Inner-city general practice population of people with schizophrenia

A survey of people with schizophrenia in an inner-city general practice was undertaken to identify levels of social disability, service receipt and patterns of care received. Contacts with general practitioners, psychiatrists, community psychiatric nurses and social workers were quantified, and the nature of the contacts assessed. Overall social disability for the group was marked (mean Health of the Nation Outcome Scales (HoNOS) rating 55.5). There was a correlation coefficient of +0.899 between the numbers of agencies involved and the overall HoNOS scores suggesting appropriate targeting of care. While there were few differences in the HoNOS ratings of the various contact subsets, there were significant differences in the extent of agencies contact with patients, the greatest number of contacts being in general practice. Limited information sharing, the absence of a formal shared care plan and sectorisation of services are thought to obstruct more effective general practice involvement in care.

Providing a good service to the chronically mentally ill in West Kensington presents sig nificant challenges. North Thames region has the highest prevalence of serious mental illness in the UK, spending 18% of its total budget on mental health. In our practice, 65% of the patients live in wards with an underprivileged area score greater than 30. Patient turnover in any one year is approximately 20%. The practice area is in four different social service localities and the practice has contact with six community psychiatric nurses (CPNs). At the time of the study very little Care Programme Approach had been implemented. Most local community men tal health teams were run from a hospital base with basic staffing levels. One area team, responsible for a population of 40 000, had three full-time CPNs, supplemented by a home nursing team which had a limited ability to provide more intensive short-term support for a small number of patients. Psychiatric admissions are generally to the Riverside Mental Health Trust.

The study
The study was undertaken in an inner London teaching general practice with six partners. There are some 10 700 registered patients.
Inclusion criteria were: (a) registered at the practice prior to 1994; (b) a diagnosis of schizophrenia or schizoaffective disorder (past or current) made by a consultant psychiatrist; (c) patient's consent for the interviews.
The Health of the Nation Outcome Scales (HoNOS) was selected as the research instru ment for evaluating social disability and func tioning. It also measures aspects of psychiatric symptomatology (e.g. delusional thought). Struc tured interview assessments are made of the patients' behaviour in the preceeding two weeks, grading being on the basis of the worst state during that time. The parameters assessed include: aggressive behaviour, suicidal ideation, non-accidental injury, problems associated with alcohol or drug misuse, problems with memory, orientation and understanding, physical illness and disability, mood disturbance, problems associated with hallucinations and delusions, other mental or behavioural problems (e.g. panic, sleep disorders, fatigue), problems mak ing supportive social relationships, problems with housing and locality, problems with employ ment, recreation and finance.
Each parameter is graded and stored on a spreadsheet for detailed analysis. Following the interview (with G.B.) the researcher, with refer ence to the findings, arrives at an overall HoNOS rating for functional disability. The range is from 0-100 (0 being no disability, 100 maximally disabled).
Service receipt (for the complete year 1994) was assessed by a variety of means, including detailed analysis of all GP, CPN and hospital psychiatric records of the patients (both manual and computer held). Each entry in the various clinical records was assessed to determine the nature of the activities recorded (e.g. mental assessment, repeat medication, issuing a sick note, discussion or help with social problems includingjobs accommodation and money, treat ment of intercurrent medical illness and emo tional support). In many instances, an individual contact resulted in more than one activity being recorded. The data on individual activities have been presented as a percentage of the total activities recorded for each provider group. Unfortunately, no volume or qualitative infor mation for social service contacts was available and data are limited to whether there had been contact during 1994. Forty-three patients satisfied the criteria for admission into the study. The data for service receipt and patterns of care relates to the whole group of 43 patients; 42 patients were recruited for the HoNOS interviews. One patient proved impossible to access, and was believed at the time to be seriously mentally ill (and possibly in prison). All data relating to social disability thus relates to this group of 42 patients.

Social disability and functioning
The mean overall HoNOS rating for the group as a whole was 55.5 (median 58; mode 21; range 10-95). Sixty-four per cent of patients had problems characterised as moderate or worse on the HoNOS score, while 24% had severe problems (Table 1). Seventy-one per cent of patients had persistent delusional thought with 59% in the moderate to severe range, in line with other studies (Melzer et al, 1991). Ninety per cent of patients had some difficulty forming supportive social relationships; employment and finance are also problematic, with only 9% of patients under pensionable age in some employment. Sixty-two per cent were receiving sickness benefit. Worries about money were common with lack of funds hindering normal social integration.

Service receipt
Detailed analysis of the data reveals interesting subsets: of the 42 patients in this group, all had been in contact with at least one agency (GP, psychiatrists, CPNs or social workers). While there were surprisingly few differences in the HoNOS scores of the different contact subsets, there were significant differences in both the proportion of patients in contact with the various specialities and their consultation rates for 1994. Not only was general practice in contact with a greater proportion of the patients, it also had the biggest volume of contacts with the highest consultation rates ( Table 2).

Nature of contacts
The clinical notes of all contacts were assessed and the activities categorised (e.g. medication renewal, mental state assessment, etc.). Indi vidual contacts, for all specialities, often resulted in more than one activity being recorded (e.g. of the 370 GP contacts, 655 activities were re corded). The data for psychiatric out-patient records is collated from 18 of the 23 patients who had been seen in 1994, five sets of clinical records being unobtainable during the period of the study. We feel the data from these 18 sets of records is large enough to be representative. Table 3 shows the comparative frequency in dividual activities were recorded, expressed as a percentage of the total number of recorded activities for that speciality.
Mental state assessment and renewal of medication are the two most common entries in the records of all three specialities. Mental state assessments are twice as likely to be recorded by the psychiatrist and CPNs than by the GPs. Treatment of intercurrent medical problems and production of sick notes account for 31 % of GP recordings but only 6% of psychiatric recordings. The CPN records suggest a greater focus on the pastoral elements of care (e.g. advice on social issues and emotional support).

Patterns ojeare
There were nine different patterns of care. Nine patients received care from all four agencies, and a further nine from three agencies (Table 4). Psychiatrists were the sole point of contact for two patients, who were both stable and had the lowest HoNOS scores. One patient, who had the highest HoNOS score of the entire group, only had contact with the social workers. Other agencies had attempted to contact her, without sucess. No patients only had contact with the CPNs. Other studies have shown that between 25 and 40% of patients have no contact with specialist services and rely exclusively on general practice for their care (Parkes & Brown, 1962;Pantelis, 1988). In our study this group was slightly smaller at 19% (eight patients). These patients were less disabled by their illness with a mean HoNOS rating of 38.7 (c.f. 55.5 for the group as a whole).

Comments
The study is clearly limited to those patients who were registered with the practice and does not include those who are unregistered or homeless. Also, given the retrospective design of the study, it would be wrong to assume that activities undertaken during contacts were limited solely to what is recorded. All contacts would be likely to involve an element of mental state examina tion. However, these assessments, if unrecorded are not available to others for future reference. Unlike other studies (Kendrick et al 1994), all patients had contact with at least one agency and no one was lost to review. There was a positive correlation coefficient of 0.899 between the HoNOS ratings and the number of agencies involved in patients' care. This finding is sig nificant and encouraging, suggesting appropri ate targeting of the local mental health services.
However, despite this multi-agency involve ment, information sharing between the agencies was partial. Analysis of correspondence in the patients' general practice records showed that only 9% of CPN contacts during 1994 resulted in a letter to the GP. The figure rises to 74% for psychiatrist contacts during the same period.  This still leaves the GP uninformed of one in four psychiatric contacts. Informal discussion with local representatives revealed most agencies had little idea of the levels of input from other agencies. It is hard to be proactive in these circumstances. Inevitably, as a consequence of not having a clear 'shared care' plan duplication of effort is likely. Analysis of the activities recorded at these contacts reveals some overlap between agencies, all having medication renewal and mental state examinations as their two most common entries.
It is generally accepted that the prevalence of serious mental illness is in part related to the social deprivation of the area (Campbell et al, 1989). The significant levels of social deprivation within the practice area will inevitably present a challenge to those delivering services to the seriously mentally ill. Where case-loads are high and demographics unfavourable, the care avail able must be optimised. This group of 42 patients received a minimum of 601 contacts. While some overlap in specialist functioning is inevitable, given the absence of effective sharing of information, this input is likely to be poorly focused and sub-optimal. A shared care plan would allow an appropriate division of tasks among the specialities, hopefully reducing dupli cation of effort.
Previous studies have shown GPs to be ambivalent about the extent of their commitment to the management of the seriously mentally ill in the community (Kendrick et al, 1991). They have shown themselves less likely to perform mental state assessments and adjust drug regimes less often than for the chronically physically ill (King, 1992). Indeed, a recent General Medical Services Committee publication cautioned GPs against becoming keyworkers in such care (General Medical Services Committee, 1996). Yet the work and contact remains predominately in general practice.
Examination of the data on patterns of care reveals there are no patients cared for by the CPNs and GPs alone. In west London, as in any other parts of the country, the psychiatric services are 'sectorised' making individual con sultant psychiatrists responsible for patients living in a particular patch. Often these patches are coterminous with social service areas and CPN services, which are also often geographically placed. This allows psychiatrists, CPNs and social services in a particular patch to develop closer working relationships.
However, it ex cludes the GPs whose practices do not neatly fit into individual social service 'areas'. In this study, the practice has patients in four different sociali service areas, contact with several psy chiatrists and six CPNs. It is difficult to develop close working relations with this many profes sionals. Yet, as the findings of this study support, for those patients with schizophrenia who are registered, it is the GPs who have by far the most contact with the schizophrenic popula tion both in terms of numbers and volume. Consideration should be given to the possibility of realigning the services allowing community teams to become more practice-area sensitive. A consultant psychiatrist might be responsible, with a team, for providing care to all patients from an appropriate number of practices. Provi sion would need to be made for unregistered or homeless patients within the practice area. Staffing levels within the GP aligned community teams could reflect this need. Indeed, with such teams, GPs might be encouraged to become more involved with this challenging group of patients. Clearly there are a variety of obstructions to increased GP involvement with the seriously mentally ill including interest, confidence and resources. Yet if the concept of a primary care led NHS is to have any weight, this sort of realign ment must be considered.
In many ways the management of schizophre nia in general practice today is similar to the management of diabetes mellitus some 15 years ago. GPs did not perceive themselves as being primarily responsible for their patients with diabetics. Diabetes was thought to be a hospi tal/consultant responsibility, and yet GPs came to realise that denying prime responsibility did not necessarily prevent complications or reduce the workload from these patients with chronic illness. The same is true for the seriously mentally ill. With a commitment to proper shared care, adequate training, resources and effective information sharing, GPs could use their sig nificant contact with these patients in a much more positive and useful manner.