Factors influencing the death of the elderly at home in an institution specializing home medical care--analysis of survey of the bereaved family

AIM: To determine factors influencing on the death of the elderly at home in an institution specializing in-home medical care (home death rate: about 80%). METHODS: This study was an anonymous mailed survey of bereaved families of patients (the caregivers) who died using the home medical care setting provided by an institution specializing in home medical care in Japan. We analyzed the relationship of demographics of the patients and the caregivers with such factors as the anxiety of the caregiver and eventual death at home, and that of such factors as perceived quality of home medical care by the caregiver, and satisfaction with home medical care, satisfaction with death bed nursing, regret for death bed nursing. RESULTS: The most significant determinant factor associated with home death was that both patients and caregivers preferred death at home (OR=19.42). Determination coefficients (R(2)) in the multiple linear regression analysis of caregiver's satisfaction with home medical care, satisfaction with death bed nursing and regret for death bed nursing were as high as 0.68, 0.55 and 0.62, respectively. Significant influential factors were: peaceful death, psychological consistency of the caregiver, a good relationship with one's physician and fulfillment of the care service system. CONCLUSIONS: Important components of home medical care and death bed nursing for the realization of death of the elderly at home are 1) peaceful death, 2) psychological consistency of caregivers, 3) a good relationship with one's physician, 4) fulfillment of the care service system.     

What determines the life satisfaction of the elderly? Comparative study of residential care home and community in Japan

Background:   Increasing lifespan has forced elderly persons and researchers alike to look at life and aging in a new way: at adding life to years rather than years to life. With increasing lifestyle choices for the elderly, the present study was undertaken to determine and compare the factors associated with life satisfaction in elderly living in a residential care home and in the community in Japan.
Methods:   This was a cross-sectional, self-reported, questionnaire-based study. Elderly persons from the residential care home in Kyoto City ( n  = 261, mean age ± SD = 79.6 ± 6.6 years, male : female = 1 : 3.1) and from the town of Urausu, Hokkaido ( n  = 733, mean age ± SD = 74.8 ± 6.8 years, male : female = 1 : 1.3), participated in the study. This represented 95.3% and 89.5% of all eligible elderly ≥ 65 years in the home and town, respectively. Activities of daily living, higher functions, medical and social history, geriatric depression scale and quality of life were studied. Using multiple logistic regression models, factors associated with high satisfaction of life were identified.
Results:   In community-dwelling elderly, relationship with friends, as opposed to relationship with family for elderly in residential care home, were factors related to life satisfaction.
Conclusions:   This study demonstrates the relative importance of social factors versus medical and functional factors as determinants of life of elderly. We have also shown how these factors differ for elderly living in different environments. Understanding these different as well as common determinants of life satisfaction from a whole array of diverse ones allows us to make effective strategy to improve the quality of life of elderly living in different conditions.     

Factors associated with ADL dependence: A comparative study of residential care home and community-dwelling elderly in Japan

Background:   Since the start of Long Term Nursing Insurance in Japan in April 2000, there has been much discussion about and emphasis on leading healthy and active twilight years, without ever being bedridden if possible. In an attempt to address some of these concerns, this comparative study examines the factors associated with dependence in the activities of daily living (ADL) in two samples of elderly people living in two different environments.
Methods:   In this cross-sectional, self-reported study, 262 and 968 elderly people from a residential care home and Yogo town completed a 65-item questionnaire. Questions pertaining to ADL, medical and social history, quality of life (QOL) and the 15-item Geriatric Depression Scale were included in the questionnaire. Using logistic regression model, factors associated with ADL dependence were determined.
Results:   In the residential care home group, age, female gender, instrumental ADL, communication-related ability, history of falls, stroke, and osteoarthropathy, depressive tendency and low satisfaction with life emerged as the factors associated with ADL dependence, after multivariate logistic regression. The corresponding factors in the community-dwelling elderly group were age, depressive tendency, history of osteoarthropathy, and low sense of health.
Conclusion:   Age and female gender were the two non-modifiable risk factors associated with ADL dependence. Stroke, followed by osteoarthropathy and falls were the main medical conditions predisposing to functional dependence. Dependence in instrumental ADL, independence in communication-related ADL, depressive tendency and components of QOL were factors associated with ADL dependence. Awareness of these factors not only helps to identify at-risk patients, to initiate preventive measures and to promote disability-delaying activities, but also helps in the holistic management of geriatric patients.     

Analysis of comprehensive geriatric assessment of elderly residents in a social welfare home for the aged compared with those in a residential care home in an urban area in Japan

Background:   The purpose of the present study was to clarify the influence of the socioeconomic factors during the middle age on the results of comprehensive geriatric assessments in later stage of life.
Methods:   A cross-sectional, questionnaire-based study was conducted of elderly residents in a welfare home for the aged in Osaka and those in a residential care home in Kyoto. Results of questionnaires pertaining to activities of daily living (ADL), quality of life (QOL) and a 15-item Geriatric Depression Scale as well as medical and social histories of the two groups were compared by unpaired T -test and χ² test.
Results:   Elderly residents in the welfare home were significantly younger and more were both male and unmarried or divorced than among those in the residential care home. Scores in ADLs and QOLs were significantly lower and the prevalence of depression was significantly higher in residents in the welfare home than in the residential care home.
Conclusion:   The ADLs, mental mood and subjective QOLs of residents in the welfare home were significantly lower than those of residents in the residential care home. The reason for these differences is suggested to be due to both the differences in the lifestyles of the residents in their middle age and the difference in the quality and quantity of care the residents are receiving. Local social welfare government should pay more attention to lower ADLs and QOLs of residents in welfare homes and increase efforts to improve the quality and quantity of care for them.     

Factors influencing death at home in terminally ill cancer patients

Aim: The purpose of this study is to investigate factors affecting terminally ill cancer patients dying at home. Material: Ninety‐two terminally ill cancer patients who were receiving home medical care services and died between April 2005 and December 2006 were included in the study. The data included patients' and caregivers' demographic characteristics, disease‐related information, place of death, and status of home care support. To identify the factors predicting the place of death, multivariate logistic regression analyses were performed. Results: Patients of families who had no preference regarding the place of death or a preference for death at home were more likely to die at home (vs preference for hospital death, odds ratio = 5.87, 95% confidence interval = 1.02–36.53; odds ratio = 90.35, 95% confidence interval = 8.15–1001.51, respectively) after adjusting for potential confounders. Meanwhile, if the patient's family preferred that the patient not die at home, the patient's place of death was not at his/her home irrespective of his/her preference. Conclusion: The results suggested the stronger involvement of families' preferences regarding the patients' place of death over patients' own preferences. Therefore, factors affecting families' preferences need to be clarified for the dissemination of death at home for terminally ill cancer patients.     

Combined effect of factors associated with burdens on primary caregiver

Background:   It is argued that a multidimensional approach is necessary for burden assessment. Reducing caregiver burden is a social problem in the ageing Japan society. We examined the combined effect of factors affecting the care burden among community-dwelling handicapped people and their caregivers.
Methods:   The participants were 49 handicapped people (aged 53–104 years) who received home-visit rehabilitation, and their 49 caregivers (age 42–85 years). Caregivers were provided questionnaires consisting of questions on social support, subjective well-being, self-efficacy with regard to care continuation, the Motor Fitness Scale and caregiver burden. Care recipients were assessed using the Bedside Mobility Scale and the Barthel Index.
Results:   We prepared the hypothesis model using structural equation modeling with the bootstrap method within outcome measures. The hypothesis model did not fit the data well. The impact of the Motor Fitness Scale was shifted from the caregiver burden to care self-efficacy and well-being, having a cooperator for care and variable of spouse caregiver or others associated with caregiver well-being in the revised model. The fit of the revised model was acceptable (goodness of fit index, 0.903; comparative fit index, 0.998; root mean square error of approximation, 0.017). In the revised model, the care recipients' disabled state was associated with caregiver burden. In addition, higher burden and poor motor fitness of caregivers might lead to lower care self-efficacy in providing continuous care and lower caregiver well-being.
Conclusion:   These findings suggested that the program to reduce caregiver burden should focus on aspects of the care recipients' disabled state, the caregivers' well-being, fitness, and care self-efficacy.     

Influence of diabetes mellitus on caregiver burden in home care: A report based on the Nagoya Longitudinal Study of the Frail Elderly (NLS-FE)

Background:   Because little attention has so far been paid to the impact of diabetes mellitus (DM) on caregiver burden in community settings, we sought to confirm the influence of DM on perceived caregiver burden among caregivers providing care to a home elderly person using data from the Nagoya Longitudinal Study of the Frail Elderly (NLS-FE).
Methods:   The NLS-FE is a large prospective study of 1875 community-dwelling elderly. A total of 1592 pairs of dependents and caregivers were included in the analysis. The data we used in this study included the Japanese version of the Zarit Caregiver Burden Interview (J-ZBI), characteristics of caregivers and dependents, and caregiving situation. The pairs were sorted into one control and three DM groups: (i) no DM; (ii) DM taking no medications; (iii) DM taking oral medication only; and (iv) DM taking insulin. The differences in dependent and caregiver characteristics among the groups were assessed.
Results:   Two hundred and twenty-eight dependents from the NLS-FE study had DM. Of these, 25% took no medication to treat it, 55% took oral medications only, and 20% used insulin. No statistical differences were found in age, gender or kinship among caregivers. No differences were found among the DM categories in levels of caregiver burden according to the J-ZBI, before and after adjusting for these baseline variables.
Conclusions:   Among the community-dwelling frail elderly, DM is not an independent predictor of caregiver burden.     

Survey of primary care physicians and home care clinicians

CONTEXT: Providing home care in the United States is expensive, and significant geographic variation exists in the utilization of these services. However, few data exist on how well physicians and home care providers communicate and coordinate care for patients. OBJECTIVE: To assess communication and collaboration between primary care physicians (PCPs) and home care clinicians (HCCs) within 1 primary care network. DESIGN: Mail survey. SETTING: Boston. PARTICIPANTS: Sixty-seven PCPs from 1 academic medical center-affiliated primary care network and 820 HCCs from 8 regional home care agencies. MEASUREMENTS: Provider responses RESULTS: Ninety percent of PCPs and 63% of HCCs responded. The majority (54%) of PCPs reported that they only "rarely" or "occasionally" read carefully the home care order forms sent to them for signature. Further, when asked to rate their prospective involvement in the decision making about home care, only 24% of PCPs and 25% of HCCs rated this as "excellent" or "very good." Although more HCCs (79%) than PCPs (47%) reported overall satisfaction with communication and collaboration, 28% of HCCs felt they provided more services to patients than clinically necessary. CONCLUSIONS: PCPs from 1 provider network and the HCCs with whom they coordinate home care were both dissatisfied with many aspects of communication and collaboration regarding home care services. Moreover, neither group felt in control of home care decision making. These findings are of concern because poor coordination of home care may adversely affect quality and contribute to inappropriate utilization of these services.     

Chewing problems and mortality in older adults in home care: results from the Aged in Home Care study

OBJECTIVES: To assess the association between chewing problems and risk of mortality in an older population receiving home care in Europe.
DESIGN: Retrospective cohort study.
SETTING: Eleven European countries.
PARTICIPANTS: Two thousand seven hundred fifty-five older adults (mean age±standard deviation 82.2±7.2) in home care.
MEASUREMENTS: Data were collected using the Minimum Data Set for Home Care. Study personnel recorded chewing problems (inability to chew food easily and without pain or difficulties, regardless of cause) that presented in the last 3 days before baseline assessment. Data on mortality were collected over a 1-year period.
RESULTS: Three hundred ninety-five participants (14.3%) presented with chewing problems. One-year mortality differed significantly according to presence of chewing problems; 303 of 2,361 (12.8%) without chewing problems (crude incident rate per person-year (p-y)=0.15) and 80 of 394 (20.3%) participants with chewing problems (crude incident rate per p-y=0.24) died during follow-up. After adjusting for potential confounders, the risk of death was significantly higher for participants with chewing problems (adjusted hazard ratio (HR)=1.45, 95% confidence intervals (CI)=1.05–1.99). This association remained after exclusion of participants with cognitive impairment (adjusted HR=1.50, 95% CI=1.03–2.20) and those with unintended weight loss (adjusted HR=1.62, 95% CI=1.12–2.34).
CONCLUSION: In older adults in home care in Europe, chewing problems are associated with greater risk of mortality.     

The Overlap Syndrome of Depression and Delirium in Older Hospitalized Patients

OBJECTIVES: To measure the prevalence, predictors, and posthospitalization outcomes associated with the overlap syndrome of coexisting depression and incident delirium in older hospitalized patients.
DESIGN: Secondary analysis of prospective cohort data from the control group of the Delirium Prevention Trial.
SETTING: General medical service of an academic medical center. Follow-up interviews at 1 month and 1 year post-hospital discharge.
PARTICIPANTS: Four hundred fifty-nine patients aged 70 and older who were not delirious at hospital admission.
MEASUREMENTS: Depressive symptoms assessed at hospital admission using the 15-item Geriatric Depression Scale (cutoff score of 6 used to define depression), daily assessments of incident delirium from admission to discharge using the Confusion Assessment Method, activities of daily living at admission and 1 month postdischarge, and new nursing home placement and mortality determined at 1 year.
RESULTS: Of 459 participants, 23 (5.0%) had the overlap syndrome, 39 (8.5%) delirium alone, 121 (26.3%) depression alone, and 276 (60.1%) neither condition. In adjusted analysis, patients with the overlap syndrome had higher odds of new nursing home placement or death at 1 year (adjusted odds ratio (AOR)=5.38, 95% confidence interval (CI)=1.57–18.38) and 1-month functional decline (AOR=3.30, 95% CI=1.14–9.56) than patients with neither condition.
CONCLUSION: The overlap syndrome of depression and delirium is associated with significant risk of functional decline, institutionalization, and death. Efforts to identify, prevent, and treat this condition may reduce the risk of adverse outcomes in older hospitalized patients.     

Effects of a care protocol on care outcomes in older nursing home patients with chronic obstructive pulmonary disease

OBJECTIVES: To evaluate the effects of a care protocol used by community nurses to support nursing home staff in the care of patients with chronic obstructive pulmonary disease (COPD). DESIGN: Matched, randomized case-control trial. SETTING: Forty-five nursing homes of the New Territories South (NTS) cluster of Hong Kong. PARTICIPANTS: Eighty-nine older people (> or =65, present resident of a nursing home in the NTS region, main diagnosis of COPD, at least one hospital admission in previous 6 months) discharged to the nursing homes from the geriatric units of two hospitals. INTERVENTION: Using a care protocol, community nurses followed up older patients in the experimental group for 6 months after their discharge from the hospitals to the nursing homes. MEASUREMENTS: Data on functional, respiratory, and psychological parameters were collected at entry to study and 6 months later with standard measures. Data on hospital service utilization, nursing home staff, and patient satisfaction were also collected at 6 months. RESULTS: Experimental group participants had significant (P =.008) improvements in psychological well-being. Nursing home staff and experimental group patients were highly satisfied with the use of the protocol. There was no significant difference between the two groups in functional and respiratory outcomes or hospital service utilization. CONCLUSION: Psychological well-being as an important factor in rehabilitation in chronic illness has been much neglected in the literature. Supporting nursing home staff in the care of COPD patients through community nursing visits can enhance older residents' psychological well-being. Psychological aspects of care should be emphasized and incorporated into the delivery of regular nursing home care.     

Comparison of two home care protocols for total joint replacement

OBJECTIVES: To examine the effect of a more-efficient home care protocol to manage total joint replacement (TJR) patients after surgery. DESIGN: A randomized trial of two home care protocols for TJR management. SETTING: A hospital-affiliated home healthcare agency in a large midwestern city. PARTICIPANTS: Medicare-eligible individuals undergoing elective total hip or knee replacement surgery (N = 136). INTERVENTION: A home care protocol that included preoperative home visits by a nurse and a physical therapist and fewer postoperative visits (range of 9-12 visits) to the home than an existing protocol (range of 11-47 visits). MEASUREMENTS: Functional status, lower extremity functioning, health-related quality of life, satisfaction with care, and use and cost of healthcare services for 6 months postsurgery. RESULTS: There were no differences in functional status, health-related quality of life, or lower extremity functioning by group at 6 months. A marginally significant gain in satisfaction with access to care (P =.059) was found in the intervention group at 6 months. Home healthcare costs were 55% lower for the streamlined group (P <.001). Other costs did not differ significantly by group. CONCLUSION: TJR patients who received the more-efficient home care protocol experienced comparable outcomes to those who received the existing protocol. An abbreviated set of home care visits resulted in more-efficient delivery of care without compromising patient outcomes.     

Age-related differences in care receipt and symptom experience of elderly cancer patients dying at home: Lessons from the DEATH project

Background:   Data on the differences between older and younger elderly cancer patients dying at home is sparse. To clarify age-related differences in symptom experience and care receipt of elderly cancer patients at end-of-life, we conducted a subanalysis study of the Dying Elderly at Home (DEATH) project, a multicenter study of 240 elderly aged 65 and older dying at home.
Methods:   We assessed the frequency of symptom experience and end-of-life care receipt in home elderly patients during the last 2 days of their lives and evaluated the differences between younger elderly (aged 65–74) and older elderly (aged 75+) cancer decedents. The general practitioners were asked to fill out a questionnaire immediately after the death of study patients. A total of 66 younger and 51 older elderly cancer decedents were included in the analysis.
Results:   Coma and dementia were common among younger and older elderly patients. Older decedents were less likely to experience anxiety, but, after adjustment for baseline characteristics, this age-related difference did not clearly appear. Older decedents were also less likely to receive opioids than younger decedents. There were no significant differences in volume of i.v. hydration between the two groups.
Conclusions:   Our results suggested that there were no differences in symptom experience and care receipt among older and younger decedents, except in opioid use, at end-of-life. These findings imply a similar need of end-of-life care for younger and older elderly cancer patients who opt for home death.     

Impact of informal care levels on discontinuation of living at home in community-dwelling dependent elderly using various community-based services

The aim of the study was to examine the effect of informal care levels on overall discontinuation of living at home, all-cause death, hospital admission, and long-term care placement for community-dwelling older people using various community-based services during a 3-year period. Prospective cohort study of 1582 community-dwelling disabled elderly and paired informal caregivers was conducted. Baseline data included the recipients and caregivers’ demographic characteristics, comorbidities, informal care levels (sufficient, moderate, and insufficient care), which were evaluated by trained visiting nurses, and the level of formal community-based service use. Among 1582 participants, 97 died at home, 692 were admitted to hospitals, 318 died during their hospital stay, and 117 were institutionalized in long-term care facilities during 3 years of follow-up. A multivariate Cox hazard model demonstrated that when compared with a sufficient informal care level, an insufficient informal care level was associated with overall discontinuation of living at home, all-cause mortality, hospitalization, and institutionalization during 3 years of follow-up (hazard ratio: 1.65, 95% confidence interval: 1.15-2.36; 1.98, 1.17-3.34; 1.56, 1.04-2.35; 2.93, 1.25-6.86, respectively). The results suggested that informal caregiving is an important factor in the prevention of overall discontinuation of living at home in a population of disabled older people.     

Substitutive "hospital at home" versus inpatient care for elderly patients with exacerbations of chronic obstructive pulmonary disease: a prospective randomized, controlled trial

OBJECTIVES: To evaluate hospital readmission rates and mortality at 6-month follow-up in selected elderly patients with acute exacerbation of chronic obstructive pulmonary disease (COPD).
DESIGN: Prospective randomized, controlled, single-blind trial with 6-month follow-up.
SETTING: San Giovanni Battista Hospital of Torino.
PARTICIPANTS: One hundred four elderly patients admitted to the hospital for acute exacerbation of COPD were randomly assigned to a general medical ward (GMW, n=52) or to a geriatric home hospitalization service (GHHS, n=52).
MEASUREMENTS: Measurements of baseline sociodemographic information; clinical data; functional, cognitive, and nutritional status; depression; and quality of life were obtained.
RESULTS: There was a lower incidence of hospital readmissions for GHHS patients than for GMW patients at 6-month follow-up (42% vs 87%, P <.001). Cumulative mortality at 6 months was 20.2% in the total sample, without significant differences between the two study groups. Patients managed in the GHHS had a longer mean length of stay than those cared for in the GMW (15.5±9.5 vs 11.0±7.9 days, P =.010). Only GHHS patients experienced improvements in depression and quality-of-life scores. On a cost per patient per day basis, GHHS costs were lower than costs in GMW ($101.4±61.3 vs $151.7±96.4, P =.002).
CONCLUSION: Physician-led substitutive hospital-at-home care as an alternative to inpatient care for elderly patients with acute exacerbations of COPD is associated with a substantial reduction in the risk of hospital readmission at 6 months, lower healthcare costs, and better quality of life.     

Families' acceptance of near death: A qualitative study of the process for introducing end-of-life care

Aim:   Providing effective end-of-life (EOL) care for the elderly with severe brain damage is difficult because patients' families find it hard to accept the condition of their loved ones as "near death". In Japan, this has become an urgent social problem. Although health-care teams sometimes expect that the elderly with severe brain damage should be treated as terminal, many find that patients' families cannot accept the condition as near death. As a result, they are not able to appropriately introduce any EOL services. It was the aim of the present study to develop a comprehensive understanding of the process by which families accept the elderly with severe brain damage as near death.
Methods:   Qualitative methodology with focus groups and semi-structured interviews were used. Twenty-three participants in two sets of male and female focus groups were interviewed in a semi-structured format. Ten participants (five men and five women) who made important medical decisions were then interviewed separately.
Results:   Eleven categories emerged: (i) family affection with wishes for continued survival; (ii) vacillation of desire for death with dignity; (iii) family members' hierarchy; (iv) awareness that others may make different decisions; (v) family members' discussion overcoming discordance; (vi) satisfaction with physicians' explanations; (vii) impressions of life-sustaining measures; (viii) entrusting important decisions to hospital physicians; (ix) significance of family members' previous experiences; (x) patient's age; and (xi) duration of medical treatment.
Conclusion:   We developed a conceptual model that is useful for evaluating which stage of the process families are currently experiencing as well as introducing EOL care in a timely manner.     

Cost-effectiveness of Veterans Administration hospital-based home care. A randomized clinical trial

A randomized design was used to examine the cost-effectiveness of a Veterans Administration hospital-based home care program that case managed inpatient and outpatient care. Patients (N = 419) with two or more functional impairments or a terminal illness were randomized to hospital-based home care (n = 211) or customary care (n = 208). Functional status, satisfaction with care, and morale were measured at baseline and at 1 and 6 months after discharge from the hospital; health care utilization was tracked for 6 months. Findings included significantly higher (0.1 on a three-point scale) patient and caregiver satisfaction with care at 1 month and lower Veterans Administration and private sector hospital costs ($3000 vs $4245) for the experimental group. Net per person health care costs were also 13% lower in the experimental group. We conclude that this model of hospital-based home care is cost-effective and that its expansion to cover these two patient groups throughout the Veterans Administration system can improve patient care at no additional cost.     

A randomized trial of family caregiver support in the home management of dementia

A randomized trial of family caregiver support for the home management of older people suffering from moderate to severe progressive irreversible dementia was conducted in an urban center in southern Ontario. Thirty caregivers were allocated to receive the experimental intervention consisting of: caregiver-focused health care, education about dementia and caregiving, assistance with problem solving, regularly scheduled in-home respite, and a self-help family caregiver support group. Thirty control subjects received conventional community nursing care. Before completion of the intervention, 18 (30%) were withdrawn, almost equally from each group. The most frequent reason was long-term institutionalization of the demented relative (n = 10). At baseline, caregivers in both groups were suffering from above-average levels of depression and anxiety. After the six-month intervention period, we found neither experimental nor control group improved in these areas. However, the experimental group showed a clinically important improvement in quality of life, experienced a slightly longer mean time to long-term institutionalization, found the caregiver role less problematic, and had greater satisfaction with nursing care than the control group.     

Preventive home visits for community-dwelling frail elderly people based on Minimum Data Set-Home Care: Randomized controlled trial

Background:   Although preventive home visits for the elderly are mandatory in Japan, there has been little research into their effectiveness. The present study used the Minimum Data Set-Home Care (MDS-HC) as the basis for assessing community-dwelling frail elderly persons.
Methods:   The present study was a randomized controlled, community-based investigation of 368 elderly people aged 65 years and older who were dependent in the instrumental activities of daily living, but independent in activities of daily living. The participants were randomly assigned to an intervention group (184) or a control group (184). For 18 months, the intervention group received scheduled home visits by public health nurses who had been instructed that the primary objective of these visits was human interaction. The MDS-HC was used to assess the elderly person at each visit. The primary outcome was the EQ-5D score of the EuroQol and secondary outcomes were five items constituting the EQ-5D, self-rated health, and health behaviors. The types of advice given and subject compliance were also analyzed.
Results:   Intervention had no effect on the EQ-5D score, and positive results were limited to some aspects of health behavior. Subgroup analyses showed that the home visits were effective for elderly people who perceived their own health as poor at baseline, and for participants who complied with advice.
Conclusions:   The preventive home visits based on the MDS-HC were effective in selected groups of frail elderly people. Further modification in the use of the MDS-HC and more focused targeting are needed to make the visits more effective.