The number of admissions to hospitals in England and Wales following deliberate self-harm (DSH) has remained in excess of 100 000 each year for two decades (Royal College of Psychiatrists, 1994). Deliberate self-harm is one of the most common reasons for general hospital admission in the UK for both men and women (Reference Hawton and FaggHawton & Fagg, 1992). Numerous guidelines have been produced for the assessment and management of DSH, including those from the Department of Health and Social Security (1984) and, more recently, the NHS Health Advisory Service (1994) and the Royal College of Psychiatrists (1994).

The Department of Health and Social Security's and the Royal College of Psychiatrists' guidelines both recommend that all individuals who have carried out DSH should have a complete psychosocial assessment by an appropriately trained health professional. A small number of studies have demonstrated that there are similarities in the content and quality of assessments carried out on DSH patients, whether these be completed by psychiatrists, non-psychiatric doctors, social workers or psychiatric nurses (Reference Gardner, Hanka and O'BrienGardner et al, 1977; Reference Blake and BrambleBlake & Bramble, 1979; Reference Newson-Smith and HirschNewson-Smith & Hirsch, 1979; Reference Catalan, Marsack and HawtonCatalan et al, 1980; Reference Gardner, Hanka and RobertsGardner et al, 1982; Reference Griffin and BissonGriffin & Bisson, 2001). These studies do not show any significant differences in the types of follow-up arrangements offered.

Within Swansea NHS Trust, south Wales, individuals requiring hospital management following DSH are taken to either Singleton or Morriston Hospitals. These two general hospitals have approximately 600 and 750 inpatient beds, respectively, and serve the catchment area of Swansea, comprising a population of about 230 000 in 1998 (Chief Medical Officer, 1999). Morriston Hospital has a large accident and emergency department, whereas Singleton has only a smaller casualty department. At Morriston Hospital, patients admitted following DSH are assessed by a community psychiatric nurse (CPN); this service is provided on a daily basis. At Singleton Hospital, junior psychiatric doctors routinely assess individuals following DSH; this service is provided on Mondays, Wednesdays and Fridays only. This is partly because of the smaller number of individuals admitted to Singleton after DSH. Both CPN and routine junior doctor services are provided during normal working hours only (9 am until 5 pm). A CPN may, if required, request a ‘second opinion’ from one of the junior doctors regarding any individual assessment.

The majority of individuals who present to the accident and emergency department in Morriston Hospital after a deliberate overdose are admitted to a hospital bed. A separate analysis of the department's computerised database showed that 78% of attendees (following overdose) were admitted to hospital. The other 22% were discharged, absconded or refused admission, and no formal psychiatric assessment was made at the time. Similar information was not available for Singleton Hospital. It is not known what follow-up is given for those individuals who are discharged directly from the accident and emergency/casualty departments.

At the time of the study, supervision and training for the CPNs and junior doctors was provided purely on an informal/as-required basis by one of the consultant psychiatrists. The doctors received an introductory seminar on DSH, and only carried out assessments if they had at least 6 months' psychiatric experience.

The aims of this study relating to the assessment of individuals following DSH were:

1. (a) to elicit differences in psychiatric referral practices following assessment by CPNs and junior psychiatric doctors;

2. (b) to identify differences in perception by nurses and doctors of the degree of mental illness of the patient, and to determine whether these findings were compatible with subsequent referral to psychiatric services.

Method

This was a prospective observational study, whereby the participants (CPNs and junior psychiatric doctors) completed a questionnaire, designed by the author, after carrying out a DSH assessment (further details available from the author on request). All 11 psychiatric nurses and nine junior psychiatric doctors involved in the daytime DSH assessment rota agreed to participate. One of the nurses was a dedicated DSH liaison nurse. The doctors comprised eight psychiatric trainees (senior house officer grade) and one staff grade doctor.

All patients with DSH admitted to Morriston or Singleton Hospitals between 1 May and 31 October 2000 were included in the study if they satisfied the following criteria:

1. (a) they were assessed by a junior psychiatric doctor or CPN within either of the two specific components of the DSH service outlined above;

2. (b) they were aged 16 years or over and had left full-time (school) education;

3. (c) the method of DSH involved self-poisoning.

Data analysis was carried out using the statistical package SPSS for Windows version 10.05. Specifically, the chi-squared (χ²) test was used to analyse demographic differences between outcomes and perceptions of mental illness by the doctor and nurse groups. All tests were two-tailed.

Results

Demographics

There were 247 eligible assessments during the 6-month study period; 175 (71%) assessments were carried out by nurses and 72 (29%) by doctors. A χ² test revealed no statistically significant differences in any demographic variables between the two groups.

Overall, 132 (53%) of the sample of 247 were female. The mean age of subjects assessed was 34.9 years, with a range of 16 to 88 years. Regarding employment status, 64 (26%) were employed, with 13 (5%) retired and 10 (4%) students. The remaining 160 (65%) individuals were not in employment. All individuals in this sample had used overdose as a means of DSH, indeed this was the sole method in 233 (94%) of cases. Alcohol was ingested during, or in close association with, the episode of DSH in 134 (54%) of cases and 12 individuals (5%) claimed to have consumed both alcohol and illicit drugs. A previous episode of DSH was either self-reported or evident from case notes in 145 (59%) of the assessments.

Diagnoses relating to a previous psychiatric history were recorded in case notes, or obtained from the individuals assessed or other informants in 100 cases (40%). Only 55 patients (22%) were under the care of psychiatric services at the time of assessment (Table 1). Among those with a single previous psychiatric diagnosis, the three most frequent were unipolar affective disorder, substance misuse and psychotic disorder (comprising 14.2%, 9.3% and 4.0% of the total study sample, respectively).

Table 1. Previous psychiatric history, current psychiatric service involvement and perceptions of mental illness

	Nurses (n=175) n (%)	Doctors (n=72) n (%)	Number of assessments (n=247) n (%)
Diagnosis relating to previous psychiatric history
Diagnosis not known	110 (63)	37 (51)	147 (60)
Single previous diagnosis	56 (32)	24 (33)	80 (32)
Multiple previous diagnoses	9 (5)	11 (15)	20 (8)
Current psychiatric service involvement
Yes	39 (22)	16 (22)	55 (22)
No	136 (78)	56 (78)	192 (78)
Perceptions of mental illness by doctors and nurses
Mental illness	86 (49)	57 (79)	143 (58)
No mental illness	89 (51)	15 (21)	104 (42)

Outcomes of assessments made by either doctors or nurses

The most common outcomes are shown in Table 2. For a variety of reasons, 18 assessment outcomes were excluded from the initial statistical analysis. There was a highly significant difference in the referral patterns of doctors and nurses (χ²=48.59, degrees of freedom (d.f.)=8, P<0.001). Doctors were significantly more likely (χ²=27.54, d.f.=1, P<0.0001) to refer individuals to services that involved further direct contact with other doctors (51 of 72 (71%), compared with 60 of 175 (34%)). Services that involved direct ‘doctor-contact’ included psychiatric out-patient clinics, admission to the psychiatric in-patient ward, second opinions from junior doctors or referral to the on-call junior doctor or consultant psychiatrist.

Table 2. Outcomes of assessments made by doctors or nurses¹

	Nurses (n=159) n (%)	Doctors (n=70) n (%)	Total number of assessments (n=229) n (%)
Discharge without follow-up	68 (43)	8 (11)	76 (33)
Out-patient clinic	27 (17)	34 (49)	61 (27)
Alcohol/drug agency	18 (11)	9 (13)	27 (12)
Community mental health team	21 (13)	3 (4)	24 (10)
Admission to psychiatric unit	11 (7)	13 (19)	24 (10)
Out-patient clinic & community mental health team	5 (3)	1 (1)	6 (3)
Counsellor	5 (3)	0 (0)	5 (2)
Out-patient clinic & alcohol/drug agency	1 (1)	2 (3)	3 (1)
Out-patient clinic & day hospital	3 (2)	0 (0)	3 (1)

Discussion

The demographic characteristics of this study, as well as the overall referral practices of health professionals, are compatible with previously reported findings (Reference Platt, Hawton and KreitmanPlatt et al, 1988; Reference HallHall, 1994; Reference YeoYeo, 1992; Reference Hawton, Fagg and SimkinHawton et al, 1997; Reference Hawton, Harriss and SimkinHawton et al, 1999; Reference Griffin and BissonGriffin & Bisson, 2001). However, there is a huge discrepancy between CPNs and junior doctors both in perceptions of mental illness and in referral practice. Either the CPNs are discharging a large number of individuals who have severe problems or the junior doctors are unnecessarily referring on to services patients who do not require treatment.

The finding that doctors were more likely to refer on to services that involved direct contact with other doctors is unlike the outcomes described by Catalan et al (Reference Catalan, Marsack and Hawton1980) and Griffin & Bisson (Reference Griffin and Bisson2001). Griffin & Bisson (Reference Griffin and Bisson2001) found no difference in the management patterns of 145 patients seen after DSH by either a psychiatric nurse or junior psychiatrist at the Cardiff poisons unit.

The differences found are unlikely to reflect true differences in the patients assessed, but could be accounted for by differences in practice between the professional groups. Significant differences existed between the levels of psychiatric experience of the nurses and doctors (mean duration of experience being 15 years for nurses compared with 18 months for doctors). Also, the type of experience may well have been different. The CPNs may have been more used to working independently and approaching issues holistically, taking account of psychosocial factors which may have resulted in increased confidence in excluding mental illness and discharging individuals who did not require psychiatric intervention.

Since completion of the study, a consultant liaison psychiatrist has been appointed, who has taken responsibility for regular monthly supervision of the CPNs. Teaching of junior doctors has also become more structured, involving initial observation of three DSH assessments, observation while carrying out at least three further assessments, discussion of all cases with the consultant for 3 months and regular monthly group supervision. This may lead to a more structured or uniform approach in conducting DSH assessments by doctors and nurses.

The study findings, in particular the huge differences in referral practices between doctors and nurses, clearly have implications for the psychiatric service as a whole. Given the apparent preoccupation by doctors with onward referral to other doctors, it is impossible to know which of the two groups of professionals is assessing most appropriately. An alternative methodology may be needed to examine these differences, such as the assessment of the same patient by both a doctor and a nurse. Without further work, including an economic appraisal, it will be impossible to determine whether the widespread introduction of nurse-led practice will prove an effective and efficient use of resources.

Declaration of interest

None.