Introduction

According to the American Heart Association, on average, a stroke occurs every 40 seconds and a stroke fatality occurs every 4 minutes [1]. In the United Kingdom (UK), stroke is a major cause of morbidity, the third-most-common cause of death for women in England and Wales, and the fourth-most-common cause of death in men [2]. In China, stroke is the most frequent cause of death and the annual incidence rate of stroke is 8.7 per 1000, resulting in costs in excess of more than $100 billion [3,4]. About 75% of Chinese stroke survivors have at least some form of disability and experience difficulty in completing day-to-day chores [5]. More than 80% of disabled stroke patients need long-term care from family caregivers [6]. Caring for patients with stroke results in burden and emotional distress, as well as hope and positive outcomes [7–9], all of which affect the QoL of their caregivers [10,11]. QoL is regarded as an established marker of biopsychosocial health and an increasingly-used measure of caregivers’ health [11,12]. Reports from the literature indicate that the QoL of caregivers can affect the physical outcomes and QoL of patients [13,14]. QoL of Chinese caregivers has been found to be poor in both community and hospital setting (in the short term as well as the long term care contexts), suggesting that these caregivers need support from health providers [15–18]. Consequently, promoting the health of stroke patients may require acknowledging and ensuring the QoL of their caregivers.

The World Health Organization (WHO) defines QoL as a broad multidimensional concept which includes subjective evaluations of both the positive and the negative aspects of life [19]. We focus here on health related QoL, which, according to the National Center for Chronic Disease Prevention and Health Promotion, includes perceptions of both physical and mental health, and their correlates on the individual level [20]. Researchers have used QoL (in the stress-coping framework of Lazarus and Folkman) as the outcome of the process of coping with stressor for caregivers when using the the stress-coping framework of Lazarus and Folkman to explain [21, 22]. White et al. [23] developed a model to evaluate the QoL of family caregivers of stroke survivors in the context of the caregiving situation, the caregiver’s characteristics, and characteristics of the environment, including the balance of positive and negative effects of caregiving. Reports from the literature indicate that the QoL of caregivers for patients with cancer was multidimensional, including psychological, social, mental and physical and behavioral components [22, 24]. However, the QoL of caregivers for patients with stroke has not been studied widely in China, possibly due to the lack of a proper assessment tool.

There are some evidence-based instruments which can be used to assess QoL, most of which are general in nature to facilitate use with different types of subjects. The Short Form-36 Healthy Survey (SF-36) is a tool which previous researchers have used to measure the QoL of healthy people, patients [25], and caregivers for patients with stroke [26]. Researchers have also employed a simplified version of the SF-36, the 12-item Short Form Health Survey [27, 28]. Jeong et al. [29] and Chuluunbaatar et al. [30] have used the WHO Quality of Life instrument (WHOQOL) to measure the QoL of stroke caregivers. These instruments demonstrate good reliability and validity in both English-language forms and Chinese versions, but they are general tools which focus more on the negative (not the positive) aspects of QoL, and are not specifically intended for caregivers of stroke patients [11]. According to Joseph et al. [31], the developers of the AC-QoL, the influences on caregivers may be more wide ranging and affect the caregivers’ entire QoL. More general instruments thus may fail to evaluate domains of particular importance to family caregivers, such as personal growth and stress due to caring.

The AC-QoL is a tool designed specifically to assess the QoL of caregivers, developed by Joseph et al. [31] using an initial panel of 385 caregivers from the UK and principal components analysis to test an set of 100 questions; as a result of their testing. They reduced the number of questions to 40 across eight domains (support for caring, caring choice, caring stress, financial implications, personal growth, sense of value, ability to care and carer satisfaction). The AC-QoL’s internal consistency/reliability is 0.94, indicating that the AC-QoL is a useful tool to assess the QoL of adult caregivers[31]. Brand et al. [11] used the AC-QoL to explore the relationship between social support, benefit finding and QoL in caregivers of patients with mental health difficulties and physical health problems, and reported a Cronbach’s alpha of 0.93 for the AC-QoL. Parker et al. [32] and Brand et al. [33] used the AC-QoL to evaluate the effects of their interventions (Parker et al. peer-support and a peer staffing model for residential mental health rehabilitation; Brand et al. benefit finding through writing) on caregivers of people with mental and physical disabilities. To the best of our knowledge, the AC-QoL has not been used in China and there are no other tools specific to stroke caregivers’ QoL. However, due to the differences in culture and language between the UK and China, we felt it essential to develop a Chinese version AC-QoL so that we could assess the QoL of Chinese caregivers. Since the psychometric properties of AC-QoL also have never been tested in stroke caregivers, to provide preliminary support for the use of the AC-QoL in cross-culture stroke caregiver research, our purpose in this study was two-fold: 1) to translate the English AC-QoL into Chinese; and, 2) to test the reliability and validity of AC-QoL in Chinese stroke caregivers.

Methods

Results

After excluding invalid questionnaires, our study population consisted of 409 stroke caregivers.The demographic and medical characteristics of the stroke patients are shown in Table 1. The stroke patients ranged in age from 22 years old to 97 years old and the average age was 64.13 years old (standard deviation [SD] 16.30). Most patients were male (64.10%), married (81.4%), and had finished primary education (33.3%), but were not employed (80%). Approximately 61.9% of the patients had health insurance at the city level. Most of the patients had an ischemic stroke (65.5%), for a majority of patients (63.3%) this was their first stroke (63.3%), and the majority of stroke events had occurred no more than 6 months previously (56.0%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic and medical characteristics of stroke patients (n = 409).

https://doi.org/10.1371/journal.pone.0186680.t001

The demographic characteristics of stroke caregivers are shown in Table 2. The caregivers ranged in age from 18 years old to 83 years old and the average age was 48.92 years old (SD 16.30). Most caregivers were female (63.8%), married (87.3%), and had completed high school or a higher level of education (52.8%). Nearly half of the caregivers (47.7%) were the children of the stroke patients. The majority of caregivers did not live with the patients (75.6%), but provided care to them for more than 12 hours per day (64.1%), and were not employed (56.0%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Demographic characteristics of stroke caregivers (n = 409).

https://doi.org/10.1371/journal.pone.0186680.t002

We tested an 8-factor model, in keeping with the structure of the original AC-QoL version. The criteria to evaluate the model fit are shown in Table 3. Our results indicated that we needed to perform EFA to modify the model for the Chinese AC-QoL.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Goodness-of-fit indices for the AC-QoL factor models.

https://doi.org/10.1371/journal.pone.0186680.t003

Our results indicated that the item total correlation values ranged from 0.317 to 0.734 (all with p < 0.001); the item total correlation value of item 19 was 0.193, and that of item 37 was 0.244, so we deleted these items before we performed EFA. Extreme group comparison t scores ranged from 5.501 to -17.976 (all with p < 0.05) and the test-retest reliability value was 0.86. The ICC value was 0.924 (95% confidence interval [CI] 0.857 to 0.960). The I-CVI values ranged from 0.90 to 1.0, and the S-CVI was 0.98.

There were five factors in the model, which we rotated using Varimax, accounting for 69.27% of the total variance (the details are shown in Table 4). Our estimates of the Kaiser-Meyer-Olkin measure of sampling adequacy (KMO) and Bartlett’s test indicated that our data was suitable for EFA (KMO = 0.908, χ2 = 9530.99, P < 0.001). We extracted five components with a variance of 67.56% according to the standard and each factor’s loading should be greater than 1 (scree plot shown in Fig 1). We considered for inclusion in the five components each item with a loading value that was higher than 0.4 (Table 4). Based on our EFA results, we named factor 1 “Caring Benefits”, while factors 2, 3, 4 and 5 were in the same domain as the original version, so we kept the same names (“Caring Stress”, “Caring Choice”, “Support for Caring” and “Money Matters”). We moved item 38 from “Carer Satisfaction” in the original version to “Caring Stress.” We removed items 1, 16, 19, 22, 24, 25, 28, 29, and 37 based on our predetermined criteria.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Scree plot of Chinese version AC-QoL.

AC-QoL: Adult Carer Quality of Life Questionnaire.

https://doi.org/10.1371/journal.pone.0186680.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Five factors with factor loading for stroke caregivers (n = 409).

https://doi.org/10.1371/journal.pone.0186680.t004

The Cronbach’s alpha of our Chinese-language AC-QoL for stroke caregivers was 0.924, Cronbach’s alpha for the subdomains ranged from 0.779–0.920, and the model-based reliability coefficient was 0.914. These results indicated that our instrument had good reliability.

Table 5 shows the descriptive statistics of our Chinese version of the AC-QoL (S2 File) in caregivers for stroke survivors. The overall mean score of caregivers on the AC-QoL was 55.20 with an SD of 15.88. Each item had low to moderate non-normality.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. The descriptive statistics of Chinese version AC-QoL in caregivers (n = 409).

https://doi.org/10.1371/journal.pone.0186680.t005

Discussion

Our work is the first to indicate a cultural variable factor structure in caregivers of patients with stroke in China. Our findings are similar to those of Ozer et al [45].

Interestingly, the results of our factor analysis suggested that some of the AC-QoL items are not intrinsic to the QoL of stroke caregivers. Nine items were not included in the factor structure of our Chinese version. This may be due to differences between the Chinese culture and the British culture. In a manner similar to the Chinese caregivers of cancer patients [24], our Chinese caregivers may have assumed that they would be the providers of care for their loved ones, whether parents or spouse. This may be related to the concept of filial piety in Chinese culture [46].

The characteristics of our stroke caregivers may also have influenced the differences between Chinese version and English version. Our caregivers were mostly younger than their counterparts in the UK. Reports from the literature indicate that advancing age in stroke caregivers puts those caregivers at an increased risk for adverse outcomes [47]. Our caregivers had also cared for their loved ones for significantly less time than their British counterparts (six months versus twelve years, on average) [31], and previous researchers have demonstrated that caregivers had lower burden and higher general self-efficacy in the first six months post-stroke [48]. In addition, the families of the stroke patients in our study were relatively well-off financially (most had a monthly income of more than 3000 Yuan), which may have influenced their responses to the questions, i.e., they may have felt less of a financial burden with respect to caregiving (in keeping with the findings of Hanratty et al. [49]), unlike caregivers from rural areas who are poor. Further exploration of these differences should be conducted in larger groups from different economic levels of society.

Our Chinese-language AC-QoL contained five domains (Caring Benefits, Caring Stress, Caring Choice, Support for Caring and Money Matters). This is similar to the Chinese version of the cancer caregiver QoL instrument, which (after cross-cultural adaptation) included domains for Burden, Disruptiveness, Positive Adaptation and Financial Concern [24]. According to White et al.’s model of QoL for family caregivers of stroke survivors [23], caring stress is a caregiver factor, whereas support for caring and money matters are environmental factors. Brand et al. demonstrated that more benefit finding by caregivers can increase their perceived QoL [11]. Thus, our Chinese version of AC-QoL covered a wide range of the aspects of QoL and focused on both the negative and the positive outcomes of caregivers.

There are a few limitations to the generalizability of our findings. First, we used a convenience sampling method and we recruited participants from hospitals only in Zhengzhou, China, which means our findings cannot be generalized to caregivers from other provinces. Second, we recruited our participants from within the hospital setting—the QoL of caregivers may be significantly different within their communities and their homes. Future researchers should recruit participants from communities. Third, we did not evaluate a criterion-related validation of the constructs in our Chinese version of the AC-QoL. Research should be conducted on the criterion-related validation using well-developed tools such as the SF-36 or WHOQOL. The validity assessment is needed in future research.

Despite these limitations, our findings indicate that our version of the AC-QoL is a reliable and valid tool which can be used to assess the QoL of stroke caregivers in mainland China, which may have some important implications. First, health care providers should consider the QoL of caregivers for stroke patients, which they can assess using the AC-QoL. Second, the AC-QoL may be useful as a potential screening tool to detect caregivers who need help. Third, assessing both the positive and negative aspects of the QoL of stroke caregivers provides a more comprehensive view of caregivers, which can help us to understand better the caregiving experience and caregivers themselves. Fourth, we can investigate more deeply the factors influencing the QoL of caregivers, including factors which may mediate or modify the interrelating domains of QoL. Fifth, the AC-QoL can be used to assess the effectiveness of interventions, which may help to improve the QoL of caregivers. Moreover, an improved understanding of the QoL of caregivers may help us understand how the QoL of caregivers may affect that of stroke patients. A better understanding of the association of caregiver factors with patient outcomes may have implications for rehabilitation research, professional practice, policy directions and resource allocation [50]. Finally, QoL can be implemented as a cost-effective measure in studies of interventions attempting to improve the outcomes of stroke survivors [51]. Reports from the literature indicate that the QoL of stroke survivors is highly correlated with that of their caregivers, and is greatly associated with the physical function changes of stroke survivors [16]. The findings from previous studies indicate that caregiver training is associated with less financial burden and a higher QoL for both stroke survivors and their caregivers [52, 53]. However, there is limited evidence that interventions for caregivers of stroke survivors are (cost) effective, thus more economic evaluations are needed [53].

In conclusion, this study provides some evidence on the reliability and validity of the AC-QoL is a reliable and valid tool which can be used to assess the QoL of stroke caregivers in mainland China. Our version of the AC-QoL has slightly different items and components when compared with the English version. We therefore suggest that instruments from developed countries may need to be modified when used in developing countries or in different cultures.

Supporting information

Acknowledgments

The authors want to thank all the caregivers of stroke who took part in this study, and thank all the experts for translation, back-translation, conceptual and content identification. We appreciate the authors of the AC-QoL for their permission to translate and use the AC-QoL, especially Professor Stephen Joseph for helping us during the whole process. We want to thank Dr. Anna Beeber and Dr. Lixin Song from University of North Carolina at Chapel Hill for their kind-hearted guide for the writing and revising. We are grateful for Dr. Mary Lynn and Dr. Wiesen Christopher from University of North Carolina at Chapel Hill for their selfless statistical consultation.