Caregiving process and caregiver burden: Conceptual models to guide research and practice

Most of the studies conducted in this field of parent well-being in the context of chronic childhood impairments involve a survey with a variety of questionnaires, or either structured or unstructured interviews conducted with caregivers. The present review was not intended to be comprehensive. The review only included studies that were published in English in peer-review journals. The search was performed on following four databases: (1) MEDLINE, (2) CINAHL, (3) PsychINFO, and (4) Sociological Abstracts. These databases were searched from their date of release to the end of 2002. The search was conducted using combinations of the search terms and included terms such as "caregiv*" or "care giv*" or "parent" or "mother" or "father" or "informal care", "child health", "child disability" and "pediatrics". In addition, we also examined the reference lists of all included papers to identify any additional articles that might have been missed by our search strategy. Table 1 summarizes studies examining the factors involved in caregiver stress. However, we have also explored the key studies from the geriatric literature to identify factors that might be relevant to the caregiving process in a pediatric population (Table 2). Evidence suggests that there is considerable variation in how parents adapt to this caregiving demand[9, 10]. Some adapt well to the challenges of caring for a child with a disability, while others do not. The typical outcomes studied relate to caregiver physical and mental health, their well-being, as well as their ability to maintain their social roles.

Many studies have sought primarily to qualify the association between caregiving and negative health outcomes. Cadman, Rosenbaum, Boyle and Offord (1991) found parents of children with disabilities were more likely to experience depression and distress than parents of children without disabilities[11]. Another study found that caregiving for a child with a developmental disability places increased demands on the caregivers[12]. Dyson (1993) confirmed significantly higher parental stress in parents of children with disabilities, who were also found to have pessimism regarding the future[13]. In addition, the parents' perception of how difficult it was to care for the child was related to feelings of depression[14]. Much of the research has focused primarily on mothers of children with a disability. It has been shown that these mothers are more stressed than mothers of children without disabilities due to the extra daily tasks which take away time from the mothers to take care of themselves[15]. Dunst, Trivette and Cross (1986) also found that mothers of children with disabilities reported poorer emotional and physical health and that they felt that there were greater demands on their time from the child[16]. Gowen, Johnson-Martin, Goldman and Applebaum (1989) similarly reported that mothers of children with disabilities had more difficulty caring for their children and that they felt that they could not find enough time for themselves due to the increased daily demands associated with caring for a child with a disability[14, 17].

The literature supports a link between context, especially socioeconomic status (SES), and health [18‐21] but results have been mixed concerning the physical and psychological health of the caregivers. In the literature concerning caregivers of the elderly, lower SES, age, and marital status of the caregiver have been shown to be linked to psychological health of the caregiver[22, 23]. Higher SES has been associated with fewer psychological life stressors[10]and better emotional well-being[9]in caregivers of children. It is of course recognized that the SES is but one of several important indicators of the context.

In terms of how child characteristics have been shown to modify the health outcomes of caregivers, most findings concern psychological health of the caregiver. King et al. (1999) found child behavior problems to be the single most important predictor of caregiver psychological well-being[24]. Children's temperament has been related to feelings of depression in the caregiver and to self-rated scores of well-being and competence in mothers of children with disabilities[25]. Furthermore, severity of the disability has been repeatedly shown to be related to parental stress [26‐30]. Other variables relevant to caregiver psychological health may include age and gender of the child, level of the child's communication ability, and the presence of specific cognitive or sensory impairments[11, 13, 27, 31, 32]. Literature examining intrapsychic factors has shown that caregivers' self-perception is related to overall caregiver health. Mastery is a concept addressing 'the general degree to which one experiences control over what goes on in his or her life' [33]. This sense of mastery has been repeatedly shown to be associated with both physical and psychological health in a variety of populations [34‐36]. Similarly, low self-esteem and 'loss of self' have been shown to be associated with maternal depression[35, 37].

A portion of the literature has investigated various aspects of coping. Friedrich et al. (1985) showed that caregivers in difficult circumstances but with relatively high social support managed better than those with less social support[26]. The level of available informal social support has consistently been shown to be associated with both fewer reports of depression and better physical health in caregivers of the elderly[22, 23, 38, 39]. Satisfaction with social support networks has been associated with more positive caregiver attitudes and personal well-being[16, 24, 40]. In a study of 48 mothers and 48 fathers of children with disabilities, Frey, Greenberg, and Fewell (1989) determined that "positive belief systems, or a non-critical family network" were associated with low psychological distress in mothers, and that social network predicted family adjustment[27]. Social support was investigated by McKinney and Peterson (1985), who identified peer support as one of the most important predictors of stress[12], while Dunst, Trivette and Cross (1986) demonstrated that parents who had more satisfaction with social supports reported fewer emotional and physical problems[16]. Family functioning and stress management have also been shown to serve as coping factors[9, 41‐43]. Freidrich (1979) showed that marital status was the best predictor of coping for mothers of children with disabilities, indicating spousal support as a coping factor[44]. This was supported by Erikson and Upsur, (1989) and Lambrenos et al. (1996)[31, 45]. Specific stress management strategies can be associated with better caregiver health [27, 46].

Most of the current broadly-based studies investigating caregiver health in the child health literature do not rely on any specific theoretical frameworks that guide research into an understanding of the mechanisms by which some caregivers experience negative consequences and other do not. The majority of the published research has focused on understanding hypotheses using traditional analytical approaches that examine the relationship of a factor to the outcome after adjusting for other variables. However this method has many shortcomings. This type of study is used primarily to estimate the "independent" or direct effects of childhood disability on the "dependent" variable, in this case caregiver health. However, single factor changes are relatively rare outside of the context of a natural or laboratory experiment, as are the assumptions of linear, additive relationships and perfect measurement which are inherent to regression analysis. Moreover, few regression-based studies look specifically for possible contingent relationships among explanatory mechanisms; regression analyses will typically assume that such interactions do not exist[47]. Although these studies have been beneficial in pointing to a wide range of possible mechanisms mediating caregiver health, they fall short of providing an integrated picture by failing to examine comprehensively the direct and indirect pathways occurring within this array of factors that may influence health outcomes. To understand the complex web of direct and indirect relationships that impact the health and well being of family caregivers, research needs to test hypotheses that are guided by theoretical frameworks, whereby these hypotheses can be tested using advanced statistical techniques such as structural equation modeling.

This model was designed to assess the informal caregiving processes affecting caregiver health[35, 49]. Similar to other models described above, Aneshensel et al. and Pearlin et al. applied the stress process model to the caregiving of older adults with dementia who at one point were independent but where the relationship had now changed to one where the caregiver was responsible for meeting all the needs of the older adult with dementia.

As mentioned earlier, the caregiving experience and associated stress process reflects a process that changes over time. Pearlin and colleagues define stressors as the "problematic conditions and difficult circumstances experienced by caregivers" that strain or supersede the individual's capacity to adapt[8]. These are conceptualized as primary and secondary in nature. The primary stressors are linked directly to the individual and the disability, while the secondary stressors arise from the demands of the caregiving role itself. The potential proliferation effects of the stress involved in the caregiving role highlight the existence of a complex stress process[8]. This conceptualization allows for numerous instances for the moderators to impact the situation. The moderators include social supports and concepts of mastery or self-efficacy, which determine how people are impacted differently by the same stressors, and may help to sustain the caregiver and lessen the effect of the stressors[8].

The manifestations of stress (outcomes) are well-being, physical and mental health and the caregivers' ability to sustain their own social roles[8, 50]. A considerable base of independent research literature supports each of these elements. Pearlin and colleagues recognize that the inter-relationships among these variables change and develop over time, however what has not been researched are the processes or linkages that join these components. The Caregiving Career/Stress Process model suggests that "life events can lead to negative changes in people's roles, changes whose persistence wears away desired elements of self-concept, and that through this set of linkages stress is aroused[51, 52]. Coping and social supports, for their part, can intervene at different points along this process, thereby mediating the outcomes."[35].

Arrows within the model were chosen with two criteria in mind: first, that there be clear evidence in the literature to support an association between variables; and second, that the relationship be defensible from a theoretical, causal standpoint. We have selected constructs and latent variables for which there is some evidence in the literature, however there might be other constructs such as availability and access to care that can easily be incorporated based on new emerging evidence. While some of the causal relationships among the factors depicted in the proposed model (Figure 1) are almost certainly bi-directional, including bi-directional effects in a model often results in practical difficulties of empirical testing of the models. For this reason, the unidirectional relationships were chosen on the basis of the direction that was more defensible in terms of a causal model, or for which there was empirical evidence (Table 1 and 2)

In a cross sectional study with caregivers of children with cerebral palsy, the proposed model (Figure 1) will use a single comprehensive Structural Equation Model (SEM) to examine simultaneously the five major constructs described in this paper and thought to be relevant to the health of caregivers. SEMs are primarily representations of causal relationships between hypothesized constructs, rather than just the simple associations between those constructs[69, 70]. Each arrow in the model is therefore meant to indicate not simply an association between variables, but rather a theoretically-derived representation of the causal nature of that association. Our future research will therefore serve to test the proposed empirical model, and provide an example of a study that may be conducted under this new framework.

Our current research efforts are directed at testing this proposed model by conducting a cross-sectional study on caregivers of children with cerebral palsy. This paper outlines the variable measures and indicators that we have chosen for our study, which are congruent with our theoretical model.

Future studies using longitudinal data under this theoretical model could potentially provide many answers to the puzzles of caregiver health. The dynamic nature of the caregiving career might then be addressed empirically, which would help to provide insight into what stages are actually experienced by caregivers, and how these stages and the transitions between them can be made easier, potentially reducing the probability of negative health outcomes related to specific points along a dynamic continuum. Conducting a longitudinal study under the guidelines of the proposed model would therefore allow for increased specificity when determining the measures and assistance that should be implemented to relieve caregiver burden.