Dr Toshiyuki Ono MD, PhD, Department of Psychiatry, Tsuruga Onsen Hospital, 41-1-5 Yoshiko, Tsuruga, Fukui 914-0024, Japan. Email: email@example.com
Background: In a previous study, we found gender differences among care recipients and suggested that elderly women living alone have difficulty receiving care from their families. We investigated the gender differences among predictors which influenced outcomes after discharge from a ward for treatment of demented elderly with severe psychotic symptoms.
Methods: We enrolled the data of 325 patients with dementia who were hospitalized between 1 April 2000 and 31 March 2007, and discharged by 31 March 2008. Two hundred and ninety-four patients were divided into men and women. We checked the gender differences of their characteristics. After that, they were subdivided into three groups to analyze the effect of each patient's characteristics and care situation on their outcomes: (i) given home care (home); (ii) became institutionalized (institution); and (iii) transferred to another hospital (hospital).
Results: In the hospital groups, the incidence of complications was high for each sex. The differences between the institution group and home group were shown by N-ADL score in men. In contrast, the predictors in women were the HDS-R score, the number of cohabitants and the caregiver.
Conclusions: It was observed that there were gender differences among outcome predictors. The outcomes of demented patients were predicted by both complications and condition on admission. Most women did not return to their homes because of the situation in which they received care. It is necessary to establish a clear system for providing care for the demented elderly, especially for women.
Many reports have shown that depending on a patient's characteristics and/or those of their caregivers, demented patients are placed in nursing homes from their homes. However, there are few reports that evaluate the predictors of outcome when they are placed in a ward for demented elderly.1,2 The main outcomes of patients in wards for demented elderly are homes, institutions and other hospitals. We investigated the predictors that influenced these outcomes in each sex. In a previous report by the present authors,3 the distress of family members caring for demented elderly and the effect of gender differences among care recipients was investigated. As a result, the care giving situation for elderly men differed from that of elderly women. Most men had been cared for by their spouses and were hospitalized because of progressive dementia. In contrast, women could be divided into two groups. The first group was living alone and most of them were cared for by their children. They were easily hospitalized and could not leave the ward for demented elderly and return to live with their families. If a woman started to live alone, then it was difficult for her to receive care from her family. Another group was living with their families and most of them were cared for by their daughters-in-law. Although they were hospitalized, they easily left the ward for demented elderly and lived together with their families again. It was assumed that hospitalization in a ward for demented elderly was a result of their own condition, but mainly caused by the conditions in which they received care.
It is important for geriatrics and caregivers to understand the psychosocial influences on placement decision. We investigated whether the situation of care giving showed differences between men and women.
Subjects and procedure
We enrolled the data of 325 demented patients who were hospitalized for psychological and behavioral problems between 1 April 2000 and 31 March 2007, and discharged by 31 March 2008 in our ward for demented elderly. This ward is for the treatment of the demented elderly with severe psychotic symptoms. They fulfilled the criteria according to the text version of the fourth edition of Diagnostic and Statistical Manual of Mental Disorders (DSM-IVTR),4 including the criteria for dementia of the Alzheimer's type (DAT), vascular dementia (VaD) and other types of dementia (OTD). We checked each patient's characteristics including age, sex, diagnosis, medical comorbidity (comorbidity), and complications during hospitalization (complications), primary caregiver (caregiver), two main reasons of hospitalization, number of cohabitants and length of hospital stay. The cognitive status was assessed by the revised version of Hasegawa's Dementia Scale (HDS-R).5 Also, the behavioral and psychological symptoms of dementia (BPSD) were evaluated by the Assessment Scale for Symptoms of Dementia (ASSD).6 Functional status was assessed by Nishimura's activity of daily living scale (N-ADL).7 These assessments were checked every 3 months. We principally used the initial scores of HDS-R, ASSD and N-ADL on admission and the final scores when they were discharged from our ward for demented elderly. If the HDS-R, ASSD and N-ADL scores were not available at the time of discharge, we substituted the scores on admission.
If a patient's severe psychiatric symptoms or behavioral disturbances improved after medical therapy, they were advised to leave the ward for demented elderly. Discharge planning was managed by the social workers at our hospital. With respect to the type of outcome after discharge, the patients were divided into five groups: (i) given home care (home); (ii) became institutionalized (institution); (iii) transferred to another hospital (hospital); (iv) changed ward; and (v) deceased.
Next, we investigated the gender differences in the characteristics, and checked the relationship between caregivers and number of cohabitants. Some patients could go home for care giving. Other patients were institutionalized or transferred to another hospital for their physical condition. The main destinations of the inpatients in the ward for demented elderly were homes, institutions and other hospitals. We investigated the predictors which influenced these three destinations in each sex. As a result, the 294 patients were subdivided into three groups (home, institution and hospital) for a comparison of outcomes; we found a high incidence of complications in the hospital groups.
After that, we investigated the gender differences in outcomes between the institution group and home group. We evaluated the effects of the characteristics on their outcomes in two stages. The model using the initial scores of HDS-R, ASSD and N-ADL on admission was named Model 1. In Model 1, we checked the effect of each patient's characteristics, without complications, on their outcomes. Model 2 used the final scores of HDS-R, ASSD and N-ADL with complications.
Informed consent was obtained from the patients and/or their caregivers. The experimental procedure was carried out in line with the Declaration of Helsinki.
The two-sample t-test, Mann–Whitney U-test for continuous variables and χ2-test for categorical data were used to evaluate the gender differences in the characteristics (age, diagnosis, the scores of HDS-R, ASSD and N-ADL, comorbidity, complications, caregiver, number of cohabitants, care distress, hospital length of stay and outcomes). The relationships between the caregivers and number of cohabitants were tested for significance using χ2-tests. One-way analysis of variance, Kruskal–Wallis test and χ2-test were used to test for significance regarding the relationships between the three types of outcomes (home, institution, hospital). Binomial logistic regression with a stepwise forward method was used to analyze the relationship between outcomes (institution group vs home group) and characteristics. In this analysis, we used characteristics including age, diagnosis, comorbidity, caregiver, number of cohabitants, care distress, and the initial HDS-R, ASSD and N-ADL scores as their cognitive, behavioral and functional status in Model 1. In Model 2, we used characteristics including age, diagnosis, comorbidity, complications, caregiver, number of cohabitants, care distress and final scores. The caregivers were divided into five categories (spouse, son, daughter, daughter-in-law and others) in the binomial logistic regression analysis. The others included relatives and caregiving staff. We had a small number of missing data. However, there were no differences of variables between the whole sample and the sample used in the binomial logistic regression analysis. Statistical analysis was carried out using spss version 17.0. We regarded P < 0.05 as a statistically significant level.
We found some differences between men and women in terms of diagnosis, comorbidity, caregiver, outcomes and hospital length of stay. However, these two groups did not differ in terms of age, HDS-R, ASSD and N-ADL scores, complications, and the number of cohabitants (Table 1). The main three reasons of hospitalization were violence (42.0%), wandering (33.0%) and care distress (32.1%) in men, and care distress (37.6%), wandering (25.8%) and hallucination-delusion (20.7%) in women. However, there was no difference in care distress in both sexes. DAT was more frequent in women (145/213) than in men (54/112). The incidence of comorbidity was significantly lower in women (66.2%) than in men (76.8%). Most men had been cared for by their spouses (66/112), but many women (56/213) lived alone and had been cared for by daughters-in-law (79/213). Most women were institutionalized after discharge, and the ratio of institutionalization of women (91/213) was higher than that of men (29/112). The length of hospital stay for women (mean 194 days, SD 228 days) was longer than that for men (mean 159 days, SD 222 days). The caregiving situation for elderly men differed from that for women (Fig. 1). There were differences between the caregiver and number of cohabitants in each sex (P= 0.000). Most men living with their wives had been cared for by them. In contrast, women could be divided into two groups. The patients of the first group had been living alone and most of them had been cared for by their children, relatives or care staff. The second group of women had been living with their families and most of them had been cared for by their daughters-in-law (Fig. 1). These results were similar to our previous report.3
Table 1. Clinical characteristics of 325 patients
Men (n = 112)
Women (n= 213)
ASSD, Assessment Scale for Symptoms of Dementia; DAT, dementia of Alzheimer' type; HDS-R, revised Hasegawa Dementia Scale; home, home caregiving; hospital, transferred to another hospital; institution, became institutionalized; N-ADL, Nishimura's activity of daily living scale; NS, not significant; OTD, other type of dementia; VaD, vascular dementia.
Initial score of HDS-R
Initial score of ASSD
Initial score of N-ADL
Final score of HDS-R
Final score of ASSD
Final score of N-ADL
Medical comorbidity (%)
Care distress (%)
Hospital length of stay (days)
Median hospital length of stay (days)
Comparing the three groups (home, institution and hospital), diagnosis, the initial score of N-ADL, final scores of HDS-R, ASSD and N-ADL, complications, caregiver and length of hospital stay were different in men. Subgroups in women differed significantly in terms of diagnosis, the final score of HDS-R, complications, caregiver, the number of cohabitants and length of hospital stay. In both sexes, a high incidence of complications was found in the hospital groups (men 97.2%, women 87.1%). Two-thirds of men (35/53) and one-third of women (27/80) with complications were transferred to another hospital (Table 2).
Table 2. Differences in the characteristics by outcome
Men n= 105
Women n= 189
P < 0.05. ASSD, Assessment Scale for Symptoms of Dementia; DAT, dementia of Alzheimer' type; HDS-R, revised Hasegawa Dementia Scale; home, home caregiving; hospital, transferred to another hospital; institution, became institutionalized; N-ADL, Nishimura's activity of daily living scale; OTD, other type of dementia; VaD, vascular dementia.
According to binomial logistic regression analysis, in men the differences in outcome between the institution group and home group were influenced by the scores of N-ADL (B, −0.099; P= 0.002; Exp (B), 0.906 (Model 1; Cox-Snell R2= 0.215) and B, −0.114; P= 0.000; Exp (B), 0.892 (Model 2; Cox-Snell R2= 0.288)). In contrast, Model 1 (Cox-Snell R2= 0.203) in women showed that the differences were linked to the initial score of HDS-R (B, −0.069; P= 0.016; Exp (B), 0.934), caregiver (others) (B, 3.002; P= 0.010; Exp (B), 20.119) and the number of cohabitants (B, −0.300; P= 0.016; Exp (B), 0.741). Model 2 (Cox-Snell R2= 0.228) in women also showed that the differences were influenced by the final score of HDS-R (B, −0.086; P= 0.002; Exp (B), 0.917), caregiver (others) (B, 3.120; P= 0.008; Exp (B), 22.637) and the number of cohabitants (B, −0.319; P= 0.012; Exp (B), 0.727) (Table 3). The stepwise backward methods showed the same results. The Cox-Snell R2 and percentage correct classification of Model 1 were lower than those of Model 2, but Model 1 and Model 2 showed similar results. The predictors of the differences in outcomes (institution group vs home group) were affected by the N-ADL scores in men and by the HDS-R scores, caregiver and number of cohabitants in women. The N-ADL scores showed a protective role regarding institutionalization in men. In women, the probability of institutionalization was higher when the patient had severe cognitive impairment and a small number of cohabitants, whereas it was lower when the caregiver was a close relative.
Table 3. Predictors of outcome between the institution and home groups
Percentage correct classification
Binomial logistic regression, stepwise forward method. HDS-R, revised Hasegawa Dementia Scale; N-ADL, Nishimura's activity of daily living scale; others, relatives and staff.
Initial score of N-ADL
Final score of N-ADL
Initial score of HDS-R
Final score of HDS-R
In Japan, the traditional family system has greatly changed and most elderly now live alone or with their spouses. In particular, most women tend to live alone. In these situations, it is difficult to care for the elderly if they need help to manage routine life. Also, the increasing number of demented elderly is a great problem, causing stress for caregivers. Most demented elderly are cared for by their relatives, especially spouses, children and daughters-in-law. Caregivers are stressed by the demented elderly who are often violent, wander and show anxiety. Some caregivers continue to give care beyond their limits. Gallicchio et al.8 reported that adequate assistance must be given to female caregivers to ensure that they are not strained beyond what is clinically healthy. Interventions should target caregivers of behaviorally disturbed patients, as well as caregivers who report poor physical health, to reduce the negative psychological impact of caregiving. In Japan, Takano and Arai9 reported that female caregivers are more likely to experience care burden than male caregivers, and showed the importance of mental health support for female caregivers. Hirakawa et al.10 carried out a study which gave a better understanding of the Japanese male caregiver burden in home care. Their study suggested that differences in caregiver burden might not necessarily exist between male and female caregivers.
A long-term care insurance system started in April 2000 in Japan. Demented patients, who are very violent and wander, are treated in wards for demented elderly. After treatment, some patients leave the ward and are cared for by their relatives at home. However, other patients cannot be cared for by their caregivers and are institutionalized or transferred to another hospital as a result of physical problems.
Previous reports11–15 regarding predictors of nursing home placement focused on the characteristics of both the patients and caregivers. Fisher and Lieberman16 showed that families scoring high on emotional closeness, high on negative family feelings and low on family efficiency institutionalized their sick elderly at a significantly higher rate than other families, especially at low levels of patient severity. These findings suggest that a careful examination of multigenerational families would increase the prediction of nursing home placement and the relevance of services offered to these families. Hebert et al.17 reported that screening caregivers for burden and depression, and designing interventions to reduce the consequences of behavioral disturbance on caregivers would be relevant avenues to explore to decrease institutionalization of people with dementia. Yaffe et al.18 suggested that patient and caregiver characteristics were both important determinants of long-term care placement for patients with dementia. Interventions directed at delaying placement, such as reducing caregiver burden or difficult patient behaviors, need to take into account the patient and caregiver as a unit. Gaugler et al.19 suggested that the caregiving indicators (i.e., burden and self-rated health), community-based service use, caregiver instrumental activities of daily living and care recipient activities of daily living were significant predictors of earlier placement. Some reports suggested the necessity of intervention.20,21
The participants of these reports were community dwelling. Few reports have investigated the outcomes of patients treated in wards for demented elderly. Mitsuyasu and Nohtomi1 suggested that understanding the condition of home care, the severity of a patient's psychiatric symptoms and the patient's ADL were useful for decision making concerning patients being discharged. They also showed that the predictors that influenced the outcome between home and institution groups were the care giving situation and the psychological symptoms on discharge. However, they did not consider gender differences. Rozzini L et al.2 suggested that living alone, the patient's functional impairment, severity of dementia and caregiver's burden were independent predictors of institutionalization. The interaction between a patient's and a caregiver's characteristics had an important effect on the rate of nursing home placement in demented patients. They also reported that BPSD was one of the major reasons for admission to rehabilitation units for dementia, but was not risk factors for institutionalization after discharge.
The present study showed a gender difference in the care situation of demented patients. Most men had been cared for by their spouses. In contrast, women could be divided into two groups. The first group had been living alone and most of them had been cared for by their children. The second group of women had been living with their families and most of them had been cared for by their daughters-in-law. In Japan, women tend to live longer than men. As a result, most men are cared for by their spouses. However, some women cannot be cared by their husbands and live alone or live with their relatives. Most women were institutionalized after discharge and the ratio of institutionalization in women was higher than that in men.
The predictors that influenced the outcomes between the hospital group and the others were mainly affected by complications in each sex. In contrast, the difference in outcomes between the home group and institution group was influenced by N-ADL in men and by the number of cohabitants, caregivers and HDS-R in women. Based on this report, the caregiving environment emerged as a reliable predictor that influenced the outcome between the institution group and home group in women. If the demented patients suffered from severe complications, they were more likely to be hospitalized. The outcome of demented patients without complications could be predicted by their situation on admission. In men, the outcome was dependent on the patient's condition, because most of them were cared for by their spouses and the care situation of the demented men was simpler than that of women.
Many reports suggested that the patient's gender was not related to nursing home placement.2,16,22,23 However, Heyman et al.24 showed that only married men had a reduced risk of placement. Smith et al.13 showed that men are at reduced risk for placement compared with women, and explained that the role of gender remains unclear.
Yaffe et al.2 reported that the relationship between the caregivers and cohabitants did not affect the nursing home placement. However, Hebert et al.17 showed that a caregiver who was not a spouse or child was a predictor of institutionalization. Smith et al.13 reported that the type of cohabitant was significant in univariate analysis. Living with a spouse was protective against placement, but living with caregivers other than spouses increased risk. However, cohabitant type might be associated with aspects of disease severity as cohabitant type did not remain a significant predictor in multivariate modeling. The nursing home placement did not depend on who the caregiver was.
Colerick and George22 reported that most demented males were cared for at home by their spouses and that they were also the group most likely to relinquish care to professionals. Female patients were cared for mostly by their employed female children, who were likely to be caring for families of their own; however, it was unclear what brought about the gender differences in outcomes.
In conclusion, the results showed that there were gender differences in outcome predictors. The outcomes of demented patients in the ward were predicted by both complications during hospitalization and the situation on admission. Most women did not return to their homes because of the situation in which they received care. It is necessary to establish a clear system to provide care to the demented elderly, especially for women.
Our sample selection was a limitation of this study, because we considered demented elderly at risk of institutionalization. The low Cox-Snell R2 in our report ranged from 0.2 to 0.3, indicating that the contribution rate of these predictors was relatively low. It is suggested that unknown factors might influence the outcomes. Therefore, the results are not generalized. Further analysis is necessary, evaluating the rate of institutionalization of patients after special ward stays.