Hong Kong J Psychiatry 2004;14(4):24-32

ORIGINAL ARTICLE

CM Yeung, L Chiu

Mr CM Yeung, Department of Psychiatry, Kowloon Hospital, Hong Kong, China.
Dr L Chiu, Department of Psychiatry, Kowloon Hospital, Hong Kong, China.

Address for correspondence: Dr Chiu Ling, Department of Psychiatry, 147A Argyle Street, Kowloon, Hong Kong, China.
E-mail: chiul@ha.org.hk

Submitted: 25 June 2004; Accepted: 11 December 2004

pdf Full Paper in PDF

Introduction

It has been established that carers of elderly people with dementia have higher stress levels and depression and are more likely to have lower levels of subjective well-being, physical health, and self-esteem than non-caregivers.^1,2 It has also been observed that neither cognitive nor functional status predict the burden of care. Behavioural disturbance, in particular, and informal support are significant determin- ing factors of the burden for carers.^3,4

All aspects of behavioural and psychological symptoms of dementia (BPSD) can be associated with caregiver burden. However, paranoia, aggression, and sleep-wake cycle dis- turbance appear to be particularly important drivers of bur- den of care.⁵

A carer support group was organised by the Kowloon Hospital Psychogeriatric Team to provide education, support, advice, and recreational activities to carers, with the aim of reducing the carers' burden.^6,7 The carers' burden was assessed according to the method of Hepburn et al⁷ before and after carers participated in the carer support group to assess the effectiveness of the group for alleviating carers' distress. The patients' cognitive function, measured by the Mini-Mental State Examination (MMSE) and selected items of BPSD were analysed according to the methods of Arai et al⁸ and Burns et al⁹ to establish the correlation with the carers' burden. Carers' demographic data were also analysed to establish any significant relationship with their burden.

Patients and Methods

Carers of elderly patients with dementia who attended the outpatient department or day hospital, or who were inpa- tients at Kowloon Hospital Psychiatric Unit were invited to join the psychogeriatric carer support group. Medical and nursing staff, occupational therapists, and medical social workers recruited the carers and organised the activities.

Carers who were interested in the support group were enrolled as members and were regarded as the study group in a questionnaire survey. Carers who were unable to attend were invited to complete the questionnaires and were con- sidered to be the control group for data comparison.

Twenty carers were approached and 12 joined the carer support group. The remaining 8 carers were interested in joining but were unable to come, either because they were working when the training sessions were organised or were unable to find somebody to care for their relative at home.

The revised Chinese version of the Zarit Burden Scale was administered to the carers before and after joining the support group. The total scores and subscores were compared. The Chinese version of the Zarit Burden Scale (Appendix 1) was administered to the 20 carers before the start of the study.

The Zarit Burden Scale was evaluated for test-retest re- liability and internal consistency. The Chinese version of the Zarit Burden Scale was first used at the Prince of Wales Hospital, Hong Kong. Questions 14 to 18 of the Chinese version were further modified by members of the psychogeriatric team at Kowloon Hospital. When all the items in the scale were finalised, a test-retest reliability study was carried out. The scale was administered to a selection of carers of psychogeriatric patients and was repeated within 2 to 4 weeks.

The carers' characteristics were recorded on a form de- vised to measure such factors as the duration of contact with the patient, presence of other carers in the family, and health status. The characteristics of the carers were analysed and compared between the 2 groups (Table 1).

The patients with dementia were administered the MMSE, and the presence of BPSD was noted using a 2-point scale before and after the carers joined the support group (Table 2). Paranoia, aggression, and sleep-wake cycle disturbance were measured and regarded as an indicator of BPSD. Presence of BPSD scored 1 and absence of BPSD scored 0.

The patients' cognitive status was screened using the MMSE to determine the presence of mild, moderate, or severe dementia. The MMSE was administered before and after the carers joined the support group.

Support Group Activities

A series of 12 monthly sessions were designed to enhance recognition and management of symptoms and associated problems. Comments and feedback from the carers were welcomed and the sessions were modified accordingly. The programme took place between July 2002 and August 2003, with a short break from March 2003 to June 2003 because of the outbreak of severe acute respiratory syndrome (SARS).

Three lectures on the recognition of presentation of de- mentia and associated BPSD were held. Three lectures were given on improvement of self-care, prevention of falls by elderly people, and enhancing communication with patients with dementia. Three sessions on the topics of stress man- agement and improving the general health of patients with dementia were arranged. The remaining sessions were ac- tivity sessions conducted with the carer and their relative to demonstrate physical exercises and activities that could be performed together.

Each session lasted for approximately 2 hours. The pa- tients were invited to participate if they were able to join the activities. The carers were encouraged to discuss the topics with the staff and among themselves during and after each session.

Statistical Analysis

The Wilcoxon signed rank test was used to evaluate the pretest and post-test results. The experimental hypothesis was that the self-help group activity programme reduced stress levels among carers of patients with dementia. The null hypothesis was that the self-help group activity programme had no effect on stress levels. A 1-tailed signifi- cance test was performed; the critical value for n(11) was 14 (p = 0.05).

Results

Validation of the Modified Chinese Version of the Zarit Burden Scale

Thirty patients participated in this part of the study, 26 of whom completed the retest. Five of these patients had changes in their medical status during the retest period and were excluded. The data for these patients were not included included in the analysis; the characteristics of the study group are summarised in Table 3. The first set of data was compared with the retest data and analysed by intra-class correlation (ICC). The ICC coefficient was 0.84 and the internal consistency by a coefficient was 0.91 (p = 0.0000).

Carers' Burden

Table 4 shows the characteristics of the carers and patients in the study and control groups. There were 8 carers in the control group (Table 5). The score in the pretest ranged from 17 to 47 (mean, 33.63 ± 11.69). The post-test was done 2 weeks later and resulted in a range of 18 to 54 (mean, 33.00 ± 14.23). The mean decrease in score was 0.63. Two of the 8 carers had a substantial decrease in stress levels of 43.59% and 55.81%. For these 2 carers, their relatives had been admitted to a care home prior to the post-test.

Among the 12 carers in the study group, the pretest score ranged from 21.00 to 78.00 (mean, 43.08 ± 16.17). The post- test score ranged from 11.00 to 64.00 (mean, 32.50 ± 16.10).

The study group had higher stress levels than the control group in the pretest. The study group had a greater decrease in stress level score than the control group (10.58 versus 0.63). Two patients were admitted to hospital and the carers' burden underwent a substantial decrease of 60 and 30 in the post-test.

The test statistic was 13, which was less than the critical value of 14, so the null hypothesis was rejected. Table 7 shows the results from the Wilcoxon signed rank test. It was concluded that the carers' support group activity programme reduced stress levels.

Data Analysis

The median split test was used to analyse the demographic and situational variables by dividing the carers into a high- burden group and a low-burden group (Table 8). The median cut-off point was 39.5. Selected variables such as carer's age, patient's age, and duration of care were assessed using unpaired t test (Table 9).

The average duration of care was 4.15 years (range, 1 to 11 years). Approximately 45% of the carers were daughters.

As shown in previous studies, 70% of carers were female.³ The majority of the carers did not have a maid to assist with the housework. Approximately 15% of carers had a maid, although the change in stress level was comparatively small, with scores ranging from 1 to 4.

Employment was postulated to play a significant part in the stress level as reflected in the data from the control group. Carers in full-time employment who had to take care of a patient had the highest scores in this group. The only carer in the study group who was working scored a contrastingly low score. However, as this was a relatively small sample, the effect of employment on a carer's burden may not be accurately reflected.

Approximately 50% of the carers in both the control and study groups were older than 60 years. Carers who were older than 60 years are likely to experience more stress as they may also have health problems. The Pearson correlation test gave a significant result of 0.543, which indicated a correlation between carers' age and stress level.

In both the control and study groups, spouses and chil- dren acted as carers, female and male relatives were involved in care, and elderly (>60 years) and younger (40 to 60 years) carers were available. The employment status, living status of the patient, and frequency of visits to patients were com- parable (Table 8).

Patients' scores were compared before and after the car- ers joined the activity programme. Paranoia, aggression, and sleep-wake cycle disturbance were measured and were re- garded as an indicator of BPSD. The results revealed no significant changes after the carer's activity programme.

The cognitive status of the patients was screened using the MMSE before and after the carer's activity programme and graded according to mild, moderate, and severe dementia. Among the study group, cognitive impairment was noted to be 11.1% for mild, 22.2% for moderate, and 66.7% for severe dementia. The control group showed a relatively higher score for moderate dementia than for severe dementia, with 14.3% for mild, 57.1% for moderate, and 28.6% for severe dementia. After the carers completed the activity programme, the change in cognitive impairment of the pa- tients in the corresponding groups did not significantly differ.

Discussion

The study supported the hypothesis that the programme pro- vided was beneficial for relieving the burden of carers. It is difficult to delineate the relative significance of accessibil- ity of professional staff; knowledge and skills acquired when caring for an elderly relative with dementia; and the mutual support fostered among the carers that contributed to the relief in carers' burden. Carers are often reluctant to join a group because of the need to take care of the relative at home. To enhance the participation of the carers who are unable to find help in caring for their relative at home, staff should be assigned to attend to their relative in rooms close to the programme venue, while the carer joins the pro- gramme. Alternatively, the training sessions could be held in conjunction with an elderly persons activity centre lo- cated in a residential area, and volunteers from the centre could help to care for the relatives.

As the training sessions offered were relatively small- scale, another programme has been planned. The intention is to recruit more carers and run the programme on a weekly basis for 3 months to increase their participation. Other ini- tiatives have been planned, including publishing educational pamphlets for each session and a weekly quiz with a small token as a reward. Another study will be conducted and it is anticipated that more positive findings can be established.

This study has several limitations that must be consid- ered when interpreting the findings. Firstly, the study group was restricted to carers of patients who had been receiving the psychogeriatric team services. Therefore, the results may not be generalised due to the localised catchment popula- tion of Kowloon Hospital only. Secondly, the study group was a convenience sample from a population selected on the basis of accessibility to the researcher rather than a ran- dom sample.¹⁰ In addition, the study started shortly before the outbreak of SARS, and the programme had to be post- poned during this period.

It can be concluded that the carer support group was effective for alleviating carers' distress. In addition, patients' cognitive function was not found to be directly related to carers' burden. The carer's age positively correlated with the burden. However, the effect of BPSD on carers⁴ cannot be properly demonstrated as the sample was too small.

Acknowledgements

With thanks to Ms I Chan, Occupational Therapist I, and Mr KC Lam, Registered Psychiatric Nurse, for their contribution in collecting the data.

References

1. Sarensen S, Pinquart M. Differences between caregivers and non- caregivers in psychological and physical health: a meta-analysis. Psychol Aging 2003;18:250-267.
2. Zarit SH, Orr NK, Zarit JM. The hidden victims of Alzheimer's disease: families under stress. New York: New York University Press; 1985:69-83.
3. Coen RF, Swanwick GR, O'Boyle CA. Coakley D. Behaviour distur- bance and other predictors of carer burden in Alzheimer's disease. Int J Geriatr Psychiatry 1997;12:331-333.
4. O'Donnell, BF, Drachman DA, Barned HJ. Incontinence and trouble- some behaviours predict institutionalisation in dementia. J Geriat Psy- chiatry Neurol 1992;5:45-52.
5. Lawlor B. Managing behavioural and psychological symptoms in dementia. Br J Psychiatry 2002;181:463-465.
6. Marriott AD, Tarrier N. Effectiveness of cognitive-behavioural family intervention in reducing the burden of care in carers of patients with Alzheimer's disease. Br J Psychiatry 2000;176:557-562.
7. Hepburn KW, Tornatore J, Centre B, Ostwald S. Dementia family caregiver training: affecting beliefs about caregiving and caregiver outcomes. J Am Geriatr Soc 2001;49:450-457.
8. Arai Y, Kudo K, Hosokawa T, Washio M, Miura H. Reliability and validity of the Japanese version of Zarit caregiver burden interview. Psychiatry Clin Neurosci 1997;51;281-287.
9. Burns A, Lawlor B, Craig S. Rating scales in old age psychiatry. Br J Psychiatry 2002;180:161-167.
10. Clifford E, Harkin L. Inferential statistics in nursing and healthcare. Open learning foundation. Edinburgh: Churchill Livingston; 1997.