Use of Health Services and Family Health Strategy Households Population Coverage in Brazil Utilização dos Serviços de Saúde e Estratégia Saúde da Família

Resumo O objetivo deste estudo é descrever o perfil de utilização dos serviços de atenção primária à saúde, estimado pela Pesquisa Nacional de Saúde (PNS), da população residente em domicílios cadastrados e não cadastrados na Estratégia de Saúde da Família (ESF), nos anos de 2013 e 2019. Estudo transversal realizado com microdados dos inquéritos nacionais de saúde entre 2013 e 2019. A amostra originou-se de uma amostra mestra, composta por um conjunto de unidades de áreas selecionadas em um cadastro. Variáveis sexo, idade, cor da pele, renda, escolaridade, autopercepção de saúde, domicílio cadastrado na ESF, atendimento médico no último ano, tipo de serviço que você procura quando está doente foram selecionados. As variáveis dependentes foram uso de serviços de saúde e uso de serviços públicos de saúde. As variáveis dependentes e independentes foram descritas com os respectivos intervalos de confiança e foi realizada regressão logística ajustada para cada desfecho analisado. Nos serviços públicos de saúde, menor renda, ter doenças crônicas (hipertensão arterial ou colesterol alto), estar grávida e ter uma autopercepção de saúde ruim estiveram associados à maior utilização de serviços de saúde nos dois períodos. Morar em domicílios cadastrados na ESF foi associado aos serviços de saúde mais utilizados (públicos ou privados). A estratégia de saúde da família é uma estratégia importante para expandir o acesso de forma igualitária. Palavras-chave Atenção Primária à Saúde, Estratégia Saúde da Família, Serviços de saúde Abstract The objective of this study is to describe the profile of use of primary health care services, estimated by the PNS, of the population living in households registered and not registered with the Famly Health Strategy FHS, in the years 2013 and 2019. Cross-sectional study carried out using microdata from national health surveys 2013 and 2019. The sample originated from a master sample, consisting of a set of units from selected areas in a register..The variables sex, age, skin color, income, education, self-perceived health, home registered with the FHS, medical care in the last year, type of service you seek when you are ill were selected. The dependent variables were use of health services and use of public health services. The dependent and independent variables were described with the respective confidence interval and adjusted logistic regression was performed for each outcome analyzed. In public health services, lower income, have chronic diseases (arterial hypertension or high cholesterol), be pregnant, and having a bad self-perception of health were associated with used more health services in both periods. Living in registered households was associated with more used health services (public or private). The family health strategy is an important strategy for expanding access equally.


Introduction
Primary Health Care (PHC) is the structural axis of a health system. This is the first level of access, organized to offer a longitudinal and comprehensive service, coordinating care within the health system itself (essential attributes). PHC also has guidelines for organizing care geared to the needs of families, communities, observing the cultural characteristics of each population. These are considered attributes derived from PHC¹.
PHC aims to balance two goals of a health system: to optimize the health of individuals and the population; provide equity in the distribution of resources, both proper to care and financial².
The Family Health Strategy (FHS) is the great Brazilian bet to structure PHC services in the country. In the last 20 years, there has been a strong expansion of the FHS in all regions of the country³. At the end of 2019, there were almost 45 thousand Basic Health Units, with 43,458 FHS teams with potential coverage of about 150 million people 4 . advances in FHS coverage have enabled a reduction in infant mortality and preventable mortality, a reduction in hospitalizations for sensitive conditions, among other advances 5,6 . On the other hand, there is a significant decrease in the speed of health gains in the face of public investment: decreased vaccination coverage, loss of speed in reducing child mortality, a large proportion of preventable hospital admissions, in addition to enormous difficulty in managing diseases chronic diseases, aging, and also coping with syphilis and HIV 7 .
National home-based surveys are important tools to study and understand the health needs of the population, coverage of health services, profile of access and use, among others. The National Health Survey (PNS), conducted in 2019, with a sample of 86,820 households, collected information on the performance of the single health system.
The objective of this study is to describe the profile of use of primary health care services, estimated by the PNS, of the population living in households registered and not registered with the FHS, in the years 2013 and 2019.

Methodology
The PNS sample originated from a master sample, consisting of a set of units from selected areas in a register, in order to meet the subsample selections for several different surveys provided for in the Integrated Home Survey System (IHS / IBGE), such as the National Household Sample Survey and the Household Budget Survey. Such units are conceptualized as primary sampling units (PSU), within the sample planning of the researches that use the master sample, as in the case of the PNS 8 .
The sampling strategy consisted of (i) a three-stage conglomerate plan, with stratification of the UPA (in this case, census sectors or set of sectors) and selection of these for the master sample, with probability proportional to the size, defined by the number of households permanent private individuals, and (ii) the selection for the PNS sample, with an equally proportional probability 8 .
The second stage consisted of the selection of households from the National Register of Addresses for Statistical Purposes, in its most recent update (carried out for the execution of the Continuous National Household Sample Survey 2019) before the completion of this stage of the sampling plan, by simple random sampling. Then, within each household, a resident aged 15 or over was randomly selected, based on the list of residents obtained at the time of the interview. To scale the sample size with the level of precision desired for the estimates, some indicators from the 2013 edition of the PNS were considered, such as data from chronic non-communicable diseases (diabetes, hypertension, depression), violence, use of health services, possession of a plan health, smoking, physical activity and alcohol consumption, among others 9 .

Data collection National Health Survey
The organization of collections and the coordination of fieldwork, carried out by Brazilian Institute of Geography and Statistics (IBGE), involved collection agents (interviewers), supervisors (supervision of data collection and management of collection agents) and coordinators (responsible for research in a given state or central unit) ) of IBGE's own staff 8 .
The training of coordinators and others involved consisted of stages: at first, the coordinators of IBGE state units were trained through a face-to-face workshop held in the city of Bento Gonçalves, Rio Grande do Sul State. Participants became multiplier agents and, upon returning to their units, passed on the training content to supervisors and collection agents. The field team (coordinators, supervisors and collection agents) participated, throughout the process data collection, simultaneous training, with the possibility of online questioning 8 .
The interviews were conducted with the use of mobile collection devices (MCD), programmed to "jump" over items in the questionnaire and for critical analysis of the variables. Upon arriving at the selected household, first the interviewers made contact with the responsible person or another resident. The agent explained to the residents the objectives, the data collection procedure and the importance of their participation in the research. At that time, a list of all individuals residing in the household was filled out, regardless of whether or not they agreed to participate in the research 8 .
Then, the resident was identified who would provide information about the questionnaires at home and of all the residents of the household, in addition to the draw of the resident of 15 years and more to answer the interview. individual. The interviews were scheduled at the most convenient time for the residents. Two or more visits were planned in each household 8 .
In 2013, records of interviews were obtained in 64,348 households and the informant of each responded to the others about the FHS coverage. Thus, valid information was collected for about 205,000 residents. For data analysis, expansion factors or sample weights were used for PSUs, for households and all residents and for the selected resident.
In 2019, records of interviews were obtained in 86,820 households and the informant of each responded to the others about the FHS coverage. Thus, valid information was collected for about 134,221 residents. For data analysis, expansion factors or sample weights were used for PSUs, for households and all residents and for the selected resident.

Variables
The households registered in the FHS were presented in proportions, and to estimate the number of people per household, the division was made: number of households registered in the family health unit / number of households interviewed in the PNS.
Brazilian major regions, census situation, Sexsex, age, skin color, income, education, self-perceived health, home registered in the FHS , use of medical care in the last year (public and private), type of service you seek when you are ill, chronic diseases and pregnancy, were independent variables selected for the year 2013 and 2019.
The independent variables were categorized: Sex (Male, Female); Age (18-29 years-old, 30-39 years-old, 40-49 years-old, 50-59 years-old, 60 years-old or more); Skin color (White, Brown, Black, Yellow, Indigenous); Family and household income in quintiles, Head of household education (none, Incomplete Elementary / Middle school, Elementary / Middle school, High school, Undergraduation or more); search for health service in case of illness (private, public, pharmacy / others). Self-reported health condition -arterial hypertension, diabetes, high cholesterol, and (yes / no); pregnancy -(yes / no); the self-perceived health variable is the categorization (very good, good, regular, bad, very bad). Brazilian major regions (North, Northeast, Southeast, South, and Midwest); census situation (urban/rural); register in a family health unit (yes/no).
The variable of use to public health services was the outcome of association analysis and was built through Tthe question was built: "In the last two weeks, looked for a place, service or health professional for care related to one's own health", where only individuals who accessed public services were considered.

Data Analysis
The proportions of households registered in the FSH were stratified according to Major Regions, Federation Units (UF). The proportion and their 95% confidence intervals (95% CI) were described; and the absolute numbers have been estimated. Descriptive analysis of the use of medical care (public and private) in the last year was also carried out, considering living in households registered in the FHS . In addition, the co-variables were used for, region, sex, age, income, education, skin color, search for health care in case of illness, self-reported health condition (diabetes, hypertension, high cholesterol, pregnant woman) and self-perceived health. Then, adjusted logistic regression was performed to establish the association between the independent variables and the outcome "use to public health services". To the co-variables investigated in this analysis were: Register in a Family Health unit, census situation, Brazil region, sex, age, skin color, family income, household wealth, schooling of the head of household, self-reported health condition (diabetes, hypertension, high cholesterol, pregnant woman) and self-perceived health.
The inclusion of independent variables in the model was performed using backward stepwise considering a p-value <0.2 for the adjusted one. The crude and adjusted odds ratio and their 95% confidence intervals (95% CI) were described.
The data were analyzed by the software Stata 16, through the survey module, which considers the effects of complex sampling.

Results
The descriptive analysis of the households registered with the FHS showed that between the years 2013 and 2019 there was an expansion of coverage. In 2013, 54.7% of households were registered with the FHS, in PNS 2019 this proportion jumped to 61.5% of the sampled households. Considering the number of people per household, the potential coverage doubled (79.610.000 to 134.121.000). Rural households had a higher proportion of registered households than urban households. The northeast region had the highest coverage of registered households in both periods analyzed, at the other extreme, the southeast region had the lowest proportion of registered households (Table 1).
It is possible to observe people who live in households registered in FSH use health services more. This is observed both in 2013 and in 2019, and for this last period the differences deepen ( Table 2). Only 23.8% of people living in households not registered in FHS in the southeast of the country did not use health services in 2019, while 79.6% of people living in registered households in the northeast used health services in the last 12 months. The results also point out that people with lower income and education have used more health services than the others, as well as people who have some chronic health condition ( Table 2).
The Table 3 explore the associated factor to used public health services (SUS). The results showed that not belonging to a registered household with the FHS decreases the chance of using public health services (SUS) OR 0.65 (0.55 -0.78) in 2013 and OR 0.86 (0.77 -0.96) in 2019. Living in the countryside, which was not associated in adjusted analysis in 2013, was associated with a decrease of use SUS services in 2019 (OR 0.87, CI95% 0.78 -0.97). In both periods, living in the northern region of the country was a factor of difficulty (risk) in using public health services.
Regarding age, despite older individuals were only associated with increase of service use in crude analyze in 2013, individuals of 40-49 years (OR 1.21, CI95% 1.01 -1.44) and 50-59 years (OR 1.27, CI95% 1.07 -1.51) were associated in adjusted model in 2019.Both in 2013 as in 2019 individuals with black, indigenous or brown skin color were associated with increase of public health service use in crude analyses (p <0.001). After the adjustments, the associations were lost. While education level was not associated in adjusted analysis, high levels of income -in both periods -were associated with decrease in public health service use (p <0.001), establishing an equity performance. The health conditions evaluated showed a higher use of public health services than those that did not present this condition (p <0.001), it should be noted that the differences reduced in 2019. Regarding self-perceived health, the high use by those who stand out that are perceived in the worst health conditions both in 2013 (OR 12.07, CI95% 6.24 -23.4) and in 2019 (OR 4.65, CI95% 3.45 -6.26).

Discussion
The results of the present article demonstrated that there was an increase (54.7% to 61.3%), during the evaluated period, of the studied population that refers to be registered in the family health units, being greater in the rural compared to the urban area. The Northeast region had the highest percentage, followed by the South, the Southeast had the lowest proportion. As for the units of the federation, there was an increase in registered households in almost all locations, except in the state of Tocantins. Living in a home registered with the FHS was an important factor for the use of public or private health services.
Adjusted analyzes showed that having health conditions that require continuous care (high blood pressure, high cholesterol, pregnancy) was a factor associated with greater use of public health services. Moreover, it was also observed that living in the rural area (in 2019), have no register in a Family Health Unit, lower income, and bad self-perception of health were associated with the outcome. The FHS is the Brazilian option for organizing Primary Health Care (PHC) services 9,10 .
In the last few years, it underwent a strong expansion of these services reaching, in 2019, 43,755 teams with a potential coverage of 64.47% 11 . PHC is an organized service to answer the most frequent health needs of the population, being the main gateway to a health system, responsible for the health of a defined population, based on longitudinal and equitable care. The PNS 2013/2019 data demonstrate that the FHS achieves this objective to the extent that it is present in all entities of the federation and facilitates the use of public health services, especially for those individuals who are historically excluded from access to health services. health as non-whites, lower income, less education and for those who live in rural areas. Previous studies have shown that women, lower income, less education, poorer self-perception of health, having a chronic health condition and advancing age are related to greater use of health services in the FHS [12][13][14] . An unusual finding for the use of health services was the protective factor found for nonwhites. It is possible to speculate that the Bra-zilian FHS has built bonding relationships and work organization capable of overcoming some discriminatory barriers that may be associated with the use of services and skin color 15 .
In addition, the increased use of health services for people with health conditions that require continuous care shows that the FHS traces a correct path for expanding access in an equitable way. However, the effectiveness of care cannot   17 in Belo Horizonte. The undeniable advances in access and management of conditions sensitive to PHC seem to be running out of breath 5,6 . Currently, the indicators approved by the new PHC financing and made publicly available show a low number with semi-annual measured arterial tension (4%), low number of diabetic patients with glycated hemoglobin tests (10%), low number of pregnant women with 6 consultations. prenatal care being the first until the 20th week of pregnancy (33%) and only 37% of Brazilian pregnant women with HIV and Syphilis tests requested in PHC 11 . These findings reveal that there is still an extensive agenda for implementing primary health care in Brazil, particularly if the impacts of the COVID-19 pandemic were considered for the near future 18,19 . PNS 2019 data consolidate the movement of expansion of the FSH as the main strategy for PHC organization in Brazil, with characteristics that guarantee equity in the provision of services. There is a clear expansion of services when comparing the results obtained by the PNS 2013-2019, with a deepening of an equity relationship in the use of services in 2019. Despite this, the observed fact of the increase in the proportion of registered households may not guarantee that individuals residing in the household are registered and making regular use of public health services (SUS). In a recent study on PHC financing in Brazil, Hazheim et al. (2020) 7 , pointed out that although there is an estimated covered population of 148,674,300 million people, there were only 90 million people registered with the FHS. Despite the strong and necessary expansion of services experienced by Brazil in recent years, it is essential that this expansion becomes more effective care. In 2019, there was an army of people potentially covered by FSH services without registration. We can speculate that these people do not use the FHS health services or used them sporadically without the possibility of forming a bond.
Some limitations are inherent to the study, such as the interviewee / resident / resident providing the information collected by all the in-    in the development of data collection activities. Many of these surveys already established by the use of data carried out by the Brazilian scientific community 14,18-20 . Without national home-based surveys that provide relevant information on the Brazilian population, such as the National Health Survey, National Household Sample Survey and the Census, the country is in the dark and public policy planning, development and evaluation are exposed to ideological obscurantism.