New community mental health team for acute psychiatric illness

Aims and method The aim was to provide effective time-limited management of acute episodes of mental illnessin the community, thus reducing dependency on the psychiatric services. A multi-disciplinary team supplied intensive psychological and pharmacological interventions. Results Fifty-nine per cent were discharged within 12 weeks to the care of their general practitioner. Twelve per cent were admitted to in-patient care. Clinical implications A time limited multi-disciplinary approach willhave benefits to patients and the service.

Panmure House is the base of a community mental health team in Dundee. A survey of people presenting to the Dundee service with acute psychiatric illness identified a group who would be most appropriately managed by a community-based acute response service. Panmure House opened on 1August 1993, staffed by a multi-disciplinary team comprising a consul tant psychiatrist, half-time clinical assistant and registrar/senior house officer, charge nurse, six staff nurses, clinical psychologist and occupa tional therapist.
Panmure House does not provide all ongoing care for people with chronic mental health problems. It opened with the novel aim to manage acute episodes of mental illness with a contact time of approximately 12 weeks. The expectation is expressed that significant change can be achieved and that dependency may be prevented or reduced. A variety of pharmacolo gical and psychological interventions are used. Patients are offered individual sessions at home or at Panmure House as well as the opportunity to attend a variety of structured groups. All referrals are tertiary from within the acute psychiatric service. At initial assessment three questionnaires are completed: the 60-item Gen eral Health Questionnaire (GHQ: Goldberg & William. 1988), the Social Functioning Question naire (SFQ: Tyrer, 1990) and the Hospital Anxiety and Depression Scale (HADS: Zigmond & Snaith, 1983). These are repeated at dis charge.
As this is an entirely new service with the aim of providing time-limited treatment of discrete mental illness episodes, we felt it was important to study the characteristics of the people seen and their contact with the services as well as aspects of service delivery.

The study
Subject contact information was recorded on a computer using a local clinical database system. Useful information was obtained from the ad ministrative secretary's ledger, notes kept by the consultant at weekly meetings, appointment diaries and copies of discharge letters. From these sources we were able to study subject demographics, referral sources, rapidity of re sponse to referral, venue of assessment, length of time in contact, diagnosis and disposal at discharge as well as symptom rating scales. The period studied was 1 August 1993 to 31 July 1994.
The source of referral was as follows: 40% (130) were referred from psychiatric out-pa tients. 21% (68) from the Emergency Assessment Clinic. 18.5% (60) from psychiatric in-patient wards, 5% (17) from the general hospital liaison service and 15.5% (49) from other sources including Tayside Alcohol Service (TAPS).The Emergency Assessment Clinic is a 24-hour clinic where emergency general practitioner (GP) and self-referrals are seen by medical staff and suitable follow-up arranged. The general hospital liaison service provides assessment of deliberate self-harm and other psychiatric disorders re ferred from general hospital-based specialists. Four per cent (12) of referrals come from TAPS, whose in-patient facility is situated 40 miles from Dundee. Liaison with Panmure House is felt to be a useful adjunct to the management of mental illness episodes compounded by alcohol misuse.
As it is the aim of this new resource team to respond to acute mental health problems the number of days between referral and first appointment offered was examined. Fifteen per cent (48) of people were seen on the day of referral with 16% (52) seen on the following day. In total, 71% (230) were seen within five working days of referral. Five per cent (17) were seen 10 or more working days after referral. In this year, 76% (246) of people were first assessed at Panmure House, 20% (64) at home and 4% (14) assessed in a psychiatric ward. Eighty-two per cent (59) of men were first seen at Panmure House and 14% (15) at home compared with 73% (157) of women being seen at Panmure House and 23% (49) at home. The larger number of women first assessed at home approached statistical significance (P=0.055).
We examined the length of time in contact for the 205 people discharged in this year. Twentysix per cent discharged within four weeks. A further 33% were discharged by 12 weeks with only 7% being discharged after more than 24 weeks. There was no significant difference between genders for length of time in contact.
Fifty-nine per cent (194) of those referred completed rating scales at first assessment with only 28% (58) returning them at discharge. For each rating scale there is a subset of people who have completed 'before and after' sets. Table 1 shows the mean scores for the paired rating scales.
We studied the disposal of the discharged subjects. Over half of the 205 subjects dis charged 54% (111) did not require further follow-up from the psychiatric services and were returned to GP care. Only 15% (30) required further out-patient supervision. Twelve per cent (24) of subjects were discharged to in-patient care. Seven per cent (14) were discharged for further psychological treatment to the psycho logy service, psychodynamic psychotherapy and behavioural psychotherapy and 3% (7) were discharged to TAPS.

Comment
Sixty-six per cent (214) of subjects were aged 20-40 years, suggesting Panmure House is seeing a relatively young population. We hope by optimis ing independent functioning at this age we might prevent over reliance on the in-patient setting in the future. We feel it is important to maintain patients within their home and family environ ment. Retrospective audit of the first 100 referrals found 46 had no previous contact with the service prior to this illness episode: 40% of referrals came from psychiatric out-patient clinics but as this was the disposal for only 15% it suggests that Panmure House is effective in managing acute mental illness episodes with intensive intervention and in a time-limited fashion. The large percentage from the out patient setting suggests that it would be appro priate in the future for Panmure House to receive referrals directly from GPs. This would allow a more rapid response to acute problems. Twentyone per cent of patients referred came from the Emergency Assessment Clinic. It was hoped this figure might be higher with the potential of reduction in admission rates to in-patient beds from the Emergency Assessment Clinic. We believe this lower rate of referral is due in part to Panmure House operating 9am-5pm, five days a week and that for Panmure House to significantly reduce admission rates it would need to operate over the weekend.
The standard of offering an appointment with in five working days was met for 71% of patients with a third seen within 24 hours. However, 29% were seen outwith the standard. We believe the following contributed to this figure: patient preference; restricted room availability: and staff  availability. At times prioritising of referrals is required. Assessment is carried out in the setting deemed most appropriate. Home assessment is often more suitable to people, targets those at high risk of default and generates useful in formation about a person's environment. We feel no firm conclusions can be drawn from the place of first assessment. It was postulated that more women than men might be seen at home due to the caring responsibilities of women and the perceived higher risk of male violence.
Fifty-nine per cent were discharged within the aimed for 12-week period. Approximately a quarter of people referred were in contact for four weeks or less. This proportion most likely reflects two effects: (a) the number of patients who fail to attend or engage at Panmure House; and (b) the large number of patients whose acute situational crises were managed and resolved within one month. Length of time in contact was difficult to measure precisely as during this first year only the date of discharge was recorded and no record was accessible for the last kept appointment. This will have artificially extended the contact for some patients. That over half of patients discharged returned to the care of their GP is extremely encouraging, especially as those referred are already within the psychiatric service.
Twelve per cent of attenders required a period of in-patient care. This is not surprising as these were patients presenting with acute psychiatric problems for whom referral to Panmure House was often the preferred alternative to admission. Management of these patients while in hospital was by the ward staff. In a study of home versus hospital-based care of serious mental illness (Mark et al, 1994), length of in-patient stay was 80% shorter for the home-based group when they were managed by the home-based staff. When responsibility for discharge was trans ferred to the ward team length of crisis admission for the home-based treatment group increased three-fold. Staff looking after patients prior to their in-patient stay will have a clearer idea of the level of functioning at which a patient can manage in their environment. It may be possible that in the future Panmure House would staff a small number of in-patient beds for the crisis admission of their patients. Disposal to the longer term psychiatric care (e.g. day hospital and community psychiatric nurse service) seems modest at 4%. Dundee has established day hospital and community psychiatric nurse ser vices for patients with chronic disorders. At present referrals to Panmure House are from staff familiar with services in Dundee and as Panmure House aims to treat discrete illness episodes the small percentage of patients dis charged to longer term follow-up would seem appropriate.
The number of patients presenting with schizo phrenia or delusional disorder is small at 3%. These tended to be first presentations of illness rather than relapses of patients with established disorders who are already engaged with the parts of the service mentioned above.
The most frequent diagnoses are of phobic/ anxiety disorder (F40.41) and stress reaction/ adjustment disorder (F43) which make up 71% of the total diagnoses. The percentage of dis charges diagnosed as affective disorder seems small at 11%. This study covered the time of transition from ICD-9 to ICD-10 classification. Using ICD-9. diagnosis of a depressive episode led many patients to fall into the category of "adjustment reaction-prolonged depressive" as symptoms did not meet the criteria for diagnosis of "manic depressive psychosis-depressed". The conversion of ICD-9 adjustment reaction to the equivalent ICD-10 diagnosis will have resulted in the 'under-diagnosing' of episodes of depression.
It is as important to look at the severity of symptoms as well as diagnosis. The small number of 'before' and 'after' rating scales was disappointing and due to a failure to return the second set of questionnaires. Patients now complete these in advance of discharge to ensure their return. Mean scores were high. 91% of the sub-group patients fulfilling caseness for GHQ-60 when referred with significant reduction of 47% at discharge. Sixty per cent of the paired sub-group met caseness for the depression component of the HADS when using the higher cut-off point of 10/11. This supports our impression that Panmure House is managing more patients with depressive illnesses than our diagnostic percentage for affective disorders would suggest. The SFQ is a less symptom specific rating tool and by its nature has no cut-off point for caseness. It does however generate a useful measure of impaired social functioning and again its mean score improved significantly.

Discussion
This is a study of a service whose aim from the onset is to provide time-limited management of acute mental illness episodes with this aim being communicated to the patient at assessment. The patient group identified is perhaps different from that expected as we have seen fewer patients with psychotic illnesses and severe mood dis orders. However, in this first year there has been the opportunity to respond rapidly and treat intensively in a community setting a relatively young population with high psychiatric morbid ity resulting in significant improvement in their symptoms. It is hoped that this treatment New community teams for acute psychiatric illness approach may have avoided chronicity with the return of patients to the care of their GP.
Our study identified difficulties and omissions in data recording. This has been radically modified as the result, particularly in recording dates of attendance, default rates, first contact to service, specific interventions, prescribing prac tice, and rating scales. It has been essential to make these improvements to our database in advance of any expansion of the service to take referrals directly from GPs or to manage a small in-patient unit. We propose further studies looking at those patients re-referred to Panmure House, as well as the subsequent functioning and use of mental health services by a random group of those patients discharged from this new service.