On busy acute wards long-stay patients may lose out to the more immediate demands of those newly admitted. The often highly stimulating environment may, on prolonged exposure, be detrimental to their recovery. Their progress may also be hindered by the lack of opportunity to rehearse appropriate social roles and practical skills (Reference Kurian, George and BallardLelliottt & Quirk, 2004). The unacceptably high number of long-stay patients stuck on acute wards has been consistently reported and the associated shortcomings in community and longer stay in-patient provision have been highlighted (Kurian et al, 1994; Reference Rowlands, Milner and WingShepherd et al, 1997; Reference Lelliott, Wing and CliffordRowlands et al, 1998). The role of rehabilitation, whether in the hospital or in another placement, has largely been neglected (Reference HollowayHolloway, 2005) during a decade that has seen national policy initiatives focus chiefly on the development of non-residential community services, notably assertive outreach and home treatment teams (Department of Health, 1999). Despite the anticipated progress in reducing bed occupancy, the limitations of these teams in meeting the needs of the most disruptive and disabled service users in the community is already apparent (Reference Commander, Sashidharan and RanaCommander et al, 2005; Reference Commander and DisanayakeCommander & Disanyake, 2006).

This 2-year longitudinal survey describes the profile of patients on acute psychiatric wards staying longer than 6 months in all acute wards in Birmingham where assertive outreach and home treatment teams are well established. It examines the reasons why they remain there for lengthy periods and compares staff appraisals of future accommodation needs with patients’ placements 2 years on.

Method

We identified individuals aged 16 years old or more and occupying an acute psychiatric bed (including intensive care) for more than 6 months (without a break of 3 weeks or more) on a nominated census day (1 June 2005); forensic and older adult services were excluded. The census was repeated 6-monthly over the following 2 years to ascertain the number of new in-patients staying in acute psychiatric wards longer than 6 months as well as to determine the final location of those recruited at the outset. The details of eligible patients were ascertained through contact with the wards and a simple pro forma was used to collect data on the in-patients at the initial census. A ward nurse familiar with the patient provided information on a range of risk behaviours derived from the Functional Analysis of Care Environments (FACE, www.facecode.com). The consultant psychiatrist gave an opinion about risk should the patient be discharged and, where relevant, the reason for any delay. In addition, both the nurse and psychiatrist identified whether the patient was appropriately placed on the ward, and if not, where they considered the patient more suitably placed (based on categories derived from the Community Placement Questionnaire; Reference CliffordClifford, 1993).

We compared our findings with the national audit of new long-stay psychiatric patients (Reference Lelliott and QuirkLelliott et al, 1994) undertaken in 1992, prior to recent reforms in mental health services. In contrast to the present study, the audit covered an age range of 18–64 years old, included individuals in any residential place fully funded by the National Health Service (NHS) and restricted entry to those who had been there less than 3 years.

Results

At the outset, 38 in-patients met the study criteria, occupying 18% of the 208 acute beds. The mean length of stay was 15 months (s.d.=9; range 6–40) compared with 16 months (s.d.=9) in the national audit (Reference Lelliott and QuirkLelliott et al, 1994). The number of patients staying longer than 6 months was 32, 38, 33 and 42 at each subsequent census. Of these, 22 (69% of the total), 14 (37%), 18 (55%) and 28 (67%) respectively were ‘new’ patients whose stay was now over 6-months duration. The mean age of patients in the study was 43 years old (s.d.=13; range 22–67) compared with 42 years old (s.d.=13) in the national audit (Reference Lelliott and QuirkLelliott et al, 1994). Nearly two-thirds had never been married and none were currently in work (Table 1). Schizophrenia was the most frequent diagnosis.

Table 1. Demographic, clinical and problem behaviour profile of long-stay in-patients in acute psychiatric wards

Variable	Birmingham, 2005 (n=38) n (%)	National audit, 1992 (n=523) %
Demographic
Male	21 (55)	58
White British	24 (63)	90
Never married	25 (66)	63
Living in house/flat/bedsit	23 (58)	68
Living alone	17 (45)	25
In work	0	6
Clinical
Primary diagnosis: schizophrenia	29 (76)	59
First admission	3 (8)	11
Contact with services > 5 years	27 (71)	0
Detained compulsorily at census	26 (68)	29
Detained compulsorily during admission	33 (87)	44
Problem behaviour1
Physical harm to others	6 (16)	19
Threats/intimidation	13 (34)	42
Inappropriate sexual behaviour	3 (8)	2
Incontinence	4 (11)	0
Fire setting/risk	2 (5)	2
Suicide risk/deliberate self-harm	2 (5)	20
Substance misuse	4 (11)	18

The majority of patients had lengthy previous involvement with mental health services and had been detained compulsorily during this admission. Fifteen patients were under the supervision of an assertive outreach team; one, an early intervention team; and three, a rehabilitation and recovery team. The remaining patients were under the care of a primary care liaison or community mental health team. They all had a substantial level of problem behaviours, notably violence and threatening behaviour (Table 1). Five patients (13%) had a conviction for a violent or sexual offence, 16 (42%) had a lifetime history of physical violence to others and 6 (16%) of setting fire; 1 patient had previously been in a special hospital. The proportion of patients considered by psychiatrists a moderate to high risk for neglect (n=25; 66%), violence (n=11; 29%) and suicide/deliberate self-harm (n=11; 29%) should they be discharged compares with the 42%, 30% and 20% respectively identified in the national audit (Reference Lelliott and QuirkLelliott et al, 1994).

Psychiatrists considered that 6 patients (16%) only initially needed to remain on an acute ward and a further 14 (36%) were deemed to require a longer in-patient stay. At baseline, there was a consensus between psychiatrists and nursing staff about the most appropriate placement for only 18 out of 38 patients (kappa=0.34 – fair; Table 2). Of the 32 patients considered by their psychiatrist to no longer require acute inpatient care, the majority (n=24; 75%) had as yet no accommodation to go to. A further 2 patients were waiting for a forensic assessment, 5 had been accepted for a place but lacked funding, and the remaining patient was waiting for adaptations to be completed on her house. The location of patients at 2-year follow up is shown in Table 2. Of the 5 patients who remained on an acute ward, 4 had been there for the duration of the study and 1 had left but had then been readmitted from their supported accommodation. One other individual was readmitted but subsequently returned to their care home. For those who did move on and at the outset were deemed to be inappropriately placed, both psychiatrist and nurses initial recommendations showed only moderate agreement with the patients final placement (kappa=0.41 and 0.43 respectively).

Table 2. Initial staff recommendations and final placement at 2 years

	Recommendation at outset (n=38)		Placement at 2 years (n=38) n (%)
Variable	Ward nurse, n (%)	Consultant psychiatrist, n (%)
Acute ward	7 (18)	6 (16)	5 (13)
Long-term secure unit	5 (13)	7 (18)	31 (8)
In-patient rehabilitation	5 (13)	7 (18)	7 (18)
Nursing/care home	13 (34)	13 (34)	11 (29)
Supported housing2	5 (13)	2 (5)	7 (18)
Independent living	3 (8)	3 (8)	53 (13)

Discussion

The profile of long-stay patients we arrived at needs to be understood within its local context and the overall pattern of services. Birmingham has a population of 1 006 500 with 30% from Black and minority ethnic groups (Birmingham City Council, 2006). The city is ranked the 15th most deprived local authority in England (Index of Multiple Deprivation, 2004). The number of available acute beds is lower than projected figures based on the National Beds Inquiry (Department of Health, 2000), while the number of rehabilitation and low secure beds (n=74) is in line with estimates in a recent review of the government's mental health policies (Reference Boardman and ParsonageBoardman & Parsonage, 2007).

The long-stay patients in this study consistently occupied around a fifth of all acute beds, which is comparable with figures reported in surveys (Kurian et al, 1994; Reference Lelliott, Wing and CliffordRowlands et al, 1998) undertaken prior to the reform agenda set out in the National Service Framework for Mental Health (Department of Health, 1999). Although encouragement can be taken from the finding that most patients had moved on from an acute ward after 2 years, this needs to be set against the fact that at any one time four out of five were considered inappropriately placed and the average length of stay was well over a year.

The enduring evidence of a substantial number of long-stay patients stuck on acute wards is of particular concern given the Department of Health's focus on delayed transfers of care reflected in an extension, in April 2006, of the reporting of such cases (via weekly situation reports) to include all mental health NHS trusts. Delayed transfers/discharges have been further highlighted in the Ten High Impact Changes for Mental Health Services (Care Services Improvement Partnership, 2006) and a good practice toolkit aimed at improving discharge from in-patient units (Care Services Improvement Partnership, 2007). These documents address the importance of local data collection and appraisal of the reasons for any delays in conjunction with problem-solving sessions involving senior managers and clinicians. They also stress the need for early discharge planning and the potential role of discharge facilitators as well as the critical contribution of home treatment services. Yet in Birmingham, despite seeking to optimise performance in the ways suggested, at any one time the majority of long-stay patients were considered inappropriately placed.

Shepherd et al (Reference Rowlands, Milner and Wing1997) found that almost two-thirds of long-stay patients on acute psychiatric wards required a specialist rehabilitation placement or a setting with higher supervision. Likewise, in Birmingham around a half of all in-patients were initially deemed to need ongoing treatment in hospital and just over a third remained there 2 years on, indicating that in-patient provision is crucial to resolving the issue of long-stay patients on acute wards. A shortage of medium-term beds was identified in the national audit (Reference Lelliott and QuirkLelliott et al, 1994) but, as in our study, it is uncertain whether this reflects a deficit in capacity or arises from lack of accessibility due to the delayed discharge of patients awaiting community placement. Certainly, where there were community service recommendations, and eventually placements, by far the majority involved nursing and care homes or supported accommodation with 24-h staffing, challenging a notion that these individuals are likely to be supported in their own homes albeit with the supervision of newly established and well-resourced community teams. Alongside the noteworthy omission of longer stay in-patient settings from national policy initiatives (Reference HollowayHolloway, 2005), there has been a dearth of the sophisticated cross-agency commissioning necessary to deliver a comprehensive range of accommodation for people with mental health problems in the community (Reference Rowlands, Milner and WingShepherd et al, 1997).

There was only fair agreement between psychiatrists’ and nurses’ opinions regarding the most appropriate placements for patients. Likewise, for those considered in a position to move on, there was only moderate concordance between the recommendations and the patients’ final location. It may be that changing circumstances and presentation partially accounts for that as may the constraints of working with a limited range of disposal options. However, it may also be relevant that half of the Birmingham in-patients were under the supervision of primary care liaison or community mental health teams and so involved staff who may have little expertise in assessing and placing such individuals, or indeed in advocating for the necessary funding. Specialist rehabilitation teams, which for many years worked to resettle individuals from long-stay wards into the community, have often had their profile changed to assertive outreach teams or simply disbanded in recent years. The government's failure to endorse their role in its new service model (Department of Health, 1999) may now need to be reconsidered (Reference HollowayHolloway, 2005).

Declaration of interest

None.