As primary care groups become established as purchasers of mental healthcare, the interface between primary care and community mental health teams is likely to become the focus of increasing attention. This interface needs to be effectively managed if the CMHTs are to reconcile the competing demands of GPs and central government directives.
Central government policy emphasises the need for CMHTs to focus their resources on providing comprehensive care for the most severely disabled client group.1 Yet GPs are also likely to demand that CMHTs adopt a more flexible approach by providing care for clients whose needs are difficult to address within primary care, such as people with a personality disorder who experience acute crisis.2
Existing procedures for managing GP referrals to CMHTs have been criticised for inefficiency. Referrals are typically made to, and picked up by, individual professionals or allocated via multidisciplinary allocation meetings in which team members volunteer to assess new referrals.
In practice, referrals are more often allocated between overburdened team members via a system of informal bartering as members are reluctant to take on extra work. These strategies reduce the possibility of overt conflict between the different professionals about their respective roles within the team. But they are not necessarily the most effective strategies for dealing with new referrals, and a very high proportion of referrals are taken on by the assessing clinicians.3
Establishing a clear organisational framework for dealing with GP referrals may be extremely helpful. Accepting responsibility for gatekeeping can be a major source of stress for professionals working in isolation. Failure to set realistic boundaries can lead to 'burnout', producing a situation in which clients receive poor-quality care.4
The mental health services of Salford introduced a new system for dealing with referrals from primary care in 1996. The duty assessment nurse system was set up to streamline the process of dealing with referrals from GPs. The nurses have the remit to assess the majority of both routine and emergency referrals of patients between the ages of 16 and 65 and to co-ordinate access to services provided by the CMHTs.
Urgent cases are seen on the day of referral after consultation with the referring GP to check whether an immediate assessment is necessary. Routine referrals are seen within two to three weeks. The initial assessment follows a standardised format incorporating a health and social history, a mental state examination and assessment of risk.
The nurse gives a brief summary of the assessment at the team's weekly meeting. This provides a forum for advice, support and discussion of plans of care.
The operational policy for the system encourages the nurses carrying out assessments to think seriously about the implications of offering clients support, but the decision to provide support is left to their clinical judgement rather than by reference to any absolute criteria.
Plans for the new system were developed jointly by managers and clinicians and agreed with purchasers. Five additional community mental health nurses were recruited, and a rota system was introduced to enable them to act as the assessor one day a week. Despite initial reservations from a few GPs about nurses taking on assessments previously undertaken by the psychiatrists, the feedback from GPs has been largely positive.
A pilot evaluation was carried out in one of the community teams supported by North West regional office. The aim was to explore the factors which influence the duty assessment nurses' gatekeeping decisions. Data obtained from patient assessment records was used to conduct a retrospective analysis of all GP referrals to one of the teams between January and June 1997. The team serves an inner city sector with a population of 44,000, consisting of four electoral wards with an average mental illness needs index score of 119, well above the national average.
Clients' demographic characteristics, primary diagnosis and the outcome of their assessment was recorded. The severity of health and social problems was rated using the Health of the Nation outcome scales version 4 (HoNOS - 4)5 and scores for clients given ongoing input were compared to those referred elsewhere. For the purpose of analysis, the 12 HoNOS items were combined to establish four sub-scales rating:
behaviour - aggression, self-harm, use of alcohol and drugs;
symptoms - depressed mood, psychotic symptoms;
impairment - cognitive and physical problems;
social functioning - relationships, activities of living, living conditions, occupation.
Records were traced for 177 (91 per cent) of the 195 referrals from GPs during the period studied. Of these, 134 were assigned to the assessing nurse - the remainder being assigned to the consultant psychiatrist, either at the specific request of the GP or for the purpose of student and junior doctor training. The mean number of GP referrals to the community team over a six-month period was 3.1.
The majority of clients seen by the assessment nurses (65 per cent) had a primary diagnosis of a mood or anxiety-related disorder; 22 per cent had a disorder related to personality or substance misuse; 4 per cent had an organic or psychotic disorder and 9 per cent had no apparent diagnosis.
After being seen by the duty assessment nurse, 18 per cent of clients were given ongoing support by members of the CMHT (defined as more than three appointments), 19 per cent were given immediate crisis support (1- 3 appointments), and 63 per cent were referred to other avenues of support.
The HoNOS-4 scores for clients who were given ongoing support by the CMHT are compared to the scores for those referred elsewhere in figure 2. Scores for the sub-scales rating the severity of symptoms and problem behaviour were generally higher than those for social functioning or impairment. The mean total score for clients given ongoing input was 7.99 compared to 5.60 for clients who were referred elsewhere.
The assessing nurses recommended a range of alternatives to clients who were not given support by the community team. These ranged from individual coping strategies to counselling via voluntary organisations such as Relate.
The pattern of GP referrals to the team was consistent with the findings of previous studies, concluding that the majority of clients referred by GPs have mood or anxiety-related disorders which are the most prevalent mental health problems.
Nevertheless, the mean number of 3.1 referrals per GP over the six-month period indicates that only a small proportion of clients who were experiencing mental health problems were referred to the community team.
The ratings for the severity of clients' symptoms and problem behaviour were generally higher than those for social functioning or impairment. This would appear to indicate that the clients had specific and immediate mental health problems, rather than having the secondary disabilities that are often associated with enduring mental illness.
The finding that the HoNOS-4 scores were higher for clients given ongoing support by the CMHT is a rough indication that the clients offered support were those judged to be in greatest need.
The results of the pilot suggest that the system fulfils two functions. It appeared to control access to the community team operating as a semi- permeable interface between primary care and the CMHT. Clients referred by their GPs were not given open access to the CMHT, but they were seen quickly. Those identified as having the most pressing needs appeared to be given highest priority irrespective of their diagnosis. The second function could be described as signposting - directing clients on to alternative services to help address their needs.
Gatekeeping systems such as this are not a substitute for lack of investment in community mental healthcare. But they may help to ensure that resources are targeted towards those identified as having greatest need, while also ensuring that the legitimate needs of clients who fall outside the remit of the CMHT are addressed. REFERENCES
1 Department of Health. Building Bridges: the health of the nation. London: HMSO, 1995.
2 Goldberg D, Jackson G. Interface between primary care and specialist mental healthcare. British J of General Practice 1992; 42, 360: 267-269.
3 Patmore C, Weaver T. Community mental health teams: lessons for planners and managers. London: Good Practices in Mental Health, 1991 .
4 0vretveit J. Co-ordinating community care:multidisciplinary teams and care management. Buckingham: Open University Press. 1993.
5 Wing J et al. Health of the Nation Outcome Scales (HoNOS): research and development. British J of Psychiatry 1998; 172: 11-18.
Phil McEvoy is a community mental health nurse and research associate, school of nursing, midwifery and health visiting, Manchester University.
Employing community psychiatric nurses to assess clients referred to community psychiatric teams can help target resources.
The assessment nurses deal both with routine and emergency referrals.
Some GPs were initially concerned about nurses undertaking assessments previously carried out by psychiatrists.
Figure 1. Pattern of referrals
Total number of referrals assigned to DANs: 134
No of GPs referring clients: 43. Mean number of referrals/GP: 3.1
Attenders: 78 (58%). Non-attenders: 56 (42%)
Routine referrals seen: 70 (90%). Urgent referrals: 8 (10%)
40 males, 38 females. Ratio = 1.05/1
Mean age: 35.1 years. Standard deviation = 12.7 years