Duration of stay and outcome for inpatients on an assessment ward for elderly patients with cognitive impairment

Duration of stay and outcome for inpatients on an assessment ward for elderly patients with cognitive impairment

Ball, Sue


The move away from the provision of long-stay beds by the NHS inevitably meant a change in function for wards for elderly patients with cognitive impairment to a more acute way of working. Literature is scarce on the role or effectiveness of the new assessment wards that have replaced them and the factors affecting outcome and the duration of stay. Evidence suggests that those patients with higher dependency levels and behavioural problems stay in hospital longer, as do those awaiting a nursing home placement.

This paper reports a prospective study of a consecutive group of 101 patients who died on or were discharged from an acute assessment ward for elderly patients with cognitive impairment. Clinical characteristics were recorded according to an in-patient dementia care pathway, which included Mini-MOUSEPAD, Crichton activities of daily living, Mini-Mental State Examination and the Burvill physical health score evaluations. Outcome measures were duration of stay, destination on discharge or death on the ward.

Most patients had cerebrovascular disease (48%) or Alzheimer’s disease (32.9%), and their average Mini-Mental State Examination score was 14.9. The mean duration of stay was 7.9 weeks. Self-funding status and lack of behavioural and psychological complications were associated with a reduced duration of stay. 22.2% of patients were successfully rehabilitated to their own homes, but 20% died. Discharge home was most strongly predicted by having a spouse at home, and the need for nursing home rather than residential care was related to the severity of cognitive impairment. This study concludes that patients can expect to stay in hospital for 8 weeks but two areas of concern are highlighted. Firstly, the importance of the funding of community rehabilitation for patients with memory disorders and, secondly, the importance of a spouse at home to look after the patient.

KEY WORDS elderly patients NHS cognitive impairment assessment wards outcome and duration of stay


In the United Kingdom, the National Health Service has been moving away from the provision of long-stay or continuing care beds for patients suffering from dementia, even though as recently as 1995 the Royal College of Psychiatrists issued a consensus statement recommending the provision of large numbers of continuing care beds for such patients (Royal College of Psychiatrists, 1995). Current NHS provision in this area is mainly in acute admission beds for functional and organic disorders. However, there is very limited literature on the effectiveness of this type of ward or the optimal length of stay. What is certain is that the closure of long-term beds has resulted in a change in the activity on such wards (Domken et al., 1995) with dementia patients spending less time in the ward.

Following the closure of all long-term dementia beds in one NHS trust in Dorset in 1998, an acute organic psychogeriatric ward was developed to assist the rehabilitation of confused elderly patients who were failing in the community. One of the main aims of the new ward was to improve the patients’ health and function as much as possible to enable their optimal rehabilitation in a community setting. An integrated inpatient dementia pathway (available from the authors on request) was used from admission to discharge. The aim for each patient was to make a diagnosis, assess cognitive and non-cognitive features as well as physical health and activities of daily living, institute and monitor a treatment programme and plan discharge and follow up.

Purpose of the study

The purpose of this study was to examine factors that influenced the discharge outcomes and duration of stay of acute inpatients on a psychiatric ward for confused elderly patients. The study took place following the closure of all long-term beds and the opening of a new assessment/rehabilitation ward.


The hypotheses were that psychiatric complications would cause increased duration of stay on the ward, but that finding nursing home placements might be influenced more by social factors than the patients’ clinical characteristics.


This was a prospective study conducted during 1999. Data were collected from an integrated inpatient dementia care pathway. A total of 101 patients admitted consecutively to the acute organic assessment ward were included in the study.

Inclusion and exclusion criteria

All patients with ICD-10 criteria for dementia (WHO, 1993) admitted to the ward were included in the study. Patients with functional mental illness as their main diagnosis or patients on respite admission were excluded from the study.

Demographic details

Demographic details were obtained from the case notes. These included age, social class, carer support, marital status and whether they were self or social service funded.

Clinical features and psychiatric characteristics

Information pertaining to the patient diagnosis and psychiatric characteristics was also recorded prospectively according to the in-patient dementia care pathway. This included the following measures: Mini-Mental State Examination (Folstein et al., 1975), measuring general cognition after one week; (Folstein et al, 1975), the Crichton Royal Behaviour scale (Robinson, 1965), measuring functional ability after one week (Robinson, 1965); the Mini-MOUSEPAD, (Alien et al, 1996) measuring psychiatric behaviour (Alien et al, 1996) at two weeks; and a global estimate of physical health made using the Burvill Physical Health Questionnaire (Burvill et al., 1990) representing severity, chronicity and disability, measured at discharge.


Data were analysed using the Statistical Package for Social Sciences (SPSS) version 12. Correlations were obtained using Pearson’s product-moment correlation coefficients for parametric variables and Spearman rank correlation coefficients for non-parametric variables. Differences between groups were calculated for continuous variables using analysis of variance (ANOVA) with adjustment for confounding variables using analysis of covariance (ANCOVA). Differences between groups of categorical variables were calculated using Chi-square tests. Forward stepwise linear multiple regression was used to examine multiple inter-related predictors of continuous dependent variables and the independent regression coefficients of the predictors are quoted. For categorical dependent variables, logistic regression was used and the odds ratios of the predictors with 95% confidence intervals were calculated.


Description of the patient population

Table 1 summarises the patient data and outcomes. In total 52% of patients were dependent upon Social Services to fund their community care at home or in residential or nursing home care. Most patients had cerebrovascular disease (48.5%) or Alzheimer’s disease (38.62%). The patients were moderately cognitively impaired, with an average Mini-Mental State Examination score of 14.9. Those with cerebrovascular disease were significantly more physically ill (F=18.2; p

Duration of stay

The mean duration of stay was 7.9 weeks. Social Service funding status significantly (p

Given that there may have been an interaction of the behavioural and psychological complications with the social factors in the prediction of duration of stay in hospital, a regression analysis was used to identify the strongest independent predictors of duration of stay in the surviving patients. Mini-MOUSEPAD score and funding status were entered as the independent predictor variables in a linear multiple forward stepwise regression analysis where the duration of stay in weeks was the dependent variable. The analysis demonstrated that both were independent predictors of duration of stay self-funding status (B=4.2, 95% CI for B=1.4-7.1; p=0.004) and psychosis (B=0.48 95%CI for B=0.07-0.88; p=0.022) were independent predictors of stay. Similarly as can be seen in Table 2, there is a significantly increased duration of stay for the Social Service funded patients even after adjustment for the behavioural complications (ANCOVA).

Community rehabilitation

In simple terms the analysis of variance shows that patients who are dependent on Social Services for funding can expect to wait an extra 5.7 weeks to be discharged from hospital, regardless of their clinical and social characteristics (Table 2).

Outcome at discharge

Outcome was categorised as death on the ward, discharge home with community care, or discharge to residential or nursing home care. Functional ability as estimated by the Crichton Royal Behaviour scale and cognitive impairment as estimated by the Mini-Mental State Examination were the factors associated with discharge outcome (Table 3 over the page).

Factors associated with death were examined. 27% of patients with cerebrovascular disease died, compared with 15.8% of those who had Alzheimer’s disease. Patients with Lewy body disease faired best with a 14.3% mortality rate during the admission. However, death rates were not significantly different across the diagnostic groups. The only significant factor associated with an increased death rate was poorer physical health (F= 13.5; p

Home versus care

Table 3 shows the physical health, functional and cognitive levels and carer availability associated with outcome and successful rehabilitation in the community. Having a spouse or a carer were positive indicators for coping at home as was a low dependency rating on the Crichton Royal Behaviour scale. Regression analysis was used to determine the independent predictors of discharge home. The results showed that having a spouse living at home was the strongest predictor (odds ratio=0.17; 95%CI for odds ratio=0.04-0.73; p=0.017) of rehabilitation home and that a (non-spouse) carer living at home did not influence chance of discharge back home. Having a low Crichton Royal Behaviour score, ie, having greater functional ability, was an additional independent predictor of successful rehabilitation at home (odds ratio=0.87, 95%CI for odds ratio= 0.76-0.98; p=0.025).

Residential home versus nursing home

The Mini-Mental State Examination score was low in the group of patients who required nursing home care (Table 3). This measure was moderately sensitive and specific. Of patients with a Mini-Mental State Examination score

Furthermore, nursing home patients had high levels of dependency, as assessed by the Crichton Royal Behaviour scale (Table 3). Forward stepwise logical regression analysis of the data from patients who were discharged showed that only the Mini-Mental State Examination score predicted nursing home as opposed to residential home placement (odds ratio=0.89; 95% CI for odds ratio= 0.81-0.98; p=0.019).


Since the National Health Service in the United Kingdom reduced the provision of NHS long-term care wards for patients with acquired cognitive impairment, new acute/rehabilitation psychogeriatric wards have evolved, with little published data on what they can offer. There are now concerns that patients have to pay privately for long-term care elsewhere from the NHS. This study evaluates such a ward and provides objective evidence for its clinical outcomes.

Two main groups of implications can be drawn from this study. Firstly, those relating to outcomes for the patients discharged from the ward and secondly, those relating to length of stay.


Descriptive outcomes suggest an expected 19% mortality, but a good chance (approximately 1/3) of rehabilitation home for surviving patients. The high death rate may be because the provision of an acute ward for confused elderly patients is likely to offer an inpatient sanctuary for patients with delirium, which by definition is due to an underlying physical cause, thereby conferring the high death rate. In support of this view we have demonstrated the very poor physical health in those patients who died. The terminal care of such patients is an expected function of an acute organic psychiatric ward. This specialised area of old age psychiatry involves the nursing and medical care of physically ill and debilitated patients who are often too disturbed to be managed on a medical ward or in a hospice setting. We have found that geriatricians and general practitioners are required to work in close liaison with psychogeriatrics to achieve good care.

A third of survivors were rehabilitated back home with a community care package. This was an encouraging finding because one current philosophy of elderly care services, based upon person-centred care, is to maintain people in their own homes for as long as is good for them. This outcome might not have been seen with a traditional model of inpatient/long-term dementia care. Our inpatient dementia care pathway promotes active rehabilitation and regular family sessions: so that feasible care at home is promoted.

We found that the chance of patients returning home is dependent on many factors, but strongly associated with functional disability and having a spouse at home. This emphasises the need for a long-term relationship of the carer with the patient to cope with the demands of dementia care at home. A recent review of carer support showed that the quality of the relationship of the spouse (Spruytte et al., 2001; Reay & Browne, 2001) with the dementia patient is an important factor in carer strain and ability to cope.

The Mini-Mental State Examination (Folstein et al., 1975) and Crichton Behaviour scale (Robinson, 1965) had a clinical application in predicting the rehabilitation outcome. Cream (2002) points out that no guidance has been given on whether a person with dementia will need nursing rather than residential care. It is known that both residential and nursing homes have high rates of cognitive impairment amongst residents (Netten et al, 2001). The Personal Social Services Research Unit (Office for National Statistics, 1999) found that 54% of those admitted to residential care had mild cognitive impairment and 25% had severe cognitive impairment. Of those admitted to nursing home beds, 38% had mild cognitive impairment and 46% had severe cognitive impairment. Severe cognitive impairment is clearly more prevalent in nursing homes. Similarly, we found that nursing home placement as an outcome was associated with low cognitive functioning on the Mini-Mental State Examination, especially when the Mini-Mental State Examination score was less than 10.


The mean duration of stay was 7.9 weeks and was most strongly influenced by Social Service funding issues and psychosis. There is other contemporary evidence that the closure of long-term care beds has resulted in a change in the activity on National Health Service (NHS) organic wards. Domken et al. (1995) described the changing patterns of discharges from a psychogeriatric inpatient ward over a period of six years, during which time there was a substantial reduction of beds in the statutory and health sector. By 1991 their mean length of stay had fallen from 12.4 to 9.4 weeks. Length of stay was much longer for those with high dependency and behavioural problems who were waiting for available community facilities that could provide the care they needed. An American study (Aisen et al., 1994) found that the duration of stay for patients in an inpatient acute old age psychiatric ward was independent of age, presence of cognitive impairment, depression, psychosis and mental co-morbidity. However, duration of stay was significantly (p=0.002) increased at 84 days for patients who required a nursing home when compared to the 41 days it took for patients to move back to their own homes. Sceptically, these findings might suggest that admission of dementia patients to acute psychiatric (NHS) wards may be achieving a simple psychosocial function – merely the provision of somewhere to stay pending a lengthy battle to secure a nursing home placement. Against this argument, Holm et al. (1999) demonstrated significant improvements in cognitive, behavioural and functional status during hospital stay, thereby confirming the benefit to patients. A study by Zubenko et al. (1992) confirmed the effectiveness of short-term psychiatric hospitalisation for patients with Alzheimer’s disease by showing that, following improvement in depression, half of those admitted from home were discharged back to their homes. The authors suggested that short-term admission reduces the need for long-term care. The ability to treat depression in dementia has been supported in community studies of elderly patients (Baldwin et al., 1993) who show response rates similar to patients with depression who have no cognitive impairment.

In the present study, the biggest factor influencing duration of stay was waiting for state Social Service funding, which amounted on average to an extra 5.7 weeks in hospital compared with those relying on private funds. Interestingly, there was no significant difference in any clinical or psychiatric characteristics in patients between the two funding groups. Even after adjusting for the confounding influence of behavioural problems that might limit nursing home choice, the ANCOVA test in Table 2 showed that there was an increase in duration of stay of 5.3 weeks to discharge of Social Service funded patients. A fair system would allow multidisciplinary teams to use the same advocacy for poor people as is available to more affluent families. Many nursing homes out-price Social Services thereby limiting the choice available to Social Service funded families. Clearly a two-tier system exists, but this begs the question of who should hold the budget and manage the funding of patients with dementia disorders who require rehabilitation in the community. The Government, through new proposals on the funding of long-term care (Department of Health, 200Ia) has tried to address this problem by funding the nursing element of care. However, it is important to include health care costs for behavioural and psychological complications and not just the physical health care in patients with dementia because mental health complications may require as high if not higher levels of health care as patients with general medical illnesses. For example dementia activities and stimulation therapies are known to improve quality of life and have effect sizes and numbers needed to treat that are similar to the anti-dementia drugs (Spector et al., 2003). These dementia therapies are a health requirement and are needed on an ongoing long-term care basis. This predicts the need for skilled long-term health care to maintain well-being in patients with chronic neurological disease. Any attempt to regulate the charges applied by private long-term care facilities so that the state can afford them is likely to result in a loss of such facilities and prohibit enterprise. Perhaps there needs to be more of a return to state provision, rather than funding, of long-term care for patients with dementia. Many parts of the UK have stopped all long-term care beds for dementia, yet as recently as 1995 the Royal College of Psychiatrists issued a consensus statement (Royal College of Psychiatrists, 1995) emphasising the need for large numbers of continuing care beds to be available from the NHS. In the present study, the delays in discharge were predominantly in the Social Service funded patients suggesting that it is not a shortage of long-term care placements that is the issue, rather it is the funding system. This idea is strengthened by the finding that both the regression analysis and ANCOVA had adjusted for other variables that might have been confounding the Social Service funded patients. Only 52% of the patients in this study required Social Service funding, yet it led to a significant impact on the dynamics of the ward. Dorset is a very affluent county in the UK and many families are easily able to pay long-term care costs. However, in other, less affluent areas of the UK, the Social Service funding issue is likely to bring the dynamics of elderly wards to a halt. Perhaps flexible multidisciplinary management of the long-term care budget could make a difference to the duration of stay. For example there is some evidence that having a dedicated social worker in the team with a dedicated rehabilitation budget (without recourse to a Social Services funding panel) can reduce inpatient bed usage (Shah et al., 2001). Some provision of NHS long-term care wards might also expedite rehabilitation and maintain the dynamics of the acute/ rehabilitation wards.


As a result of this study we have concluded that an acute assessment ward for elderly patients with cognitive impairment serves more than the simple psychosocial function of providing somewhere to stay while awaiting a nursing home placement. A third of survivors go home and discharge back home is positively influenced by having a spouse at home and lower functional disability. However, patients who need Social Services to fund their rehabilitation have a prolonged stay in hospital. If other similar research replicates our findings of slow rehabilitation for patients by Social Services, there should be strong arguments to restructure the financial administration of dementia rehabilitation or to further the development of state funded long-term dementia care places.

In this study, the Mini-Mental State Examination (Folstein et al., 1975) and Crichton Behaviour scales (Robinson, 1965) used as part of an inpatient dementia care pathway were our most useful tools to help predict rehabilitation outcome, although having a spouse at home makes the most difference in whether the patient can return home.


Aisen PS, Giblin KE, Packer LS & Lawlor BA (1994) Determinants of length of stay in geropsychiatry. American Journal of Geriatric Psychiatry 2 (2) 165-168.

Alien NHP, Gordon S, Hope T & Burns A (1996) Manchester and Oxford universities scale for the psychopathological assessment of dementia (MOUSEPAD). British Journal of Psychiatry 169 293-307.

Baldwin RC, Benbow SM, Marriott A & Tomenson B (1993) Depression in old age. A reconsideration of cerebral disease in relation to outcome. British Journal of Psychiatry 163 82-90.

Burvffl P, Mowry B & Hall W (1990) Quantification of Physical Illness in Psychiatric Research in the Elderly. International Journal of Geriatric Psychiatry 5 161-170.

Cream J (2002) Dementia Bulletin. London: Hayward Medical Communications.

Department of Health (200Ia) Guidance on free nursing care in nursing homes. Available from: www.dh.gov.uk

Department of Health (200Ib) Survey ofselffunders in nursing homes in England. Available from: www.dh.gov.uk

Domken M, Bothwell R & McKeith I (1995) Discharge patients from a psychogeriatric inpatient unit. International Journal of Geriatric Psychiatry 10 (1)41-46.

Folstein MF, Folstein SE & McHugh PR (1975) Mini mental state: a practical method for grading the psychiatric state of patients for the physician. Journal of Psychiatric Research 12 189-198.

Holm A, Michel M, Stern GA, Hung TM, Klein T, Flaherty L & Michel S (1999) The outcomes of an inpatient treatment program for geriatric patients with dementia and dysfunctional behaviours. Gemntologist 39 (6) 668-676.

Netten A, Darton R, Bebbington A et al. (2001) Residential and nursing home care of elderly people with cognitive impairment: prevalence, mortality and costs. Aging and Mental Health 5 (1) 14-22.

Office for National Statistics (1999) Social Focus on Older People. London: The Stationery Office.

Reay AM & Brown KD (2001) Risk factor characteristics in carers who physically abuse or neglect their elderly dependants. Ageing and Mental Health 5 56-62.

Robinson RA (1965) The Organisation of a Diagnostic and Treatment Unit for the Aged in a Mental Hospital. Psychiatric Disorders in the Aged (ppl 86-205). Manchester Geigy UK for the World Psychiatric Association.

Royal College of Psychiatrists (1995) Consensus statement of the assessment and investigation of an elderly person with suspected cognitive impairment by a specialist old age psychiatry service CR49. Available from: www.rcpsych.ac.uk

Shah A, Wuntakal B, Fehler J & Sullivan P (2001) Is a dedicated specialist social worker working exclusively with psychogeriatric inpatients and an associated dedicated domiciliary care package cost-effective? International Psychogeriatrics 13 (3) 337-346.

Spector A, Thorgrimson B, Woods, B, Royan L, Davies S, Butterworth M & Orrel M (2003) Efficacy of an evidence-based cognitive stimulation therapy programme for people with dementia: Randomised controlled trial. British Journal of Psychiatry 183 248-254.

Spruytte N, Van Audenhose C & Lammertyn F (2001) Predictors of institutionalisation of cognitively impaired elderly cared for by their relatives. InternationalJournal of Geriatric Psychiatry 16 1119-1128.

WHO (1993) WorldHealth Organisation Collaborative Centre for Drug Statistics Methodology: Guidelines for daily drug doses. Oslo, Norway: World Health Organisation.

Zubenko GS, Rosen J, Sweet RA, Mulsant BH & Rifai AH (1992) Impact of psychiatric hospitalisation on behavioural complications of Alzheimer’s disease. American Journal of Psychiatry 149 (11) 1484-1491.

Sue Ball is Associate Specialist Psychiatrist, Stewart Wing, Yeatman Hospital, Dorset.

Steve Simpson is a Consultant Psychiatrist, Forston Clinic, Dorset.

Diane Beavis is a Research Nurse in Old Age Psychiatry, Forston Clinic, Dorset.

John Dyer is Nurse Specialist in Neuropsychiatry Blandford Community Hospital, Dorset.

Copyright Pavilion Publishing (Brighton) Ltd. Oct 2004

Provided by ProQuest Information and Learning Company. All rights Reserved