Evaluation of family caregiver satisfaction with a mental health inpatient service

Objetivo: Avaliar a satisfação de familiares cuidadores com um serviço de internação em saúde mental no Brasil. Métodos: Trata-se de um estudo transversal, com abordagem quantitativa. Uma amostra de 80 familiares cuidadores respondeu à Escala de Avaliação da Satisfação de Familiares com os Serviços de Saúde Mental (SATIS-BR), além de questionário sociodemográfico. As variáveis categóricas foram expressas como freqüências e porcentagens, e as variáveis quantitativas, como médias e desvios padrão. As interações entre variáveis e índices da escala foram analisadas utilizando o test t de Student, correlação de Pearson e análise de variância. Resultados: Os resultados indicaram escore médio de satisfação geral elevado quando considerada a categorização dos itens da escala, tendo ocorrido maiores índices de satisfação no fator ‘Resultados do tratamento’ e menores índices nos fatores ‘Acolhida e competência da equipe’ e ‘Privacidade e confidencialidade’. Na comparação das amostras estudadas, foram observados maiores escores de satisfação geral e por fator no modelo de atendimento residência médica em relação ao modelo psiquiatra assistente. Não houve diferenças significativas quanto à satisfação dos familiares em relação às variáveis sociodemográficas. Conclusão: A satisfação dos familiares foi elevada. Foram evidenciadas necessidades de melhoria nos aspectos relacionados à infraestrutura dos serviços. Esta pesquisa aponta para a importância de serem realizadas avaliações contínuas e regulares dos serviços, tendo como foco a satisfação dos usuários e familiares para uma melhor compreensão dos fatores que contribuem para a qualidade do atendimento. Descritores: Pesquisa sobre serviços de saúde, escalas, hospitalização, cuidadores, serviços de saúde mental. Abstract


Introduction
Psychiatric hospitalization is a thorough procedure of great importance.Currently, it is indicated in severe cases that characterize situations of risk for the patient or third parties and when the resources for out-ofhospital treatment have been used. 1 The treatment offered in all services should aim, as a permanent purpose, for the social reintegration of patients in their environment. 1,2though an important part of the care provided to patients with mental disorders continues to be carried out in psychiatric hospitals, it is possible to observe, in the last decades, a consistent transformation in mental health care in Brazil, as a result of the psychiatric reform. 2 With the implementation of the new model of care, there was a severe reduction of hospital beds and the creation of substitute services to the asylum model, in addition to national programs of care provided to patients with mental disorders.Initially, the experience of the new service was not evaluated, making it difficult to analyze the processes and the results of implementations. 3,4[7] Besides the fact that services lack resources, there is also the possibility of reproduction of some negative characteristics of the psychiatric hospital context (institutionalism, segregation, abandonment) in the substitute services, justifying the need for supervision through evaluation of the quality of care. 8e World Health Organization (WHO) 6 makes relevant recommendations regarding the mental health reform process, highlighting the importance of greater investment in resources for this sector, in the quality of services, with a progressive reduction in the need of hospitalization and improvement of the attention given to patients with mental disorders.In this context, the current period is characterized by two simultaneous movements: the construction of a mental health care network to replace the model centered on hospital admission; and the evaluation and supervision of existing psychiatric services. 2e importance of an adequate evaluation of the quality of care is related to a greater approval of the treatment provided and reduction of both dropout rates [9][10][11][12] and number of hospitalizations. 13,14 this way, the evaluation of services must be a continuous activity, carried out periodically.Among the recommended actions are the monitoring of services, with indicators reflecting the quality of access, the adequacy of care, the quality of preventive and therapeutic interventions, in addition to the assessment of user satisfaction. 6 the 1970s, the first studies of health quality assessment appeared, making the user's judgment of services become an object of investigation. 15Despite the recommendations of the health regulatory agencies, most studies have evaluated mental health services focusing on patient satisfaction and placing little emphasis on the opinion and satisfaction assessment of other groups involved in the health context, such as family members and professionals.
In the WHO Report on Mental Health in the World, 6 10 recommendations were set forth for mental health planning and practice.The involvement of family members in the treatment of psychiatric patients is one of them, emphasizing that the participation of the family improves the treatment of mental disorders and that family participation is of utmost importance in the evaluation process proposed by the services. 16,17s well as family participation, assessment of the quality of mental health services by the users' families was also neglected in studies conducted before the psychiatric reform; rather, pre-reform research was limited to investigating the influence of the family in the determination of mental illness. 17,18Currently, the impact of mental disorders on the family has become an important area of research.
0][21][22][23][24][25] In particular, only one study has been found that evaluates the satisfaction of family members in the context of inpatient units, 22 and it was conducted in community services.Therefore, this study aims to evaluate the satisfaction of family caregivers with a mental health inpatient service in Brazil.

Sample and design
This was a cross-sectional, descriptive, analytical  26,27 and Bandeira & Silva. 28he scale comprises eight quantitative questions

Procedures
The data were collected through structured interviews that lasted approximately 20 minutes and were performed at the units of the mental health service after certifying that all the inclusion criteria were met.Contact In the inferential analysis, the following tests were performed for situations of comparison of means between groups: Student's t test for independent samples in situations where the independent variable had two categories; analysis of variance (ANOVA) in situations where the independent variable had more than two categories.
For the comparison of the indexes of the scale (OS -overall satisfaction; F1 -treatment results; F2reception and staff competence; and F3 -privacy and confidentiality), repeated measures ANOVA was used followed by Bonferroni post-hoc test in the presence of statistical significance.Pearson correlation coefficient was used to analyze the continuous variables.In all the inferential analyses, significance was set at p < 0.05.
The treatment of qualitative information used the technique of thematic or categorical analysis.According to Bardin,29 this technique is based on operations of breaking up the text into registration units and, later, regrouping it into classes or categories.

Ethical considerations
This study was approved by the research ethics committee of Escola de Saúde Pública do Ceará (protocol 1.481.613/2016).All caregivers signed an informed consent form before they were interviewed.

Family satisfaction with the service
Table 2 presents the mean family satisfaction scores for the global scale and for the three subscales of SATIS-BR and the corresponding p-values.The mean overall satisfaction score was 4.05 (SD = 0.66), ranging from 1.12 to 5, indicating that, in general, the patients' caregivers were satisfied with the service.3).
Regarding satisfaction scores, the items with the lowest indexes were privacy (item 5), understanding of the type of help that the patient needed (item 2), and understanding of the professional who admitted the patient (item 1).In the comparison between satisfaction scores according to the different care models, the research group found statistically significant differences with higher satisfaction scores in the medical residency modality in relation to overall satisfaction and to the factors 'Treatment results' and 'Privacy and confidentiality.'No statistically significant differences were found when overall satisfaction and by-factor scores were compared among the hospitalization units of the service.Multiple comparisons in the Bonferroni Post-hoc test: A -significant difference between OS-F1; B -significant difference between F1-F2; C -significant difference between F1-F3.

Factors associated with family satisfaction
For the analysis of factors associated with family satisfaction, some categories of variables were grouped according to the answers given to subjective questions.
Only seven participants did not respond or answered "I do not know" to one or more qualitative items of the scale after being told that answering was optional.The main factors that contributed to a high level of family satisfaction were patient care and improvement observed with treatment, especially professional care, in addition to services offered such as food and cleaning of the units.
As for the aspects that the subjects disliked in the service, the predominant contents were issues related to security in the units, infrastructure, comfort and appearance of the service, and insufficient number of professionals.
In the analysis of sociodemographic variables of family members related to satisfaction (Tables 5 and 6), only gender presented statistically significant differences in the subfactor 'Treatment results,' demonstrating that male relatives were more satisfied with the improvement observed in the patients.In the study of the variables   However, age did not present statistically significant correlations with the subfactors of the scale.

Discussion
The results of this study made it possible to evaluate a mental health service from the perspective of family caregivers and to highlight the importance of assessing the quality of service from the point of view of satisfaction.
The overall mean satisfaction score among the family members participating in this study was 4.05, a value that can be considered high.0][21][22][23][24][25] The results also corroborate those found by the National Program of Hospital and Psychiatry Assessment (Programa Nacional de Avaliação dos Serviços Hospitalares -PNASH 2012/2014), in which the hospital where this study was performed received high scores after being evaluated for technical requirements and user satisfaction. 302][33] Several theories are proposed to explain the differences observed in the assessment of user satisfaction with health services.
Among them, social acceptability stands out, which emphasizes the importance of the agreement between what is offered by the services and the preference of the users, being an important factor for the quality of the service. 34Another example is the contrast that holds in the expectation model, which suggests that subjects evaluate a given service based on a comparison made between a current experience and previous experiences, which could point to advantages of some of the services analyzed. 350][21][22][23][24][25] These aspects were also widely cited in qualitative data as factors of higher family satisfaction.However, the 'Privacy and confidentiality' factor, associated with physical conditions and comfort of the service, showed lower levels of satisfaction, and these factors were more often reported when the participants were asked about the aspects that would need to be improved in the units.
These results resemble those found by Gigantesco et al. 37 and Perreault et al. 31 It is worth mentioning the result that evidenced the difference in relation to overall and by-factor satisfaction in the medical residency modality when compared to attending psychiatrists, with higher satisfaction indexes associated with the former.These data, to the present date, had not been evaluated in satisfaction studies.
One of the factors that may have contributed to the differences found is the greater frequency of care offered to the patients by resident physicians, a fact cited in the subjective answers regarding the aspects that the family member liked the most.
This result may also reflect the quality of care, as attending psychiatrists bear a greater workload in terms of number of patients assisted as compared to medical residents, which reduces the number of weekly evaluations, does not allow for adequate contact between family members and the medical team to answer questions, and limits treatment results, thereby decreasing satisfaction scores.
The family members presented a sociodemographic profile with predominance of the female gender, married, and aged 30 to 59 years.,25 With regard to education level, high school education was slightly predominant, a result that is higher than that found in most studies.It is worth noting that all illiterate family members were excluded from the present study, a fact not observed in other investigations and that has contributed to increase the educational level in our sample.
Regarding the degree of kinship of relatives with patients, siblings were predominant in this sample, differing from other national and international studies, in which parents were the main caregivers.22][23][24][25]32,38 As observed by Bandeira et al., 19  The present study has some limitations.The results have a reduced potential for generalization because they refer to a single service, with its peculiarities.
The sociodemographic and clinical characteristics of the patients were not evaluated, which may have had an influence on family satisfaction.In addition, it is worth mentioning the inclusion of a non-random sample of family members.Non-randomized samples may influence the results in favor of a higher degree of family satisfaction with services.

Conclusion
The data highlighted a high percentage of family member satisfaction with the mental health service, reflecting a positive view of family members regarding the service at which the patients are being treated.
Among all the factors associated with satisfaction, those related to results of the treatment stood out, reflecting the benefits related to the care provided and improvements observed.
The results point to the need for improvements in the service regarding aspects related to professional care and infrastructure of the units.In this way, the importance of greater investments in these areas becomes evident, since these deficiencies can compromise the quality of the service and its results.
There were no differences in the overall satisfaction between hospitalization units or in relation to sociodemographic variables.As far as care is concerned, the greatest overall and by-factor satisfaction was observed in the medical residency care model vs.
the attending psychiatrist model, which may reflect the greater overload of the latter group, thereby compromising satisfaction with care.
Family members did not show resistance to participate in this research, which translates as a stimulus to promote their involvement in the treatment of patients and the evaluation of mental health services.
Although widely recommended by the WHO, so far only few national studies have been conducted to assess mental health services from the perspective of the family.The present investigation points to the need to carry out continuous and regular assessments of services and to extend these studies, focusing on the satisfaction of users and their families, for a better understanding of the factors that contribute towards the quality of care.

with answers arranged on a 5 -
point Likert-type ordinal scale, where 1 = very dissatisfied, 2 = unsatisfied, 3 = indifferent, 4 = satisfied, and 5 = very satisfied.The abbreviated version of SATIS-BR/Family is distributed into three subscales: 1) satisfaction with treatment results; 2) satisfaction with reception and staff competence; and 3) satisfaction with privacy and confidentiality.There are also three qualitative questions regarding the perception of family members about various aspects of the services received.A sociodemographic questionnaire was also applied to evaluate the characteristics of family members, including the following variables: gender, age, marital status, educational level and degree of kinship with the patient.The characteristics related to the care received by the evaluated patients included hospitalization time and care modality (divided into 1 -medical residency, represented by a team of physicians attending a firstyear residency program and supervisors; and 2attending psychiatrists).
with the participants was made through an appointment scheduled by phone or through personal contact with the researcher or previously trained workers (medical students), during patient visiting hours, established in two alternating weekly shifts, until the sample size was complete.Family member understanding of the questions contained in the satisfaction scale was verified by means of a pilot test performed with 10 family members prior to the beginning of data collection.Data analysisThe data were input and analyzed using the Statistical Package for the Social Sciences (SPSS) version 15.0.To describe the characteristics of the sample, absolute and relative frequencies were used for categorical variables, and means and standard deviations (SD) for continuous variables.The mean and respective SD of the overall satisfaction scores (all scale items) and of each subscale score were also calculated based on the validation study performed by Bandeira et al.27 The highest satisfaction factor was obtained in the 'Treatment results' subscale, with a mean score of 4.25.Lower satisfaction indexes were found for the subscales 'Reception and staff competence' (3.95) and 'Privacy and confidentiality'(3.91).These data sets were submitted to a comparative analysis using repeated measures ANOVA.Multiple comparisons indicated a significant difference between overall satisfaction and the 'Treatment results' factor and between this factor and the 'Reception and staff competence' and 'Privacy and confidentiality' factors.Regarding the data with percentages for each specific item of the scale, most of the relatives reported being satisfied (score 4) or very satisfied (score 5) with the service.The data showed that 81.25% of the relatives had high satisfaction scores in relation to the benefit of patients with the care received.Likewise, most of the family members considered themselves satisfied or very satisfied in relation to the other seven items on the scale.The variables of the scale that obtained the highest scores were: benefit with care (item 8), help the patient to deal more effectively with the problem (item 3), and satisfaction with professional competence (item 7) (Table and 'Overall satisfaction' (r = -0.57,p = 0.614) and between age and subfactors 'Treatment results' (r = -0.66,p = 0.56) and 'Privacy and confidentiality' (r = -0.118,p = 0.297); and a positive correlation with 'Reception and staff competence' (r = 0.038, p = 0.741).
levels of satisfaction have been found in the studies even when expectations regarding the service were negative, an effect known in the literature as elevation of satisfaction rates.36Differences in satisfaction indexes may also be related to the structure and organization of the services, cultural characteristics of the location and participants, as well as the model of data collection used for the study.The analysis of the factors of the SATIS-BR scale allowed to identify 'Treatment results' as the factor that brought greatest satisfaction to the family, with levels between satisfied and very satisfied.The subscales 'Reception and staff competence' and 'Privacy and confidentiality' had lower indexes of satisfaction (mean satisfaction scores of 3.95 and 3.91, respectively); both were categorized as intermediate levels of satisfaction.
who evaluated family member satisfaction in three mental health services in Belo Horizonte, Brazil (mental health outpatient clinics), none of the sociodemographic variables investigated in the relatives were associated with overall satisfaction with mental health services; only the male gender, in this study, correlated with higher satisfaction scores in the 'Treatment results' factor.The qualitative data collected allowed to add information and to provide suggestions of what aspects could improve satisfaction with the service.A low rate of response abstention was observed, indicating the family members' desire to participate.Among the factors mentioned in the subjective questions, professional care was highlighted, characterized as a quality by some interviewees, and at the same time, recognized as insufficient in relation to the number of professionals and the service offered.Lack of qualified professionals can compromise care and treatment outcomes.Another aspect widely remembered by the family members were failures related to the security of the service.These situations could be minimized by reducing the number of patients per unit, making adjustments to the professional staff (number and workload) and the qualifying mental health workers.

Table 1 -
Sociodemographic characteristics of family members according to inpatient units at Hospital de Saúde Mental Professor Frota Pinto, Fortaleza, Brazil Data presented as n (%).

Table 2 -
Overall satisfaction and factor scores of family members from inpatient units at Hospital de Saúde Mental Professor Frota Pinto, Fortaleza, Brazil, according to the SATIS-BR scale SATIS-BR = Brazilian Mental Health Services' Family Satisfaction scale.* p < 0.05 in repeated measures analysis of variance.

Table 3 -
Percentages of response and mean satisfaction for each item on the SATIS-BR scale of family members from inpatient units at Hospital de Saúde Mental Professor Frota Pinto, Fortaleza, Brazil SATIS-BR = Brazilian Mental Health Services' Family Satisfaction scale.* Answers 4 and 5. Data presented as n (%).

Table 4 -
Overall satisfaction of family members from inpatient units at Hospital de Saúde Mental Professor Frota Pinto, Fortaleza, Brazil, according to care modality * Significant difference in the Student's t test for independent samples (p < 0.05).

Table 5 -
Comparison of overall satisfaction scores of family members from inpatient units at Hospital de Saúde Mental Professor Frota Pinto, Fortaleza, Brazil, according to gender SD = standard deviation.* Significant difference in the Student's t test for independent samples (p < 0.05).

Table 6 -
Comparison of overall satisfaction scores of family members from inpatient units at Hospital de Saúde Mental Professor Frota Pinto, Fortaleza, Brazil, according to sociodemographic variables