Emergency work at an Inner London psychiatric hospital: a study of assessments made over six months

Following the introduction of a 24 hour Emergency Clinic at the Maudsley Hospital in 1952, psychiatric hospitals across the UK have come to offer a similar but often smaller service. UK literature describing such services includes only one substantial series of 2390 assessments over six months in Camberwell (Lim, 1983). Lim reported the activity of a large and well-staffed clinic with access to facilities for "guest" admissions and out


Background
The catchment hospital is a 60-bedded psychiatric unit with 28 acute adult beds. The catchment area, the postal district of London SW1, includes a major travel terminus and several famous public buildings.
The resident population numbers only 28,000 (1991 census) but there is a large migrant population of tourists and workers. The area contains the second largest population of homeless persons in London (Black et al, 1991).
A number of community services are based in the catchment hospital. There is a multidisciplinary team caring for the homeless mentally ill, a CPN service and a crisis assessment team which meets twice a day to provide immediate domiciliary assessments of mentally ill persons by open referral. Independent substance misuse services and HIV clinics also operate nearby. A liaison psychiatry team operates during office hours at the nearby general hospital. Each of these services might be expected to contain or avert crises and so lessen the demand for emergency assessments.

The emergency service
The service is provided by a duty doctor, an SHO or registrar on site. The duty doctor makes all initial assessments, backed up by clerical staff, hospital social workers and a senior registrar or a consultant during the day. Out of hours, a duty senior registrar and consultant and an approved social worker are available by telephone.
The service is open to referrers, including casual attenders. Cover is also provided for the casualty department of the nearby general hospital and for the wards outside office hours.

The study
All assessments made between 1Juneand 1December 1991were retrospectively reviewed. Information was obtained from a duty log in which a brief summary of each assessment was recorded.
Errors in diagnosis are likely where inexperienced junior staff make single assessments. We attempted to minimise this effect by adopting broad diagnostic criteria based on ICD-9 (see Table I).
Where more than one diagnosis was recorded, the acute clinical condition responsible for attendance was chosen. Otherwise, a "hierarchical" approach was adopted, psychotic conditions taking precedence over neurotic conditions and substance misuse, each taking precedence over a diagnosis of personality disorder.

Data analysis
The original data were not recorded with future complex analysis in mind. Accordingly, simple percentages are offered. Where trends of interest appeared, the x2test was used.

Findings
Four hundred and eighty-nine clients made a total of 626 visits. Men predominated, requiring 58% of the assessments. Unexpected repeat visits were made by a total of 55 clients who made a total of 192 visits (range 2 to 20, mean 3.5). Fifty-nine per cent of assessments took place out of hours. There was no evidence that men or women attended at particular times of the day. Referral source. Thirty-six per cent referred them selves. A quarter of all referrals came from hospital wards or casualty. CPNs referred only 2% of attenders and GPs referred less than 5% of men and 10% of women. Police referrals, by contrast, were frequent at 15% of the total. There was a statistically significant tendency for police to bring females to the hospital informally as opposed to formal detention of males (0.02 < P < 0.01). Diagnosis. Table I shows the diagnoses recorded as a percentage of the number of different clients and as a percentage of all repeat visits. Outcome. Regarding outcome, 17% of males and 22% of females were admitted, totalling 165 admissions, a third of these under a section. Over a half of attenders were offered psychiatric out-patient follow-up, mainly at the same hospital. Referrals back to GP care constituted a minute fraction of overall outcome. Eleven per cent of patients absconded or took their own discharge.

Comments
The emergency clinic contributed 165out of 267 total admissions in the study period, i.e. 62% of the total. This compares with 50% of Camberwell admissions (Lim, 1983).
At 6.9%, the rate of GP referrals to the service was dramatically lower than the referral rate outside London which may be as high as 65 to 75% (Huckle & Nolan, 1992), similar to the figure recorded in Camberwell in 1965(Brothwood, 1965. It is com parable to the figure of 4.1% reported by Lim in 1983 85 and confirms a decline in emergency referrals by GPs to Inner London psychiatric hospitals.
It is apparent from our study that much responsi bility falls onto the police, who provided 14.7% of referrals, and the local Accident and Emergency department which referred 15.5%. The fraction of patients referred back to the care of their GP was only 0.6%. While our study did not record the number of clients who were actually registered with a GP, this figure suggests extremely poor awareness of the mental health capabilities of primary care ser vices on the part of psychiatrists as well as that of patients. This area requires further study.
Police referrals included 66 Section 136 assess ments, an extremely high figure. Preliminary analysis confirms that over 90% of these referrals were judged appropriate by the receiving doctor.
As for diagnosis, higher rates of psychosis and lower rates of neurosis were found than in compar able studies (Lim, 1983, Brothwood, 1965. Rates of substance misuse were broadly comparable to the above reports. The single most common diagnosis made was "personality disorder". In our study, 54 clients, constituting only 11.2% of the total, made a total of 122 visits, i.e. 20% of the total work. As the diagnosis was strictly one of exclusion of other disorders and permitted no consideration of comorbidity, we consider these figures to be an under estimate. Katschnig (1983) has drawn attention to the importance of the social environment in deter mining the individual's response to a "psychic crisis". More work is necessary to investigate the environment and resources of a group who are dis proportionately large consumers of psychiatric services.

Conclusions
We have demonstrated a considerable demand for emergency psychiatric services in a deprived area of Inner London. Despite a variety of communityoriented services, the clinic is busy and provides the majority of acute admissions. There is evidence that many attenders bypass the primary care services. Of equal concern is apparent evidence that psychiatrists are inclined to do the same. We have presented evi dence of frequent use of the service by clients with a primary diagnosis of personality disorder, suggesting a need to determine why such clients arc unable to use other services effectively.
Regarding service provision, we have demon strated a significant need for acute psychiatric care in an area with a small catchment population. We suggest that data such as these are urgently required to prevent levels of staffing and funding based on capitation causing a serious shortfall in service provision. BROTHWOOD, J. ( 1965)  Many mental disorders are associated with an increased risk of fire setting. As a result, its assess ment and management is an important part of the management of people with psychiatric disorders. As more patients are managed in the community, away from the protection of high levels of hospital staff and fire regulations, such assessments are vital to ensure appropriate and safe placement. The presence of fire regulations in hostels offers only limited protection. Further, as residences with four people or less are not subject to fire regulations, assessment by psychiatric staff may have to extend to the fabric of the placement. Despite psychiatric hostels tending not to accept patients who are 'a fire risk', figures suggest that fires in hostels are common. One health authority reported 26 fires in hostels over 12 months (Moxom, personal communication). Psychiatric research and clinical attention has tended to focus on deliberate fire setting, usually within a forensic setting (e.g. Soothill, 1990). Such behaviour has been described in patients suffering from a wide range of psychiatric disorders, and is frequently related to patient's social deficits (Geller, 1987). A small proportion of deliberate fire setters fulfil the specific DSM-III-R diagnostic criteria for pyromania (American Psychiatric Association, 1987), which include having a fascination with fire and intense pleasure, gratification or relief, when setting fires. In marked contrast, accidental fire setting in psychiatric patients is widely ignored and standard psychiatric textbooks pay scant, if any, attention to the assessment of fire risk outside the forensic setting. A recent study (Barker et al, 1991) demonstrated that accidental fire setting is a common problem among patients in hospitals. The authors examined fire incidents in two psychiatric hospitals over four years. During this period 49 (34%) of the 143fires were accidental. In comparison, 59 (41%) of the fires were deliberate and in the remaining 35 (25%) fires the cause was not known. Comparative figures are not available for patients living outside hospital, but it is to be expected that accidental fires constitute a similar if not larger proportion of the total fires among patients in community settings.

Assessment
We suggest an approach to assessing fire risk that examines both individual and environmental factors (see Table I). When assessing individual fire risk, the emphasis should be on behaviour and disabilities rather than on diagnostic categories. This approach helps to draw attention to the increased risk in all psychiatric patients. It is vital that the impact of physical, psychological, and social disabilities on at risk behaviours is assessed in the context of the indi vidual's natural environment. This assessment may be readily included in the assessment of activities of daily living activities. Involvement of the members of the multidisciplinary team is useful to ensure that the full range of activities is covered. Attention must also be paid to any previous history of deliberate or