Mediating effects of burden on quality of life for caregivers of first-time stroke patients discharged from the hospital within one year

Most stroke survivors rely on family caregivers to assist in managing stroke-related deficits and reintegration into the community [1, 2]. Numerous studies have been conducted to explore the quality of life (QoL) of the family caregivers of stroke patients during various stages of treatment or rehabilitation in different countries [3‐6]. Additionally, both positive and negative impacts of stroke survivors on caregivers have been explored through cross-sectional and longitudinal studies [7, 8].

The relationships between characteristics of the stroke patient or the caregiver, caregiver burden, and caregiver QoL appear related to others, yet the direction of these relationships remains unclear among various populations. Several studies have shown that caregiver burden is either a strong determinant of caregiver QoL [9‐11] or an outcome instead of a predictor [12]. As such, caregiver burden could have a mediating effect between the patient or caregiver characteristics and the caregiver's QoL. Mediation analysis explores the process underlying how one variable influences another variable [13]. Caregiver burden with a mediating effect is consistent with limited earlier studies [3, 14]. Caregiver characteristics (spouse, caregiver's health status) and family resources (family income) were identified as significant predictors of caregiver QoL with a mediating effect of caregiver burden among stroke survivors receiving rehabilitation therapy [3]. The association between the characteristics of both patients and caregivers, as well as the caregiver's burden and QoL, revealed various results in different stages of stroke patients undergoing rehabilitation [14]. The caregiver's burden and QoL were associated with patient and caregiver characteristics (age, gender or mood), as well as family resources (family networks). Of these variables, only patient and caregiver characteristics affected caregivers' burden and QoL at the early stage of post-stroke. Furthermore, patient dependency and family resources were additional independent predictors for caregivers' well-being at a later stage [14]. However, only a limited number of studies have investigated QoL and associated factors affecting family caregivers for first-time stroke survivors within the critical one-year period of home care after being discharged from the hospital. Understanding the association between caregivers' QoL and influencing factors could be beneficial for mediating caregiver burden and improving caregiver QoL. The results from these investigations will help health providers develop caregiver interventions for post-stroke patients from the early stage, especially for first-time stroke families.

We hypothesized that there is an association among variables related to first-time stroke patients and caregivers, family resources, and caregiver QoL, and this relationship can be mediated by reported caregiver burden. Thus, to determine which patient or caregiver factors directly affect caregiver QoL, which predictors are simultaneously shared with caregiver burden and QoL, and which factors affect caregiver QoL through caregiver burden, the aim of this study was to explore the predictors of caregiver QoL by testing this hypothetical relationship and determining whether caregiver burden was a mediator for the QoL of caregivers of first-time stroke patients discharged from the hospital less than a year previously.

This was a cross-sectional study. We recruited subjects from lists of first-time stroke patients who had been discharged within 1 year. These lists were provided by a medical center, three regional hospitals, and a local registration system in a public health center located in northern Taiwan. We initially screened caregivers by phone for the following inclusion criteria: age > 20 years old and being the major family caregiver who provided the most hours in taking care of stroke patients. Caregivers could either be full- or part-time and were able to communicate verbally. After ethical approval was obtained from the Research Ethics Committee of Medical Foundation Institutional Review Board (No. 95-0489B), phone calls were made to individuals on the list to explain the purpose of the research, request participation, and screen for potential caregivers. Oral agreement was obtained before a home visit, and written informed consent was obtained from all participants before data collection. Participants completed questionnaires by face-to-face interview at caregivers’ homes. A total of 184 families agreed to participate; however, only 126 primary caregivers fit the selected criteria and completed questionnaires.

The independent variables were patient characteristics, caregiver characteristics, and family resources. The characteristics of stroke survivors consisted of demographic data and health information, including age, gender, marital and employment status, past medical history, duration of hospitalization, use of tracheostomy tubes, nasogastric (NG) tubes, Foley catheters, assistive devices for walking, activities of daily living (ADL) (Bartel Index, BI), and cognitive status (Short Portable Mental Status Questionnaire, SPMSQ) at the time of data collection. The BI is a widely used measure of ADL. The range of scores is 0-100, with a lower score indicating a higher dependent status of the subject [15]. Cognitive status was measured by the Chinese version of the SPMSQ [16], which has good reliability [17]. The total number of correct answers was adjusted by the educational level of subjects to produce the SPMSQ score (range 0-10). Caregiver characteristics consisted of demographic data and health information, including age, gender, education, employment and marital status, living with other family members, and self-rated health status. Additionally, family resources included monthly family income, the identity of the person responsible for the majority of the patient’s medical fees, the number of family members living together, the number of family members helping with caregiving tasks, and the presence of a hired non-family caregiver.

Another independent variable was the Caregiver Strain Index (CSI) total score. CSI was selected to measure the caregiver burden in five domains: employment, financial, physical, social and time stress [18]. The index consists of 13 items with yes (score = 1) / no (score = 0) answers for a total score range of 0-13, with a higher score indicating a higher caregiver burden. The internal consistency of the CSI Chinese version was good [19], and the reliability coefficient of Kuder-Richardson Formula 20 in this study was 0.93. In our study, the total CSI score was treated as a mediating variable between caregivers’ characteristics and their QoL.

The dependent variable was the total score of the Chinese version of the Caregiver Quality of Life Index (CQLI), which was chosen to measure the caregiver QoL. This index was developed by McMillan and Mahon [20], with four items on a 0-to-100-mm visual analogue scale to measure the emotional, social, financial, and physical aspects of QoL. Several studies revealed that the CQLI had good internal consistency [21‐23]. In this study, Cronbach's alpha was 0.75.

IBM SPSS Statistics version 23 was used for data analysis. Kolmogorov-Smirnov test results showed that the variables were normally distributed (p > 0.05). Univariate analyses (t-test, ANOVA and Pearson’s correlation) and multiple-regression analyses were conducted to identify predictors of caregiver QoL. We calculated power from the primary outcome of the multiple-regression. With our sample size of 126, power = 0.99 was computed under the effect size of 0.59, which meant our study achieved a large effect size [24], and our sample size was more than enough. All significance levels were set to 0.05 (2-tailed tests). To test whether caregiver burden had a mediating effect on caregiver QoL, a simple mediation model (one mediator existed) in the PROCESS SPSS computational tool was used. PROCESS is a modeling tool for SPSS and SAS that integrates many of the functions of statistical tools for mediation and moderation analysis [25]. Here, CQLI was treated as the outcome variable, and CSI was a mediator variable, while the other predictors were used as independent variables or covariates. Bootstrap analysis with 5,000 bootstrap samples was used to identify indirect effects. A 95% confidence interval (CI) was used to test the indirect effects of independent variables on the dependent variable via the mediator (caregiver burden, CSI). Mediation was considered significant if the bootstrapped 95% CI between the upper limit (BootULCI) and lower limit (BootLLCI) did not contain zero [26].

Our findings supported our hypothesis, and the financial factors, including lower monthly family income and the spouse paying the patient’s medical fees, had direct negative effects on caregiver QoL without mediation by caregiver burden. Moreover, the caregivers’ poor self-rated health and low educational background directly contributed to the caregivers’ poor QoL and their higher caregiver burden, and these predictors affect caregiver QoL indirectly with caregiver burden as a mediator as well. Therefore, interventions aimed at managing financial issues and alleviating caregiver burden by enhancing caregiver health and decreasing their gaps in stroke-related knowledge and caregiving skill may be effective in promoting caregiver QoL.

To improve caregiver QoL, family caregiver issues should be incorporated into healthcare system designs and policy implementations [36]. The Taiwanese government has promoted long-term care (LTC) policies in three stages since 2008. The first two stages focus on providing patients a home-care service system and community health care. However, the financial relief for caregivers is still limited [37]. In the near future, the third stage of LTC policies, a National compulsory long-term care insurance (LTCI) program, will include an assessment of caregiver strain index (CSI) and provide more medical care services and cash pay out to caregivers. LTCI will remit more caregivers’ financial burden than the first two stages. According to our research and experiences from other countries, we make several suggestions below for policy makers and health providers to promote the QoL of post-stroke family caregivers.

First, regarding caregivers’ financial issues, the following policies may be taken into consideration: (1) Cash benefits, medical benefits or life subsidy; in Germany and the Netherlands, the family receives cash benefits or medical benefits from their LTCI [38]. These cash benefits and contributions to their mandatory pension funds may directly improve caregivers’ financial QoL and encourage caregivers to take care of their family [39]. Additionally, health care insurance provides a subsidy for caregivers on leave or unemployed in Germany [40]. This approach could especially benefit primary caregivers or the major-payer for patients’ medical fees. (2) Expansion of institutionalization: Naomi et al. suggests that institutional care for a patient with a high degree of dependence could cost less than home care for families’ long-term care expansion in a remote area in Japan [41]. This type of care may help stroke patients with high dependency or those living in rural areas when providing subsidy to their institutions and support facilities are needed. (3) Coordinating multiple coverage between health and social policies: the Japanese LTCI scheme covers comprehensive health care and welfare services to relieve the financial burden on family caregivers [39]. LTCI and service delivery in South Korea provides multiple help, such as in-home benefits, facility benefits, family-care cash benefits and hospitalizations attending cash benefits [42]. These multiple financial supports may apply to post-stroke families and include discharge care plans and social and health policies systems as a whole, especially customized for lower-income families to provide appropriate help.

Second, regarding caregivers’ education level and health issues, we recommend providing those caregivers with easy-to-follow care instructions, tailored consultation, and free nursing-care training courses through discharge care plans, workshops or support groups. Additionally, health providers should consider not only patients’ conditions but also caregivers’ health status when providing community health services.