The difficulties that the specialist community mental health services have encountered in dealing with the volume of referrals from primary care have led to suggestions that service tiers (Reference Paxton, Shrubb and GriffithsPaxton et al, 2000) or different levels of entry (Reference Lovell and RichardsLovell & Richards, 2000) are necessary to utilise resources efficiently. All patients do not need the same type and level of intervention (Reference HaagaHaaga, 2000) and it makes sense to target specialist resources in accordance with the National Service Framework for Mental Health (Department of Health, 1999). Recent evidence suggests that the specialist community mental health services are becoming more focused and utilising case management approaches to manage patients with severe mental health problems (Reference Brooker and WhiteBrooker & White, 1998; Reference Kendrick, Burns and GarlandKendrick et al, 2000). However, the effectiveness of the procedures that are used to match resources to the level of patient's needs have rarely been evaluated. This paper compares how consultant psychiatrists, senior house officers (SHOs) and community psychiatric nurses (CPNs) dealt with the cohort of referrals from primary care to community mental health teams (CMHTs) in the Salford area in the calendar year 1997.

The study

A survey was carried out on a cohort of patients referred to four sectorised CMHTs in the Salford area, using patients' records as the primary source of data. The CMHTs in the area studied provide specialist mental health care for the adult population (16-65 age-group). The catchment area had a population of 220 000 at the time of the 1991 Census and the average Mental Illness Needs Index (MINI) score for the electoral wards in the area is 112.

The total number of referrals from primary care during the year was 1814 but 742 referrals were excluded from the analysis; either because their records were unable to be traced, they failed to attend their initial assessment appointment or because they were assessed by assessors that were not included in the study. This gave a total sample of 1072. CPNs who operate a nurse-led system for dealing with referrals from primary care, called the Duty Assessment Nurse (DAN) System (Reference McevoyMcEvoy, 1999), assessed the majority of referrals seen (n=874). The remainder of the referrals were seen by consultant psychiatrists (n=129) and SHOs (n=69) in outpatient clinics. The severity of presenting problems were rated using the Health of the Nation Outcome Scales, version 4 (HoNOS—4) (Reference Wing, Beevor and CurtisWing et al, 1998) and contacts with clinicians following the initial assessment were retrospectively tracked. Four types of support were identified.

1. • ongoing support — more than three follow-up contacts with a mental health professional

2. • crisis support — three or fewer follow-up contacts with a mental health professional or a limited number of sessions in a therapeutic group

3. • referral to another service such as a specialist drug, alcohol or psychotherapy service

4. • referral back to the general practitioner (GP) with recommendation to pursue a range of alternative therapeutic options.

The data were analysed using the Statistical Package for Social Sciences, version 7.5. The Pearson's χ² test was used to compare the characteristics of patients seen by the consultants, SHOs and CPNs and analysis of variance (ANOVA) procedures were used to compare the mean total HoNOS—4 scores.

Results

The consultants saw a significantly higher proportion of patients with a diagnosis of psychosis (10% compared to the overall average of 3%, χ²=11.9, d.f.=2, P=0.003). However, there were no other significant differences in the diagnoses of the patients seen by the consultants, SHOs or CPNs. The mean HoNOS—4 scores were highest for the patients seen by the consultant psychiatrists, 6.67 compared to 6.37 for the patients seen by the SHOs and 6.43 for the patients seen by the CPNs, but these differences were not statistically significant (F=0.34, d.f.=2, P=0.71, NS). The overall proportion of patients given ongoing support by one or more members of the CMHTs ranged from 35% for the patients seen by the consultant psychiatrists to 25% for the patients seen by the CPNs. Patients seen by SHOs were by far the most likely to be given short-term crisis support (see Table 1).

Table 1.

Type of support provided by assessor

	Referred back to GP n (%)	Other specialist n (%)	Crisis support n (%)	Ongoing support n (%)	Total n (%)
Consultant (n=7)	16 (12)	7 (5)	61 (47)	45 (35)	129 (12)
SHO (n=12)	5 (7)	2 (3)	432 (62)	19 (27)	69 (6)
CPN (n=23)	2751 (31)	55 (6)	325 (37)	219 (25)	874 (82)

For patients with sub-threshold mental health problems and patients with personality disorder the type of support given was comparable. Patients with sub-threshold disorders were most likely to be referred back to their GP and patients with personality disorder were likely to be given crisis support, although a significant minority of the patients with personality disorder were referred to a specialist psychotherapy service.

Differences in the type of support were apparent for patients with a primary diagnosis of drug and alcohol misuse. Patients seen by the CPNs were more likely to be referred to the specialist alcohol and drug services, whereas they were more likely to be given crisis support if they were seen by the consultants and SHOs. The CPNs also referred a higher proportion of patients with less severe affective disorders and neuroses back to their GP.

There was a significant association between the mean HoNOS—4 scores and the level of intervention given to patients seen by the CPNs (see Table 2). However, even though the trend in the HoNOS—4 scores for the patients seen by the consultants and SHOs reflected the level of interventions, these associations were not statistically significant (possibly because of the lower numbers).

Table 2.

Comparison of the mean HoNOS—4 scores for patients offered the different types of support by assessor

	Referred back to GP	Other specialist	Crisis support	Ongoing support
Consultant	5.37	5.57	6.69	7.27	F=1.7, d.f.=3
					P=0.2 NS
SHO	5.80	7.50	6.14	6.95	F=0.6, d.f.=3
					P=0.6 NS
CPN	5.05	7.38	6.52	7.80	F=37.5, d.f.=3
					P <0.001

Discussion

The findings of this small retrospective study indicate that patients with more severe problems were being channelled towards the consultant psychiatrists prior to their initial assessment. Patients with psychoses were more likely to be given an initial assessment by a consultant and the mean HoHOS—4 scores were also highest for the patients seen by the consultants. However, access to ongoing support was broadly similar irrespective of whether the consultant psychiatrists or the CPNs saw the newly referred patients. Patients with psychoses were most likely to be given access to ongoing support, but the severity of patients' mental health and social problems also appeared to be taken into account when decisions were made about the appropriate level of service provision.

The findings of the study highlight two issues that may need to be considered if more consistent service boundaries are to be established. First, the differences in the type of support given to patients with a primary diagnosis of alcohol or drug misuse and patients with less severe affective disorders and neuroses, suggest that it may be necessary to clarify the remit of local CMHTs. There is no definitive answer to the question of where particular patients are most appropriately treated in order to obtain the best outcomes and clinicians have to respond flexibly to take into account the local configuration of services. However, closer liaison with commissioning bodies may help to establish clearer boundaries for clinicians who are responsible for gatekeeping access to the general psychiatric services.

Second, the differences in the type of support offered to patients by the SHOs in comparison to those offered by the consultants and CPNs suggests that it may be necessary to re-examine the organisational context within which SHOs work. SHOs who are new to psychiatry and unfamiliar with local services can gain valuable clinical experience by giving short-term follow-up support to referrals with mild/moderate mental health problems. Nevertheless, it is also important for them to be prepared for the realities of the practice environment (Reference Hoge, Jacobs and BelitskyHoge et al, 2000), in which secondary mental health services receive far more referrals than they can deal with. SHOs may need clearer guidance and support if they are to make greater use of alternative resources in the local community. This issue is important given the present shortage of consultants because a potential benefit of establishing clearer service boundaries is that it may help to make general psychiatry a more attractive career pathway for SHOs contemplating their future.