The Royal College of Psychiatrists college report on individuals who enter old age with a psychosis such as chronic schizophrenia (Royal College of Psychiatrists, 2002) highlighted an unmet need in this area and recommended local surveys to address this. One, undertaken in a Scottish service (Reference McNulty, Duncan and SempleMcNulty et al, 2003), found high levels of disability and significant unmet need among people with schizophrenia aged over 65 years old. These individuals are often admitted to care homes but it is not known in what way they differ from those who continue to reside in their own homes. Factors known to contribute to morbidity in late-life psychosis include level of psychiatric symptomatology, cognition, medical morbidity and disablement (Reference Jolley, Kosky and HollowayJolley et al, 2004). In a single catchment area we compared these domains in two groups of older people with either chronic schizophrenia or delusional disorder, one in care homes and the other living independently in the community. The hypothesis we wanted to test was whether or not those in care homes would have greater severity of psychiatric symptoms, more medical morbidity and poorer cognition than those living independently at home.

Method

Inclusion criteria were: aged over 65 years old, under the care of one psychiatrist (R.B.) who had a defined catchment area, a primary diagnosis of schizophrenia or delusional disorder according to ICD-10 criteria (World Health Organization, 1993) and living in the community. People excluded from our study were those with a primary diagnosis of dementia or affective disorder, in-patients of acute in-patient psychiatric units and those who refused to take part. Mental capacity and consent was undertaken by R.B. Where a person lacked the capacity to give informed consent, their caregivers and/or relatives were approached. The research was approved by the local research ethics committee and Manchester Mental Health and Social Care Trust was the sponsor of the study.

We used the Positive and Negative Symptom Score (PANSS), which assesses positive, negative and general psychotic symptomatology on a 30-point scale (Reference Kay, Fiszbein and OplerKay et al, 1987). Cognition was assessed using the Mini Mental State Examination (MMSE; Reference Folstein, Folstein and McHughFolstein et al, 1975) and dementia was diagnosed based on the ICD-10 criteria for dementia (World Health Organization, 1993). The Burvill Physical Illness Scale was used to quantify morbidity from physical illness (Reference Burvill, Mowry and HallBurvill et al, 1990) and the Instrumental Activities of Daily Living Scale (IADL; Lawton et al, 1969) was used to assess self-maintenance (6 categories, rated 0 or 1) and more complex daily-life tasks (8 categories, rated 0 or 1). Interviews were conducted at the participant's residence in compliance with the Trust's safe visiting policy. Sources of information included the participant, caregivers and case notes.

The data was analysed using SPSS version 11.5 for Windows, with χ²-test for categorical data and t-test for continuous data.

Results

The inclusion criteria were met by 52 individuals. Four were excluded because they died or left the area before the study commenced, two refused to take part and one was diagnosed as having psychotic depression rather than schizophrenia. Forty-five patients were therefore included, 23 living independently in their own homes and 22 living in care homes. There were three specialist homes for the elderly mentally infirm (n=14) and three residential homes (non-specialist, n=8).

Mean age was 74.3 years old, with 23 men and 22 women and with no significant age or gender differences between the groups (Table 1). The duration of illness ranged from 2 years to 68 years, with mean of 22 years (16 years in own home-living group and 30 years in residential home-living group, borderline significance). In 78% of the participants, onset was at age less than 65 years old, with no significant difference between groups. Three times as many (72% v. 28%, P<0.01) of those living in care homes than those living in their own homes were prescribed more than one psychotropic drug, but although higher by 22 mg, the calculated mean chlorpromazine equivalent dosage (using a formula by Reference WoodsWoods, 2003) did not differ significantly between groups.

Table 1. Differences between the two groups

	Living in own home (n=23)	Living in a care home (n=22)
Mean age, years	74.3	74.6
Gender
Male	13	10
Female	10	12
Duration of psychosis, years	16	30
PANSS total score	21.00	38.9**
Positive	7.3	9.2
Negative	5.6	14.2**
General	7.9	16.4**
MMSE	25.2	17.5*
Medical morbidity (Burvill; the higher, the worse)	4.4	5.6
Mean daily chlorpromazine equivalents, mg	152.7	174.4
IADL total score (the lower, the worse)	11.01	2.82**
Subscale (day to day)	5.481	0.77**
Physical maintenance	5.521	2.00**
Dementia (clinical diagnosis), n	22	3

Those in care homes had higher total PANSS scores but not on the positive symptom sub-scale. Their daily-life activities were reduced across both basic and complex sub-scales. On the Burvill scale, medical morbidity did not differ between groups. Using the MMSE, cognition was significantly lower in the care home group but the overall rate of dementia defined clinically (ICD-10) did not differ.

Discussion

The strengths of the study are its delineation of all individuals from a single catchment area, low refusal rate and groups well-matched by age. Its limitations include its small size and difficulty in generalising to other services. With these limitations, this study suggests markedly higher morbidity and disablement among older people with chronic schizophrenia living in care homes compared with those living in their own homes. However, the original hypothesis was only partly supported. Individuals in care homes had higher levels of psychiatric symptomatology but this was largely from negative or deficit symptoms rather than positive ones. Their higher scores on the ‘general’ sub-scale of the PANSS may have been owing to poorer cognition as this sub-scale includes items such as disorientation and confusion. People in care homes had far greater impairment in their daily-life activities, which included basic tasks such as bathing, eating, toileting and food preparation as well as more complex daily tasks. Four participants scored 0 on the IADL scale, suggesting total reliance on carers for all of their needs. This disability was not accounted for by those in care homes having more medical morbidity as we had hypothesised. Rather, the combination of deficit symptoms and poor cognition appeared to be responsible. Poor cognition with or without dementia is reported in older people with schizophrenia (Reference Harvey, Sterman and RichardHarvey et al, 1999). Thus the disablement of care home residents seems likely to be intrinsic to their psychotic disorder, although the effects of long-term institutionalisation cannot be ruled out as an additional factor.

There exists no national service framework that would adequately address the needs of older people with psychosis (Royal College of Psychiatrists, 2002). The two probably most relevant here are not directly concerned with this group - the National Service Framework for Older People (Department of Health, 2001) focused on dementia and depression, and the National Service Framework for Mental Health (Department of Health, 1999) on adults of working age. Considering the complex handicaps arising from their disorder, detailed care planning and review by specialist psychiatric services are imperative for older people with schizophrenia residing in care homes. Despite this, few of those in care homes that took part in this study received input from any member of the Old Age Psychiatry Community Mental Health Team other than the consultant, whereas a majority of those in their own homes received some input from those services. It is not known whether this is typical of other services or not, but it warrants further research. What also need to be examined are the pathways of entry into care homes among this population group.

Declaration of interest

None.