| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Joseph E. Gaugler, PhD, Department of Behavioral Science, University of Kentucky, 110 College of Medicine Office Building, Lexington, KY 40536. E-mail: jgaugle{at}uky.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Nursing home placement • Alzheimer's disease • Family caregiving • Transitions • Informal care
Duration of Dementia Care
The process underlying the progression of family caregiving appears to be complex, as informal care for disabled older adults is not attached to a defined social position or well-established norms of behavior. Despite theoretical inferences about the dynamic, longitudinal nature of caregiving, initial descriptive research attempted to explore the wear-and-tear hypothesis. The wear-and-tear hypothesis suggests that the longer a caregiver remains in her or his role, the more likely negative outcomes, such as caregiver distress or even care-recipient institutionalization, will occur. Many early efforts were cross-sectional in design and attempted to correlate duration of care indices (i.e., how long have you been providing care to your relative?) with various measures of caregiver burden, depression, and other forms of psychosocial distress. More recent research suggests that duration of care rarely appears as an independent, significant predictor of psychosocial outcomes or institutionalization when it is considered with other time-varying covariates (e.g., Gaugler, Kane, Kane, Clay, & Newcomer, 2003; Walker et al., 1996). These prospective longitudinal studies instead support more complex views of caregiving duration. It appears that duration of care operates in concert with contextual variables such as educational and social roles (Moen, Robison, & Dempster-McCain, 1995) and amount of care provided (Walker et al., 1996) to account for changes in important psychosocial outcomes over time. In fact, several panel studies of dementia caregiving have found that caregivers who remain in follow-up intervals report stability or decreases in negative emotional and psychological outcomes, such as burden and depression (Townsend, Noelker, Deimling, & Bass, 1989; Zarit, Todd, & Zarit, 1986). These analyses have supported counterhypotheses in the literature that emphasize the ability of families to tolerate once stressful situations and to adapt to and successfully manage chronic caregiving demands. However, the role of attrition in accounting for these trends is rarely addressed.
More recent work has examined how circumstances early in the caregiving role influence longitudinal outcomes. For example, several research efforts have explored how the transition from noncaregiver to caregiver, or, from providing no activity of daily living (ADL) assistance at one time point to providing help with at least one of these tasks in a subsequent interval, affects stress and negative mental health (e.g., Lawton, Moss, Hoffman, & Perkinson, 2000). In general, care recipients' functional dependencies appeared to influence the onset of informal caregiving (e.g., a health crisis) and caused distress initially, followed by a sense of stabilization in emotional and psychological outcomes. In an analysis of a subset of spousal caregivers from the Caregiver Health Effects Study, noncaregivers who transitioned into heavy ADL provision were more likely to suffer from depression after assuming the caregiving role than individuals who offered less intensive care to older adults (Burton, Zdaniuk, Schulz, Jackson, & Hirsch, 2003). Building on this research, a longitudinal study of dementia caregivers indicated that those family members who provided intensive assistance to their relatives prior to symptom recognition and diagnosis were less likely to institutionalize care recipients or suffer increases in emotional and psychological distress over a 3-year period than caregivers whose entry was predicated on symptom recognition or diagnosis (Gaugler, Zarit, & Pearlin, 2003). These empirical relationships were apparent even when important changes in care-recipient function and other covariates were accounted for. In sum, recent findings have implied that wear and tear or longitudinal adaptation in the caregiving career may have more to do with how the role is acquired than with length of time as caregiver.
Research Focus
Current research has demonstrated the importance of role entry in predicting caregiving stress and institutionalization (e.g., Burton et al., 2003; Gaugler, Zarit, et al., 2003), but few studies examine how care demands that occur early in dementia care affect long-term outcomes, such as care-recipient institutionalization. For example, one demand ranking among the strongest empirical predictors of institutionalization is problematic behavior (e.g., Gaugler, Kane, et al., 2003). Family members who deal with stressful, frequent behavior problems early in the role may be expected to have the most difficulty managing care demands over time. In contrast, family members who have the opportunity to gradually ease into care responsibilities and adapt to fewer behavioral outbursts may be less likely to relinquish in-home care responsibilities. Classic sociological research on role transitions has emphasized that difficult transitions into roles are more likely to cause distress or exit (Cottrell, 1942). Similarly, it is expected that family members who must deal with intense care demands earlier may expedite their exit from in-home caregiving.
In the present study we utilize 3-year data to determine how duration of care potentially moderates behavior problems' influence on care-recipient institutionalization. Strengths of this data include its multiregional design; dementia patients and their family caregivers were recruited from eight catchment areas (Rochester, NY; Urbana, IL; Memphis, TN; Portland, OR; Cincinnati, OH; Parkersburg, WV; Minneapolis, MN; and Miami, FL). Although the sample is not nationally representative of dementia caregivers, the multiregional design makes the findings from this study more generalizable than those from prior research. Using such data, in this study we test the following hypothesis: If individuals are caring for a relative with severe behavior problems earlier in the caregiving role, they will be more likely to institutionalize their relative.
Although we anticipated that those caregivers who had provided assistance for shorter periods of time and under more intense care demands would institutionalize sooner, the literature has established other variables that may account for such outcomes. For example, the stress process model (Aneshensel, Pearlin, Mullan, Zarit, & Whitlatch, 1995) for dementia caregivers offers a multidimensional, theoretical framework to analyze predictors of institutionalization (see Figure 1). Important components of the model include background characteristics of the caregiver and care recipient (e.g., age, kin relationship of caregiver to care recipient, and living arrangements) and several types of stressors that potentially affect caregiver outcomes (e.g., primary objective stressors, such as ADL dependencies and cognitive impairment; and primary subjective stressors, or caregivers' emotional reactions and appraisals of care demands). Moreover, the stress process model considers variables that directly mitigate the deleterious implications of caregiving stress, such as assistance provided by other family members or friends or community-based long-term-care services. Finally, the stress process model includes global outcomes, such as depression or caregiver health, that are a result of the cumulative proliferation of stress that is due to care responsibilities. These components of the stress process model have been found to influence cognitively impaired older adults' time to placement (Gaugler, Kane, et al., 2003), and in the subsequent analysis we incorporated the stress process when we examined duration of care, behavior problems, and institutionalization.
|
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Initial findings from MADDE suggested that the expanded case-management protocol had little or no influence on care-recipient institutionalization (Miller et al., 1999). Specifically, the provision of case management and subsidized community long-term-care services to randomly assigned caregivers in MADDE did not have significant effects on nursing home entry rates when compared with usual care controls. A series of subgroup analyses appeared to confirm the main findings as well. For these reasons, we included both the treatment and control conditions from MADDE in subsequent analyses.
Sample
The baseline sample included 5,311 care recipients and their primary caregivers who received an initial assessment prior to random assignment and were scheduled to receive interviews every 6 months thereafter. Information on duration of care was collected at the first interview subsequent to baseline and then throughout the study. Because information on duration of care was not collected in the baseline interview, we do not include those individuals who died or were institutionalized during the initial 6 months of MADDE (n = 550) in this analysis. These exclusion criteria resulted in a final sample of 4,761 primary caregivers and their care recipients. (For statistical comparisons between participants who provided duration of care for this analysis and those who did not, please contact the authors.)
Measures: Independent Variables
Duration of Care
Caregivers were asked how long, in months, they provided help to the care recipient because of his or her dementia. As Table 1 shows, caregivers had provided assistance to cognitively impaired care recipients for 48.08 months prior to baseline, on average (SD = 48.86; range = 1.00–1003.00).
|
Behavior Problems
We utilized the 19-item behavior problems measure developed by Zarit and associates (Zarit, Orr, & Zarit, 1985). The behavior problems measure assessed a range of behavioral disruptions such as asking repetitive questions and being suspicious or accusative (responses are 0 = no and 1 = yes and were summed). The behavior problems measure showed good reliability at baseline (
=.93). In order to capitalize on the prospective follow-up data available in MADDE, we used the average behavior problems score for intervals the caregiver remained in the study. Descriptive information is presented in Table 1.
Measures: Dependent Variable
Institutionalization
MADDE obtained institutionalization dates for stays that the caregiver reported as permanent. Any stays initially paid for by Medicare that ended in the care recipient's death (rather than discharge to the community) were also classified as permanent stays. Nursing home entry dates for short stays (i.e., fewer than 60 days) were compiled, but these stays were not counted as permanent nursing home days. Efforts to establish data quality in MADDE and corroborate caregiver reports with demonstration-financed reimbursement claims were successful (Newcomer, Spitalny, Fox, & Yordi, 1999). By the end of the 3-year study period, 2,185 caregivers (45.9%) had institutionalized their care recipients; among care recipients who were institutionalized, average length of time in the study was 473.44 days (SD = 238.44; range = 31.00–1094.00).
Measures: Stress Process Covariates
We included a number of covariates to control for other potential influences of the stress process on institutionalization. Table 1 presents key descriptive baseline information for the sample.
Context of Care
Care-recipient demographic variables included site, gender, race, age, Medicaid status, living arrangement, and whether the care recipient was assigned to the MADDE treatment or control condition. Caregiver demographics included gender, caregiver relationship to care recipient, age, income, employment status, and education.
Primary Objective Stressors
Functional and cognitive status variables included care-recipient dependence on personal assistance with ADL tasks (walking, wheeling, transferring from bed or chair, grooming, bathing, dressing, eating, using the toilet, having bowel or bladder accidents, and using transportation out of walking distance; Katz, Ford, Moskowitz, Jackson, & Jaffee, 1963) and dependence on personal assistance with IADL tasks (meal preparation, shopping, routine housework, money management, laundry, telephone, medications, heavy chores–home maintenance; Lawton & Brody, 1969). Responses included 0 = no difficulty, 0.5 = some difficulty, and 1 = maximum difficulty. The ADL and IADL measures showed strong reliability at baseline ( =.89 and =.89, respectively). Case managers administered the 30-item Mini-Mental State Examination, which assesses orientation, recall, and ability to name objects (Folstein, Folstein, & McHugh, 1975; baseline =.95). We assessed the number of hours that caregivers typically spent managing care recipient's functional and cognitive needs during a typical week. We also summed caregivers' unmet needs (i.e., not enough help indicated by the caregiver) with care recipients' ADL and IADL limitations (baseline =.79).
Primary Subjective Stressors
We measured caregiver burden by using the seven-item version of the Zarit Burden Scale (Zarit et al., 1986). Items included the following: Do you feel stressed between caring for the care recipient and meeting other family responsibilities, or stressed between caring for the care recipient and having enough time for yourself? angry around the care recipient? tense or anxious because of your involvement in caregiving? that your health or your social life has suffered because of caregiving? that you have lost control of your life since the care recipient's illness? burdened in caring for the care recipient? Responses ranged from 0 (never) to 4 (almost always) and demonstrated strong internal reliability at baseline ( =.87).
Resources
We asked primary caregivers to identify, from a fixed list of options, the services they had used in the past 6 months and how often they relied on these services. Because three community-based services accounted for 80% of all community-based care use in the MADDE samples (chore, personal care, and adult day care), we included these services in the analysis. Caregivers also reported the number of times overnight hospital services were used in the 6 months prior to the baseline interview. Comparisons of self-reported service use with demonstration-reimbursed claims in the MADDE analyses found that 93% of the individuals could correctly identify that they were or were not receiving a service (Miller et al., 1999). Although the reporting of actual service units was less reliable, we found no systematic bias. Secondary caregiving hours were measured by asking respondents how many hours per week they typically received help from other family members or friends.
Global Outcomes
We measured caregiver depression by using the 15-item Geriatric Depression Scale; responses include 0 = no and 1 = yes (Yesavage, Rink, Rose, & Aday, 1983; example items include the following: Do you often get bored? Are you hopeful about the future? Do you feel downhearted and blue?). Baseline reliability was strong ( =.98). We summed caregivers' ADL and IADL dependencies (baseline =.84, =.75, respectively), and we included a single-item self-rating of health (item responses included 1 = poor, 2 = fair, 3 = good, 4 = excellent) to further assess global caregiver functioning.
Analysis: Timing of Care Demands and Care-Recipient Institutionalization
Analysis 1
To initially determine whether duration of care and behavior problems interact to influence care-recipient institutionalization, we utilized a multinomial logistic regression. A multinomial logistic regression is used to predict a multiple discrete outcome based on a set of predictor variables that are continuous, dichotomous, or discrete. Because this analysis includes an outcome that has four discrete and conceptually distinct levels (care recipient remained in the community, care recipient was institutionalized, care recipient died, and caregiver or care recipient was lost to follow-up), a multinomial logistic regression was an appropriate method for the first analysis. Baseline covariates included variables from the other dimensions of the stress process model (i.e., context of care, primary objective stressors, primary subjective stressors, resources, and global outcomes). We also entered the main effects of duration of care (log transformed) and average behavior problems during the course of MADDE in the multinomial logistic model; the independent variable of interest was an interaction term of Duration of care (log transformed) x Behavior problems. We anticipated that those individuals who provided care for a shorter period of time and for care recipients with more frequent behavior problems would be more likely to institutionalize care recipients when compared with respondents who were in different stages of caregiving and assisted relatives with varying ranges of behavior problems.
Analysis 2
In order to further capitalize on the 3 years of data available in the MADDE study design, we conducted an event-history analysis. An event-history analysis (also called survival analysis or hazards modeling) examines whether a particular event occurs (e.g., institutionalization) and, if so, when. We utilized a specific type of event-history analysis, a Cox proportional hazards model, to analyze the empirical effects of the onset experience on the probability that caregivers would institutionalize care recipients during the 3-year course of the study. Like other types of event-history analyses (such as life tables or Kaplan–Meier analyses), a proportional hazards model corrects for bias in estimates caused by censoring (e.g., a care recipient dies, is lost to follow-up, or remains in the study for the 3-year duration). By utilizing the Cox proportional hazards model approach, the subsequent analysis incorporated censored cases as well as those individuals who experienced institutionalization, and was thus able to include the entire sample of 4,761 caregivers and care recipients with dementia. Specifically, the dependent variable in an event-history analysis is a combination of time and whether the event occurs or not. We assigned censored observations in this study a length of time representing their duration of participation in the analysis (i.e., from 0 to 1,096 days), but the persons in these cases were not coded as having experienced institutionalization.
The Cox model included a wide range of baseline covariates, including measures of context of care, primary objective stress, primary subjective stress, resources, and global outcomes. Following entry of the average behavior problems and duration of care (log transformed) variables, the independent variable of interest was the interaction between behavior problems and duration of care. As in Analysis 1, we anticipated that those who assumed care responsibilities for relatives with frequent problematic behavior earlier would be more likely to expedite nursing home placement. The number of months to nursing home placement from baseline MADDE interviews was the dependent variable of interest.
Preliminary bivariate analyses of potential predictors of placement suggested that two important covariates, ADL dependencies and MMSE, had curvilinear relationships with placement decisions. For these reasons, we partitioned baseline MMSE and ADL scores to examine at what particular level care recipients were most likely to enter nursing homes. An additional analytical issue was potential multicollinearity between key covariates, particularly baseline measures of behavior problems and other primary objective stressors, primary subjective stressors, and depression. A preliminary correlational analysis demonstrated that the behavior problems variable was weakly associated with ADLs (r = –.10), IADLs (r =.15), MMSE (r =.03), and caregiver depression (r =.13). The correlation between behavior problems and burden was slightly higher (r =.21). All of these correlations were significant at the p <.05 level; however, the significance of the bivariate analyses may be due as much to the very large sample size as to any clinically relevant relationships. As is evident in a review of the covariates in Tables 2 and 3, these associations did little to dilute the predictive power of ADL, IADL, MMSE, or burden scores on time to institutionalization.
|
|
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Institutionalization
As shown in Table 2, a number of stress process covariates were associated with care-recipient institutionalization during the 3-year study period. Consistent with other research findings on nursing home placement, predictors included Medicaid eligibility, MMSE, care-recipient ADL dependencies, and burden (e.g., Gaugler, Kane, et al., 2003). The Duration of care x Behavior problems interaction term was not related to institutionalization.
Death
A number of stress process covariates were also predictive of care-recipient death during the 3-year study duration. Factors reliably associated with care-recipient death included care-recipient gender, MMSE score, care-recipient age, and ADL dependencies (see Table 2). The Duration of care x Behavior problems interaction term was a significant predictor of care-recipient death; B = –.04, SE =.02, exp(B) =.93, p <.01.
Loss to Follow-Up
We found several baseline variables to predict loss to follow-up, including site, Medicaid eligibility, and secondary support (see Table 3). The Duration of care x Behavior problems interaction term was not a significant predictor of loss to follow-up over the 3-year study period.
Analysis 2: Predicting Time to Nursing Home Placement
In order to take advantage of the 3-year data available and provide a more complete analysis of timing and institutionalization, we conducted a Cox proportional hazards model (see Table 3). Similar to findings from this and prior research, a number of stress process covariates were key predictors of time to nursing home placement in the 3-year study. However, the Duration of care x Behavior problems interaction term was not a significant predictor of time to institutionalization.
Main Effects Analysis
Because the interaction terms were not statistically significant in either the multinomial or Cox proportional hazards models, we conducted main effects analyses that excluded the interaction term. As demonstrated in the main effects multinomial logistic model in Table 2, both the behavior problems and duration of care variables had strong direct effects on institutionalization; respondents who reported a greater duration of care were less likely to place [B = –.27, SE =.05, p =.00; exp(B) =.76], whereas caregivers who assisted loved ones with greater behavior problems were more likely to institutionalize [B =.10, SE =.01, p =.00; exp(B) = 1.10]. Similarly, the main effects Cox proportional hazards model in Table 3 found that caregivers who entered their roles more recently institutionalized care recipients earlier [B = –.18, SE =.03, p =.00; exp(B) =.84]. Care recipients who engaged in more frequent problem behaviors were also placed in nursing homes sooner [B =.09, SE =.01, p =.00; exp(B) = 1.10].
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Data on caregivers' personality attributes also could have offered greater insight into the role of duration of care and behavior problems on time to institutionalization. As prior research has indicated, caregiver neuroticism and personality often have important effects on caregivers' perceptions of burden, depression, and physical health (Hooker, Monahan, Bowman, Frazier, & Shifren, 1998). Other work has shown that spousal caregivers with higher neuroticism and lower levels of mastery are more likely to report frequent and intense behavioral disruptions in older care recipients (e.g., Bookwala & Schulz, 1998). Although some of these empirical findings are based on cross-sectional data which make it difficult to infer the potentially complex causal relationships between caregiver personality and care-recipient behavior problems (e.g., personality traits and mannerisms of the caregiver may elicit more frequent or severe behavior problems in the care recipient, or intense behavioral outbursts may affect caregivers' feelings of mastery and neuroticism), it is possible that caregivers with such personality traits may also exit the caregiving role sooner, providing an alternative explanation to the data presented here.
Although the empirical results did not support the main study hypothesis (i.e., individuals who managed frequent behavior problems earlier in their caregiving roles would institutionalize sooner), duration of care had a strong, negative, and direct effect on time to nursing home placement. This finding is important, because prior research efforts report variable relationships between duration of care and other key caregiving outcomes. Because of the utilization of a multiregional sample, 3-year data measured at 6-month intervals, the strong conceptual framework, and consistent findings across different analyses, the empirical results lend confidence to the meaning of duration of care and its potential effects on important transitions in the dementia caregiving career. Specifically, the results illustrate the potential challenges families must deal with early in their roles. It appears that, for families who have assumed care responsibilities more recently, the demands associated with providing at-home assistance to individuals with dementia (regardless of disease severity) are particularly pronounced. Caregivers who are less experienced and attempt to manage the typical functional and cognitive dependencies of someone suffering from dementia (such as problem behaviors, ADL dependencies, or memory impairments) may have difficulty establishing care plans or routines to successfully address these problems. Without the opportunity to acclimate, negotiate, or manage responsibilities, care demands may become more difficult to manage for recent dementia caregivers. Subsequently, there is a need for external formal assistance, and institutionalization emerges as a practical option.
The findings also emphasize the important conceptual and methodological issues related to the analysis of dementia caregiving over time. In contrast to postulations that lengthy duration of care is associated with negative caregiving outcomes (i.e., wear and tear), the empirical findings instead lend support to other models of informal long-term care, where those who remain in their caregiving roles for longer periods of time may actually demonstrate an ability to adapt to dementia care. Given the importance of duration of care as a direct predictor of institutionalization, conceptual models of stress and adaptation should address length of caregiving when attempting to longitudinally describe and intervene on behalf of dementia caregivers. Understanding how events affect caregivers in different stages or time points appears integral to subsequent outcomes on the part of family caregivers (e.g., Burton et al., 2003; Gaugler, Zarit, et al., 2003; Moen et al., 1995). Although considering how and when dementia caregivers are most susceptible to negative outcomes is complex, it appears increasingly vital to characterizing the individual trajectories of role entry and exit that each family caregiver experiences. Caregiving research has yielded comprehensive models of caregiving adaptation that take into account how stress proliferates from care responsibilities to other life domains that influence global outcomes (e.g., Aneshensel et al., 1995). However, contextual and life-course issues such as circumstances surrounding caregiving onset and how this role progresses have yet to receive adequate attention. This refined focus, coupled with emerging research emphasizing that the demands of informal caregiving do not abate following the conclusion of in-home care responsibilities (e.g., nursing home placement; see Tornatore & Grant, 2002), would add considerably to current conceptual models that describe the dynamics of dementia caregiving stress over time and the importance of caregiving duration.
The results raise other issues to be addressed in future studies. Although much can be learned about the effects of duration on key caregiving outcomes from the retrospective approach adopted here, there is a need to develop prospective research that explores the progression of informal caregiving to a greater extent. The complex and rich prospective experiences that potentially define care responsibilities over time (e.g., the quality of the caregiver–care-recipient relationship, assumption of the primary caregiver role, care-recipient health status–deterioration, and family system support) can extend retrospective research on the longitudinal ramifications of caregiving. A prospective research design of dementia caregiving from onset to termination would be difficult to implement, however. The identification of a representative sample of families who are not yet providing intensive assistance as a result of a loved one's dementia but will do so in the foreseeable future is challenging. One possibility is the inclusion of at-risk control samples recruited at several Alzheimer's disease research centers (e.g., Schmitt, Wetherby, Wekstein, Dearth, & Markesbery, 2001). In addition, measurement points in prospective designs must be compressed enough to identify families who, because of abrupt circumstances, assume caregiving responsibilities quickly. Another intriguing aspect of timing and dementia caregiving that warrants further research attention is the inclusion of the care recipient's perspective. An increasing number of studies are recognizing that dementia patients, even in moderate stages of the disease, can provide information on their mood and quality of life (Brod, Stewart, Sands, & Walton, 1999). Such methods could be incorporated in subsequent studies of the early stages of caregiving and their longitudinal implications.
The findings also have practical significance. A greater understanding of duration of care and its effects can help refine the targeting and design of respite programs. Prior research on community-based long-term care emphasizes the difficulty inherent in targeting programs to delay nursing home placement or achieve other desired outcomes (e.g., Weissert & Hedrick, 1994). Similar issues occur in the implementation and evaluation of respite and psychosocial services for dementia patients and their caregivers; a common lament in the literature is that many families utilize services late in their particular caregiving careers, making it difficult for respite or psychosocial programs to exert positive benefits (Gaugler & Zarit, 2001). As the results here suggest, if community-based respite and psychosocial services are delivered to caregivers soon after role acquisition, these programs may have a better chance of exerting positive, beneficial effects for both the older client and family caregivers. Moreover, the findings emphasize the wide-ranging effects of behavior problems on caregiving outcomes (e.g., Pinquart & Sorensen, 2003). Clinical interventions that help caregivers manage and deal with the unpredictable behavior problems that disintegrate the informal care network may prove successful in not only alleviating caregiving burden and exhaustion but possibly delaying placement as well. Novel psychosocial caregiver interventions such as the multisite Resources for Enhancing Alzheimer's Caregiver Health project (see Schulz et al., 2003) have demonstrated effectiveness in reducing caregiver stress by helping family members manage and cope with behavior problems through counseling, training, and similar techniques. Interventions that target the consequences of behavior problems may be particularly effective in delaying nursing home placement as well.
The results also offer some insight into attrition and its influence on longitudinal adaptation among dementia caregivers. As prior studies have indicated, dementia caregivers often report decreases in burden, depression, or other indicators of stress over time, suggesting that caregivers actually adjust to their roles and responsibilities (e.g., Aneshensel et al., 1995; Townsend et al., 1989). However, it is possible that early exit from the caregiving role by means of institutionalization may selectively bias any longitudinal panels of care recipients and their family members. The findings of the present study imply that length of time in the caregiving role helps to determine whether caregivers relinquish in-home care responsibilities in favor of institutional placement. Future analyses of caregivers' psychosocial adaptation must account for attrition and its potential impact on any subsequent longitudinal results.
|
Footnotes |
|---|
1 Department of Behavioral Science, University of Kentucky, Lexington.
2 School of Public Health, University of Minnesota, Minneapolis.
3 Clay Software and Statistics, Ashland, OR.
4 Department of Social and Behavioral Science, University of California–San Francisco.
Decision Editor: Linda S. Noelker, PhD
Received for publication July 1, 2003. Accepted for publication December 20, 2003.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. J. McClendon, K. A. Smyth, and M. M. Neundorfer Long-term-care placement and survival of persons with Alzheimer's disease. J. Gerontol. B. Psychol. Sci. Soc. Sci., July 1, 2006; 61(4): P220 - P227. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. T. Buhr, M. Kuchibhatla, and E. C. Clipp Caregivers' Reasons for Nursing Home Placement: Clues for Improving Discussions With Families Prior to the Transition Gerontologist, February 1, 2006; 46(1): 52 - 61. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Gaugler, R. L. Kane, R. A. Kane, and R. Newcomer Early Community-Based Service Utilization and Its Effects on Institutionalization in Dementia Caregiving Gerontologist, April 1, 2005; 45(2): 177 - 185. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
