Psychiatric involvement in an Edinburgh hospice

Palliative care is a growing speciality. The terminally ill surfer high levels of psychiatric morbidity. The involvement of one senior registrar in setting up a liaison psychiatry service to a Marie Curie Hospice in special interest sessions is described. A liaison-consultation model developed and 29 referrals are described over sixmonths.


Service and referrals
The Marie Curie Hospice in Edinburgh is one of three city hospices. It has 37 beds and 500 annual referrals with in-patient, day unit and out-patient home care. One consultant and one senior house officer join other part-time medical support. Clinically, it has nursing, physiother apy, occupational therapy, social work, phar macy, aromatherapy and chaplaincy services with a weekly multi-disciplinary meeting.
Training in liaison psychiatry at the Depart ment of Psychological Medicine in the Edinburgh Western General Hospital, I arranged a two session input to this hospice. A liaison-consul tation model developed with cases first discussed at the multi-disciplinary meeting. I contribute to teaching, run an open patient relaxation group and developed a protocol for the diagnosis and management of delirium, the biggest hospice psychiatric problem.
Twenty-nine patients were referred in the first six months. Fifty-two per cent female with a mean age of 70 years (range=46-83). All but one were in-patients with 86% coming from medical staff; 72% were seen the same day. Reasons for referral are shown in Table 1. Two requests were to arrange transfer to psychiatric in-patient care and two requests for insomnia treatment. Psy chiatric consultations were recorded on a stan dardised form with ICD-10 diagnoses recorded at assessment (Table 1).
All had cancer, the primary being breast in 24%, lung in 21% and prostate in 10%. Cord compression occurred in 17%, brain mÃ©tastases were known in 7%. Pain was a problem in 70%, fatigue in 34%, nausea in 24% and constipation in 17%.
Forty-one per cent lived alone and the same number had social supports beyond family: 31% had a negative past experience of death of a close relative: 31% had a past psychiatric history, depression in 14%, and schizoaffective disorder in one patient. Thus referrals were particularly vulnerable to psychological problems.
Psychiatric involvement led to an increase in antidepressant (38 to 55%) and antipsychotic usage (34 to 58%) with a reduction in steroid and opiate prescription. On two occasions antibiotic and on one anti-parkinson medication were recommended after assessment.
Other input included advice on investigation and behavioural management, emotional support of patients and relatives and relaxation training. Thirty-four per cent were seen once with an average of two follow up visits. The majority died during admission, only 17% were discharged home.

Comment
The difficulties in diagnosing depression in cancer patients due to shared symptoms have been described (Carroll et cd, 1993). It is interesting that referring physicians in this study overdiagnosed depression, misattributing mood disorder from delirium, steroid use and alcohol withdrawal while never diagnosing adjustment reaction. They seemed biased to 'functional' rather than 'organic' psychiatric diagnoses.
Psychiatric involvement resulted in the use by physicians of higher doses of antidepressants and drugs from other classes with prescribing individually tailored. The ideal antidepressant would have low cardiotoxicity, few side-effects, and be mildly sedative. Paroxetine was often the best antidepressant in these respects and its rapid elimination allowed its quick removal if delirium developed. Trazodone has helpful seda tive qualities and venlafaxine was tried as a safer antidepressant with less side-effects than amitriptyline for neuropathic pain. Although an unlicensed indication, the similar neurotransmitter effects of venlafaxine to amitriptyline suggests it may be analgesic.
Haloperidol was mainly used as an antiemetic prior to psychiatry input. Disturbed, delirious patients were sedated with benzodiazepines as a midazolam subcutaneous infusion. A protocol suggested a plan of simple investigations with increased antipsychotic use to reduce arousal. Staff are often dismayed by the development of delirium soon after admission. Rather than a failure of care this reflects the high levels of medical illness in a distressed elderly population on large amounts of medication whose environ ment is changed. The cause of delirium is seldom found as it can be inappropriate to perform more than simple investigations.
Psychiatric input involves behavioural strategies for patient safety and re-orientation, with explanation to relatives. Improved levels of staff expertise have meant that detention under the Scottish Mental Health Act (1984) has not been needed for two years.
Hospice work is emotionally demanding. Staff have to weather the storm of continual bereave ment with distressed, suffering patients. They can feel useless, guilty and impotent, at-risk from staff burn-out (Ramirez et al, 1995). Pressure is reduced by the pleasant working environment, high staffing levels, education and training and mutual support. Psychodynamic defence mechanisms of displacement, projec tion, denial and intellectualisation are used by staff as well as patients and relatives. These are often better supported than broken down, unless maladaptive, negatively influencing patient care. The service seems valued by patients, staff and relatives. This is a rewarding training experience.