Changes in the pattern of health services use in Brazil between 2013 and 2019

Abstract

This study aimed to investigate changes in the health service use pattern based on information from the 2013 and 2019 National Health Surveys (PNS). The two outcomes, “Seeking health-related care in the past two weeks” and “Medical visit in the last twelve months”, were analyzed according to socioeconomic, geographic and health conditions characteristics. Multivariate Poisson regression models were used to investigate the factors associated with seeking care due to a health problem or prevention. The prevalence of chronic diseases increased from 15.0% to 22.5% between 2013 and 2019. The proportion of seeking care increased from 15.3 to 18.6%, and medical visits from 71.2% to 76.2%, ranging from 61.4 to 75.8% and 68.0 to 80.6% between the North and Southeast regions. There was no significant association of seeking care due to a health problem with per capita income, after controlling for the other covariates. We conclude by saying that, despite the expanded coverage of health service use, the persistent regional inequalities indicate unmet health needs among residents of the less developed regions. Health care models focused on prevention and health promotion are required.

Key words:
Health service use; Surveys; Health needs; Inequalities; Brazil

Resumo

Este estudo teve o objetivo de comparar os padrões de utilização de serviços de saúde, a partir das informações das edições da Pesquisa Nacional de Saúde (PNS), 2013 e 2019. Os dois desfechos “Procura de atendimento relacionado à saúde nas últimas duas semanas” e “Consulta médica nos últimos doze meses” foram analisados segundo fatores socioeconômicos, geográficos, e condições de saúde. Foram usados modelos multivariados de regressão de Poisson para investigar os fatores associados à procura de atendimento de acordo com o motivo (problema de saúde ou prevenção). Entre 2013 e 2019, a prevalência de doenças crônicas aumentou de 15,0% a 22,5%. A proporção de busca de atendimento cresceu de 15,3 a 18,6%, e de uso de médico, de 71,2% a 76,2%, com amplitudes de variação de 61,4-75,8% e 68,0-80,6% entre as regiões Norte e Sudeste. Para atendimento por problema de saúde, não houve associação significativa com rendimento per capita, após o controle das demais covariáveis. Conclui-se que apesar da expansão da cobertura de utilização de serviços de saúde, as persistentes desigualdades regionais indicam necessidades de saúde não atendidas entre os residentes das regiões menos desenvolvidas. Modelos de atenção focados na prevenção e promoção da saúde são necessários.

Palavras-chave:
Uso de serviços de saúde; Inquéritos; Necessidades de saúde; Desigualdades; Brasil

Introduction

In Brazil, the national health system’s organization is based on having a public system, responsible for providing free and universal care to any Brazilian in need, and supplementary health, defined by all private health care, performed or not through an agreement with a health insurance plan.

The SUS establishment by the 1988 Constitution was an essential change in the pattern of organization of health services in the country, strengthening primary health care and decentralizing services to increase accessibility to the health system and stepping-up health prevention and promotion actions¹. Geographic targeting of care was established through family health teams in socially disadvantaged areas and population subgroups² with the implementation of the Family Health Program (PSF) in the 1990s.

In the mid-2000s, as the main gateway to the health system, primary health care started to coordinate care in the service network and implement the integration of the service network³. The current guidelines for the integration of care involve a regular demand service, the provision of primary care and health promotion services, and the guarantee of access to different care levels through referral mechanisms to ensure healthcare continuity⁴.

Health regions and strategies were established to integrate different care levels as part of the decentralization process. The municipalities developed proposals to facilitate and streamline the implementation of new care models, with organizational criteria defined by the federal government and other municipal management criteria, imposing the need to evaluate the implantation and implementation of interventions and their population’s health results⁵.

The research aimed at assessing health performance from the user population’s perspective has been increasingly gaining relevance and was applied globally^6-8. The user’s perception has been overly valued and was an essential step for the planning of actions aimed at securing the quality of health care and subsidizing decisions meeting the needs of the population⁹. In Brazil, the periodic generation of population-based information has been an indispensable tool to monitor access to and use of health services and the health conditions of the population, allowing the evaluation of public health policies and providing subsidies for its improvement^10-12.

In this sense, the search for health-related care and medical visits are positive indicators of access to the health system insofar as they can significantly impact the health of the population, preventing the occurrence of diseases, reducing morbimortality, and increasing healthy life expectancy¹³. Additionally, studies on the use of health services can contribute to care organization, facilitating the establishment of inequalities in coverage levels according to sociodemographic features and place of residence and identifying the most vulnerable population groups¹⁴.

This study aimed to investigate the profile of health services utilization in 2019 in Brazil and analyze its trend compared to 2013, using data from the National Health Survey (PNS). Demographic and socioeconomic factors and care need indicators are considered in the analysis. Regional and social inequalities in the use of health services are compared using data from the two editions of the PNS.

Methods

The National Health Survey is a household-based and a nationwide survey carried out by the Ministry of Health in partnership with the Brazilian Institute of Geography and Statistics (IBGE) in 2013 and 2019. The surveyed population includes Brazilian permanent dwellers of private households, except those located in special census tracts (barracks, military bases, halls of residence, settlements, boats, penitentiaries, penal colonies, prisons, jails, asylums, orphanages, convents, and hospitals) and in indigenous lands.

The PNS sample is a subsample of IBGE’s Master Sample of the Integrated Household Survey System (SIPD), which consists of a set of units from areas selected to meet the various SIPD surveys, following four-criteria stratification: administrative, according to the administrative division of Brazil (Federation Units, capitals, metropolitan regions); geographic (districts and subdistricts of large municipalities); urban/rural situation of the area units; statistical, to obtain more accurate estimates¹⁵.

The PNS employed a three-stage cluster sampling plan, stratifying primary sampling units (PSU). In the first stage, the selection of primary sampling units (census tracts or tract composition) was carried out by simple random sampling, with the same stratification as the Master Sample. In the second stage, a fixed number of permanent private households was randomly selected in each PSU selected in the first stage from the National Register of Addresses for Statistical Purposes (CNEFE). In the third stage, a resident (aged 18 or older in 2013 and aged 15 or older in 2019) was selected within each household in the sample, with equiprobability, from a list of eligible residents built at the time of the interview, to answer the individual interview¹⁵.

The National Research Ethics Commission (CONEP) approved the 2013 PNS in July 2013 under opinion N° 328.159, and the 2019 PNS in August 2019 under opinion N° 3.529.376.

The PNS questionnaire is divided into three parts: the household, all household residents, and the individual. Home-based questionnaires and for all household residents are answered by a resident of the household who knows how to inform about the socioeconomic and health status of all residents. A selected resident answers the individual questionnaire with equiprobability among all adult household residents. The 2013 PNS included residents aged 18 or over to answer the individual interview, while residents aged 15 or over were included in the 2019 PNS.

The IBGE was responsible for the fieldwork of the two editions of the PNS. In 2013, at the end of fieldwork, 69,994 households were occupied, and 64,348 household interviews were held. In 2019, 108,525 households were visited, and 94,114 interviews were conducted. The non-response rates were 8.1% and 6.4%, respectively.

The expansion factors were calculated by the inverse of the selection probability product at each stage, including a correction factor for the losses. A calibration process based on population projections for Brazil and Federation Units was carried out after weighing the bases by natural expansion factors. IBGE recalibrated the expansion factors of the 2013 PNS¹⁶ to allow comparisons between the two editions of the PNS.

The construction of the 2019 PNS questionnaire followed the logic of providing a comparison with the 2013 PNS to monitor health indicators that are space-time¹⁷ monitored. In this study, information from the questionnaire of all household residents regarding the demand and use of health services was analyzed and compared.

Two outcomes were considered. The first referred to the “Seeking health-related care in the last two weeks before the survey”, based on the following question: “In the last two weeks, did you search for a place, service, or health professional for care related to your health?” The reasons for seeking were also considered, and the indicator was broken down into two others: “Seeking care due to illness or health problem” and “Seeking preventive care”. The second referred to the “Medical visit in the last twelve months before the survey”, which was based on the following question: “When was the last time that you visited a doctor?” and individuals with at least one medical visit in the last 12 months before the survey were considered.

Outcomes were analyzed according to demographic indicators (gender, ethnicity/skin color, age range), regions of residence, and urban/rural situation in the residence tract. The analysis also included classes of per capita income concerning minimum wage (MW) and having health insurance.

Health care need indicators were also considered, established by the following variables: health status assessment (good, fair, poor); having a diagnosis of a chronic, physical or mental illness, or long-term illness with a duration of more than six months (no, yes, without limitations on usual activities, and yes, but with limitations on usual activities); and health problem in the last two weeks, based on the question “In the last two weeks, how many days did you fail to perform your usual activities, due to your health?”.

Additionally, in the analysis, the indicator “Having a usual source of care” was considered based on the following question: “Do you usually look for the same place, same doctor, or same health service when you need health care?” If so, the types of a usual source of care were categorized as: primary care unit (UBS); emergency care unit (UPA); another public service; private office; private emergency care; other.

The statistical analysis consisted of comparing the distributions of the individuals surveyed in 2013 and 2019 according to sociodemographic, geographic, and health conditions factors by the estimates of the proportions and respective 95% confidence intervals in each category of the variables considered in the study.

In order to verify possible changes in health service utilization indicators between 2013 and 2019, the outcomes “Demand for health-related care in the last two weeks” and “Use of medical visit in the last year” were analyzed according to sociodemographic, geographic, and health conditions factors, estimating the prevalence and respective 95% confidence intervals for all categories of each study variable. Student’s t-test of independent samples was used considering significant differences at the 1% level to compare prevalence. Borderline differences were considered when p-values were close to 1% (descriptive level of significance of the test).

Considering as response variables “Seeking care due to illness or health problem” and “Seeking preventive care”, and “Medical visit in the last twelve months before the research”, multivariate Poisson regression models were used to identify the factors most associated with outcomes, with PNS data from 2013 and 2019, among individuals aged 30 years or older.

First, we used covariates age, gender, regions of residence, urban/rural situation, having a usual source of care, and per capita income. A logarithmic transformation was adopted to control the per capita income variance. Subsequently, the per capita income variable was replaced by having a health insurance plan. Due to the collinearity between per capita income and having a health insurance plan, the two covariates were used separately in the Poisson regression models. Prevalence ratios were estimated, and significance tests were performed at the 1% level. The sampling design of the two PNS editions was considered in the data’s statistical analysis, taking into account the sample weights, and the conglomeration effect. We used Software for Statistics and Data Science (StataCorp LP, CollegeStation, Texas, United States), version 14.0, module “survey”.

Results

In 2013, 205,546 people participated in the PNS, 48.1% male and 51.9% female. In 2019, 279,382 individuals participated, 47.8% men and 52.2% women. Comparing the distributions by age group in 2013 and 2019, we can see the Brazilian population’s aging: in 6 years, the percentage of older adults increased from 13.2% to 16.4%. A significant increase in the percentage of self-declared black people was observed. No statistically significant differences were noted between 2013 and 2019 regarding the distribution by per capita income class. Also, comparing distributions by region of residence and urban/rural situation did not show statistically significant differences (Table 1).

Table 1
Proportional distribution (%) of the population according to demographic and socioeconomic characteristics, region of residence, indicators of health conditions, and usual source of care. Brazil, National Health Survey, 2013 and 2019.

Variables	Category	2013		2019
		%	95% CI	%	95% CI
Age group	0-4	6.4	6.2-6.5	6.1	6.0-6.3
	5-14	15.4	15.1-15.7	13.5	13.3-13.8
	15-29	24.7	24.3-25.0	21.7	21.4-22.0
	30-44	22.3	22.0-22.6	22.6	22.3-22.9
	45-59	18.1	17.8-18.4	19.6	19.3-19.9
	60+	13.2	12.8-13.5	16.4	16.0-16.8
Gender	M	48.1	47.9-48.4	47.8	47.6-48.0
	F	51.9	51.6-52.1	52.2	52.0-52.4
Ethnicity/skin color	White	45.9	45.2-46.6	43.4	42.8-44.1
	Black	8.6	8.2-9.0	10.4	10.1-10.7
	Brown	44.2	43.6-44.9	44.9	44.3-45.5
	Other	1.2	1.1-1.3	1.3	1.1-1.4
Region of residence	North	8.4	8.1-8.6	8.6	8.3-9.0
	Northeast	27.6	27.1-28.1	27.2	26.6-27.8
	Southeast	42.2	41.6-42.9	42.2	41.3-43.0
	South	14.3	13.9-14.8	14.3	13.9-14.7
	Center-West	7.5	7.3-7.7	7.7	7.4-8.0
Situation	Urban	85.2	84.7-85.7	85.4	85.0-85.8
	Rural	14.8	14.3-15.3	14.6	14.2-15.0
Per capita income	Up to ½ MW	26.2	25.5-27.0	26.5	25.9-27.1
	½ to 1 MW	29.9	29.1-30.6	29.2	28.6-29.8
	1 to 2 MW	25.8	25.1-26.4	25.9	25.4-26.4
	2 to 3 MW	8.0	7.6-8.4	8.2	7.9-8.5
	3 MW and over	10.1	9.5-10.8	10.2	9.6-10.7
Health insurance plan	Yes	27.9	27.1-28.8	28.5	27.8-29.2
Health status	Good	74.1	73.6-74.6	73.0	72.6-73.4
	Fair	21.6	21.1-22.0	22.3	22.0-22.7
	Poor	4.3	4.2-4.5	4.7	4.5-4.8
Health problem in the last two weeks	Yes	7.0	6.8-7.2	8.1	7.9-8.3
Chronic disease	No	85.0	84.6-85.4	68.5	68.1-68.9
	Yes, without limitations	8.1	7.8-8.5	16.2	15.9-16.5
	Yes, with limitations	6.9	6.7-7.1	15.3	15.0-15.6
Usual source of care	Yes	77.8	77.0-78.6	76.5	75.9-77.1
Source type (among those with a usual source of care)	UBS	47.8	46.7-49.0	46.8	45.8-47.9
	UPA	4.1	3.7-4.6	14.1	13.4-14.9
	Other public services	19.1	18.3-20.0	8.9	8.5-9.4
	Private clinic	20.2	19.4-21.1	22.9	22.1-23.6
	Private emergency care	4.9	4.4-5.4	4.4	4.0-4.8
	Other	3.9	3.6-4.2	2.8	2.6-3.1

Concerning the health conditions variables, significant differences were found for two of the three indicators considered in the study between 2013 and 2019: the percentage of people with a health problem in the last two weeks increased from 7.0 to 8.1%; the proportion of people diagnosed with a chronic disease with no limitations to perform their usual activities increased from 8.1 to 16.2%, and those with chronic disease with limitations hiked from 6.9 to 15.3%. Borderline differences were found for health status assessment: the regular evaluation proportion increased from 21.6 to 22.3% and poor assessment from 4.3 to 4.7% (Table 1).

No significant differences were found regarding having health insurance, with proportions ranging from 27.9% in 2013 to 28.5% in 2019. Likewise, the percentages of people with a usual care source when they need it did not vary significantly (77.8% in 2013, and 76.5%, in 2019). As for the type of public service as the usual source of care, while UBS proportions remained similar, at 47.8% and 46.8%, in 2013 and 2019, respectively, an apparent increase was observed in using UPAs, from 4.1% to 14.1%. A significant increase in private practice was noted in the private sector, from 20.2 to 22.9% (Table 1).

Table 2 shows the differences in the proportions of people who sought health-related care in the last two weeks before the survey, according to the study’s indicators. Between 2013 and 2019, the percentage of demand for care increased from 15.3 to 18.6%: it increased from 11.5 to 14.7% due to illness or health problem, and from 1.9 to 3.7% for preventive care. The largest relative increase was for preventive care.

Table 2
Proportion (%) of people who had health-related care in the last two weeks before the survey, according to demographic and socioeconomic characteristics, region of residence, indicators of health conditions, and usual care source. Brazil, National Health Survey, 2013 and 2019.

Variables	Category	2013		2019
		%	95% CI	%	95% CI
Total		15.3	15.0-15.7	18.6	18.3-19.0
Reason for use	Disease	11.5	11.2-11.8	14.7	14.4-15.0
	Prevention	1.9	1.8-2.0	3.7	3.5-3.9
Age group	0-4	18.6	17.5-19.7	21.7	20.6-22.8
	5-14	10.1	9.4-10.7	12.6	12.0-13.3
	15-29	10.9	10.4-11.4	13.3	12.8-13.8
	30-44	14.2	13.7-14.8	17.2	16.6-17.7
	45-59	19.0	18.2-19.7	22.0	21.3-22.6
	60+	24.9	24.0-25.9	27.5	26.8-28.3
Gender	M	11.8	11.5-12.2	14.8	14.5-15.1
	F	18.5	18.1-19.0	22.1	21.7-22.6
Ethnicity/skin color	White	17.0	16.4-17.5	20.2	19.7-20.7
	Black	14.7	13.7-15.7	19.1	18.3-19.9
	Brown	13.7	13.2-14.1	17.0	16.6-17.5
	Other	17.5	15.0-20.2	18.9	15.6-22.7
Region of residence	North	10.0	9.5-10.7	13.7	13.1-14.3
	Northeast	13.4	12.8-13.9	16.6	16.1-17.0
	Southeast	17.1	16.4-17.7	20.9	20.3-21.6
	South1	17.9	17.0-18.9	19.8	19.1-20.5
	Center-West	13.7	13.1-14.4	16.7	16.0-17.4
Situation	Urban	15.8	15.4-16.2	19.3	18.9-19.6
	Rural	12.4	11.7-13.1	14.8	14.3-15.4
Per capita income	Up to ½ MW	12.8	12.2-13.4	15.3	14.8-15.9
	½ to 1 MW	14.8	14.2-15.3	17.8	17.3-18.4
	1 to 2 MW	16.1	15.5-16.8	19.4	18.7-20.0
	2 to 3 MW	18.0	16.8-20.5	21.8	20.7-22.9
	3 MW and over	19.3	18.2-20.5	25.0	24.1-25.9
Health insurance plan	No	13.7	13.3-14.1	16.4	16.1-16.8
	Yes	19.5	18.8-20.2	24.1	23.5-24.8
Health status	Good	11.6	11.3-12.0	14.6	14.2-14.9
	Fair	23.6	22.9-24.4	26.8	26.1-27.4
	Poor	36.9	35.2-38.7	43.2	41.7-44.6
Health problem in the last two weeks	No	11.9	11.6-12.2	14.8	14.5-15.1
	Yes1	61.0	59.5-62.6	62.4	61.4-63.5
Chronic noncommunicable disease	No1	12.5	12.1-12.8	13.0	12.7-13.3
	Yes, without limitations1	25.2	23.9-26.5	24.5	23.7-25.2
	Yes, with limitations1	38.6	37.2-40.0	37.6	36.7-38.4
Usual source of care	No	12.1	11.5-12.7	15.8	15.2-16.3
	Yes	16.2	15.8-16.7	19.5	19.1-19.9

• ¹ Non-significant differences in the level of 1%.

The analysis of the demand for health-related care in the last two weeks by age group shows a higher prevalence in extreme groups, from 0-4 years and 60 years or more, both in 2013 and in 2019, and the most significant growth occurred among children under five years old (3.1%). The comparison of prevalence by gender shows greater demand for care among women, but significant increases for both genders, between 2013 and 2019. Regarding ethnicity/skin color, while white individuals are more prevalent, growth was more pronounced for black people, exceeding four percentage points in the period (Table 2).

The region of residence analysis points to the substantial regional inequalities in the health care demand in the past two weeks. In 2013, the prevalence ranged from 10.0% in the North Region to 17.9% in the South Region, and in 2019, from 13.7% in the North Region to 20.9% in the Southeast Region. As for the urban/rural situation, the most significant increase in prevalence occurred among residents of urban areas, escalating inequalities (Table 2).

While significant increases were observed in the prevalence of seeking health care in all per capita income classes in the period studied, income inequalities remain very pronounced. In 2013, the percentage of people who sought health care ranged from 12.8%, among those with the worst per capita income, to 19.3%, among those with the highest income, with a difference of 6.5% between extreme classes. In 2019, the difference was even more pronounced, from 9.7%, with a gradient of 15.3 to 25.0%, in the worst to highest income classes. The prevalence increased by 4.6 percentage points among people with health insurance, and by 3.3 percentage points among people with a usual care source (Table 2).

The highest proportions of demand for health-related care in the last two weeks, both in 2013 and in 2019, occurred among people in need of care: over 60% among those who had a health problem in the last two weeks, without significant differences between 2013 and 2019; about 25% and 38%, among those who have a chronic disease without and with limitations in usual activities, also without significant differences in the period; and significantly increasing from 36.9 to 43.2%, between 2013 and 2019, among those with poor health status.

In Table 3, the proportions of people who saw a doctor in the last 12 months were analyzed according to the same indicators, comparing the results obtained in 2013 and 2019. The percentage of people who saw a doctor in the last year increased from 71.2 to 76.2%. The highest percentages of medical visits are for children under five years of age and older adults, reaching 90.9% and 86.9% in 2019, respectively. The lowest prevalence corresponds to children over five years old, adolescents, and young adults.

Table 3
Proportion (%) of people who visited a doctor in the last 12 months before the survey according to demographic and socioeconomic characteristics, region of residence, indicators of health conditions, and usual care source. Brazil, National Health Survey, 2013 and 2019.

Variables	Category	2013		2019
		%	95% CI	%	95% CI
Total		71.2	70.7-71.7	76.2	75.8-76.5
Age group	0-4	87.7	86.8-88.6	90.9	90.2-91.6
	5-14	66.5	65.5-67.5	72.7	71.7-73.6
	15-29	62.7	61.9-63.6	67.0	66.3-67.7
	30-44	69.0	68.2-69.8	73.1	72.5-73.7
	45-59	74.6	73.7-75.4	78.6	78.0-79.2
	60+	83.5	82.8-84.2	86.9	86.4-87.4
Gender	M	63.9	63.2-64.6	69.4	69.0-69.9
	F	78.0	77.5-78.5	82.3	82.0-82.7
Ethnicity/skin color	White	74.8	74.1-75.5	79.4	78.9-79.9
	Black	69.5	68.1-70.9	74.8	73.9-75.7
	Brown	67.7	67.1-68.4	73.3	72.8-73.8
	Other	73.9	71.0-76.6	78.7	75.0-82.0
Region of residence	North	61.4	59.9-62.8	68.0	66.9-69.0
	Northeast	66.3	65.5-67.1	71.9	71.3-72.5
	Southeast	75.8	74.8-76.8	80.6	79.9-81.2
	South	73.8	72.6-74.9	77.7	76.9-78.5
	Center-West	69.5	68.4-70.6	73.7	72.6-74.8
Situation	Urban	72.7	72.1-73.2	77.5	77.1-77.9
	Rural	62.8	61.5-64.0	68.6	67.8-69.4
Per capita income	Up to ½ MW	64.6	63.6-65.5	69.7	69.0-70.5
	½ to 1 MW	69.3	68.4-70.1	74.6	74.0-75.2
	1 to 2 MW	73.7	72.7-74.6	78.1	77.5-78.7
	2 to 3 MW	77.1	75.6-78.5	82.4	81.4-83.5
	3 MW and over	83.1	82.1-84.1	87.6	86.8-88.4
Health insurance plan	No	65.9	65.3-66.6	71.6	71.2-72.1
	Yes	84.8	84.0-85.5	87.6	87.1-88.1
Health status	Good	67.7	67.1-68.4	73.0	72.5-73.4
	Fair	79.9	79.1-80.6	83.3	82.8-83.8
	Poor	87.8	86.5-88.9	91.8	91.0-92.5
Health problem in the last two weeks	No	69.6	69.0-70.1	74.6	74.2-75.0
	Yes1	92.8	92.0-93.5	94.0	93.5-94.5
Chronic disease	No	67.8	67.2-68.4	69.7	69.2-70.2
	Yes, without limitations1	88.8	87.9-89.7	88.0	87.5-88.6
	Yes, with limitations1	92.4	91.6-93.1	92.6	92.2-93.0
Usual source of care	No	63.2	62.2-64.3	69.7	68.9-70.5
	Yes	73.5	72.9-74.1	78.2	77.8-78.6

• ¹ Non-significant differences in the level of 1%.

The differences by gender regarding doctors’ use in the last year persisted but decreased in 2019, due to the more significant increase for males. As for differences by ethnicity/skin color, white individuals had significantly higher prevalence rates, although a lower ratio of inequalities between the prevalence of use of “whites” and “blacks” was observed in the period analyzed (Table 3).

The region of residence analysis shows large discrepancies in the proportion of people with a medical visit in the last 12 months, ranging from 61.4% in the North Region to 75.8% in the Southeast Region in 2013, and from 68.0% to 80.6% in 2019, respectively. The prevalence of medical visits is significantly higher among residents of urban areas, with differences of around ten percentage points in 2013 and nine percentage points in 2019 (Table 3). Inequalities by per capita income class are very pronounced, both in 2013 and in 2019. In the last year, the prevalence of medical visits ranged from 64.6 to 83.1% in 2013 and from 69.7 to 87.6 % in 2019, reaching differences of approximately 18 percentage points between the lower and upper levels of the income classes.

Sharp and significant differences in having health insurance were also found in the two years analyzed: in 2013, the percentage of people who visited a doctor in the last year ranged from 65.9 to 84.8%, and from 71.6 to 87.6% in 2019, corresponding to variations of 19 and 16 percentage points, respectively, among those having and not having health insurance (Table 3).

Considering the health conditions indicators, the proportions of individuals who visited a doctor in the past year grow with increasing needs. According to the health status, the proportions ranged from 67.7 to 87.8% among people with a good and bad evaluation in 2013, and from 73.0 to 91.8%, in 2019, respectively. The prevalence reached more than 92%, both in 2013 and 2019, considering the individuals with health problems in the last two weeks. In the last year, significant prevalence levels of medical visits are found among people diagnosed with some chronic disease, without and with limitations in usual activities of about 88 and 92%, respectively (Table 3).

Table 4 shows the results of the multivariate Poisson regression models for individuals aged 30 years or older. The following were found considering “Demand for care due to illness or health problem” as the answer variable both in 2013 and 2019: higher prevalence of demand with age; prevalence ratios 50% higher among women compared to men; 12% higher among residents of urban areas than those in rural areas; and about 25% higher among those with a usual source of care. The differences between the Southeast and South and North regions are significant and pronounced in the two years considered. The association of the per capita income variable with the outcome is not statistically significant after controlling for age, gender, urban/rural situation, the region of residence, and having a usual care source, both in 2013 and 2019.

Table 4
Factors associated (considering per capita income as a covariate) with the demand for health-related care in the last two weeks before the survey according to the reason (health problem or prevention) among individuals aged 30 and over. Brazil, National Health Survey, 2013 and 2019.

Seeking care due to health problems
Variables	Category	2013		2019
		Adjusted PR	P-value	Adjusted PR	P-value
Age		1.02	<0.001	1.01	<0.001
Gender	M	1.00	-	1.00	-
	F	1.50	<0.001	1.46	<0.001
Region of residence	North	1.00	-	1.00	-
	Northeast	1.04	0.420	1.13	<0.001
	Southeast	1.37	<0.001	1.31	<0.001
	South	1.53	<0.001	1.28	<0.001
	Center-West	1.09	0.085	1.15	<0.001
Situation	Urban	1.12	0.002	1.12	<0.001
	Rural	1.00	-	1.00	-
Usual source of care	No	1.00	-	1.00	-
	Yes	1.28	<0.001	1.25	<0.001
Per capita income (logarithm)		0.98	0.081	1.00	0.608
Seeking preventive care
Variables	Category	2013		2019
		Adjusted PR	P-value	Adjusted PR	P-value
Age		1.00	0.866	1.01	0.001
Gender	M	1.00	-	1.00	-
	F	2.43	<0.001	1.68	<0.001
Region of residence	North	1.00	-	1.00	-
	Northeast	1.94	<0.001	1.22	0.004
	Southeast	1.73	<0.001	1.23	0.007
	South	1.68	0.001	1.14	0.101
	Center-West	1.45	0.013	1.01	0.861
Situation	Urban	1.23	0.088	1.46	<0.001
	Rural	1.00	-	1.00	-
Usual source of care	No	1.00	-	1.00	-
	Yes	1.18	0.083	1.07	0.214
Per capita income (logarithm)		1.21	<0.001	1.29	<0.001

Also, in Table 4, comparing the two outcomes, “seeking health-related care due to a health problem” with “seeking preventive care”, shows higher prevalence ratios for gender and urban/rural situation, and with significant differences between Northeast and Southeast regions compared to the North. As for having a usual source of care, the prevalence ratios become insignificant, while the associations with per capita income are significant at the level of 1% (adjusted PR=1.21, in 2013, and adjusted PR=1.29, in 2019).

Table 5 shows the multivariate models’ results, replacing the variable per capita income with having a health insurance plan. Regarding the demand for care due to illness or health problems, the associations with age, gender, urban/rural situation, having a usual source of care, and the differences by region were similar to those found in the models shown in Table 4. However, the prevalence of seeking care was significantly higher among people with a health insurance plan, with adjusted prevalence ratios of 1.11 and 1.16 in 2013 and 2019, respectively. Concerning seeking preventive care, the adjusted prevalence ratios are much higher, at 1.77 in 2013 and 2.10 in 2019; that is, individuals with health insurance have a prevalence twice as high of seeking preventive care.

Table 5
Factors associated (considering health insurance plan as a covariate) with the demand for health-related care in the last two weeks before the survey according to the reason (health problem or prevention) among individuals aged 30 and over. Brazil, National Health Survey, 2013 and 2019.

Seeking care due to health problems
Variables	Category	2013		2019
		Adjusted PR	P-value	Adjusted PR	P-value
Age		1.02	<0.001	1.01	<0.001
Gender	M	1.00	-	1.00	-
	F	1.50	<0.001	1.45	<0.001
Region of residence	North	1.00	-	1.00	-
	Northeast	1.04	0.447	1.13	<0.001
	Southeast	1.33	<0.001	1.28	<0.001
	South	1.48	<0.001	1.25	<0.001
	Center-West	1.07	0.212	1.13	0.001
Situation	Urban	1.09	0.031	1.00	-
	Rural	1.00	-	1.08	0.002
Usual source of care	No	1.00	-	1.00	-
	Yes	1.27	<0.001	1.24	<0.001
Health insurance plan	No	1.00	-	1.00	-
	Yes	1.11	0.001	1.16	<0.001
Seeking preventive care
Variables	Category	2013		2019
		Adjusted PR	P-value	Adjusted PR	P-value
Age		1.00	0.675	1.01	<0.001
Gender	M	1.00	-	1.00	-
	F	2.37	<0.001	1.62	<0.001
Region of residence	North	1.00	-	1.00	-
	Northeast	1.90	<0.001	1.20	0.007
	Southeast	1.67	<0.001	1.20	0.015
	South	1.68	0.001	1.17	0.056
	Center-West	1.43	0.015	1.04	0.645
Situation	Urban	1.18	0.164	1.37	<0.001
	Rural	1.00	-	1.00	-
Usual source of care	No	1.00	-	1.00	-
	Yes	1.15	0.145	1.04	0.487
Health insurance plan	No	1.00	-	1.00	-
	Yes	1.77	<0.001	2.10	<0.001

Discussion

The results of this study show growth both in demand and in the use of health services, with increases in the proportion of individuals who sought health-related care in the two weeks before the survey and in the proportion of people who visited a doctor in the last 12 months before the survey, when the 2019 PNS data are compared to those of the 2013 PNS. The benefits of expanding the coverage of the use of health services are duly recognized since the increased prevention, diagnosis, and early treatment of diseases can result in lower levels of premature mortality, increased life expectancy with quality, and reduction of unnecessary hospitalizations¹⁸. However, the interpretation of the increased use of health services is not immediate since it depends on several factors, such as the need, the sociodemographic characteristics, the provision of services, doctors’ availability, health care financing, and the users’ perception of the care provided¹⁹.

This study’s findings show the aging of the Brazilian population between 2013 and 2019 and the increased prevalence of chronic noncommunicable diseases, generating more health care needs, limitations in activities of daily living, and more significant hardships in accessing health services among older adults²⁰. Challenges related to minimizing access barriers among older adults, offering more services, and providing financing for preventing and treating chronic noncommunicable diseases have also been faced in other countries^21-23.

It is interesting to note that there was a significant increase in the percentage of people who self-declared black between 2013 and 2019, corroborating results of the analysis of birth cohorts with data from the National Household Sample Survey, which show that cohorts become less “white” each year, especially after the 2000s. Variations in proportional distributions by ethnicity/skin color are attributed to changes in skin color self-identification during life, in a movement consistent with a higher percentage of declaring black skin color to the detriment of white²⁴.

Regarding the differences in the use of health services by gender, higher percentages of seeking care and using a doctor were found among women. In general, studies that address morbidity and the use of health services report that women have a worse perception of health status and are more likely to use health care¹³. As for age, the highest frequencies of seeking health-related care occur among those with the greatest need for care, children under five years of age, and older adults, corroborating findings from previous studies^25,26.

Having a health problem continued to be the main reason for seeking services in 2013 and 2019. However, a more significant relative increase was observed in the proportion of preventive visits, suggesting that the aging of the Brazilian population and the increased need for health care possibly explain the greater use of health services, but other factors may have influenced the expanded search for care. The National Health Promotion Policy (PNPS) was implemented in the mid-2000s and provided greater availability of prevention services in PHC. A study conducted in 2014 showed that health promotion programs were in place in most PHC units²⁷.

Having health insurance showed no significant increase between 2013 and 2019, with more than 70% of the population using the public system²⁸. Having a usual source of care was a significant factor in seeking health care. The main source of care in the SUS was the primary network, with about 60% of the visits in public establishments carried out in the UBS in the two years considered, as already pointed out in a publication with data from the 2013 PNS²⁹. However, significant growth was observed in the use of UPAs as a usual source of care from 2013 to 2019, showing that the expanded access was relevant for obtaining care, but it is not always adequate if it does not fully meet users’ needs³⁰. A study carried out in the U.S. shows that the provision of PHC through alternative services can overcome some barriers to access, but the lack of integration in the service network may result in users’ dissatisfaction with the quality of the service provided³¹.

The proportion of individuals who visited a doctor in the last year increased considerably, by five percentage points, reaching 76% in 2019, a level similar to that of some developed countries³². The prevalence was higher among individuals with health insurance plans, reaching 88% in 2019, but the increase in doctors’ use (5.7 percentage points) was higher among SUS users than in those with a health insurance plan (2.8 percentage points).

Having a health insurance plan can mitigate possible financial barriers when consuming services and lead to a timelier response with reduced waiting times for care and greater user satisfaction¹⁹. As discussed in other national papers, a health insurance plan increases the number of medical visits^14,28 and determines the greater use of preventive services³³.

The analysis of multivariate models, both in 2013 and in 2019, showed that, in the situation of seeking care due to illness or health problem, no significant association was observed with the income level, after controlling for age, gender regions of residence, urban/rural situation, and having a usual source of care. These results indicate that health services are accessible to individuals with a perception of their health problems, regardless of income level, and are particularly relevant in the context of reducing socioeconomic inequalities in health.

However, the multivariate analysis of the data from the two PNS editions related to the search for care by region of residence shows that regional inequalities are still pronounced, even controlling the effects of income and having a health insurance plan, regardless of the reason for the demand. Regional inequalities in the use of health services have been pointed out since 1998, with the highest prevalence in the Southeast and South and the lowest in the North³⁴. Differentiated access to services and inequality in the geographic distribution of available resources have been considered the main explanatory mechanisms^35,36.

This work shows the good performance of the national health system in expanding the coverage of health services, allowing to meet changes in the population’s health needs, despite the chronic underfunding of the SUS and austerity policies after 2016. However, regional inequalities indicate that the health needs of the population living in less developed regions are not being met and should be the subject of further studies to support the planning of strategies aimed at overcoming inequalities. Care models more focused on prevention and health promotion are required³⁷ given the Brazilian population’s aging, the increased obesity and prevalence of chronic diseases, and the urgency of adopting healthy behaviors by the Brazilian population.

One of the limitations of this work refers to the differences in the questionnaires of the two surveys. In 2019, the module on the evaluation of medical care was excluded, and it was not possible to know the reason for the medical visit in the last year (health problem or prevention), hindering a multivariate analysis similar to the one presented for the outcome “Seeking health care in the 15 days before the survey”.

Another limitation is that health service use indicators are based on reported data, which are subject to memory bias. The questions in this PNS module can also be answered by the household’s key informant, increasing the possibility of errors in measuring past events. Finally, due to the large sample size in the two editions of the PNS, small differences can be considered statistically significant, and the associations found in this work should be examined in the light of this limitation.

Referências

• ¹
  Paim J, Travassos C, Almeida C, Bahia L, Macinko J. The Brazilian health system: history, advances, and challenges. Lancet 2011; 377(9779):1778-1797.
• ²
  Escorel S, Giovanella L, Mendonça MHM, Senna MCM. The Family Health Program and the construction of a new model for primary care in Brazil. Rev Panam Salud Publica 2007; 21(2-3):164-176.
• ³
  Giovanella L, Mendonça MH, de Almeida PF, Escorel S, Senna MC, Fausto MC, Delgado MM, Andrade CL, Cunha MS, Martins MI, Teixeira CP. Family health: limits and possibilities for an integral primary care approach to health care in Brazil. Cien Saude Colet 2009; 14(3):783-794.
• ⁴
  Castro MC, Massuda A, Almeida G, Menezes-Filho NA, Andrade MV, Souza Noronha KVM, Rocha R, Macinko J, Hone T, Tasca R, Giovanella L, Malik AM, Werneck H, Fachini LA, Atun R. Brazil's unified health system: the first 30 years and prospects for the future. Lancet 2019; 394(10195):345-356.
• ⁵
  Lima LD, Albuquerque MV, Scatena JHG, Melo ECP, Oliveira EXG, Carvalho MS, Pereira AMM, Oliveira RAD, Martinelli NL, Oliveira CF. Regional governance arrangements of the Brazilian Unified National Health System: provider diversity and spacial inequality in service provision. Cad Saude Publica 2019; 35(Supl. 2):e00094618.
• ⁶
  Peltzer K. Patient experiences and health system responsiveness in South Africa. BMC Health Serv Res 2009; 9:117.
• ⁷
  Birch S, Gafni A. Achievements and challenges of medicare in Canada: Are we there yet? Are we on course? Int J Health Serv 2005; 35(3):443-463.
• ⁸
  Macinko J, Guanais FC. Population experiences of primary care in 11 Organization for Economic Cooperation and Development countries. Int J Qual Health Care 2015; 27(6):443-450.
• ⁹
  Gouveia GC, Souza WV, Luna CF, Souza-Júnior PR, Szwarcwald CL. Health care users' satisfaction in Brazil, 2003. Cad Saude Publica 2005; 21(Supl. 1):109-118.
• ¹⁰
  Travassos C, Viacava F. Utilização e financiamento de serviços de saúde: dez anos de informação das PNAD. Cien Saude Colet 2011; 16(9):3646-3647.
• ¹¹
  Szwarcwald CL, Malta DC, Pereira CA, Vieira ML, Conde WL, Souza Júnior PR, Damacena GN, Azevedo LO, Azevedo E, Silva G, Theme Filha MM, Lopes CS, Romero DE, Almeida WS, Monteiro CA. Pesquisa Nacional de Saúde no Brasil: concepção e metodologia de aplicação. Cien Saude Colet 2014; 19(2):333-342.
• ¹²
  Malta DC, Silva MMAD, Moura L, Morais Neto OL. The implantation of the Surveillance System for Non-communicable Diseases in Brazil, 2003 to 2015: successes and challenges. Rev Bras Epidemiol 2017; 20(4):661-675.
• ¹³
  Silva ZP, Ribeiro MC, Barata RB, Almeida MF. Perfil sociodemográfico e padrão de utilização dos serviços de saúde do Sistema Único de Saúde (SUS), 2003-2008. Cien Saude Colet 2011; 16(9):3807-3816.
• ¹⁴
  Pilotto LM, Celeste RK. Tendências no uso de serviços de saúde médicos e odontológicos e a relação com nível educacional e posse de plano privado de saúde no Brasil, 1998-2013. Cad Saude Publica 2018; 34(4): e00052017.
• ¹⁵
  Souza-Jr PRB, Freitas MPS, Antonaci GA, Szwarcwald CL. Desenho da amostra da Pesquisa Nacional de Saúde 2013. Epidemiol Serv Saúde 2015; 24(2):207-216.
• ¹⁶
  Instituto Brasileiro de Geografia e Estatística (IBGE). Nota Técnica - Informações referentes à revisão do plano tabular da PNS 2013. [nota técnica internet]. 2020 Set [acessado 2020 Nov 03]; [2p.]. Disponível em: https://ftp.ibge.gov.br/PNS/Documentacao_Geral/Nota_Tecnica_Revisao_Plano_Tabular_PNS2013.pdf
  » https://ftp.ibge.gov.br/PNS/Documentacao_Geral/Nota_Tecnica_Revisao_Plano_Tabular_PNS2013.pdf
• ¹⁷
  Stopa SR, Szwarcwald CL, Oliveira MM, Gouvea ECDP, Vieira MLFP, Freitas MPS, Sardinha LMV, Macário EM. National Health Survey 2019: history, methods and perspectives. Epidemiol Serv Saude 2020; 29(5):e2020315.
• ¹⁸
  Macinko J, Dourado I, Aquino R, Bonolo PdeF, Lima-Costa MF, Medina MG, Mota E, Oliveira VB, Turci MA. Major expansion of primary care in Brazil linked to decline in unnecessary hospitalization. Health Aff (Millwood) 2010; 29(12):2149-2160.
• ¹⁹
  Szwarcwald CL, Damacena GN, Souza Júnior PR, Almeida WS, Malta DC. Perception of the Brazilian population on medical health care. Brazil, 2013. Cien Saude Colet 2016; 21(2):339-349.
• ²⁰
  Louvison MC, Lebrão ML, Duarte YA, Santos JL, Malik AM, Almeida ES. Desigualdades no uso e acesso aos serviços de saúde entre idosos do município de São Paulo. Rev Saude Publica 2008; 42(4):733-740.
• ²¹
  Zhang T, Liu J, Liu C. Changes in Perceived Accessibility to Healthcare from the Elderly between 2005 and 2014 in China: An Oaxaca-Blinder Decomposition Analysis. Int J Environ Res Public Health 2019; 16(20):3824.
• ²²
  Barr ML, Welberry H, Comino EJ, Harris-Roxas BF, Harris E, Lloyd J, Whitney S, O'Connor C, Hall J, Harris MF. Understanding the use and impact of allied health services for people with chronic health conditions in Central and Eastern Sydney, Australia: a five-year longitudinal analysis. Prim Health Care Res Dev 2019; 20:e141.
• ²³
  Jayathilaka R, Joachim S, Mallikarachchi V, Perera N, Ranawaka D. Chronic diseases: An added burden to income and expenses of chronically-ill people in Sri Lanka. PLoS One 2020; 15(10):e0239576.
• ²⁴
  Soares AS. A demografia da cor: a composição brasileira de 1890 a 2007. In: Theodoro M, organizador. As políticas púbicas e a desigualdade racial no Brasil: 120 anos após a abolição. Brasília: IPEA; 2008.
• ²⁵
  Oliveira BLCA, Moreira JPL, Luiz RR. The influence of the Family Healthcare Strategy in the use of healthcare services by children in Brazil: an analysis using the Propensity Score Matching (PSM) method of National Health Survey data. Cien Saude Colet 2019; 24(4):1495-1505.
• ²⁶
  Malta DC, Bernal RTI, Lima MG, Araújo SSC, Silva MMAD, Freitas MIF, Barros MBA. Noncommunicable diseases and the use of health services: analysis of the National Health Survey in Brazil. Rev Saude Publica 2017; 51(Supl. 1):4s.
• ²⁷
  Ramos LR, Malta DC, Gomes GA, Bracco MM, Florindo AA, Mielke GI, Parra DC, Lobelo F, Simoes EJ, Hallal PC. Prevalence of health promotion programs in primary health care units in Brazil. Rev Saude Publica 2014; 48(5):837-844.
• ²⁸
  Stopa SR, Malta DC, Monteiro CN, Szwarcwald CL, Goldbaum M, Cesar CLG. Use of and access to health services in Brazil, 2013 National Health Survey. Rev Saude Publica 2017; 51(Supl. 1):3s.
• ²⁹
  Dourado I, Medina MG, Aquino R. The effect of the Family Health Strategy on usual source of care in Brazil: data from the 2013 National Health Survey (PNS 2013). Int J Equity Health 2016; 15(1):151.
• ³⁰
  Lima-Costa MF, Turci MA, Macinko J. A comparison of the Family Health Strategy to other sources of healthcare: utilization and quality of health services in Belo Horizonte, Minas Gerais State, Brazil. Cad Saude Publica 2013; 29(7):1370-1380.
• ³¹
  Reed C, Rabito FA, Werthmann D, Smith S, Carlson JC. Factors associated with using alternative sources of primary care: a cross-sectional study. BMC Health Serv Res 2019; 19(1):933.
• ³²
  Van Doorslaer E, Masseria C, Koolman X; for the OECD Health Equity Research Group. Inequalities in access to medical care by income in developed countries. CMAJ 2006; 174(2):177-183.
• ³³
  Malta DC, Bernal RTI, Vieira Neto E, Curci KA, Pasinato MTM, Lisbôa RM, Cachapuz RF, Coelho KSC, Santos FPD, Freitas MIF. Noncommunicable diseases, risk factors, and protective factors in adults with and without health Insurance. Cien Saude Colet 2020; 25(8):2973-2983.
• ³⁴
  Viacava F, Porto SM, Carvalho CC, Bellido JG. Health inequalities by region and social group based on data from household surveys (Brazil, 1998-2013). Cien Saude Colet 2019; 24(7):2745-2760.
• ³⁵
  Oliveira RAD, Duarte CMR, Pavão ALB, Viacava F. Barriers in access to services in five Health Regions of Brazil: perceptions of policymakers and professionals in the Brazilian Unified National Health System. Cad Saude Publica 2019; 35(11):e00120718.
• ³⁶
  Nogueira MC, Fayer VA, Corrêa CSL, Guerra MR, Stavola B, Dos-Santos-Silva I, Bustamante-Teixeira MT, Silva GAE. Inequities in access to mammographic screening in Brazil. Cad Saude Publica 2019; 35(6):e00099817.
• ³⁷
  Paim JS. Thirty years of the Unified Health System (SUS). Cien Saude Colet 2018; 23(6):1723-1728.

Edited by

Chief editors:

Maria Cecília de Souza Minayo, Romeu Gomes, Antônio Augusto Moura da Silva

Publication Dates

• Publication in this collection
  14 June 2021
• Date of issue
  2021

History

• Received
  11 Nov 2020
• Accepted
  01 Dec 2020
• Published
  03 Dec 2020