelandic to English and some language nuances may exist.

Table 3. Subscale intercorrelations (Spearman’s rho) of the CCRQ subscales (N = 499).

	Respect and attentiveness	Responsiveness to needs and preferences	Education and enablement	Interaction with significant others
Respect and attentiveness
Responsiveness to needs and preferences	0.585**
Education and enablement	0.704**	0.492**
Interaction with significant others	0.355**	0.318**	0.474**

** p < 0.001.

Age and gender comparisons of rehabilitation needs and preferences

Overall, the subscale Respect and attentiveness represented client-centred rehabilitation the most, with the mean 4.49 (5 being the highest possible score). The subscale Responsiveness to needs and preferences had a mean of 4.29 and Education and enablement a mean of 3.98. Interaction with significant others had mean of 3.23 that is considerably lower than other subscales and therefore represents client-centred rehabilitation the least.

Table 4 presents descriptive statistics for the participants’ age and gender, mean (M) and standard deviation (SD) for continuous variables and counts and proportions for categorial variables. Mann–Whitney test indicated significant differences by gender in one of the four subscales, i.e., Interaction with significant others (U = 22230; p < 0.001). When comparing age groups Kruskal-Wallis test revealed an age difference (p < 0.001) on all subscales except for Respect and attentiveness (p = 0.077). A post hoc test with adjusted p-values for the subscale Responsiveness to needs and preferences showed that the people in the age group 39 or younger found the service to be less client–centred than people in other age groups. For the subscale Education and enablement, responses of people in the age group 39 or younger were significantly lower than for the age groups 55–64 years old and 65 or older. The post hoc for the subscale Interaction with significant others showed significantly lower responses in the age groups 39 or younger and 40–54 years old than for the other two age groups.

Table 4. CCRQ-is subscale scores for age and gender (N = 499).

		Subscales
		Respect and attentiveness	Interaction with significant others	Responsiveness to needs and preferences	Education and enablement
Gender	Age	Mean (Sd/n)	Mean (Sd/n)	Mean (Sd/n)	Mean (Sd/n)
Women	≥ 39	4.37 (0.71/81)	2.93 (1.01/81)	3.85 (0.96/81)	3.71 (0.83/81)
	40–54	4.51 (0.45/102)	2.97 (0.80/102)	4.26 (0.70/102)	4.00 (0.67/102)
	55–64	4.59 (0.37/73)	3.21 (0.84/73)	4.52 (0.47/73)	4.04 (0.66/73)
	65 ≥	4.52 (0.39/35)	3.56 (1.07/35)	4.47 (0.45/35)	4.22 (0.62/35)
Men	≥ 39	4.38 (0.41/28)	3.01 (0.89/28)	4.20 (0.47/28)	3.73 (0.81/28)
	40–54	4.50 (0.37/49)	3.34 (0.73/49)	4.35 (0.50/49)	4.02 (0.54/49)
	55–64	4.57 (0.40/46)	3.74 (0.78/46)	4.46 (0.51/46)	4.23 (0.65/46)
	65 ≥	4.51 (0.35/72)	3.53 (0.69/72)	4.41 (0.44/72)	4.00 (0.65/72)

The similarities and differences in responses of persons with different health issues as measured by the CCRQ-is

As shown in Table 5, significant differences between participants with different health issues were obtained on three dimensions. Participants with cardiac/or pulmonary health issues (38.6% of the sample) found the service to be client-centred to a more extent on three subscales than participants with other health issues. A clear gender difference was obtained for those with cardiac/or pulmonary health issues as altogether 57.7% of male participants reported such issues compared to 25.5% of female participants. Conversely, participants who had mental health issues (27.1% of the sample) found the rehabilitation service to be client-centred to a lesser extent on the subscale Interaction with significant others. (see Table 4). Again, a gender difference was detected as 16.5% of all male participants and 33.8% of all women participants reported having mental health issues. People with mental health issues had a mean of 2.83 on the subscale Interaction with significant others compared with 3.43 for those with cardiac/or pulmonary health issues.

Table 5. CCRQ-is responses of participants with different health issues.

CCRQ-is - Subscales	Mean(sd/n)
	Cardiac and/or pulmonary health issue	No cardiac or pulmonary health issue	p-value^a
Respect and attentiveness	4.55 (0.39/189)	4.45 (0.51/301)	0.074
Interaction with significant others	3.43 (0.81/189)	3.08 (0.91/301)	˂0.001^b
Responsiveness to needs and preferences	4.45 (0.46/189)	4.19 (0.74/301)	˂0.001^b
Education and enablement	4.10 (0.66/189)	3.90 (0.71/301)	˂0.001^b
	Mental health issue	No mental health issue
Respect and attentiveness	4.43 (0.58/133)	4.51 (0.42/357)	0.458
Interaction with significant others	2.83 (0.90/133)	3.36 (0.84/357)	˂0.001^b
Responsiveness to needs and preferences	4.15 (0.80/133)	4.34 (0.59/357)	0.065
Education and enablement	3.87 (0.79/133)	4.01 (0.65/357)	0.086

^aMann-Whitney U-test. SD: standard deviation. ^bSignificant difference.

Discussion

We studied the factor structure and reliability of the CCRQ-is and then evaluated the experience of clients with different personal characteristics and health issues of the rehabilitation service they receive, as measured by the CCRQ-is. Below our results are discussed in more detail. We first address the psychometric properties and conceptual dimensions of the instrument. Then we discuss the experience of users with different backgrounds and health issues. Lastly, we highlight the need for an expanded and more holistic focus within rehabilitation.

Conceptual dimensions and factor structure of the CCRQ-is

Our results indicate that the CCRQ-is is a valid and reliable tool for assessing the client-centredness of rehabilitation services in Iceland. A thorough development process focusing on conceptual criteria, acceptability of questions and semantics (e.g., words and phrases) was carried out to ensure that the CCRQ-is reflected the phenomenon of client-centred rehabilitation for persons with mental health as well as with physical health conditions. Previous steps in this process are thoroughly described in our earlier publications [19,20].

Through confirmatory and exploratory factor analyses this present study confirmed a four component instrument with 24 questions as opposed to 31 questions in the CCPQ-mh [19] and 33 in the CCRQ. Thus, the CCRQ-is contains four subscales of client-centred rehabilitation; Respect and attentiveness; Interaction with significant others; Responsiveness to needs and preferences, and Education and enablement. Each of the four subscales showed strong internal consistency [29]. Similarly internal reliability for the whole instrument was very good. The confirmatory and exploratory analyses undertaken in this study are based on a large group of rehabilitation users with various health issues and supports the underlying theoretical dimensions of the CCRQ-is.

As aforementioned studies examining the factor structure of the CCRQ in Australia [12], Germany [16] and Spain [17] have presented contrasting results. The participants in our study had more varied health issues than those in these three studies. Similarly, we had already addressed many of the recommendations these authors raised in our prior studies, such as by reverting questions with negative orientation to positive and removing the option “does not apply”. Similarly double barrelled questions had already been simplified and questions relating to two different time periods clarified after the cognitive interviews performed at an earlier stage [19,20]. Two of the questions we omitted from the original CCRQ (one after the cognitive interviews and the other following the factor analysis) were also omitted in the CCRQ-e [17]. It should be noted that according to the author of the CCRQ (Cott et al. 2006) the initial subscales were not verified through factor analysis.

Overall, the CCRQ-is appears to be a valid and psychometrically sound tool that can be used with rehabilitation users with a broad range of health issues within the Icelandic context – and possibly elsewhere. It should nevertheless be noted that the questionnaire reflects a Western perspective on client-centredness, and we acknowledge that understandings of the construct may vary across cultures.

The experience of clients with different personal characteristics and health issues

Our second aim in this study was to explore the experience of clients with different personal characteristics and health issues of the rehabilitation service they receive as measured by the CCRQ-is. For the survey as a whole the highest overall scores for women and men in all age groups were obtained for the dimension Respect and attentiveness, followed by Responsiveness to needs and preferences, and Education and enablement. The dimension Interaction with significant others had the lowest scores by far. Differences by gender were found on one subscale, i.e., Interaction with significant others, reflecting that men found the service to be more client-centred than did women. Differences in relation to age were found on all subscales except for Respect and attentiveness reflecting that people in the youngest age-group considered the service to be least client-centred. Similarly, the oldest participants in the study by Capdevila [17] in Spain, found the services to be more client-centred than those who were younger.

The differences in the extent to which people with different health issues considered the rehabilitation service to be client-centred is notable as well as the interaction effects of age and gender. More specifically, people with cardiovascular conditions considered the service to be more client-centred than people with mental health issues. Simultaneously many more men than women reported having cardiovascular conditions. In contrast, more women than men reported having mental health issues and they also considered the service to be less client-centred on the subscales Interaction with significant others and Education and enablement. For the subscale Interaction with significant others the mean was 2.83 which is noticeably low.

The gendered aspects of access to and perceptions of rehabilitation have been described but gendered analyses of experiences of rehabilitation programs are nevertheless reported to be scarce [30]. Wiklund et al. [31] in Northern Sweden found that access to specialty pain rehabilitation depended on gender, ethnicity, social status, and the character of diagnosis received (medical versus psychiatric or social). The participants in their study generally regarded women patients as being less valued and holding a lower status in the healthcare system than men. Although the intersection between gender and various other social determinants of health inequalities is increasingly in focus [32] as well as how they may interact at different lifetime stages, we found no studies that focused specifically on user- or client-centredness of rehabilitation programmes in relation to these aspects. Capdevila and colleagues in Spain [17] found no significant association in relation to gender variables but then again people with mental health issues were not included in their sample. Over half of our sample reported more than one health issue and many reported both physical and mental health challenges. This is not surprising as physical and mental health problems commonly co-exist and health conditions that appear to be mainly physical often have mental components involved [33,34]. The gender and age differences found in this study will be further explored in our on-going qualitative follow-up study with rehabilitation users in Iceland. Similarly, aspects of power relationships in rehabilitation encounters will be in focus [35] and the ways in which these may play out in services from users’ with different health conditions point of view.

For all dimensions and age-groups the lowest scores were consistently found on the dimension Interaction with significant others. This is interesting when considering the key role often played by families during and after the rehabilitation process. Evidence suggests that family involvement and readiness to provide support (physical, mental, and emotional) is associated with better outcomes [7,36]. Bamm and colleagues [7] found that family involvement in care and decision making improved both clients’ and their families’ experiences of rehabilitation, especially for patients who had difficulty articulating their questions and concerns. Education and information for both the client and the family were mentioned by both parties as being the main strategies to help them to develop a clear understanding of the rehabilitation client’s condition and prognosis [7]. Although people’s situation and preferences may certainly differ regarding the extent to which they want their significant others to be involved in their rehabilitation process the literature indicates that families and rehabilitation users generally appreciate family participation. It should be noted that family involvement in rehabilitation appears to be an underexplored and neglected field [37]. As pointed out by Bezmez Shakespeare and Yardimci [38] scholarly work on family involvement often focuses on children or family members as care-providers, rather than the experience of the rehabilitation user.

The need for an expanded rehabilitation focus

Inspired by the increased criticism of the normative and biomedical framings that have characterized rehabilitation [39–41] we set out to investigate client-centred practice within rehabilitation services for people with various personal characteristics and health conditions. We were amazed by the striking lack of dialogue between physical and mental rehabilitation as the coverage about user- or client-centred aspects appears to be almost completely separate for each field. This separation is bound to reduce opportunities for discussions about what user-centredness may in fact entail and diminish possibilities for the holistic emphasis increasingly called for [21,41]. This separation probably also reflects the current gap in praxis. As an example, in a recent draft policy for rehabilitation services in Iceland, mental health rehabilitation was left out “for practical purposes” [42]. Following vast criticism from key stakeholders the policy is now being refined to include a more holistic view.

Recently the concept of ultrabilitation was promoted in order to expand the scope of rehabilitation and counteract the current normative focus on recovery and restoration, especially in regard to disabled people [41,43]. Perhaps ultrabilitation may help connect the fields of physical and mental health rehabilitation and provide insights into what user-centred practice within both fields may entail. By looking beyond assumptions of normality and difference, and downplaying the emphasis of recovery and restoration of bodily functions, ultrabilitation may also help bridge the current gap that exists between disability studies and rehabilitation studies [44] and promote dialogues between the two. In the Convention of Rights of People with Disabilities [45] it is recognized that disability results from the interaction between persons with impairments and attitudinal and environmental barriers that hinders their full and effective participation in society on an equal basis with others. Accordingly, aspects of the environment may have to be given greater weight in rehabilitation [40], not the least since environmental aspects may at times be more amenable to change than bodily functions [46].

Strengths and limitations

The large number and diversity of participants was essential for calculation of the factor structure of CCRQ-is and thus adds to the validity of our findings. To our knowledge few if any studies have examined client-centred services as experienced by rehabilitation users with health conditions spanning such different body systems. This broad focus is important because client-centred rehabilitation should be directed towards the individual person and not a specific type of rehabilitation user who has a certain impairment or disease. Increasingly users of rehabilitation have complex and composite health issues [21], thus the focus needs to be expanded from physical rehabilitation only.

Limitations to this study are first and foremost that no information was gathered about personal factors which may have affected participants’ experiences of rehabilitation services such as their socio-economic status, education, financial resources, and family situation. Also, no measures were taken to ensure the access of people with specific accessibility needs. For instance, participants could not choose from different survey formats (such as answering by phone or through e-mail) and no support was provided to those who did not speak or understand Icelandic or would have needed assistance answering the questions.

Conclusion

This large-scale survey provides an important overview of the experiences of rehabilitation users in Iceland as well as the commonalities and differences between the people with different personal characteristics and a wide variety of health issues. Our results confirm that client-centredness, as measured by the CCRQ-is, has joint conceptual and psychometric underpinnings. Since client-centredness appears to be a generic construct that applies across rehabilitation settings and populations, client-centred rehabilitation should be directed towards the individual person and not a specific type of rehabilitation user who has a distinct impairment or disease.

Acknowledgements

The authors thank the participants of this study. Thanks are extended to the contact persons at the rehabilitation units. This study was funded by the University of Akureyri and the University of Iceland Research Fund.

Disclosure statement

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

Notes

1 In this paper we mainly use the terms client-centred and user-centred approach, where client/user stands for the person/client/patient or family (when the latter are involved).