Trainees ' forum Liaison psychiatry in Addenbrooke ' s Hospital : six months experience

order to gain closer compliance with the Royal College Guidelines (1988), arrangements were made for one of the registrar posts associated with this unit to form part of a liaison psychiatry team. This post provides training and experience in liaison psychiatry in Addenbrooke's Hospital on a half-time basis, the other half of the time being spent with the Academic Psychiatric Unit within Addenbrooke's Hospital, principally involved in the

care of patients with eating disorders.Supervision of the liaison work is provided by the senior registrar based on the Addenbrooke's Hospital Psychiatric Unit, and by a designated consultant psychiatrist.Because of the part-time nature of the liaison post and the large size of Addenbrooke's Hospital, the responsibilities of the liaison registrar are limited to providing liaison/consultation cover to two of the four acute medical wards, the Rosie Maternity Hospital, the Accident and Emergency Department (during working hours) and to any other units with out specific liaison links.The other two acute medical wards, general surgery, neurology and neurosurgery and certain other departments have liaison psy chiatry links with different consultant psychiatrists and their teams.Outside normal working hours emergency consultations are dealt with by the duty psychiatric registrar and senior registrar.There is also a part-time psychiatric social worker in Addenbrooke's Hospital whose primary function is to provide help and support to patients who have harmed themselves and whose problems are felt to be mainly social rather than psychiatric.Any of the medical or surgical teams are free to refer cases to her regardless of liaison links.Direct referrals of patients to the local Drinking Problems Service and Drug Dependency Clinic are also possible.
The purpose of this report is to examine the experi ence obtained by the liaison psychiatry registrar from 1 August 1989 to 31 January 1990.

The study
For each patient seen a variety of information was systematically recorded, including basic demo graphic details, source of, and reason for referral, the reason the patient was in Addenbrooke's Hospital, past psychiatric history, clinical diagnosis according to the ICD-10 classification and outcome/disposal.

Age and sex distribution
During the period under study, 56 patients were referred.The patients had a mean age of 36.6 years (range 18-82 years) and there was no significant age difference between the males (n = 25, mean age = 39.2 years, range 19-75 years) and the females (n = 31, mean age = 34.5 years, range 18-82 years).The greatest number of referrals for both sexes involved patients aged 20-39 years, which is in agreement with previous studies (e.g.Anstee, 1972;Tiller, 1973;Brown & Cooper, 1987).The sex ratio in our patients (F:M = 1.24:1) was similar to that found by Anstee (1972)and White & Bloch ( 1970),but lower than that of other published series (e.g.Fleminger & Mallett, 1962;Lipowski & Wolston, 1981;Brown & Cooper, 1987).

Sources of referral
Analysis of the information regarding sources of referral revealed that 58% of the patients were referred by the two acute medical teams, 16% came from Accident and Emergency and 9% from Obstet rics and Gynaecology.We had predicted that the majority of the referrals would indeed come from the medical wards, and took these findings as evidence that the liaison links were functioning efficiently.

Reasons for referral
Forty-six per cent of the referrals to the Liaison Psy chiatry Team were made for assessment of deliberate self-harm, principally overdoses.This figure is simi lar to that found by Anstee (1972) and Tiller (1973), but much lower than that found by Brown & Cooper (1987).
In Addenbrooke's Hospital, following the work of Gardner et al (1977Gardner et al ( , 1978Gardner et al ( , 1982) ) patients admitted following deliberate self-harm are initially assessed by the medical teams responsible for their care, and only those patients in whom psychiatric disturbance is felt to be present are referred for further assess ment.Thus patients referred following deliberate self-harm often had either overt psychiatric symp toms, were at high risk of repetition or had past histories of previous suicide attempts or psychiatric illness.
In one fifth of referrals, the question being asked was, "is this patient depressed?",usually because of their behaviour on the ward, sometimes interpreted in the knowledge that the patient had had previous treatment for depression."Management difficulty" was the reason for referral in 18% of cases.This usually consisted of aggressive or disruptive behav iour, fairly often occurring in a patient with a history of psychotic illness, and was the category associated with the greatest demand for urgency of attendance by the psychiatrist!In 7% of the referrals the question being asked was either, "is this patient psy chotic?" or "could you help us with the management of this psychotic patient?"

Reason for admission to Addenbrooke's Hospital
Analysis of the reasons why patients referred for a psychiatric opinion had initially been admitted to Addenbrooke's Hospital revealed that in 57% of cases this was for the treatment of deliberate selfharm, in 16% for "medical" reasons and in 9% for treatment of obstetric or gynaecological conditions.

Clinical diagnosis
For all patients seen, a diagnosis was made on clinical grounds, and this was subsequently classified accord ing to ICD-10.Of the first-order ICD-10 diagnoses obtained, 29% of the patients were diagnosed as suffering from a mood disorder, 25% had some form of personality disorder, 13% were suffering from neurotic, stress-related and somatoform disorders (principally anxiety states, somatisation disorders and occasional dissociative states) and 11% from schizophrenia.Drug and alcohol abuse was the diag-nosis in 5%, and an organic brain disorder in 4% of cases.Eight per cent of referrals were not felt to be psychiatrically unwell, and in 5% of cases there was insufficient evidence to come to any satisfactory diag nosis.Although the exact figures obviously differ, the rank order of these diagnostic categories is in broad agreement with previous studies (e.g.Fleminger & Mallett, 1962;Anstee, 1972;Tiller, 1973)except that in all these series there were a significant proportion of patients with organic brain syndromes (12-19%) whereas in our sample only 4% of the referrals fell into this diagnostic category.The probable reason for this is that in Addenbrooke's Hospital referrals of elderly patients (who are especially likely to have either acute or chronic organic brain syndromes) are made directly to the Psychogeriatric Department, thus bypassing the liaison team.The fact that neurology/neurosurgery referrals also go directly to a different consultant psychiatrist, and that patients with alcohol and drug problems also tend to be directly referred for specialist treatment, would also reduce the number of patients with organic brain syndromes likely to be seen by the liaison service.

Past psychiatric history
Of the 56 referrals, we were surprised to find that 68% of them had had previous contact with the psychiatric services; 43% as in-patients and 25% as out-patients.Indeed, it was felt that a previous psychiatric history was a major criterion in deciding upon psychiatric referral.Approximately 35% of the patients seen had been in contact with a mental health professional during the preceding 12 months.
There was no difference between those with and without a past psychiatric history with regard to the reason for referral to the liaison team or the reason for admission to Addenbrooke's Hospital (being de liberate self-harm in both cases).However, those with a past psychiatric history were much more likely to have a diagnosis of affective disorder or person ality disorder, and this difference was statistically significant (x2= 23.4,d.f.= 6, P=0.0007).This result was also true for the diagnosis of schizophrenia, although the numbers involved were very small in this case.Thus it appears that patients with more "serious" forms of mental disorder presenting to the general hospital are very likely to have had previous contact with the psychiatric services.

Management
Of the 56 referrals, 22 (40%) were admitted to an inpatient psychiatric bed.Of these, 6(11% of the total) were admitted to the Academic Psychiatric Unit at Addenbrooke's Hospital and 12 (22%) to the appro priate sector ward at Fulbourn Hospital.Four patients (7%) were admitted to other psychiatric facilities.Twenty-eight patients (50%) received out patient psychiatric follow-up.Of these, 13 (23% of the total) were seen by the liaison psychiatry registrar and 7 (13%) were referred back to the consultant psychiatrist already involved in their care.Eight patients (15%) were referred as out-patients to other psychiatric services, principally the Drinking Prob lems Service and to Clinical Psychology.In three cases (5%) advice only was given and only two patients (4%) declined any follow-up.
Our rate of follow-up in the psychiatric out-patient clinic was slightly higher than the range of 30-48% reported by other studies (e.g.White & Bloch, 1970;Anstee, 1972;Tiller, 1973 andBrown &Cooper, 1987)but our rate of in-patient psychiatric admission at 40% was much higher than the rates of 12-28% reported in the above studies.We feel that this is a reflection of the selective referral of patients with deliberate self-harm, the medical teams having dealt satisfactorily themselves with the majority of over dose patients, and referring on only those patients at higher suicidal risk and with a high proportion of psychiatric problems.

Comment
In conclusion, we feel that the establishment of a liaison psychiatry registrar post in Addenbrooke's Hospital, together with the supervision this necess arily entails, has been successful both in providing excellent teaching and experience in liaison psychiatry as recommended by the Royal College of Psychiatry guidelines (1988)and also in providing readily access ible psychiatric help within the Addenbrooke's site.
An analysis of the, admittedly small, number of patients seen during the six months studied has shown little difference from other published series, apart from those presumed to be a result of the selec tive referral criteria operating within Addenbrooke's Hospital.In particular, the scheme of overdose assessment designed and instituted after thorough research by 'Gardner et al (1977'Gardner et al ( ,1978'Gardner et al ( , 1982))appears to be working well.
In the future, we plan to continue with systematic recording of data regarding all liaison referrals which we will add to a computerised data base, thus enabling further research and medical audit to be undertaken easily.Other plans include the involvement of the more senior members of the team in liaison attachments with specialist units (e.g.Radiotherapy and Oncology) where they will have an important role to play in areas such as staff support, in addition to more clinical input.

A cknowledgemen t
I would like to thank Drs A. R. K. Mitchell, H. Middleton and I. Markova for their most helpful advice and comments during the preparation of this paper.