Characteristics and outcomes of patients admitted to a psychiatric intensive care unit in a medium secure unit

Abstract

Aims and Method

There have been no reports on psychiatric intensive care units (PICUs) in medium secure psychiatric facilities. Using case files, we retrospectively examined the characteristics and outcomes of 73 patients who were admitted to a PICU in a medium secure unit between 1 July 1994 and 30 April 1998.

Results

The PICU population was predominantly male, suffering from illness and detained under Part III of the Mental Health Act, 1983. Although the mean length of stay was 75 days, the majority were ultimately transferred to less intensive nursing environments and only nine required transfer to maximum security. In 10% of cases PICU admission was owing to lack of appropriate facilities elsewhere.

Clinical Implications

Although the PICU was intended as a crisis facility for the management of challenging behaviours, its function was affected by the lack of clear admission and discharge criteria and appropriate facilities for patients with diverse mental, physical and security needs.

Psychiatric intensive care units (PICUs) were designed to create a safe and controlled environment for the management of acutely disturbed psychiatric patients on a short-term basis, with high staffing levels and a limited number of beds. Admission and discharge criteria are usually clearly defined and the majority have locked doors (Michalon & Richman, 1990; Hyde & Harrower-Wilson, 1996). The average length of stay ranges from 2.6 days (Hafner et al, 1989) to 30 days (Citrome et al, 1994), although Rachlin (1973) reported that 20% of his patients stayed over 2 months.

The majority of PICUs reported in the literature provide care and treatment for non-offender patients with mental illness who cannot be managed in open wards. In the UK, intensive care for mentally disordered offenders is provided by the secure psychiatric services. Problems in the movement of patients through different levels of security, however, has led to the development of PICUs in some medium secure facilities. As far as we are aware this is the first report on the characteristics and outcomes of a cohort admitted to a PICU in a medium secure unit (MSU) in Britain.

The study

Using a precoded form, data were extracted from case notes on the characteristics and outcomes of 73 patients who were admitted to a PICU in a MSU in the North-West of England between its opening in July 1994 and April 1998. The five bed PICU, which was set up and run by nursing staff, was an annex to one of the three wards in the 60-bed MSU — Edenfield Centre. Unlike other ward areas, the PICU did not have a designated seclusion room. Staff to patient ratios were a minimum of 1:1, unlike other ward areas where staff to patient ratios were much lower. The unit was intended as a crisis facility to deal with patients with challenging behaviours who could not be managed on less staffed wards, but were unlikely to require maximum security because of the likely brevity of their disturbed behaviour. Assessments for transfer to and from the PICU were made by nurses from the relevant areas, and patients requiring PICU care at its inception were considered unsuitable for unescorted leave. The unit did not have explicit admission or discharge criteria in terms of patient characteristics, but was seen solely as a facility for managing disturbed behaviour.

Data were collected on the demographic details, medical/psychiatric history and criminal history of all subjects. Details pertaining to the current admission, for example, presenting problems (index offence), management problems and outcome following the PICU admission, were examined. Psychiatric diagnoses were based on DSM-III-R criteria (American Psychiatric Association, 1987).

Findings

During the study period 73 patients (one-third of the MSU population at that time) were admitted to the PICU. These patients accounted for 78 PICU admission episodes. Five patients were admitted on two occasions. The sample characteristics are shown in Table 1. The majority were male, Caucasian, single and referred from prison. The mean age of the sample was 33.2 years (s.d.=9.2). Most patients (48, 66%) were detained under Part III of the Mental Health Act (MHA) 1983 and the legal category of mental illness (65, 89%). Twenty (27%) patients were on hospital orders with restrictions (Section 37/41, MHA 1983).

Table 1. Characteristics of the sample

Demographic	n=73 (%)
Gender (male)	59 (81)
Caucasian	37 (61)
Afro-Caribbean	16 (27)
Asian	8 (13)
Married/cohabiting	13 (18)
Source of referral	n=73 (%)
Prison	37 (51)
District general hospital	18 (25)
Community	17 (23)
Maximum security	1 (1)
Offences	n=73 (%)
Assault	23 (32)
Arson	9 (12)
Homicide	8 (11)
Acquisitive	8 (11)
Sex offences	6 (8)
Motoring	3 (4)
Primary diagnosis	n=73 (%)
Psychosis	56 (77)
Affective disorder	10 (14)
Personality disorder	5 (7)
Neurosis	1 (1)
Asperger's syndrome	1 (1)
Secondary diagnosis	n=34 (%)
Brief reactive psychosis	4 (11)
Affective disorder	8 (21)
Neurosis	2 (5)
Personality disorder	9 (24)
Substance abuse	27 (71)
Dementia	1 (3)
Learning difficulties	1 (3)
Eating disorder	1 (3)
Physical disability	n=21 (%)
Respiratory problem	6 (29)
Cardiovascular	5 (24)
Epilepsy	4 (19)
Musculoskeletal	4 (19)
Skin/wound problems	3 (14)
Brain damage	3 (14)

Psychiatric/medical profiles

The majority of patients had a primary diagnosis of psychotic illness (56, 77%). Secondary diagnoses were common (38, 52%), with substance misuse and personality disorder predominating (see Table 1). Over one-third of the sample had concomitant physical illness, particularly cardiac and pulmonary disease (see Table 1). Three patients were pregnant during their stay on the PICU.

Criminal history

Fifty-seven patients (78%) had been charged with or convicted of at least one criminal offence. The majority were for violent offences, including homicide (Table 1).

Reason for admission to the PICU

Admission to the PICU usually followed a deterioration in mental state or behaviour on another ward. Thirty-four (47%) admissions were owing to threatened/actual assault on others, 10 (14%) because of self-harm, six (8%) because of threatened/actual arson and five (7%) because of socially unacceptable (sexual) behaviour.

Seven admissions (10%) were not related to behaviour or mental state abnormalities. In four cases this occurred because of a lack of beds elsewhere on the unit — particularly related to special hospital rehabilitation cases and elderly patients with cardiac problems and in three other cases patients were admitted because of physical care needs, particularly pregnancy.

Outcomes

Incidents during PICU admission episode

Thirty-seven (51%) patients were involved in at least one incident during their PICU stay. The mean number of incidents per patient was 2.74 (s.d.=6.86; range 0-35) with a mean of 1.16 incidents per patient per month of the PICU admission episode (s.d.=2.19). Incidents included: property damage (16, 43% of patients involved in incidents); assaults on staff (15, 41%) and patients (13, 35%); threatened assault on staff (14, 38%) and patients (3, 8%); breaches of security (10, 27%); verbal aggression (8, 22%); self-harm (7, 19%); and threatened arson (1, 3%).

Twenty-one of these patients (57%) were involved in incidents requiring the use of restraint. Incidents resulted in injury to other patients in five cases (14%), to the perpetrator in 14 (38%) cases and to staff in 11 (30%) cases.

Length of stay

The mean time spent on the PICU during an admission was 75 days (s.d.=106, range 2-622 days). PICU length of stay did not significantly correlate with MSU length of stay, but did correlate positively with frequency of incidents (r=0.44, n=73, P<0.001) and mean monthly incident rate (r=0.91, n=73, P<0.001).

Placement after leaving the PICU

Of 73 admission episodes that had terminated, that is, the patients were no longer resident on the PICU, 51 (65%) had moved to other clinical areas within the unit. Transfer was first agreed by the clinical team and effected via ‘day-time visits’ to the ‘receiving ward’ until non-PICU staff were satisfied, transfer was appropriate and a bed was available. Surprisingly, four cases (5%) were discharged directly to their family home, two (3%) to hostel accommodation and five (6%) to district psychiatric hospitals. These cases were all non-restricted patients who presented with acute psychosis-related behavioural disturbance that responded well to treatment. One patient (1%) admitted from prison was transferred to his catchment area MSU following improvement in his mental state. Nine (12%) cases required transfer to maximum security owing to escalating violence in the PICU and three (4%) sentenced prisoners were returned to prisons following stabilisation on medication.

Discussion

As far as we are aware this is the only study reporting on a PICU in a forensic setting, perhaps because most MSUs have sufficient staffing and security measures to deal with difficult or challenging patients and these units are a rarity.

The majority of admissions to the PICU were as a result of a significant deterioration in mental state/behaviour and most cases were eventually successfully transferred to less secure environments on the unit. On the surface this suggests that for the most part the unit operated as a crisis facility for patients with acute disturbance. However, our finding that one-third of MSU patients were admitted to the PICU and relatively few (five) cases were repeat admissions by the same individuals points to the PICU being used in an unintended manner. This may well reflect the lack of clear admission criteria at its inception. The mean length of stay was considerably longer than is reported in non-offender PICUs and could be explained by the forensic nature of the population and the observed relationship between incident rates and length of stay.

Contrary to the admission policy for the unit we note that 10% of cases (particularly female patients and those with physical health care needs) were placed on the PICU because it provided a safe environment, that is, high levels of staffing and observation. This phenomenon probably reflects the lack of appropriate facilities elsewhere (in less secure environments) on the MSU. At the time of study there were no single sex wards or suitably staffed ward areas for those with physical health needs on the unit and our findings highlight the need for such specialist services.

In terms of the immediate outcomes following discharge from the PICU, the majority of cases were relocated on the unit without much difficulty when beds became available — the latter being the most significant rate-limiting factor. Perhaps because of the bed crises inter- or intra-disciplinary arguments were rare and it was accepted that admission and discharge from the PICU could not be based on strict operational criteria. Despite this we were surprised by the number of cases (11) that were discharged directly from the PICU to the community and district services without pre-discharge relocation to less intensively nursed areas on the MSU. In all of these cases patients had shown striking improvements in mental state and behaviour with treatment and discharge directly from the PICU clearly occurred because of a lack of available beds elsewhere on the unit. These patients were also (contrary to PICU policy) having unescorted leave in the community, which highlights the difficulties of running the PICU in a strict sense when there are insufficient beds to meet a variety of patients' needs.

In conclusion, this study demonstrates the difficulties in running a PICU in a true sense in a MSU when there are no clear criteria for admission and discharge and there is a shortage of beds on the MSU as a whole. Since this study the PICU has closed, the ward is now a female only facility, a physical health care nurse has been appointed and a pre-discharge hostel ward opened.