Comparison of mental health assistance in primary care settings with or without Family Health Program team

Abstracts

INTRODUÇÃO: O objetivo deste estudo foi comparar o perfil de assistência em saúde mental realizado por unidade básica de saúde (UBS) com equipe de Programa de Saúde da Família (PSF) e sem equipe de PSF. MÉTODO: Estudo observacional, avaliando pacientes encaminhados por UBS da área de abrangência de um serviço especializado de saúde mental no período de abril de 2003 a março de 2006. RESULTADOS: A UBS com equipe de PSF apresentou melhor padrão global de registros de dados, maior responsabilidade exclusiva do médico em suas referências ao nível especializado (p = 0,000), menor capacidade de retenção dos usuários na UBS (p = 0,099), maiores taxas de abandono de tratamento em nível secundário (p = 0,060) e menor percentual de contrarreferência pela equipe especializada (p = 0,028). A taxa de concordância diagnóstica global foi semelhante entre os dois modelos de UBS, com razoável nível de concordância (índice kappa de 44,5 e 43,0%, respectivamente, para UBS com e sem equipe PSF). CONCLUSÃO: A UBS com equipe de PSF não apresentou resultados compatíveis com o que seria de se esperar, em função de sua hipotética melhor qualidade de estrutura.

Saúde mental; atenção básica a saúde; avaliação de serviços de saúde; Programa Saúde da Família

ORIGINAL ARTICLE

Comparison of mental health assistance in primary care settings with or without Family Health Program team*

Mário Sérgio Ribeiro^I; Márcio José Martins Alves^II; Priscila Matthiesen e Silva^III; Eveline Maria de Melo Vieira^IV

^IPhD in Philosophy. MD. Associate professor of Psychiatry, School of Medicine, Universidade Federal de Juiz de Fora (UFJF), Juiz de Fora, MG, Brazil.

^IIPhD in Community Health. MD. Associate professor of Community Health, School of Medicine, UFJF, Juiz de Fora, MG, Brazil.

^IIIMedical student, research assistant, scholarship holder, Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG).

^IVMedical student, research assistant.

This study was carried out at the Laboratory of Researches on Personality, Alcohol and Drugs (LAPPDA), School of Medicine, Universidade Federal de Juiz de Fora (UFJF), Juiz de Fora, MG, Brazil.

Correspondence

ABSTRACT

INTRODUCTION: The objective of this study was to compare the profile of mental health assistance provided at primary care units (PCUs) with and without a Family Health Program (FHP) team.

METHOD: Observational study evaluating patients referred by PCUs located in the coverage area of a specialized mental health institution between April 2003 and March 2006.

RESULTS: The PCU with a FHP team presented better global standards for data recording, higher exclusive participation of medical doctors on their referral to specialists (p = 0.000), lower capacity of patient retention (p = 0.099), higher rates of treatment dropout in secondary level (p = 0.060), and lower percentage of counter-referral by the specialized team (p = 0.028). The overall index of diagnostic agreement was similar for both types of PCU model, with a reasonable level of agreement (kappa index of 44.5 and 43.0%, respectively, for PCUs with and without a FHP team).

CONCLUSION: The PCU with a FHP team did not present results compatible with what would be expected based on its hypothetically better quality.

Keywords: Mental health, primary health care, health services evaluation, Family Health Program.

Introduction

In Brazil, there has been a constantly renewed and restarted process of changes in the focus of the assistance provided to individuals with mental disorders: institutions have been changed — or, at least, their names have been changed —, assistance policies have been changed, theories have been changed and technical proposals have been changed. However, we should ask ourselves if the assistance delivered has been really transformed.¹

The World Health Report,² focused on mental health, emphasizes that the control and treatment of mental disorders, within the context of primary health care (PHC), are crucial for: providing a larger number of people with easy access to health services; improving the quality of health care; reducing the financial losses due to unnecessary diagnostic investigations and unspecific or inappropriate treatments.

Although there is evidence that mild mental disorders can be treated in a more effective manner at primary health care settings, many general practitioners and even some patients prefer the referral to specialized mental health teams.³ Ten years ago, Kapczinski et al.⁴ pointed out that the core problem of delivering general health care in Brazil lay in the fact that these services were not meeting the goal of increasing the population's level of physical, mental and social health. Even though it is possible to consider that every health problem is always a mental health problem⁵ and that every mental health intervention is also aimed at promoting health,⁶ the situation described by Kapczinski et al.⁴ seems to remain the same to this date.

The Family Health Program (FHP), which was implemented by the Ministry of Health in 1995, followed the successful Program of Community Health Agents, which was implemented in 1991. These programs are focused on families, not on individuals, and their health professionals must actively contact the population they are responsible for.^7,8 Since 1996, the FHP has served as a tool used in the reorganization of the Public Health System (SUS) in Brazil, contributing to promote a change in the paradigm of the prevailing care practice.⁹

A study conducted in the state of Minas Gerais found that, among the individuals detected by the FHP health professionals as possibly presenting psychological problems, 37.8% showed evidence of nonpsychotic mental disorders.¹⁰ Similarly, a diagnostic questionnaire administered to patients seen by FHP teams in the city of Petrópolis, state of Rio de Janeiro, identified high prevalence rates of common mental disorders — characterized by unspecific depressive and anxious complaints — in the population assisted by these teams: 56% of the subjects had a common mental disorder and, among them, approximately 33% might have a more severe mental disorder.¹¹

Such findings emphasize the need that the PHC teams — and mainly the FHP teams — be capable to deliver health care to people with less severe mental disorders, instead of simply referring them to specialized services.^10,12

Nevertheless, in the daily practice there is still low effectiveness in terms of the care delivered to patients with mental health complaints at the primary level of health systems. One of the factors that decreases the quality of such assistance is professionals' lack of capacity to establish correct diagnosis and to treat these mental pathologies detected in their usual clientele. The presence of mental suffering in the patients seen in the primary health care network is usually unnoticed in spite of its high prevalence.^11,13,14

The consultation-liaison model³ is a model of interface work between primary care and secondary mental health services highly focused on the development of close relationships between the PHC team and mental health specialists with the purpose of: a) reduce the number of referrals for treatment of mild disorders; b) supporting the referral of severe mental disorders; and c) increasing the general practitioners' competence to detect and to manage mental problems. According to Bower & Gask,¹⁵ this model has significant advantages in comparison with other models of mental health care and has already been considered quite effective in terms of primary mental health care.

Based on the consultation-liaison perspective, a reformulation of the model that was used in mental health care was initiated in Juiz de Fora, state of Minas Gerais, in 1997. The Mental Health System of the city of Juiz de Fora (SMSM-JF) was basically focused on the insertion of the mental health subsector in the regionalized, decentralized and hierarchically organized system, spreading the mental health interventions to the city's Primary Care Units (PCUs).¹⁶

The SMSM-JF includes 11 different Regional Mental Health Reference Centers (CRRESAM), each one of them delivering care to a group of 41 PCUs of a certain area of the city of Juiz de Fora. CRRESAM-West, which was assessed in this study, was responsible for the patients referred by five PCUs of the area known as "uptown." In the SMSM-JF, PCUs are responsible for performing the patients' initial evaluation using the instruments and the conduct guidelines suggested in a manual designed for that purpose — the Conduct Protocol of SMSM-JF.¹⁷ If necessary — due to the severity of the case —, the health professionals who work at these units refer the patient to a specialized team responsible for seeing the patients and supervising the teams of its coverage area, providing training and permanent education to the technicians working at the PCUs. All health professionals working at the PCUs within the coverage area of CRRESAM-West, with and without FHP teams, had been previously trained, and the teams of each PCU were regularly supervised by technicians from the secondary health care service with the purpose of continually improving the assistance process. It is worth noting that, according to this assistance model, PCUs are expected to refer to the secondary level those patients with more severe disorders who are motivated and available to be treated in a more intensive manner by the specialized team.

After the evaluation by the specialized team, its health professionals may counter-refer those patients for follow-up at the PCUs or refer them to special programs that provide care to specific groups.¹⁸ Within this system, the new tasks carried out by the professionals of the secondary level, as well as the change in the style of specialized practice required from these professionals, are a concrete illustration of the necessary change in the working pattern suggested by different authors who are dedicated to the study of primary mental health care.^19,20

The objective of the present study is to compare the profile of mental health assistance provided by two different PHC models — the FHP model, which is represented by one of the PCUs, and the traditional model, without FHP teams, represented by four PCUs. All PCUs included in this study were linked to the same reference center and their professionals (physicians, nurses and social workers) were similarly trained to work in the SMSM-JF.

While the four PCUs representing the traditional model participated in the pilot project of the SMSM-JF implementation in 1997, the PCU with a FHP team was installed in the year 2000, with a complete FHP team from the start. With the purpose of assessing the quality and effectiveness of the assistance process,^21,22 we specifically analyzed: capacity of patient retention at PHC level; quality of medical records; integration of PCU teams; agreement between diagnostic hypotheses suggested by primary health care professionals and the diagnoses established by the secondary level professionals; and patients' treatment dropout or adherence.

Method

Study design and inclusion criteria

We started by conducting an observational study, without intervention by the investigators in terms of subject allocation or therapeutic procedures. We assessed the medical records of patients referred by the PCUs located in the coverage area of the secondary level services in three different time periods — from April 2003 to March 2004, from April 2004 to March 2005, and from April 2005 to March 2006. All patients that attended at least one medical visit with the specialized team, regardless of age, were included, in a total of 1,397 subjects. Fifteen patients who were referred for reevaluation within one of the three time periods were not included.

Variables and parameters

The characteristics evaluated are part of the semi-structured clinical medical records provided by the SMSM-JF, as well as the data collected by the investigators throughout the duration of treatment delivered by the secondary level professionals. The insertion of data in the database and the analysis of the results were performed using the Epi-Info 6.04d and the Statistical Package for the Social Sciences (SPSS) for Windows 14.0 (use license CFOP 9656438).

Main diagnoses were established by the specialists according to the International Classification of Diseases (ICD-10),²³ the investigators grouped the diagnoses between F50 and F69 as a unique category of "other psychiatric disorders" and between F80 and F99 as "child or adolescent mental disorders." The diagnostic hypotheses established by the technicians of the PCUs must be inserted in the semi-structured referral chart. It is important to point out that the technicians belonging to all professions at the primary level were trained to suggest basic psychiatric diagnostic hypotheses using the conduct protocol of SMSM-JF.¹⁷ In order to adapt to the assistance standard established in this assistance protocol, the variable "treatment status" of the patient — that is, his/her localization in the assistance system — was assessed 2 months after each time period studied.

For assessing the quality of the referral to the secondary level, we used five parameters: 1) percentage of patients referred compared to the total basic health interventions effectively accomplished by each PCU; 2) percentage of failure by the PHC professionals to record data; 3) percentage of referrals by each profession of technicians of the primary health care team; 4) agreement rates between the diagnostic hypothesis suggested by the PCUs technicians and the diagnosis established by the secondary level professionals; and 5) variable "treatment status," that makes possible to quantify those patients who remain being treated by the specialized team, those who were counter-referred to the PCUs, those referred to special programs or the ones who dropped out from treatment at the secondary level. Using such parameters, we estimated: the capacity of patient retention in the PHC service; the adequate recording of medical data; the integration of the PCUs team while evaluating those patients referred; the capacity of the PCU professionals to identify mental disorders (diagnostic hypothesis) and their capacity to make adequate referrals to the secondary level by comparing their diagnostic hypotheses with the diagnoses established by the specialist team; and the treatment dropout rates or treatment continuity rates.

The variable "responsible for referral" was defined considering the three professions involved in the PHC (physicians, nurses and social workers) who took responsibility for the referrals of each patient by signing the semi-structured medical chart.

Statistical analysis

In order to describe the patients' characteristics (Table 1), we present, for each variable, the percentages related to the category in which the group of the PCU with a FHP team had higher response rates, considering the data effectively recorded. The percentages of missing data regarding the characteristic assessed in terms of the total number of subjects evaluated are shown between parentheses.

Table 1 - Click to enlarge

For performing the cross-tabs procedures (contingency tables), the variables were grouped into two categories. In order to do that, we used: 1) categories "above" or "below" the mean value for quantitative variables; and 2) restriction of categories to two opposing possibilities — such as being catholic or being non-catholic —, for categorical variables. Subjects for whom the effective recording of data was not available for each specific variable (missing cases) were not taken into consideration in the contingency tables. Pearson's chi-square test was used to assess the significance of the results found in cross-tabs comparing the PCU with a FHP team to all the PCUs without FHP teams. P-values were considered statistically significant when £ 0.05 and marginally significant when £ > 0.05 up to 0.1. The charts show the regrouped variables and the tables show the variables that were not regrouped; in such case, the chi-square test was calculated considering each category in relation to all the others.

The kappa index was used to check the agreement between the diagnostic hypothesis suggested by the PCU technicians and the diagnosis established by the secondary level professionals. Kappa values between 100 and 80% are considered excellent; between 80 and 60% are good; between 60 and 40% are fair; between 40 and 20% are minimal; and lower than 20% are considered negligible.²⁴ Between April 2003 and March 2004, the patients who dropped out from treatment were not appropriately distinguished from those counter-referred; therefore, it was not possible to assess the dropout category during this period.

Ethical aspects

The present study was approved by the Research Ethics Committee of Hospital Universitário of Universidade Federal de Juiz de Fora (UFJF), process no. 377.062.2004, and by the Research Ethics Committee of UFJF (report 193/2007) and it is in accordance with the ethical principles included in the Declaration of Helsinki.

Results

We found high percentages of missing data for both models of organization of the PHC services. A larger number of female patients were referred. These patients informed to be white, catholic and their city of origin was Juiz de Fora. Their mean age was 39.2 years old. There was a higher prevalence of single subjects (or individuals who did not have a stable relationship), who had children, who lived with more than three people, who worked in sales or providing services, who were currently unemployed and had low family or personal income. Most of the subjects were economically dependent on another person and were not the breadwinner.

The number of patients who were referred and actually attended the secondary mental health service was evaluated in relation to the number of basic health interventions performed by the PHC teams (Table 2). We found that the number of patients who were referred and attended the specialized mental health care service is quite small in relation to the number of basic health interventions performed at the PCUs. It is important to consider that, even though each referral indicates one specific patient, the total number of basic interventions does not represent the number of different patients seen at the PCUs. Nonetheless, we noticed that the PCU with a FHP team had a higher percentage of referrals than the PCUs with a traditional team. Even if the percentages of referrals were similar — 0.78 and 0.70% for, respectively, the PCU with and without a FHP team —, the large number of subjects included in the present study enabled us to identify a marginally significant difference between both models studied (p = 0.099).

Table 2 - Click to enlarge

The rate of patients referred without the appropriate identification of the "professional responsible for the referral" was 17.9% for the traditional PCUs and 11.5% for the PCU with a FHP team (p = 0.009). Table 3 shows that the percentages had a statistically significant variation for all professions: the PCU with a FHP team was responsible for the largest number of referrals made exclusively by physicians, whereas the traditional PCUs referred the highest percentage of patients with more than one professional of the team signing responsible.

Table 3 - Click to enlarge

The actual registration of a "diagnostic hypothesis" was more frequent by the PCU with a FHP team than by the traditional PCUs — respectively, 73.90 and 65.8% (p = 0.013). For the "main diagnosis" established by the team of specialists, we found that this information was missing from the medical records only in 6.0% of the cases (Table 4). With regard to the diagnostic hypotheses, only organic mental disorders showed a marginally significant difference (p = 0.072) between the two PCU models. In terms of the diagnosis established at the secondary level, we found a marginally significant difference for the mental retardation group (p = 0.053).

Table 4 - Click to enlarge

Table 5 shows the levels of agreement between the diagnostic hypothesis and the main diagnosis. We found that the hypothesis of a disorder due to the use of psychoactive substance was responsible for the best agreement rates: kappa coefficient²⁴ of 78.7 and 81.9%, reaching an excellent level at the PCU with a FHP team. The overall agreement rate was similar between both PCU models, with a reasonable level of agreement (44.5 and 43.0%); however, for the psychotic conditions, the agreement rate found for the PCU with a FHP team was much lower than for the PCUs without a FHP team (14.0 and 31.5%, respectively). We could not calculate the kappa coefficient for personality disorders at the PCU with a FHP team; for the traditional PCUs, the agreement was negligible (8%).

Table 5 - Click to enlarge

Table 6 shows the localization of the patients within the mental health system after being evaluated at the secondary level, that is, the number of patients who dropped out from treatment at the secondary level, those who continued to be treated at this level, those counter-referred to the PCUs or referred to special programs. In a marginally significant manner (p = 0.060), treatment dropouts at the secondary level were more frequent among the patients seen at the PCU with a FHP team, whereas the counter-referrals for treatment at the primary health care services were significantly higher for the traditional PCUs (p = 0.028). The percentages of patients who continued to be treated at the secondary level or were referred to specialized programs were almost the same for both PCU models.

Table 6 - Click to enlarge

Discussion

Sociodemographic and economic aspects

Both subgroups had similar characteristics in terms of sociodemographic aspects; except for the significantly larger number of homes shared by more than three people (p = 0.042) among the patients seen at the PCU with a FHP team. Nevertheless, the occupational and economic aspects were different in each group: in a significant manner, the patients seen at the PCU with a FHP team had worse work conditions and lower income; even though, at a higher relative percentage, they were not economically dependent on another person (p = 0.063). The failure to record sociodemographic, economic and occupational data was similarly high in both PCU models; however, a consistently prevailing pattern was not detected.

Capacity of patient retention at the primary health care service

According to data from the Production Report SIA-SUS (Outpatient Information System of the Public Health System) and SIAB (Primary Care Information System), 196,593 basic health interventions were performed by the PCUs during the period of study, whereas 1,397 patients were referred and attended the first medical visit at the secondary mental health service, which accounts for 0.71% of the total basic health interventions. Since each referral is related to one specific patient, while the number of basic health interventions does not represent the total number of different patients seen at the PCUs, such figures must be considered only as a criterion for comparison of both organizational models and cannot be used to assess the incidence and prevalence rates of mental disorders in this population. Therefore, we noticed that the PCU with a FHP team had a higher percentage of referrals than the traditional PCUs. Even though the percentages of referrals were similar — 0.78 and 0.70% for, respectively, the PCU with and without a FHP team —, the large number of subjects studied enabled us to identify a marginally significant difference between both PCU models (p = 0.099). In spite of the fact that the PCU with a FHP team had a higher percentage of referrals than the traditional PCUs, it is not possible to assume that the PCU with a FHP team referred an excessive number of patients to secondary mental health services: its team might have identified in a more effective manner the patients who needed specialized mental health care or this higher percentage of referrals might mean that there was a higher prevalence of mental disorders among its patients. According to Wang et al.,²⁵ the different economic characteristics of the subjects may have an influence on the demand for mental health care; however, a deeper discussion of this issue goes beyond the limits of the present study.

Medical notes recorded by the PHC team

In spite of the high percentage of failure to record data related to the professional "responsible for the referral" and to the "diagnostic hypothesis" in both PCU models studied, the PCU with a FHP team showed significantly better results: the percentage of missing data about the professional "responsible for the referral" was 11.5% for the PCU with a FHP team and 17.9% for the other PCUs (p = 0.009); the failure to record the "diagnostic hypothesis" was 26.1% for the PCU with a FHP team and 34.2% for the traditional PCUs (p = 0.013). At first, such findings might suggest both a poor capacity of the traditional PCU professionals to deal with the diagnosis of mental disorders and a reduced level of compliance with the standard procedures.

Such lack of compliance with the assistance protocols of SMSM-JF seems to represent the quality of the service delivered by the PHC professionals of the city of Juiz de Fora: according to Ronzani et al.,²⁶ a certain "work overload" — as identified and mentioned by PHC technicians — would lead the team to resist to any changes in the assistance routine, since these changes are regarded only as new tasks. In spite of the fact that these data have been clearly standardized in the SMSM-JF, we should take into consideration that the responsibility of recording mental health interventions is not included among the requirements of the Ministry of Health for the assessment of the service provided by PHC teams, which may contribute to the high percentage of missing data. The difference found is corroborated by the results of the study by Facchini et al.,²⁷ who suggested that the PCUs with a FHP team used specific protocols during their assistance activities more often than the traditional PCUs.

A recent study published by Weinmann et al.²⁸ demonstrated that the failure to use clinical protocols is quite common and generalized. According to these authors, the simple distribution of protocols does not guarantee that the protocols will be implemented: even when the general practitioners are aware of the protocols, they do not necessarily use them. According to Leucht,²⁹ the study by Weinmann et al.²⁸ should not be seen as the "last word" on this topic, since it is also important to identify the reasons for such resistance. There is, however, empirical evidence that the adequate use of the assistance protocols contributes to improve the quality of the services delivered also in the mental health area.³⁰

Integration of the PCUs team in the assessment of referred patients

The results related to the responsibility for referrals could indicate that traditional PCUs work in a more integrated manner; on the other hand, these results could also mean that there is greater involvement of physicians who work in the FHP with the assistance routine of the PCU. We should consider that the PCU with a FHP team was established in 2000 — more recently than the other PCUs, which participated in the pilot project of implementation of SMSM-JF in 1997 — and its team was not complete by then. The presumption that the lack of integration among the team members contributed to the results found in our study is coherent with the evidence that an efficient team work is not easily achieved, at least due to the fact that it involves communication among professionals with different training.²¹

Capacity of PCU professionals to identify mental disorders

Among the patients referred, the most frequent "diagnostic hypothesis" was that of anxiety disorder, which was in agreement with the main diagnosis established by the team of specialists, being also the disorder with the highest prevalence in the general population.^11,31 The significantly higher percentage of patients with diagnostic hypothesis of organic mental disorders referred by the PCU with a FHP team may indicate a higher prevalence rate of these disorders in the coverage area of this PCU. However, considering the fact that the diagnostic hypothesis of mental retardation is not included in the semi-structured clinical medical records of the SMSM-JF, it is possible to assume that the patients with such diagnosis were referred under the rubric of organic mental disorders. The results of the evaluation of the agreement between the diagnostic hypothesis and the main diagnostic support this hypothesis: for the organic mental disorders, we found a higher percentage of patients who were not recorded in the interactions related to the PCU with a FHP team than in the interactions related to the PCUs without a FHP team — respectively four of 13 patients (30.8%) and nine of 34 patients (26.5%).

It was observed that the hypothesis of a disorder due to the use of psychoactive substance was responsible for the best agreement rates regarding the diagnosis. The ambivalent behavior of the PHC professionals with regard to such disorders has been established: on one hand, there is low sensitivity to the non-specific demand — that is, patients with signs and symptoms without direct reference to the use of substances³² — and, on the other hand, there is excess of sensitivity to the mental symptomatology associated with the explicit reference to the use of substances.⁶

The very low diagnostic agreement rate related to personality disorders may reflect either a conceptual limitation of the PCU technicians or a poor characterization and lack of explicit criteria used to define such disorders. Personality disorders have been historically related to concepts of psychopathic personality and psychopathy and, to these days, they are usually confounded with psychosis and considered as a synonym of mental disease.³³

The difficulties related to the reliability of psychiatric diagnoses have been discussed even by specialists.³⁴ Even after the changes that were implemented in the diagnostic criteria during the last 2 decades, there is still evidence that the physicians' lack of capacity to establish a correct diagnosis and to treat the mental pathologies presented by their usual clients is an important factor related to the low effectiveness of the assistance of patients with problems and/or complaints in the mental health area at the primary level.^5,14

In terms of psychotic disorders — which should be quickly identified and referred to treatment with specialists —, the results presented suggest that the PCU with a FHP team detected a lower percentage of true psychotic patients (lower diagnostic specificity). In a similar manner, this PCU did not identify as psychotic a relative larger number of patients (three out of five against 10 out of 23 of the traditional PCUs) that were diagnosed with psychosis by the team of specialists (false-negatives). It is worth to note that, in the case of such pathologies with a higher risk of unfavorable outcome, it is always better to use excessive caution instead of not detecting a case of psychosis; or, in other words, sensitivity is always better than specificity. Our results suggest that if the professionals of the PCU with a FHP team were really concerned about referring psychotic patients, they failed to identify them more often.

Assessment of treatment dropout and continuity

The fact that we did not find significant difference in terms of percentage of patients who continued to be treated by the team of specialists or were referred to super specialized programs indicates that the teams of both PHC organizational models identified and referred a similar relative number of patients with real need of specialized treatment. In a marginally significant manner, treatment dropouts at the secondary level were more frequent among the patients seen at the PCU with a FHP team, whereas the counter-referrals for treatment continuity at the primary level were significantly more numerous for the patients seen at the PCUs without a FHP team. Considered as a whole, these data suggest that the PCUs with traditional teams referred patients who were more motivated to be treated by specialists.

Even though definite results have not been found yet, studies conducted during the last 15 years have identified several factors that are statistically associated with absence from medical visit and/or treatment dropout in mental health, both at primary and secondary levels. Some studies^35-37 have found significantly high dropout rates among single patients, while the opposite has been found by other authors.^38,39 Younger patients had higher dropout rates in the studies by Young et al.,³⁸ Edlund et al.⁴⁰ and Rossi et al.³⁷ Melo & Guimarães⁴¹ found significantly higher dropout rates in patients who did not receive transportation passes to go the the medical appointment with specialists. In general, such authors analyzed few economic variables and found differences mainly in terms of demographic characteristics. Another study involving the same group of patients analyzed in our study — covering a larger number of demographic and economic variables — identified significantly higher treatment dropout rates among unemployed patients; in a marginally significant manner, this result was also identified among patients who were not breadwinners or who reported an income lower than one minimum salary.⁴² Within the context of the present study, we cannot rule out the possibility that the economic differences found among the patients contributed to a higher dropout rate among the patients referred by the PCU with a FHP team, since, as mentioned above, such patients had lower income and worse work conditions. In addition, it is important to bear in mind that the patients from the area covered by the PCU with a FHP team, as opposed to the inhabitants of other areas, do not have a bus route available to take them to the secondary mental health service.

Considering the high treatment dropout rates often found in mental health programs, there might be important distortions in the assessment of these services if the information related to the patients who dropped out from treatment are not included.³⁸ In fact, an assistance proposal that is rejected by its users is considered ineffective.⁴³ The high percentage of failure of the PCU teams to record data impose limitations to the reliability of the information and its analysis; on the other hand, the record of data by the team of specialists can be impaired by patients' early dropout.

Motivation for the treatment with specialists,³⁹ fear of stigma,²⁵ possible resistance against the treatment with psychiatrists,⁴⁴ type of treatment employed,^40,42 and clinical⁴⁵ and evolutive⁴⁶ characteristics, specific of certain mental disorders, must be correctly evaluated by primary care professionals, since they have an influence on the attendance and continuity of treatment at the secondary level. Melo & Guimarães⁴¹ demonstrated that those patients who spontaneously sought mental health specialized services had higher treatment dropout rates.

Comparisons between different diagnostic groups did not reach statistic significance in the comparison between both PCU models. Indeed, conclusive patterns of association between psychiatric diagnosis and treatment dropout have not been identified yet by different investigators.^{36-38,44,45,47}

The consultation-liaison model adopted by the SMSM-JF, while valuing the communication among professionals and between specialists and the PHC team, should result in a decrease in the number of referrals to treatment of mild disorders and should promote the referral of severe mental disorders.³ According to Bower & Gask¹⁵ there is still a gap between what has been demonstrated as efficient in research settings and the effectiveness in the routine practice, as it seems to be the case for assistance proposals in other areas of community health.²¹

Among the limitations of the present study — which apply to all studies using a similar design — we should mention the reduced possibility of generalizing its results, mainly due to the fact that the referral of the subjects to the specialized service complied with the conduct protocols of the SMSM-JF,¹⁷ without using any other selection or exclusion criterion. However, it is important to note that the variables used in the present study can be easily reproduced in other assistance contexts, which facilitates its replication. The large number of subjects assessed — rarely found in Brazilian studies — assures the reliability of our results. We also point out that studies that involve "real life" situations contribute in an objective manner to the assessment of the effectiveness of the practices of community health.^48,49 Furthermore, we believe that, due to the simple fact that this situations are real, they can contribute to improve the quality of the assistance or, at least, prevent it from getting worse.

Conclusion

Therefore, our results suggest that the PCU with a FHP team presented a better standard for data recording, higher exclusive participation of physicians in its referrals to specialized care, poorer capacity of patient retention, higher dropout rates and lower percentage of counter-referral by the specialized team. Both PCU models had low diagnostic agreement rates, with higher rates of wrong diagnosis for the PCU with a FHP team. Our results corroborate the hypothesis that the quality of the structure is not directly correlated with the quality of the process or results and emphasize the importance of having structured systems in which the implementation of the interface between the primary and secondary health care services is supported by a consistent process of supervision and continual training of the professionals involved.

References

• 1. Ribeiro MS, Teixeira LS. Por uma ética das práticas em saúde mental. J Bras Psiquiatr. 1997;46(6):331-7.
• ²
  WHO/SUPPORT. Relatório Mundial da Saúde 2001. Genebra: World Health Organization; 2001. http://www.who.int/whr/2001/en/whr01_po.pdf Acessado 05 jun 2006.
• 3. Gask L, Sibbald B, Creed F. Evaluating models of working and the interface between mental health services and primary care. Br J Psychiatry. 1997;170:6-11.
• 4. Kapczinski F, Quevedo J, Madruga M, Busnello JV, Busnello EDA. Psiquiatria social e comunitária. R Psiquiatr RS. 1998;20(3):69-76.
• 5. Prince M, Patel V, Saxena S, Maj M, Maselko J, Phillips MR, et al. No health without mental health. Global Mental Health. Lancet. 2007;370(9590):859-77.
• 6. Ribeiro MS. O uso funcional e disfuncional de substâncias psicoativas. In: Ribeiro MS, org. Ferramentas para descomplicar a atenção básica em saúde mental. Juiz de Fora: Universidade Federal de Juiz de Fora; 2007. p. 55-80.
• 7. Senna MCM. Eqüidade e política de saúde: algumas reflexões sobre o Programa Saúde da Família. Cad Saude Publica. 2002;18(Suplemento):203-11.
• 8. Viana ALD, Poz MRD. A reforma do sistema de saúde no Brasil e o Programa de Saúde da Família. PHYSIS (Rio J.). 2005;15(Supl):225-64.
• ⁹
  Brasil, Ministério da Saúde. Saúde da Família: uma estratégia de organização dos serviços de saúde. Brasília: Secretaria de Assistência à Saúde, 1996.
• 10. Bandeira M, Freitas LC, Carvalho Filho JGT. Transtornos mentais comuns em usuários do Programa de Saúde da Família. J Bras Psiquiatr. 2007;56(1):41-7.
• 11. Fortes S, Villano LAB, Lopes CS. Perfil nosológico e prevalência de transtornos mentais comuns em pacientes atendidos em unidades do Programa de Saúde da Família (PSF) em Petrópolis, Rio de Janeiro. Rev Bras Psiquiatr. 2008;30(1):32-7.
• 12. Rosa WAG, Labate RC. A contribuição da saúde mental para o desenvolvimento do PSF. Rev Bras Enferm. 2003;56(3):230-5.
• 13 Mari JJ. Morbilidad psiquiátrica en centros de atención primaria. Boletín de la Oficina Sanitaria Panamericana. 1988;104(2):171-81.
• 14. Kates N, Craven M, Crustolo AM, Nikolaou L, Allen C. Integrating mental health services within primary care. A Canadian program. Gen Hosp Psychiatry. 1997;19(5):324-32.
• 15. Bower P, Gask L. The changing nature of consultation-liaison in primary care: bridging the gap between research and practice. Gen Hosp Psychiatry. 2002;24(2):63-70.
• 16. Ribeiro MS, Stroppa ALPC, Salomão Neto A, Bastos MO, Costa DMT. Reforma psiquiátrica e atenção primária à saúde: o processo de implementação do sistema municipal de saúde mental de Juiz de Fora. Revista APS. 2003;6(1):19-29.
• 17. Ribeiro MS. Protocolos de Conduta do Sistema Municipal de Saúde Mental de Juiz de Fora. Juiz de Fora: SUS; 2000.
• 18. Ribeiro MS, Heckert U. Universidade e o cuidado à saúde: o "caso" do Serviço de Psiquiatria e Psicologia Médica do HU-UFJF. Rev Saude Dist Fed. 2005;16(1-2):35-47.
• 19. Bland RC. Psychiatry and the burden of mental illness. Can J Psychiatry. 1998;43(8):801-10.
• 20. Jenkins R. Linking epidemiology and disability measurement with mental health service policy and planning. Epidemiol Psichiatr Soc. 1998;7(2):120-6.
• 21. Barbara S. Atenção primária: equilíbrio entre necessidades de saúde, serviços e tecnologias. Brasília: Ministério da Saúde; 2002.
• 22. Salvador-Carulla L, Haro JM, Ayuso-Mateos JL. A framework for evidence-based mental health care and policy. Acta Psychiatr Scand Suppl. 2006;(432):5-11.
• 23. Caetano D. Classificação de transtornos mentais e de comportamento da CID-10: descrições clínicas e diretrizes diagnósticas. Porto Alegre: Artmed. 1993.
• 24. Jekel JF, Elmore JG, Katz DL. Epidemiology, biostatistics and preventive medicine. Philadelphia: Saunders; 1996.
• 25. Wang PS, Aguilar-Gaxiola S, Alonso J, Angermeyer MC, Borges G, Bromet EJ, et al. Use of mental health services for anxiety, mood, and substance disorders in 17 countries in the WHO world mental health surveys. Lancet. 2007;370(9590):841-50.
• 26. Ronzani TM, Ribeiro MS, Amaral MB, Formigoni MLOS. Implantação de rotinas de rastreamento do uso de risco de álcool e de uma intervenção breve na atenção primária à saúde: dificuldades a serem superadas. Cad Saude Publica. 2005;21(3):852-61.
• 27. Facchini LA, Piccini RX, Tomasi E, Thumé E, Silveira DS, Siqueira FV, et al. Desempenho do PSF no Sul e no Nordeste do Brasil: avaliação institucional e epidemiológica da atenção básica à saúde. Rev Bras Saude da Família. 2007;13:28-41.
• 28. Weinmann S, Koesters M, Becker T. Effects of implementation of psychiatric guidelines on provider performance and patient outcome: systematic review. Acta Psychiatr Scand. 2007;115(6):420-33.
• 29. Leucht S. Psychiatric treatment guidelines: doctors non-compliance or insufficient evidence? (Editorial). Acta Psychiatr Scand. 2007;115(6):417-9.
• 30. Smith JL, Rost KM, Nutting PA, Elliott CE. Resolving disparities in antidepressant treatment and quality-of-life outcomes between uninsured and insured primary care patients with depression. Med Care. 2001;39(9):910-22.
• 31. Almeida Filho N, Mari JJ, Coutinho E, França JF, Fernades JG, Andreoli SB, et al. Estudo multicêntrico de morbidade psiquiátrica em áreas urbanas brasileiras (Brasília, São Paulo, Porto Alegre). Rev ABP-APAL. 1992;14(3):93-104.
• 32. Ribeiro MS. A inserção da saúde mental na Atenção Básica de Saúde. In: Ribeiro MS, org. Ferramentas para descomplicar a atenção básica em saúde mental. Juiz de Fora: Universidade Federal de Juiz de Fora; 2007. p. 15-26.
• 33. Mendes Filho RB, Morana HCP. Transtornos específicos da personalidade: semiologia em psiquiatria forense. Psychiatry On-line Brazil. 2003;8(12). http://www.polbr.med.br/ano03/artigo1203_b.php Acessado 03 mar 2007.
• 34. Caetano R. Causas de baja confiabilidad del diagnóstico em psiquiatría: Una revisión critica. Acta Psiquiatr Psicol Am Lat. 1978;24:115-30.
• 35. Olfson M. Primary care patients who refuse specialized mental health services. Arch Intern Med. 1991;151(1):129-32.
• 36. Matas M, Staley D, Griffin W. A profile of the noncompliant patient: a thirty-month review of outpatient psychiatry referrals. Gen Hosp Psychiatry. 1992;14(2):124-30.
• 37. Rossi A, Amaddeo F, Bisoffi G, Ruggeri M, Thornicroft G, Tansella M. Dropping out of care: inappropriate terminations of contact with community-based psychiatric services. Br J Psychiatry. 2002;181:331-8.
• 38. Young AS, Grusky O, Jordan D, Belin TR. Routine outcome monitoring in a public mental health system: the impact of patients who leave care. Psychiatr Serv. 2000;51(1):85-91.
• 39. Gonzalez J, Williams JW Jr., Noël PH, Lee S. Adherence to mental health treatment in a primary care clinic. J Am Board Fam Pract. 2005;18(2):87-96.
• 40. Edlund MJ, Wang PS, Berglund PA, Katz SJ, Lin E, Kessler RC. Dropping out of mental health treatment: patterns and predictors among epidemiological survey respondents in the United States and Ontario. Am J Psychiatry. 2002;159(5):845-51.
• 41. Melo APS, Guimarães MDC. Fatores associados ao abandono do tratamento psiquiátrico em um centro de referência em saúde mental em Belo Horizonte. Rev Bras Psiquiatr. 2005;27(2):113-8.
• 42. Ribeiro MS, Alves MJM, Vieira EMM, Silva PM, Lamas CVD-B. Fatores associados ao abandono de tratamento em saúde mental em uma unidade de nível secundário do Sistema Municipal de Saúde. J Bras Psiquiatr. 2008;57(1):16-22.
• 43. Perkins R. What Constitutes Success? [Editorial]. Br J Psychiatry. 2001;179:9-10.
• 44. Grunebaum M, Luber P, Callaham M, Leon AC, Olfson M, Portera L. Predictors of missed appointments for psychiatric consultations in a primary care clinic. Psychiatr Serv. 1996;47(8):848-52.
• 45. Killaspy H, Banerjee S, King M, Lloyd M. Prospective controlled study of psychiatric out-patient non-attendance. Characteristics and outcome. Br J Psychiatry. 2000;176:160-5.
• 46. Siqueira MM, Garcia MLT, Souza RS. O impacto das faltas às consultas em um programa de dependentes de álcool. J Bras Psiquiatr. 2005;54(2):114-9.
• 47. Percudani M, Belloni G, Contini A, Barbui C. Monitoring community psychiatric services in Italy: differences between patients who leave care and those who stay in treatment. Br J Psychiatry. 2002;180:254-9.
• 48. Bao Y, Duan, N, Fox SA. Is some provider advice on smoking cessation better than no advice? An instrumental variable analysis of the 2001 National Health Interview Survey Health Serv Res. 2006;41(6):2114-35.
• 49. Haro JM, Suarez D, Novick D, Brown J, Usall J, Naber D, et al. Three-year antipsychotic effectiveness in the outpatient care of schizophrenia: observational versus randomized studies results. Eur Neuropsychopharmacol. 2006;17(4):235-44.

*

FAPEMIG: Projeto EDT 3322/06. As acadêmicas receberam bolsas do SUS-JF, UFJF e FAPEMIG.

Publication Dates

• Publication in this collection
  24 Aug 2009
• Date of issue
  2009

History

• Accepted
  19 Oct 2008
• Received
  13 Sept 2008