Factors associated with satisfaction at work in Psychosocial Care Centers1

OBJECTIVES: to analyze the prevalence of satisfaction at work and identify associated factors in Psychosocial Care Centers. METHOD: cross-sectional study involving 546 workers from 40 Psychosocial Care Centers in the South of Brazil. The satisfaction was identified based on the Assessment Scale of Satisfaction in the Mental Health Team and a logistic regression model was used for the adjusted data analysis. RESULTS: the prevalence of satisfaction at work corresponded to 66.4%. Factors directly associated with satisfaction: higher-level function (except physicians and psychologists), work time of six months or less, making a larger number of home visits, good supervision by the team, possibility to make collective choices and take courses. CONCLUSIONS: the satisfaction is associated with the work organization and conditions and demonstrates the need to invest in team supervisions, in process that democratize the services and in the workers' training.

Among the Brazilian regions, the South presents the best CAPS coverage, with an indicator of 0.87 CAPS/100,000 inhabitants, higher than the Brazilian average of 0.66 CAPS/100,000 inhabitants (1) .
The consolidation of the substitutive service network presupposes the problematization and organization of its work processes to consider the exchange of knowledge and practices among the professionals, as well as the valuation of creative potentials and individual competences (2) . Deinstitutionalized care comes with challenges that stimulate workers, and can also cause suffering, burden and lack of satisfaction at work.
The repercussions of work derive from the work conditions (physical, mechanical, chemical and biological pressures of the work station) as well as the organization of work (prescribed operation mode, distribution of responsibilities, hierarchy, command modalities and socioprofessional relations, among others) (3) . In mental health, daily contact with people in mental suffering is added to these factors, constituting a set that affects the workers' satisfaction and, consequently, their well-being and mental health, whose influence is perceived in the quality of user care and, therefore, of the services (4) .
Brazilian quantitative studies adopt concepts that relate satisfaction and dissatisfaction at work as opposite phenomena and use scales to verify them.
The use of validated instruments favors the assessment of satisfaction at work in practice (5) . Nevertheless, the prevalence of satisfaction at work in Brazilian community mental health services has not been assessed. In the USA, it corresponded to 59% according to the Job Satisfaction Survey, in a sample of 176 technicians (6) , and 90% according to the Minnesota Job Satisfaction Questionnaire, in 98 professionals (7) .
The factors related to greater satisfaction at work in community mental health services include: greater autonomy (8)(9) , observation of rapid changes at the service, benefits of teamwork, keeping the clients outside the hospital (9) and organizational support (10) .
And lesser satisfaction has been associated with administrative tasks and a large number of cases (11) , inappropriate physical structure, lack of human and material resources (12)(13) , devaluation in the workplace (8) and greater burden at work (8,(13)(14) .
To get to know the repercussions of work in community mental health services, the objectives in this study were to analyze the prevalence of satisfaction and identify the associated factors in workers from Psychosocial Care Centers in the three states that constitute the South of Brazil.

Method
Cross-sectional study and excerpt of the research CAPSUL II * , undertaken in 2011 to assess the mental health care offered at the CAPS in the South of Brazil, funded by the Ministry of Health. The data studied here were collected through a questionnaire that was selfapplied to the workers of the 40 services drafted for the study, among the 308 existing CAPS, and obtained between August and December 2011. The CAPS sample was structured according to the service supply in the three states within the region, the population concentration per geographic macro-region and the guaranteed presence of the capitals, besides the different CAPS models (I, II, III and excluding CAPSchildren and CAPSalcohol and drugs).
At the 40 CAPS included in the study, all 658 active workers were invited to participate and 546 answered the questionnaire, which permitted estimated a 66% prevalence of satisfaction, with a 4.0 error margin and 95% Confidence Interval. To calculate associations, an alpha coefficient of 5% was used, with 80% statistical power to detect a minimal relative risk of 1.5 in the exposures and a 2:1 index between not-exposed/ exposed.  proportions were verified (13,15) and groups 4 and 5 were identified as the presence of satisfaction.
The remaining variables were organized according to six hierarchical levels, according to the theoretical determination model (16) ; in which the level furthest from the outcome consisted of demographic variables At the fourth level, the workers' assessment of the supervision were explored, subdivided in: by the secretary of health, by the team and by the community, on a scale from 0 to 10; after collecting these variables, they were categorized as bad (0/3), intermediary (4/6) and good (7/10). This same level includes the characteristics of CAPS work, represented by: lack of tools for work, possibility to make collective choices, possibility to take courses.
The fifth level consisted of variables related to cases of absence in a six-month period and to the health conditions, which were: self-referred health problems and suspicions of minor psychiatric disorders, using the Self Report Questionnaire 20 (SRQ 20). The SRQ 20 consists of 20 questions with a yes-or-no answer, translated and validated to Portuguese (17) ; to define the prevalence of suspected minor psychiatric disorders, the cut-off point was set as eight or more positive answers for women and six or more positive answers for men.
The level that was closest to the outcome satisfaction at work, the sixth, consisted of the assessment of the work burden, measured using the Assessment Scale of the Impact of Work in Mental Health Services (IMPACTO-BR), developed by WHO and validated in Brazil (5) ; it presents good item homogeneity and high internal consistency with α = 0.87. IMPACTO-BR contains 18 items, each with answers arranged on a five-point Likert scale, where 1 = not at all, 2 = not much, 3 = more or less, 4 = a lot, 5 = extremely, based on which the global average was calculated and stratified according to the original five-point scale (12) .
The data were analyzed in the statistical program STATA 9.0. The bivariate analysis examined the prevalence of satisfaction in each variable studied.
The associations were tested using the chi-square test, and differences with p ≤ 0.05 were considered significant. The logistic regression model was applied, the gross and adjusted odds ratios were calculated with a 95% confidence interval (95% CI) and backward selection. Variables with p ≤ 0.10 were maintained in the model.

Results
The study participants were 546 workers (83% of the 658 workers at the 40 services), mostly women (79.7%), with a mean age of 37.5±10.8 years, who held a higher education degree (54%) and a mean length of experience at the services of 39.6±45 months. The prevalence level of satisfaction found in the study sample corresponded to 66.4% and the mean global satisfaction was 3.6 (range from 1 to 5).
The following variables showed to be statistically associated with satisfaction at work in the gross analysis: age; education, with a downward trend as   Lapischies SRC, Jardim VMR, Kantorski LP.
The variables that showed to be inversely associated with satisfaction after adjustments were: working at CAPS between seven and 24 months, which reduced the chances of being satisfied by 70% when compared to workers with six months or less of experience at the services; lack of tools for work, which dropped the chances of satisfaction by 62%; and work burden, which reduced the chances of being satisfied at work in Psychosocial Care Centers by 52% (Table 3). The variables training, employment contract and minor psychiatric disorders were maintained in the model, although they did not demonstrate significance, as they are potentially confounding.

Discussion
In 3.59 using SATIS-BR (13)(14) , classified as bordering on indifference, on a scale from 1 to 5 points. Other studies identified higher mean satisfaction rates (4.05 and 4.02), but their samples characterize a single service (15,18) . Intermediary satisfaction scores were identified in studies in Italy, which used a non-validated questionnaire (19) and the Minnesota Satisfaction Questionnaire (20) .
Nevertheless, the Brazilian studies did not assess the prevalence of satisfaction at work in community mental health services, in this study identified as 66.4%.
In the USA, the prevalence rate corresponded to 59% when using the Job Satisfaction Survey, in a sample of 176 technicians (6) , and to 90% when using the Minnesota Job Satisfaction Questionnaire, in 98 professionals (7) .
The short insertion period in deinstitutionalized services for people in mental services permits less exposure to the daily reality at the CAPS. Consequently, the workers may not have experienced a range of situations to permit an assessment, like workers with longer experience at CAPS. These evidences support the findings that associated five to ten years of work experience in Mental Health with lower satisfaction levels (13) ; and differ from a study in the United Kingdom that associated five or more years of work with higher satisfaction levels (11) .
When the CAPS team uses the home visit strategy, it mainly intends to enable the family to use its own resources, include it in the treatment process and increase the possibilities of bonding with the professionals (21) .
The workers identified with the enhanced view of the madness phenomenon and with deinstitutionalized and territorial care aim to reintegrate the users with their family and community and, thus, do home visits, is one of the frequent work strategies.
Many professionals get unique interdisciplinary work experience in the CAPS. Higher-education professions in mental health, with different work processes, can articulate them and organize them in a specific manner (2) . The results support the identification of the association between satisfaction and autonomy in community health service work in the United Kingdom and the USA (8)(9)11) but, nevertheless, did not identify an association with profession/function (11) and a difference in satisfaction between physicians and nurses (19) .
The results indicate that greater importance is attributed to the relations established in the work team, through supervision and collective choices, as observed in a study of 209 workers at CAPS and Home-Based Therapeutic Services, in which team difficulties or problems were identified as the most prevalent situation that bothered the workers among the situations mentioned (22) .
Besides the situations defined in the teams' internal arrangements, factors associated with satisfaction can be established based on service management, such as the possibility to make collective choices, the possibility to take courses, lack of tools for work. These results are in accordance with findings in the literature that associated satisfaction with organizational support (10) , rapid changes in the services (9) and physical structure, human and material resources (12)(13) .
The burden may be related to service management, in the form of an insufficient number of workers to provide user care within the territory, as well as to work organization difficulties. Daily work in CAPS presupposes commitment to psychosocial care and creativity in the search to reinsert the user but, besides a conquest, these premises may represent an obligation and even a burden for the service workers (23) .
Satisfaction may be overestimated due to the possible dissatisfaction among the workers who chose not to manifest their opinion and the workers absent from their workplace; also, the use of logistic their functioning yet.

Final Considerations
The results show a strong association between satisfaction and factors related to the organization and It is also important to expand and qualify the supervision by the team, including opportunities for discussion and planning in daily service work, as well as decentralized and democratic management processes.
Another aspect that should be considered is the workers' training, articulated with the conceptual frameworks and training areas in the CAPS, stimulated and promoted by the managers.
The transition between models is ongoing. The