Originally published in Psychiatric Bulletin 1991; 15: 735-6


D B Double and T I Wong



All 438 inpatients from the rehabilitation and long-stay service at Middlewood hospital, Sheffield, in 1982 were traced eight years later. None were homeless. Only 23% of the original sample were in non-hospital settings, 35% having died and 42% remaining within the service or in other wards or hospitals. Of those discharged, the majority were in residential or group homes with only 26% living on their own or with relatives. One person was in prison. These findings may raise less concern about the rundown of the traditional psychiatric hospital than studies of the amount of mental illness among the homeless.

Summary Table
Whereabouts of patients after 8 years





In hospital



In rehabilitation and longstay





In other wards or hospitals










Death as inpatients





Death after resettlement







Nonhospital setting



Residential care for the elderly





  Local authority















  Sheffield hostel for homeless men





  Private psychiatric homes





  Group homes





Own address or with relatives


















The rundown of the traditional psychiatric hospital tends to be blamed for the apparent high level of mental illness among the destitute (Weller 1989). However, in York, most long-stay patients discharged into the community were still living in fairly institutional settings, such as local authority and private homes, or group flats and lodgings (Jones et al 1986). Moreover, the first results from the Team for the Assessment of Psychiatric Services (TAPS) studying the outcome of long-term patients discharged as part of the closure plan for Friern and Claybury hospitals found only three out of 161 patients impossible to trace after one year, and they had a former lifestyle as vagrants (Dayson 1989). Consistent with these findings, a survey of discharged long-stay patients by Cheltenham and District and Gloucester Health Authorities found them generally to be well cared for.

Similar optimistic conclusions that most long-stay patients are functioning well in the community have been reported from New South Wales (Andrews et al 1990). None of 208 patients discharged into supported accommodations had drifted to a refuge for the homeless at 21 months follow-up. In a separate study, 21% of men in a shelter for the homeless were diagnosed as schizophrenic but only three of the 22 schizophrenic men had had a prolonged admission to a psychiatric hospital (Teesson and Buhrich 1990). Rather than blame deinstitutionalization, it is suggested that housing policy has led to low cost accommodation becoming difficult to find, increasing the use of the refuge.

Evidence bearing on the issue of what happens to patients from long-stay psychiatric wards in Britain is scarce. The results of a particularly well-planned resettlement programme may not be representative of the outcome from general policies of discharge from all psychiatric hospitals. The present study was undertaken in Sheffield to find the whereabouts of patients from long-stay wards after 8 years, a longer period than usually studied previously. The expectation was that difficulties in tracing patients could be overcome with sufficient time and effort.


The study

Middlewood Hospital is a traditional psychiatric hospital serving the Sheffield Health District.

Before computerization of medical records an occasional census of inpatients - about every two years - was undertaken to produce the nominal roll of patients. Information from the census taken on 12th January 1982 had been kept intact and formed the starting point for follow-up of patients on wards designated as long-stay. Psychogeriatric wards were not included.

A clear line of continuity from these long-stay wards to the rehabilitation and long-stay service at the time of follow-up in 1990 could be traced. The number of wards in the service has been reduced from 17 to 11 (including two smaller rehabilitation hostel wards). Only four of these wards were used for long-stay patients in both 1982 and 1990.

The whereabouts of all patients was ascertained as at 12 January 1990, ie. exactly 8 years later. Obviously certain patients were easier to trace than others. Some were still inpatients, others had died in hospital or been discharged. Information about those discharged was often the most difficult to obtain but persistence in following leads from staff contacts and medical records enabled all patients to be found.



The number of inpatients in the rehabilitation and long-stay service at the time of the census in 1982 was 438. At follow-up eight years later the number of inpatients had been reduced to 232 (a 47% reduction in inpatient numbers). Of these 232 patients, 171 (74%) had been inpatients eight years previously and 61 (26%) had been newly admitted.

The average age of patients in the sample was 62 years, which is less than the average age of 65 years of patients in the rehabilitation and long-stay service at the time of follow-up.

The whereabouts of all patients was traced. The highest proportion (42%) were still in hospital, not necessarily Middlewood; 35% had died, although not necessarily in hospital; and 23% were in non-hospital settings.

Of those in hospital, 171 were still in the rehabilitation and long-stay services. Fourteen were in other wards or hospitals: two in psychogeriatric wards in Middlewood; one was on an acute admission ward at the time of follow-up; four were in secure facilities (two at St Andrews hospital, Northampton and two in Rampton); one was in a psychiatric hospital in Carlisle; and six were in hospital in Rotherham (5 on psychogeriatric wards, one on a rehabilitation ward), whose health authority had agreed to take res ponsibility for these long-stay patients after health service reorganization.

Of the 100 patients living in non-hospital settings, the majority were in residential or group homes. Only 26 of those outside hospital (6% of original sample of 438 patients) were living at their own address or with relatives. The greatest number (36) were in residential homes for the elderly (Part III), fifteen of whom had been resettled in the few months before the survey as part of a project transferring funding to the local authority. Twenty were in private residential care for the mentally ill. The others were in private residential care for the elderly (1), local authority homes for the mentally ill (4), Sheffield hostel for homeless men (2) and group homes (10). One person was in prison but none were homeless.


This study has demonstrated that with perseverance discharged patients from long-stay psychiatric wards can be traced, at least in Sheffield. The situation may be different in larger cities such as London, for example. Only two patients from the sample had moved outside the local area.

No patients were homeless at follow-up. This finding may suggest some caution is needed in interpreting studies that suggest a high level of mental illness among the homeless as implying that the rundown of the psychiatric hospital should be slowed. A poorly reasoned polemic from Weller (1989) can find its way into the prestigious columns of Nature, revealing the current bias against deinstitutionalization in Britain. The present study gives some picture of what actually happens to patients after discharge.

It is important to note the high rate of death among this patient group. The long-stay population in psychiatric hospitals is elderly. The rehabilitation and long-stay service in Sheffield at follow-up eight years later has a slightly increased average age. If there are no further admissions a significant attrition rate through death can continue to be expected. It is important to remember that the decline in the number of patients in psychiatric hospitals is not only due to discharge but to death expected because of the age distribution of the long-stay population.

Of those discharged from hospital the majority were still living in fairly institutional surroundings. A significant proportion of this care in Sheffield is provided in private residential homes for the mentally ill. The person who was in prison was there for attempted murder and was not regarded as mentally ill.

The quality of life issue for those out of hospital compared with those remaining in hospital has not been answered by this study. Nor could it easily be answered because of the difficulty measuring quality of life, obtaining adequate control groups and the lack of randomization. Those who are resettled are selected to some extent, making the interpretation of any comparison with those remaining in hospital difficult.

If problems are being caused by the rundown of the psychiatric hospital, it may be because of the limited facilities for the new long-stay, rather than because of the resettlement of the older long-stay patients. Any argument about community care and the preservation of the traditional psychiatric hospital should distinguish between the needs of people who can no longer get in to the psychiatric system, as opposed to the needs of the people who may just have been thrown out. However, a substantial proportion of the rehabilitation and long-stay wards in this study at follow-up were occupied by newly admitted patients over the eight years since the census.

Government policy is moving in the direction of smaller units for continuous care of psychiatric patients. The relatively small group of patients placed outside hospital over eight years could suggest there is scope for more, not less, resettlement of patients from the traditional psychiatric hospital.


Contrary to the popular conception of psychiatric hospitals discharging patients "onto the streets", none of the 438 patients on the long-stay wards in Sheffield 8 years ago were found to be homeless at follow-up. Hospital numbers have been reduced significantly by death as well as discharge. Of those discharged, most remain in fairly institutional settings, with only 6% of the total sample living on their own or with relatives. One person was in prison. Blaming deinstitutionalization for an apparent high level of mental illness among the destitute may not be appropriate. To do so politically may be inept, as it could misdirect attention away from housing policy and the need for low cost accommodation.


Andrews G, Teesson M, Stewart G and Hoult J (1990) Follow-up of community placement of the chronic mentally ill in New South Wales. Hospital and Community Psychiatry 41 184-188
Dayson D (1989) The administrative outcome of the long-term mentally ill in the community. In Moving long-stay psychiatric patients into the community: first results (TAPS Fourth Annual Conference). North East Thames Regional Health Authority
Jones K, Robinson M and Golightley M (1986) Long-term psychiatric patients in the community. British Journal of Psychiatry 149 537-540
Teesson M and Buhrich N (1990) Prevalence of schizophrenia in a refuge for homeless men: a five year follow-up. Psychiatric Bulletin 14 597-600
Weller M P I (1989) Mental illness - who cares? Nature 339. 249-252

 Return to publications

 Return to Community and Social Psychiatry

 Visit my homepage