The social adjustment of patients with schizophrenia : implications to the mental health policy in Brazil *

MENEZES, P.R. & MANN, A.H. The social adjustment of patients with schizophrenia: implications to the mental health policy in Brazil. Rev. Saúde Pública, 27: 340-9, 1993. A sample (n=124) of schizophrenic patients from a defined catchment area of the city os S.Paulo, Brazil, who had been consecutively admitted to hospital, was assessed for psychopathological status and social adjustment levels. Sociodemographic, socio-economic and occupational characteristics were recorded: almost 30% of the subjects had no occupation and received no social benefit, more than two-thirds had a monthly per capita income of US$ 100.00 or less. Sixty-five percent presented with Schneiderian firstrank symptoms. Nearly half the sample showed poor or very poor social adjustment in the month prior to admission. The most affected areas of social functioning were participation in the household activities, work and social withdrawal. The current mental health policy of promoting extra-mural care as an alternative to the previous hospital-based model will then mean the investment in a network of new community-based services, that give effective treatment and support to patients and their families. The need of further research into the current picture of mental disorders in the country is stressed.


Introduction
Schizophrenia is an important public health problem in developed countries, with prevalence rates of 5/1,000 adult population 24 .Its costs to society are high, and comprise direct treatment costs (in-and-out-patient hospital care, residential care, community-based services), indirect costs due to lost of production, and intangible ones, such as human suffering due to the illness.Direct costs, which are most easily calculated, were estimated as £ 1,670 per year in the U.K., representing 1.6% of the total health and personal social expenditure in that country in 1987 14 .In Australia, the cost per new case was calculated as being U.S.$ 131,333, and the total direct costs reached as much as half the amount spent on myocardial infarction 4 .There is a lack of information about the burden schizophrenia represents to society in developing countries, but extrapolations from incidence rates and demographic trends suggest that the number of cases under care in these countries will grow by 87% from 1975 to the year 2000 28 .
Despite the more optimistic views about the outcome of schizophrenia in recent years, it continues to be a disabling condition for at least one third of those affected by it 21 .Attention has been given to the evaluation of social disability as a specific consequence of the illness, only partially associated with severity of psychopathological status 12,16 .Such investigations have showed that there is at least some degree of social disability in two-thirds of the individuals suffering from schizophrenia.Adequate knowledge of the types and intensity of this disability is necessary for the planning of services and programmes targeted at providing extramural care for schizophrenic patients 37 .

Schizophrenia in Brazil
Although to date there has been no epidemiological study aimed specifically at evaluating the prevalence of schizophrenia in Brazil, the rate seems to be similar to those figures found for other countries, being around 5/1,000 adult population 1,7,32 .The diagnosis of schizophrenia accounts for about one-third of all psychiatric admissions 36 .
Regarding social disability, Shirakawa 38 investigated a sample of 44 schizophrenic patients who had been attending a private psychiatric office for many years, and assessed their social adjustment using a standardized instrument.The sample had high levels of social disabilities.Less than 15% of them were receiving salaries that could allow them to live independently.Relatives expected higher levels of social functioning from males than from females.However, this sample probably consists of a group that is socially and culturally different from patients attending the psychiatric hospitals of the public health system in Brazil.

Mental health policy in Brazil
During the 1960's and 1970's the mental health policy in Brazil was characterized by the hospital-based model, with an increasing number of psychiatric beds throughout this period 2,31 .In recent years there has been a shift in the mental health policy towards a communityorientated care for mentally ill people 2,22 .The number of psychiatric beds fell from 105,765 in 1984 to 85,037 in 1991, a 20% reduction 2 .
This new policy will have important implications for the patients, their families and the communities they live in however, the current lack of information on the characteristics of those patients that are being affected by change in policy, in particular their social disability levels, makes difficult adequate planning for services and programs, and for the training of professionals properly qualified to deal with them and their families.
In the present study, a sample of schizophrenic patients who lived in a defined catchment area of the city of S. Paulo, and who were consecutive admissions to the psychiatric hospitals of that area, was assessed for inclusion in a prospective study on patterns of contact with health care services and outcome of schizophrenia.This present paper aims to describe their characteristics, with special reference to their social adjustment, and to discuss the findings in the light of the current mental health policy in Brazil.

The catchment area
The catchment area chosen is a regional health district located in the Northern region of the city of S. Paulo, identified as the "Sistema Descentralizado e Unificado de Saúde -Região 6 (SUDS-R6)".The psychiatric health care system in this area is organized in accordance with the principles of regionalization, with hierarchic levels of complexity and integration between services, characteristics that facilitate the execution of epidemiological studies 36 .The area is served by apsychiatric ward in a general hospital, with 20 beds, and 3 psychiatric hospitals with a total of about 1,500 beds.There is no shortage of psychiatric beds for individuals living in the area.All psychiatric admissions are centrally recorded and managed by a psychiatric register, on a daily basis.The register records the name, address, sex, age, the psychiatric diagnosis at admission, the unit that referred the patient for admission and the hospital which admitted the patient.
The aim of case-finding was to select individuals presenting with specific symptoms and behavioral abnormalities indicative of schizophrenia, definite or probable, regardless of pre-existing diagnostic labels.The psychiatric register was scrutinized to identify all possible cases of schizophrenia that were living in the study catchment area, were aged between 15 and 44, who had had an admission to one of the psychiatric hospitals of the area during the period between 1 June and 22 August 1991, with a diagnosis of either 'schizophrenia', 'paranoid psychosis' or 'other functional psychoses', recorded according to the 9th version of the International Classification of Diseases 46 .
Medical records were then checked, and exclusion and inclusion criteria were applied.Exclusion criteria were: alcohol or drug dependence in the last year, severe mental retardation, organic psychoses, epilepsy, hearing or speech difficulties that could impede the administration of interviews, and planning to leave S. Paulo after discharge from the psychiatric hospital.Inclusion criteria were at least one of the following symptoms: mood-incongruent delusions, definitely inappropriate or unusual behaviour, hallucinations, disorders of thinking or language, other than acceleration or retardation; or at least two of the following: psychomotor disturbance other than acceleration or retardation, social withdrawal, overwhelming fear, and self-neglect.
If patients were still eligible for entry into the study, they were then interviewed for the assessment of psychiatric symptoms, after giving their consent to participate in the study.The interview normally took place during the first two weeks of admission.That procedure was followed by an interview with a key informant, generally a relative of the patient, when the scale for evaluating social adjustment was applied.

Assessments
Assessments were made in a similar way for all patients.They included data on sociodemographic variables (sex, age, ethnic group, marital status, religion, educational level, occupational status, profession, living arrangements, and monthly economic income), psychiatric history and diagnosis, psychopathologic status and social adjustment in the month prior to admission.
For the assessment of psychiatric symptoms a standardized psychiatric interview was utilized, the 9th version of the Present State Examination (PSE) 44 .A classification based on the cross-sectional symptomatology was given by the CATEGO, a computer program developed for use with the PSE 45 .The investigating psychiatrist was trained in the use of the instrument by the MRC Social Psychiatric Unit, in London, and used a Brazilian translation of the instrument into Portuguese 8 .By adding a 30-item questionnaire covering longitudinal aspects of the disorder, diagnoses according to the "Diagnostic and Statistic Manual of Mental Disorders" 3 (DSM-III-R) were also obtained.
For the assessment of social disability the WHO Psychiatric Disability Assessment Schedule (DAS) 47 was used.This scale has been specially developed for the purpose of assessing social functioning in a variety of different settings and cultures, is available in many languages, and has been used in several countries in a series of multicentre studies coordinated by the WHO.
In the present study a translation into Portuguese 10 was applied by one of 2 interviewers, a psychiatrist and a psychologist, to a key-informant of the patients, generally a relative.This interview was carried out independently of the psychiatric examination.Sections 1,2 and 5 were used (see Appendix).Section 1 deals with overall behavior, and gives the degree of disability for activities of daily living and relates to the immediate physical environment.Section 2 is an inventory related to the perfomance of social roles, such as work, parental duties, sexual relationships.Section 5 is a global evaluation of social functioning based on all the information obtained for sections 1 and 2. Scores for each item in sections 1 and 2 range from 0, 'no dysfunction', to 5, 'maximum dysfunction'.Scores for section 5 range from 0, 'excellent or very good adjustment', to 5, ' severe maladjustment'.The baseline against which the patients behaviour is measured is the presumed 'average' or 'normal' functioning of a person of the same sex and of comparable age and socio-economic environment.An inter-rater reliability assessment was carried out by means of 20 simultaneous interviews carried out by the 2 interviewers.

Analysis
Data handling and analysis were carried out using the computer program for epidemiological studies Epi Info 5 17 .The distribution of socio-demographic and socio-economic variables, psychopathological status, and social disability levels were examined.Chi-square tests were carried out to examine differences between groups for categorical variables, t-tests for differences in continuous variables, and correlation coefficients were used to examine the relationship between continuous variables.Inter-rater reliability was assessed through the calculation of intra-class correlation coefficients 5 .

Results
Of 217 patients screened during the period between 1 June and 22 August 1991,124 were included in the study (69 men and 55 women).The remaining 93 subjects were excluded, the most common reasons being non-catchment area (n = 20) and alcoholism (n=19).
The mean age was 32.6 years, ranging from 19 to 44 years (SD = 6.9).The sample was predominantly white, single, Catholic, born outside S. Paulo, and with more than 4 years of education (Table 1).The monthly per capita income ranged from US$ 16.50 to US$ 750.00, with 71.7% of the individuals having a montly per capita income of US$ 100.00 or less (Figure 1).The proportion of subjects with any occupation at the time of the admission was low (26.6%), and only 24 were working regularly, either as employees or self-employed.Twenty-seven subjects were on social benefits, males predominating in this group (20 vs. 7 females, x 2 = 4.75, l d.f., p = 0.03).

Psychiatric characteristics
Ninety-four subjects (75.8%) had had one or more admissions to psychiatric hospitals in the past, and 27.6% (n = 34) had had 10 or more psychiatric admissions; 30 subjects were first admissions.The average length of illness was 8.3 years (S D = 6.9), randing from 0 to 27 years.Men had more years of illness than women (9.5 vs. 6.7,p = 0.021).
The distribution of diagnoses according to the DSM-III-R and the frequency of CATEGO classes are given in Table2.It can be seen that 71.8% of the subjects were classified as 'schizophrenia' by both criteria, if one considers as such classes S and P for the CATEGO and schizophrenia, schizophreniform and schizoaffective psychoses for the DSM-III-R are include in this broad category.Only 6 subjects (4.8%) were not classified as schizophrenia by either CATEGO or DSM-III-R.

Social adjustment
The WHO/DAS was applied to 115 patients; it was not possible to obtain information in 3 cases, due to lack of contact with relatives prior to admission, and 6 cases had been at home for less than 15 days between the index admission and the previous one.The inter-rater reliability of the WHO/DAS was very good, with intra-class correlation coefficients ranging from 0.88 to 1.00.Further details on the reability study may be found elsewhere 33 .
The distribution of section 5 (Global Evaluation) scores in shown in Figure 2; 81.6% of the subjects were assessed as having at least some impairment in social functioning (score 1 or above), and 50% were assessed as having poor or very poor social adjustment (scores 3 and 4, respectively) in the month prior to admission.Global scores were positively correlated with length of illness (r = 0.26,0.08< r < 0.42, f = 7.92, p < 0.01), when the later was grouped as 0-1,2-5,6-10 and 11 or more years of illness and coded as an ordered categorical variable.
The analysis of the frequency distribution of scores for each item of the schedule allows the identification of those areas of social adjustment that were the most affected by the illness (Table 3).Participation in household activities was the item with the highest proportion of subjects scoring 1 or more(73.9%),followedclosely by social withdrawal (73.1%).If items for work performance (2.7) and interest in getting a job (2.8) are considered to represent dysfunction in the same area and are added together, 72.2% of the patients had been given a score of at least 1.The same was done for items on sexual role for married and unmarried subjects (items 2.3 and 2.5, respectively); 66.9% had a score of 1 or above.On the other hand, social contacts (27.8%), slowness (33.1%), and behavior in emergencies (37.4%) were the items where fewer subjects were reported to have any disability.
In terms of severity of disability, when considering work performance and interest in getting a job taken together, this was the area with the highest proportion of subjects showing marked degrees of disability, with 45.2% of them scoring 4 or 5. Participation in the household followed, with 40% of subjects scoring 4 or 5. Sexual relationship (married and unmarried together), interest in information and social withdrawal also had high proportions of scores 4 or 5 (39.0%, 37.4% nad 35.7%, respectively).

Discussion
This is the first study carried out in Brazil in which a sample of schizophrenic patients from a defined catchment area, who had been recently admitted to psychiatric hospitals of the public health system, were assessed systematically with standardized instruments for mental state and social adjustment.Almost 30% of the subjects had no occupation or social benefit.More than two-thirds had a monthly per capita income of US$ 100.00 or less.Sixty-five percent presented with Schneiderian first rank symptoms.Nearly half the sample showed poor or very poor social adjustment in the month prior to admission.

Methodological considerations
The sample was limited to patients aged 15-44 but those over this age range represent only 20% of the admissions for non-affective functional psychoses in the study catchment area.The exclusion criteria limited the sample to patients without concurrent drug or alcohol dependence.This subgroup, which represents something around 15% of all patients with diagnosis of non-affective functional psychoses who live in the area and are admitted psychiatric hospitals, may have worse levels of social adjustment and deserves further investigation.The sample is also restricted to those patients who use the public health system, but only a very small proportion of the population can afford pychiatric in-patient treatment on a private basis 22 .Apart from these limitations, the distribution of clinical diagnoses, age and sex in the sample are similar to those observed during a 1-year period (01.07.90 to 30.06.91) for both the catchment area and a neighbouring area that uses the same psychiatric hospitals as their in-patient resources, according to data from the register.Therefore, the sample is very likely to be representative of all patients with schizophrenic psychoses in the age range 18-44 years who have a fixed address and are admitted to psychiatric hospitals of the public health system in S. Paulo without the associated conditions listed in the exclusion criteria.
The assessment of psychiatric symptoms and subsequent diagnostic classification present a considerable degree of subjectivity, but reliability between different investigators has been markedly improved by the use of standardized instruments 42 .This is the case with the PSE, which has showed good reliability coefficients, and has been widely used in psychiatric research all over the world 44 .
Similar difficulties exist for the evaluation of social adjustment levels 25 .The schedule chosen for this study, the WHO/DAS, was developed specifically for the assessment of patients behaviour and social functioning in the particular social and cultural settings.Acceptable levels of reliability have benn achieved, and its items have proved to be valid for the purpose of measuring social disability 16 .

Levels of social disability in the sample
The proportion of patients with any social disability was very high.More than 80% of the sample showed some degree of social disability (global evaluation scores 1 to 4), and almost 30% were severely disabled in their global social functioning (score 4).Ciompi 11 ,in a long-term follow-up study of 289 schizophrenic patients in Switzerland, found that the overall social adaptation appeared to be good or fair in only one-third old the sample by the end of the study.In a study of 70 first episode schizophrenics in Germany, Schubart et al. 37 reported that two-thirds had DAS global scores of 3 (poor adjustment) or more at 2-year follow-up.In the study by Shirakawa 38 , in Brazil, 43.2% of the sample had low scores (poor) and another 47.7% had intermediate scores in the assessment of global social functioning .The positive correlation between global evaluation scores and length of illness confirm the general view that the longer the condition lasts, the worse is the outcome 13 .
Regarding specific areas of social functioning, it becomes clear that some areas are more affected than others, mainly participation in the household activities, social withdrawal, work and sexual relationships.De Jong et al 16 , in a two-year follow-up study of 82 recent onset schizophrenics in Holland, showed that there is a hierarchical disruption in social roles and behaviours, the most affected areas being work, heterosexual and social contacts, while self-care remains relatively preserved.Westermeyer and Harrow 41 , in a review on the course and outcome of schizophrenia, pointed out that impaired social and family relationships are frequently observed in these patients, and that work impairment is a central feature of the illness.

Implications for a community-based mental health policy
The mental health policy in Brazil has now been orientated towards dehospitalization and communitybased care, following a trend observed in some developed countries for more than 30 years now 6,40 .Positive results for this policy in other countries have been published, both in terms of patient outcome 20,39 and of economic costs 19,43 .However, the negative consequences of such a policy of dehospitalization not followed by a corresponding increase in the investment in community care has also been described 26,30,35 .Many investigators have stressed the need of well planned extra-mural facilites and programs, if such a policy is to succeed 9,26,28 .
Community-based services will have to provide comprehensive treatment schemes, including pharmacological, psychological and social management 23 .The usefulness of neuroleptic drugs in the treatment of schizophrenia is well documented 15 , but under normal treatment conditions results are as not as good as those observed in clinical trials 27 .Almost 50% of the patients have a relapse within one year of their last episode, and one of the main reasons is the lack of compliance with neuroleptic medication.Therefore, the active monitoring of patients' contacts with extra-hospital services in order to ensure compliance with medication is essential.
Rehabilitation programs are another necessary dimension of the community-care strategy, and their need in Brazil can be seem by the high levels of disability in social functioning found in this study.They should be provided by multidisciplinary teams, for patients' needs are of various orders 23 , occupational skill training, actual help in seeking and adequate housing provision being among the main issues for the proper maintenance of schizophrenic patients in the community 34 .
Thirdly, patients' relatives cannot be forgotten, whether they are those real carers for the patients or not 29 .The burden relatives face when they have a schizophrenic patient at home has been well documented, and involves a damaging effect on social and leisure activities, financial and employment difficulties, and mental and physical health problems.Moreover, they are generally neglected by mental health professionals involved in the patients' treatment, even in the provision of information about the nature of the illness and how best to deal with these patients in their family surroundings.Adequate support for relatives is then absolutely necessary, and can even be cost-effective, in the measure in which it reduces future morbidity in both patients and the relatives who care for them.
The creation of population-based case registers make a valuable resource for the provision of detailed, longitudinal, inter-agency statistics that allow the study of the incidence, prevalence, course and outcome of schizophrenia, as well as service utilization patterns and evaluation of specific forms of care 18 .They are also very helpful in identifying those patients who lose contact with their treatment centres.
The implementation of such comprehensive care in Brazil will require careful assessment of the needs of each individual patient and good liason between the services and facilities patients need.Even if such services are provided, there is still the need of careful and continuous monitoring of their adequacy and effectiveness in meeting their original targets 20 .
A community-based network for the care of schizophrenic patients will require considerable re-allocation of resources, from spending on hospital care to the extra-mural system, extra money beig at the outset necessary for the implementation of the new network.However, in the area where this study was carried out there were only two out-patient clinics for a population of about one milion inhabitants, and no new community mental health facility had been created for the last 5 years.There has been a constant loss of professional staff, due to the low salaries and work overload.With the exception of a few isolated cases, there is no reason to imagine that the situation is too different in other regions of the country.
The Mental Health Section of the Brazilian Health Ministry has recently stressed its commitment to a community-based care strategy, established the change in financing from hospital beds to community-based services as a priority action, and proposed a joint action with the State Health Secretariats in order to implement such a policy 2 .Whether in practice health administrators will accept the 20% reduction in psychiatric beds only as a helpful economy measure resulting from the new policy in a moment of deep economic crisis or whether they will recognize the need to invest in community services and reallocate resouces remains to be seen.