A Population-Based Assessment of the Impact and Burden of Caregiving for Long-term Stroke Survivors
Background and Purpose Little attention has been focused on the demands on caregivers in stroke outcome research. A major aim of this study was to identify factors in patients associated with emotional distress in caregivers 1 year after stroke.
Methods One-year stroke survivors with residual handicap (defined by the Oxford Handicap Scale) and their chief caregivers were interviewed as part of the follow-up activities for patients (n=492) registered with the Perth Community Stroke Study. We assessed emotional distress in caregivers using the Hospital Anxiety and Depression Scale and the 28-item General Health Questionnaire. Appropriate sections of the Social Behaviour Assessment Schedule were used to assess the patient’s behavior and the impact on the caregiver’s life. Other aspects of the patient’s functional state were assessed with the Barthel Index, the Mini-Mental State Examination, the Frenchay Activities Index, and the Psychiatric Assessment Schedule (at 4 months after stroke).
Results Of 241 patients who survived to 1 year after stroke and were living outside of an institution, 103 patients (43%; 95% confidence interval, 37% to 49%) were handicapped. Eighty-four patient/caregiver units were assessed from this latter group. Almost all caregivers reported adverse effects on their emotional health, social activities, and leisure time, and more than half reported adverse effects on family relationships. Forty-six caregivers (55%) showed evidence of emotional distress on either of the two screening instruments, particularly if they were caregiving for patients with dementia and/or abnormal behavior. There was no significant relationship between emotional illness among caregivers and the degree of patients’ physical disability.
Conclusions In this population, the high level of emotional distress among caregivers of stroke patients suggests that many caregivers have unmet needs. Community services need to focus attention on the neuropsychological aspects of stroke patients and the social functioning of caregivers who support them.
It is important to determine the specific characteristics of patients that make the greatest demands of caregivers and how this burden can be best managed, either by specific treatments for the patient or by measures aimed at providing support (for example, home help and financial aid) and relief (for example, sitters, day centers, and respite care) for caregivers.1 This arises in part from increasing financial constraints on the healthcare system, which has prompted a shift from institutional care to community care, and in part from concern about a reduction in the contribution of caregivers to the management of chronic diseases because of the increasing proportion of elderly in the population, the growing number of women in the workforce, and possible reluctance on the part of some family members to provide care.
Although stroke is one of the most common causes of disability and handicap in the community, only a few outcome studies have addressed the consequences of stroke on family members and other informal caregivers.2 Most studies are based on selected groups of patients admitted to the hospital or attending rehabilitation centers and report a high proportion of caregivers with adverse health and well-being.3 4 5 6 7 8 9 10 In contrast, the only community-based study of the topic to date found “surprisingly little measurable psychological stress” among caregivers, but no other aspects of caregiving were studied.11 Sample bias is one obvious explanation for this difference and has an important bearing on the generalization of stroke outcome data.12 Other methodological problems associated with studies of caregivers include differences in the definition of caregiver, small sample sizes, inadequate number of health variables studied, and differences in the timing and criteria for assessments.13 All of these issues make it difficult to know the extent and specific nature of the emotional and physical problems associated with caregiving and, from a public health perspective, to determine where practical interventions can be applied cost-effectively.
The aim of this study was to provide information on the effects of caregiving for stroke survivors identified from a population-based stroke register undertaken in Perth, Western Australia. We chose to measure the effects of caregiving when survivors were seen 1 year after the onset of stroke. Specifically, we examined the health and social consequences of caregiving (impact) and the relationship between emotional distress among caregivers (burden) and various characteristics of patients.
Subjects and Methods
Subjects for this study were recruited from a register of acute cerebrovascular events compiled as part of the Perth Community Stroke Study (PCSS). As described elsewhere,14 the PCSS identified and documented 536 stroke events among 492 members of a geographically defined and representative segment (population, 138 708) of Perth, Western Australia, during an 18-month period in 1989 and 1990. Seventy-two percent of patients had experienced their “first-ever-in-a-lifetime” (first-ever) stroke, and 86% of patients had CT, MRI, or necropsy. All patients were followed up for 1 year after the index stroke event. Only patients who survived to 1 year after stroke and were living outside of an institution (nursing home or hostel) on follow-up are considered here.
A caregiver was defined as “the person most closely involved in maintaining a person’s ability to live independently at home.” Essentially, this definition included persons providing care to patients with handicap as defined by categories 3, 4, or 5 of the Oxford Handicap Scale (OHS).15 In most cases a caregiver was a family member in frequent contact but not necessarily living with the patient; this did not include formal caregivers from community services. In cases in which multiple persons were involved in caregiving, the patient was asked to nominate the chief caregiver for interview. The study was approved by local ethics committees. After permission was obtained from patients, those caregivers who agreed to participate underwent a semistructured interview that included measures of their own physical and emotional health and the behavioral characteristics of patients, as outlined below.
Multidimensional outcome measures were used to assess the physical and cognitive function, psychiatric morbidity, and symptomatic behavior of patients and the effects of caregiving on the lives of caregivers at uniform intervals during the 1-year follow-up period (Table 1⇓). The first author assessed all patients as soon as possible after the onset of stroke using a standardized interview and physical examination. Information obtained from each patient at baseline included data on associated illnesses, risk factors for cardiovascular diseases, and the patterns of disability within the immediate premorbid period. He then assessed the outcome of survivors at 4 and 12 months after their index event. The self-care activities of daily living (ADL) level was measured according to the modified version of the Barthel Index (BI),16 and a patient’s actual level of function was recorded rather than his or her potential ability. Higher levels of physical function or “non-ADL” activities were assessed by means of the Frenchay Activities Index (FAI),17 which is based on the frequency with which a patient performs 15 separate activities (for example, shopping, reading books, and cooking) and gives a score between 15 (no activities) and 60 (full activities). If the patient was unable to communicate or otherwise not assessable, information was obtained from the patient’s closest relative or another reliable proxy. Significant cognitive impairment was defined as a score of 23 or less on the Mini-Mental State Examination (MMSE).18 Data were obtained on the use of community services by survivors at each follow-up review.
At the 4-month review, patients were also interviewed, usually within 1 week of the first author, by two coinvestigators (both psychiatrists) without knowledge of the neurological assessments using the Psychiatric Assessment Schedule (PAS)19 supplemented by additional questions. The PAS allowed the documentation and measurement of a wide range of psychiatric symptomatology to derive clinical diagnoses for depression (either major depression or dysthymia) and anxiety disorders (generalized anxiety, panic attacks, agoraphobia, and simple phobia). Dementia among 1-year survivors was defined according to the Diagnostic and Statistical Manual of Mental Disorders (Revised), edition 3 (DSM-III-R) criteria20 and with the use of the MMSE and caregivers’ reports.
The third stage of outcome assessments involved a specially trained nurse (J.L.) visiting caregivers in their own home after the first author’s 12-month review of patients. This assessment covered the areas of impact and burden associated with caregiving that have been defined elsewhere.13 Impact, the relatively objective changes in caregivers’ lives, was assessed with the relevant section of the Social Behaviour Assessment Schedule (SBAS).21 The SBAS was originally developed to measure the “symptomatic and social role performance of patients with psychiatric illness” based on information obtained from an interview with a key informant, usually a spouse or parent. It includes a rating scale for 22 behaviors (for example, misery, forgetfulness, worry, and odd behavior) that are commonly seen among psychiatric patients. For this study, caregivers were asked whether any of these behaviors had occurred among patients after the stroke. The SBAS also includes an evaluation of the adverse effects of the illness on the informant (that is, the caregiver) within nine areas (physical health and emotional ill-health, social life, leisure time, disposable income, work or performance, time off work, disruption of life, and disruption of household relationships). All items are rated on 3-point scales—“never,” “occasionally,” or “frequently”—whether or not the caregiver attributes the effect to the patient or to another source. Scores may be summed to produce an “impact score” with a maximum possible score of 18. An assessment of “patient-relatedness” (as perceived by the caregiver) was made on a separate 3-point scale for each effect rated as present.
The SBAS also includes an assessment of the caregiver’s subjective distress (burden) by rating his/her reported distress caused by disturbances in the patient’s behavior, by the patient’s limited social performance, or by objective burden. Studies show moderate-to-high reliability and validity of the SBAS for measuring the impact and burden on caregivers’ lives.22 However, we chose to measure burden or emotional distress among caregivers using the Hospital Anxiety and Depression Scale (HADS)23 and the 28-item General Health Questionnaire (GHQ-28).24 The HADS is a brief scale that was originally designed as a self-administered questionnaire for use in the detection of mood disorders among patients admitted to the hospital. It appears useful as a screening instrument because it includes two subscales, one for depression and the other for anxiety, each comprising seven items relating to symptoms. Each item is scored (0 to 3) according to severity, with a maximum possible score of 21 on each subscale. The original validation study suggested that a score of 8 or more could be used to define a case of emotional illness on either subscale. The GHQ-28 is a widely used self-administered screening questionnaire for the detection of minor diagnosable psychiatric disorders covering symptoms of anxiety, depression, somatic disturbance, and social dysfunction. Scores of 5 or more have been shown to indicate probable cases of psychiatric disorder.
Analyses were done with the use of the Student’s two-tailed t test and nonparametric statistical tests (Mann-Whitney test for comparisons of ranked scores between groups and χ2 test for comparisons of categorical variables) on spss for Windows software.25 Data are reported with 95% confidence intervals (CI), and probability values are two-tailed.
Of the 492 patients registered in the PCSS, 305 (62%; 95% CI, 58% to 66%) survived to 1 year after the onset of stroke. The ages of these surviving patients ranged from 21 to 96 years (mean, 71 years). Fifty-five percent were male, and 74% of these patients had experienced their first-ever stroke. Cerebral infarction accounted for 253 events (83%; 95% CI, 79% to 87%); the remainder included primary intracerebral hemorrhage (in 9.8%), subarachnoid hemorrhage (3.6%), and undetermined stroke (3.6%). Those 1-year stroke survivors who required institutional care on follow-up were significantly older (≥75 years), had lived alone or were “unpartnered” at the time of the stroke, and were more physically disabled (ADL) and had greater cognitive impairment (MMSE) after stroke compared with those patients who were living in private residences at 1 year (all P<.001). Two hundred forty-one (79%; 95% CI, 74% to 84%) of the 1-year survivors were living outside of a nursing home or hostel, and 103 (43%; 95% CI, 37% to 49%) of these patients were in OHS categories 3, 4, or 5.
Of the 103 handicapped stroke survivors, 3 lived alone and received help only from community services, and 5 would not allow their relatives to be interviewed. The caregivers of 11 patients either declined to participate or were unavailable. Thus, 84 handicapped stroke patients and their respective caregivers were assessed at 1 year after stroke.
Characteristics of Handicapped Stroke Patients
The 84 patients in the study group (62% men) were aged 36 to 89 years (mean, 70 years). Twenty-three percent were living alone, and the majority (78%) were “mostly independent in their daily activities” (ie, OHS grade 3) at 1 year after stroke. Table 2⇓ shows the proportional frequencies of patients who had failed to undertake various non-ADL activities after stroke at the level that they had performed them in the period immediately before the stroke. All work-related activities were affected by stroke, but light housework was less affected than others such as shopping, heavy housework, and cooking. Only two of nine patients in full-time employment at the time of the stroke were able to return fully to work. Although many patients required assistance with ADL, it was interesting to note that more than half of the patients continued to engage in regular social visits to friends or family members after the stroke. This may reflect the input of caregivers and community services in supporting these patients at home. All other leisure activities were significantly affected by the stroke.
Table 3⇓ shows that dressing (52%), bathing (48%), and feeding (36%) were the areas of ADL in which assistance was most often given to patients. Although 93% of patients were independent in mobility about the house, only 37% walked for longer than 15 minutes outside the home at least weekly during the few months before the follow-up interview at 1 year. Twenty-nine patients (34%) had difficulty or were unable to negotiate stairs or steps, which necessitated the installation of rails or ramps for 21 patients. Twenty-seven patients (32%) were dysphasic, but in only 5 patients was it the principal cause of handicap. Only 6 patients were either totally immobile or required a wheelchair for mobility.
Thirty-eight patients (45%; 95% CI, 35% to 55%) had significant cognitive impairment according to the MMSE 23/24 cutoff point. Five of these patients were either dysphasic (2 patients) or had a low education (3 patients), whereas another 5 patients with MMSE scores of 24 or more were assessed as suffering from mild dementia according to the DSM-III-R criteria. Thus, adjusting for education and other limitations of the MMSE, we diagnosed dementia in 38 stroke survivors (45%) at 1 year. According to information obtained at baseline, only 9 of these patients had had a dementia syndrome before the index stroke, and these included 4 patients with previous strokes.
Results from the PAS at the 4-month review were available for 74 patients. Forty-two (57%; 95% CI, 46% to 68%) of these patients were found to have a psychiatric illness that included dysthymia (in 41%), major depression (34%), agoraphobia (32%), simple phobias (20%), generalized anxiety disorder (7%), and one case of panic disorder. Table 4⇓ lists the abnormal behaviors of stroke patients during 1 year that were covered by the SBAS and reported by caregivers. Many of these symptoms may have been related to poor health, cognitive impairment, and disability rather than to depressive illness. However, caregivers’ reports of approximately half of the patients being frankly miserable, withdrawn, anxious, fearful, and irritable after the stroke support the frequency of psychiatric illness among the patients.
Characteristics of Caregivers: Impact
Table 5⇓ shows the characteristics of caregivers. They included 49 spouses; 27 children (20 coresidents); and 7 other relatives including siblings, nieces, nephews, and grandchildren (4 coresidents); and one neighbor. Their ages ranged from 30 to 80 years (mean, 58 years), and 82% were female. The Figure⇓ shows the relative impact of caregiving on different aspects of caregivers’ lives that were covered by the SBAS when administered at the 1-year review. Emotional ill-health (in 79%), disruption of social activities (79%), and leisure time (55%) were the areas most affected by caregiving for stroke patients. Thirty-five percent of caregivers reported adverse effects on family relationships for a variety of reasons including tension, misunderstandings, or feelings of neglect among family members because of the physical and emotional demands of the patient. At other times caregivers’ frustration and anger with respect to patients were displaced onto others. Although financial and work-related problems were most relevant among families in which the main wage-earner had experienced the stroke, these problems were also seen among some children who supported disabled parents. Since 20% of caregivers were aged 70 years or older, it was not surprising that any physical illness was usually related to conditions that were already present at the time of the patient’s stroke. However, the physical aspects of caregiving were associated with symptoms, particularly arthritis, or interfered with management of other conditions, such as hypertension or cardiac disease.
Characteristics of Caregivers: Burden
Although many caregivers received additional help from other family members (in 58%) and community services (42%), 74 caregivers (88%) still reported adverse effects on their lives, with 46 caregivers (55%) above the cutoff score for emotional illness on either of the two screening instruments, HADS and GHQ-28. A variety of adverse emotional reactions that related to patients were specifically expressed by caregivers. These included anxiety (58%), depression (50%), fear (35%), frustration (32%), resentment (29%), impatience (25%), and guilt (10%). More than half of caregivers said that they felt anxious about leaving the patient unattended for all or part of the day because of fear that the patient might fall or experience another stroke. This often meant these caregivers undertook work-related activities such as shopping and paying bills when patients were cared for elsewhere (for example, day hospitals or day centers). In these cases, fatigue and distress were further compounded by the need to have dependent patients ready for transport, often at earlier times in the morning than on other days, and then to be available for their return home in the mid-afternoon. While most caregivers had to reduce their leisure time at home and were unable to visit friends and family, this was not compensated by an increase in social contact at home.
Table 6⇓ compares the levels of abnormal behavior (measured by SBAS), physical function (BI), cognitive impairment (MMSE), and psychiatric illness (PAS) among patients of caregivers themselves defined as “cases” or “non-cases” on the basis of scores on either the HADS or the GHQ-28. The trend was for abnormal behavior and dementia to be related to emotional illness among caregivers, but there was no significant difference in the frequency of psychiatric morbidity in patients between the normal and emotionally distressed subgroups of caregivers. However, when the 9 patients with a history of dementia before the stroke were excluded from the analysis, a significant relationship emerged between demented stroke patients and emotional distress among caregivers (χ2=5.2, P=.02). Among caregivers who were spouses, 7 of 8 husbands showed emotional distress compared with only 22 of 41 wives, but this was not statistically significant (Fisher’s exact test, P=.12). The trend was for assistance with core items of ADL to be related to emotional distress in caregivers: bathing (P=.09), incontinence (P=.05), feeding (P=.09), and grooming (P=.08).
Only 6 patients received community services before the stroke. A total of 44 patients (52%) received community services at the 4-month review after the onset of stroke, namely attendance at a day hospital (45%), home help (43%), district nursing (41%), and meals on wheels (14%). Thirty-seven (84%) of these 44 patients were still in receipt of any of these services at 1 year. At the latter review, 7 patients had had respite admissions to either a geriatric hospital or nursing home in the last 6 months, and 1 patient had a regular community day sitter. The provision of community services was related to dependence in ADL (P<.0001) and MMSE score (P<.05) but not to the presence of depression, anxiety, dementia, patients’ SBAS score, or emotional distress in caregivers as defined by the GHQ-28 or HADS.
The purpose of this study was to describe the health and social consequences of caregiving for long-term survivors of stroke with residual handicap. It is clear from the results that the burden of stroke cannot be assessed simply in terms of the effect on patients; it is also associated with a high risk of disrupting the integrity of families and the quality of life of caregivers. The data suggest that dementia and behavioral abnormalities, rather than physical disability, are the characteristics of stroke patients that are most predictive of emotional illness among caregivers. These findings support the implementation and evaluation of broader strategies to improve the well-being of informal caregivers.
Although stroke carries a high case fatality, the present findings confirm data from other community-based studies5 7 9 11 showing that the majority of survivors continue to live at home with relatively mild degrees of physical disability. In this population-based sample of stroke survivors at 1 year, roughly one fifth of patients living at home required help with dressing and bathing, both of which require adequate balance and dexterity, whereas less than 7% were incontinent or dependent in core degrees of mobility. In terms of social disadvantage (handicap), however, the effects of stroke were profound, with almost one half of these patients relying on the support of family members and other persons because of various impairments and disabilities that affected higher levels of physical functioning, cognition, and emotional reactions.
The willingness and ability of caregivers to support victims of stroke appear to have a significant influence on the patient’s emotional reactions, on the success of rehabilitation, and on whether patients can remain outside institutional care.26 However, as many studies have shown, family members—commonly the spouse or closest female relative—are affected by the patient’s illness from the outset. They must overcome the initial shock of stroke, reassess their values, and readjust their own lives and those of their families so that they can provide what is often an extraordinary level of care. In undertaking this caregiving role, however, caregivers often experience a significant burden, which, if excessive, can lead to breakdown in the support provided. Thus, caregivers merit attention in their own right as part of various rehabilitation interventions designed to reduce the incidence of handicap after stroke.
The extent of disruption to the lives of caregivers studied in Perth was profound. Almost 80% reported emotional ill-health and disruption of their social life, and half experienced a reduction in their leisure time as a direct consequence of caregiving. While there were a variety of reasons for these effects, fear and anxiety about leaving the patient unattended at home because they may become more disabled or die were prominent. Overprotection and overcaring were other ways in which caregivers reacted, perhaps in an attempt to alleviate feelings of guilt that their actions or negligence caused or contributed to the disability.
Although social activities outside the home appeared to be less affected by stroke than other social activities, this rather positive finding may have been confounded by the input provided by community services and the limitations of the measures used to assess the quality of activities, contacts, and events. It was common for caregivers to report that patients were apathetic, indifferent, lacked motivation, and were irritable after the stroke. This behavior meant that they either refused to go out alone or limited the frequency and duration of outings with others. Consequently, the effort that caregivers experienced in preparing patients for outings often far outweighed the anticipated enjoyment of the event.
Since most of these caregivers were middle-aged or elderly, they often had preexisting illnesses and functional limitations that restricted their ability to manage the physical aspects of caregiving, particularly bathing. While community services were able to provide care to those patients who were most severely disabled, there was no association between the provision of services and levels of emotional distress among caregivers. This highlights one of the major deficiencies of the current structure and distribution of community services. In contrast to the physical aspects of care, problems related to behavioral disturbance, emotional illness, and marital disharmony are less readily recognized and addressed by professionals and the lay public. Although these aspects of stroke outcome were the source of much distress among caregivers, few people were able to find support and reassurance from other family members or their general practitioners. Because caregivers are often expected by patients and others to provide care, many of them found it difficult to express their concerns and anxieties and were reluctant to seek help through fear that such a request would represent a sign of failure or inadequacy. Consequently, many caregivers expressed feelings of isolation and exhaustion.
To overcome some of the methodological shortcomings of previous studies and to provide further information of the consequences of stroke for caregivers, we studied a relatively unselected group of people and examined the role of a variety of factors that might be related to the impact and burden of caregiving. Nonetheless, there are limitations to the conclusions that can be drawn. First, the effects of caregiving were measured only at 1 year after the stroke. At this time, optimal recovery of physical function should have been achieved, and both patients and their caregivers should have adjusted to the acute event and resettlement after a period of rehabilitation. However, caregiving is a dynamic process, and we did not include an assessment of persons who took on the caregiving role soon after onset but had to abandon it at some point. Had they been included as part of a longitudinal study, the results might have been different. Furthermore, the direction of a potential causal pathway cannot be determined because antecedent information on family dynamics and other prestroke variables were not available. Because we did not compare caregivers with noncaregivers or with other caregivers involved with other disabling illnesses, we cannot conclude that stroke-related disability has a differential effect on caregivers. Finally, many of the measurement scales used have not been validated with stroke patients or their caregivers. Since impact and stress are subjective and intertwined measures, one cannot exclude the possibility that associations between variables were spurious or that some were overlooked because of various confounding effects of the interview.13
Caregivers form a heterogeneous group with respect to certain characteristics that may be systematically related to the potential for experiencing emotional distress. These characteristics include the relationship to the patient, frequency of contact with the patient, coping mechanisms, personality, living arrangements with respect to the patient, and the quality of interpersonal relationship with the patient.27 This study pointed to an association between emotional illness among caregivers and dementia among stroke patients. It is understandable that patients with poor memory, confusion, neglect, and apraxia should need extra attention at home. There was also a strong tendency for behavioral disturbance, but not symptoms that satisfied DSM-III-R criteria for depression and anxiety, to be associated with emotional distress among caregivers. These findings imply that many of the emotional reactions and cognitive impairments experienced by patients are not severe and numerous enough to place them above a certain diagnostic threshold for psychiatric illness. Other studies have shown that stress among caregivers is often associated with emotional distress, mood disturbance, and behavioral abnormalities among patients with dementia.10 28 29
What can be done to help caregivers cope with the crisis of stroke and at the same time enable them to mitigate the impact on themselves of the disability and behavioral disturbance experienced by patients? Although a variety of community services were available in this population, such as district nursing, domiciliary care, day centers, day hospitals, stroke groups, and respite care, the findings from this study suggest that the needs of many caregivers were unmet. A comprehensive stroke service should provide continuous and coordinated care across various sites through an integrated multidisciplinary team.30 Because the transition from hospital to home is often a major dislocation to the patient and caregiver, the delivery of acute or subacute care for stroke that shifts its focus from short-term physical recovery to social and psychological functioning at home is imperative if we wish to achieve optimal stroke outcomes. Caregivers need information about potential stroke-related problems, and they need to know how to obtain the necessary help and support when they occur. Assessment of the caregiver’s own medical, social, and emotional needs could be undertaken by a single accessible “key person” as part of routine follow-up activities of a stroke service. A contentious aspect of the support for caregivers includes the provision of day hospitals and intermittent respite care. These forms of support are widely used to maintain the function of frail and disabled people and to reduce stress on caregivers, but clear evidence of substantial benefit to caregivers is lacking.31 Enabling patients with stroke-related disability to live at home for as long as possible depends on adequate support for caregivers. Innovative rehabilitation programs that have a community-based orientation and include the caregiver’s experience of the patient’s illness need prompt implementation and thorough evaluation.
This study was supported by grants from the National Health and Medical Research Council, the Australian Brain Foundation, and the Medical Research Foundation of Royal Perth Hospital. This study would not have been possible without the cooperation of numerous doctors and, of course, the patients and their families. We are indebted to the other members of the study team: Professor P. Burvill and Dr G. Johnson (Department of Psychiatry and Behavioral Science, University of Western Australia); Associate Professor T. Chakera (Department of Radiology, Royal Perth Hospital); and Dr K. Jamrozik, L. Nussbaum, R. Tuohy, J. Forrester, F. Gout, J. Lim, J. Linto, and S. Forbes (Department of Public Health, University of Western Australia).
Reprint requests to Dr Craig S. Anderson, Department of Medicine (Rehabilitation and Ageing Studies Unit), Flinders University of South Australia, Repatriation General Hospital, Daws Rd, Daw Park, South Australia, 5041.
- Received September 9, 1994.
- Revision received February 7, 1995.
- Accepted February 7, 1995.
- Copyright © 1995 by American Heart Association
Brocklehurst JC, Morris P, Andrews SK, Richards B, Laycock P. Social effects of stroke. Soc Sci Med. 1981;15:35-39.
Ebrahim S, Souri F. Caring for stroke patients at home. Int Rehabil Med. 1987;21:171-173.
Sillman RA, Wagner EH, Fletcher RH. The social and functional consequences of stroke for elderly patients. Stroke. 1987;18:200-203.
Greveson GC, Gray CS, French JM, James OFW. Long-term outcome for patients and carers following hospital admission for stroke. Age Ageing. 1991;20:337-344.
Wade DT, Legh-Smith J, Langton Hewer R. Effects of living with and looking after survivors of a stroke. Br Med J. 1986;293:418-420.
Dura JR, Kiecolt-Glaser JK. Sample bias in caregiving research. J Gerontol. 1990;45:P200-P204.
Holbrook M, Skilbeck CE. An activities index for use with stroke patients. Age Ageing. 1983;12:166-170.
Folstein MF, Folstein SE, McHugh P. Mini-mental state: a practical method for grading the cognitive state of patients for the clinician. J Psych Res. 1975;12:189-198.
Dean C, Surtees PG, Sachidharan SP. Comparison of research diagnostic systems in an Edinburgh community sample. Br J Psychiatry. 1983;142:247-256.
American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (Revised). 3rd ed. Washington, DC: American Psychiatric Association; 1987.
Platt S, Weyman A, Hirsch S. Social Behaviour Assessment Schedule (SBAS). 3rd ed. Windsor, England: NFER-Nelson; 1983.
SPSS for Windows V5.0.1. Chicago, Ill: SPSS Inc; 1992.
Glass TA, Matchar DB, Belyea M, Feussner JR. Impact of social support on outcome from stroke. Stroke. 1993;24:64-70.
O’Connor DW, Pollitt PA, Roth M, Brook CPB, Reiss BB. Problems reported by relatives in a community study of dementia. Br J Psychiatry. 1990;156:835-841.
Homer AC, Gilleard CJ. The effects of inpatient respite care on elderly patients and their carers. Age Ageing. 1994;23:274-276.