ABSTRACT

OBJECTIVE

To compare the profile and prevalence of hospitalizations in Brazil based on estimates from the National Health Survey (PNS), 2013 and 2019.

METHODS

A cross-sectional study that used data from the 2013 PNS and the 2019 PNS. The outcome was having been hospitalized for 24 hours or more in the last 12 months. We calculated the proportion of the population in different categories of age group, presence or absence of chronic diseases, and perception of health status. We estimated the total number of hospitalizations and the proportion corresponding to each category of age group, chronic disease, and perceived health status. We calculated the prevalence of hospitalization according to geographic, socioeconomic, and health conditions. We compared the estimates of two editions of the PNS using Student’s t-test for independent samples. We considered significant differences when the p-value was less than 0.01. And finally, we compared hospitalization estimates with administrative data to assess data consistency.

RESULTS

We observed that the proportion of chronically ill people in the population increased from 15.04% to 31.48%. This group was responsible for 36.76% of the total number of hospitalizations in 2013 and 57.61% in 2019. The prevalence of hospitalizations increased significantly between the two surveys and the increases were higher in the Southeast region and among people who have private health insurance. A discrepancy was found between administrative data and survey estimates. Obstetric hospitalizations and health insurance hospitalizations were underestimated.

CONCLUSION

There was an increase in overall hospitalization rates in the period between the PNS 2013 and PNS 2019, especially among people with better access to health services. The hospitalization profile also changed—in the 2013 PNS, hospitalizations of people without chronic diseases predominated. This was reversed in PNS 2019.

Hospitalization; Hospitals; Hospital Statistics; Epidemiological Surveys; Cross-sectional Studies

INTRODUCTION

The hospital is the most expensive piece of equipment in a health system and can represent a significant source of wasted resources, to the point of compromising the efficiency and effectiveness of this system¹1. Silva JP. Análise da eficiência de hospitais regionais em um estado do Nordeste. Saúde Debate. 2019 Mar;43(120):84-97. https://doi.org/10.1590/0103-1104201912006
https://doi.org/10.1590/0103-11042019120... ,²2. Souza FE, Nunes ED, Carvalho BG, Mendonça FD. Hospitais de cidades pequenas: inserção e pactuação no Sistema Único de Saúde (SUS). Medicina (Ribeirão Preto). 2020 out;53(3):300-8. https://doi.org/10.11606/issn.2176-7262.v53i3p300-308
https://doi.org/10.11606/issn.2176-7262.... . In 2019, spending on hospital and outpatient care accounted for 49.94% of public health expenditures in Brazil, more than double what was spent on primary care in the same year³3. Controladoria-Geral da União (BR). Portal da transparência. Saúde. Brasília, DF:Controladoria-Geral da União; 2019 [cited 2020 Dec]. Available from: http://www.portaltransparencia.gov.br/funcoes/10-saude?ano=2019
http://www.portaltransparencia.gov.br/fu... . However, despite this large amount of resources, necessary to maintain hospital care, one cannot imagine doing without the effectiveness provided by the technological apparatus concentrated in a hospital, especially in critical health conditions.

It is likely that the spending on hospital care in Brazil would have been proportionally higher without the expansion of investments in primary care that occurred after the creation of the Sistema Único de Saúde (Unified Health System – SUS) and, mainly, after the implementation of the Family Health Strategy. From these changes, there was a significant improvement in health care. The majority place of care migrated from the outpatient clinic to the health post/center⁴4. Viacava F, Oliveira RA, Carvalho CC, Laguardia J, Bellido JG, Viacava F, et al. SUS: oferta, acesso e utilização de serviços de saúde nos últimos 30 anos. Volume 23. Cienc Saúde Coletiva.2018 jun;23(6):1751-62. https://doi.org/10.1590/1413-81232018236.06022018
https://doi.org/10.1590/1413-81232018236... and, gradually, the hospital has ceased to be the gateway to the health system. However, in 2019, more than 25% of people still sought the hospital as a place of first health care⁵5. Instituto Brasileiro de Geografia e Estatística. Pesquisa nacional de saúde: 2019: informações sobre domicílios, acesso e utilização dos serviços de saúde: Brasil, grandes regiões e unidades da federação. Rio de Janeiro: IBGE; 2020..

Hospital and primary care complement each other within a health system. Primary care has the function of comprehensive and longitudinal care for the individual’s health. Within its scope, preventive and educational actions are promoted, to rehabilitation or even palliative care⁶6. Giovanella L, Mendonça M. Atenção Primária à Saúde. In: Giovanella L, Escorel S, Lobato LVC, Noronha JC, Carvalho AI, orgs. Políticas e sistema de saúde no Brasil. 2nd ed. Rio de Janeiro: Fiocruz; 2012. p. 493-545.. In turn, the hospital must resolve certain complex situations, preferably infrequently. Therefore, the aim is a health system with robust and effective primary care, which provides the lowest possible hospitalization rates⁷7. Caminal Homar J, Casanova Matutano C. La evaluación de la atención primaria y las hospitalizaciones por ambulatory care sensitive conditions. Marco conceptual. Aten Primaria. 2003 Jan;31(1):61-5. https://doi.org/10.1016/S0212-6567 (03)70662-3
https://doi.org/10.1016/S0212-6567 (03)7... .

Hospitalization rates can be influenced by social, economic, behavioral, environmental, and demographic factors, the same factors that influence a population’s health⁸8. Carrapato P, Correia P, Garcia B. Determinante da saúde no Brasil: a procura da equidade na saúde. Saúde Soc. 2017 jul-set;26(3):676-89. https://doi.org/10.1590/s0104-12902017170304
https://doi.org/10.1590/s0104-1290201717... . Conditions such as unemployment, low income, restrictions and lack of family involvement are, for example, predictors of a higher frequency of children’s hospitalization⁹9. Araújo YB, Santos SR, Neves NT AT, Cardoso ELS, Nascimento JA. Modelo preditor de internação hospitalar para crianças e adolescentes com doença crônica. Rev Bras Enferm; 2020;73(2):e20180467. https://doi.org/10.1590/0034-7167-2018-0467
https://doi.org/10.1590/0034-7167-2018-0... . In the United States, 11% of the variability in hospitalization due to influenza is determined by social factors¹⁰10. Chandrasekhar R, Sloan C, Mitchel E, Ndi D, Alden N, Thomas A, et al. Social determinants of influenza hospitalization in the United States. Influenza Other Respir Viruses.2017 Nov;11(6):479-88. https://doi.org/10.1111/irv.12483
https://doi.org/10.1111/irv.12483... . Groups that have worse socioeconomic indicators, such as low income and education, are 2.4 times more likely to be hospitalized or have more severe respiratory diseases¹¹11. Antunes FP, Costa MC, Paim JS, Vieira-da-Silva LM, Cruz ÁA, Natividade M, et al.Desigualdades sociais na distribuição espacial das hospitalizações por doenças respiratórias. Cad Saude Publica. 2013 Jul;29(7):1346-56. https://doi.org/10.1590/S0102-311X2013000700009
https://doi.org/10.1590/S0102-311X201300... . The supply of health services is another factor that strongly affects hospitalization rates. Castro et al.¹²12. Castro MS, Travassos C, Carvalho MS. Efeito da oferta de serviços de saúde no uso de internações hospitalares no Brasil. Rev Saude Publica. 2005 Apr;39(2):277-84. https://doi.org/10.1590/S0034-89102005000200020
https://doi.org/10.1590/S0034-8910200500... showed in 2005 that “the greater the average number of beds per population, the greater the chance of hospitalization, and the greater the number of physicians per population, the lower the chance of hospitalization” (our translation). Likewise, there are other factors associated with hospitalizations, albeit weakly, such as the percentage of illiteracy, the proportion of SUS beds, urbanization, and coverage by health insurance plans¹³13. Pazó RG, Frauches DD, Maria DC, Cade NV. Modelagem hierárquica de determinantes associados a internações por condições sensíveis à atenção primária no Espírito Santo, Brasil. Cad Saude Publica. 2014 Set;30(9):1891-902. https://doi.org/10.1590/0102-311x00099913
https://doi.org/10.1590/0102-311x0009991... .

Although hospitalization rates are influenced by several variables, in different dimensions, the main factor associated with hospital admissions is the need for health. People who are hospitalized have a higher number of chronic diseases, worse health status, worse functional status, and more restrictions on carrying out usual activities¹⁴14. Castro MS, Travassos C, Carvalho MS. Fatores associados às internações hospitalares no Brasil. Cien Saude Colet. 2002;7(4):795-811. https://doi.org/10.1590/S1413-81232002000400014
https://doi.org/10.1590/S1413-8123200200... . Likewise, the prevalence of hospitalization in people with multimorbidities is double that of people without multimorbidities¹⁵15. Souza AS, Braga JU. Trends in the use of health services and their relationship with multimorbidity in Brazil, 1998-2013. BMC Health Serv Res. 2020 Nov;20(1):1080. https://doi.org/10.1186/s12913-020-05938-4
https://doi.org/10.1186/s12913-020-05938... . These relationships between chronic diseases, comorbidities, and hospitalizations should produce important changes in hospitalizations in Brazil, since the country is undergoing a demographic and epidemiological transition with an increase in chronic diseases¹⁶16. Marinho F, Passos VM, França EB. Novo século, novos desafios: mudança no perfil da carga de doença no Brasil de 1990 a 2010. Epidemiol Serv Saude. 2016 out;25(4):713-24. https://doi.org/10.5123/S1679-49742016000400005
https://doi.org/10.5123/S1679-4974201600... .

The population’s demographic and epidemiological profile influences the demand for hospitalization. In Brazil, since 1980, fertility and birth rates have decreased while life expectancy at birth has increased¹⁷17. Mendes AC, Sá DA, Miranda GM, Lyra TM, Tavares RA. Assistência pública de saúde no contexto da transição demográfica brasileira: exigências atuais e futuras. Cad Saude Publica. 2012 May;28(5):955-64. https://doi.org/10.1590/S0102-311X2012000500014
https://doi.org/10.1590/S0102-311X201200... . With fewer births, the demand for obstetrics and pediatrics beds has been decreasing and naturally there has been a reduction in supply. However, the prevalence of chronic diseases increases with an aging population. Paradoxically, this reduction in the supply of beds and decrease in the hospitalization rate has also been observed in medical and surgical practice, although to a lesser extent¹⁷17. Mendes AC, Sá DA, Miranda GM, Lyra TM, Tavares RA. Assistência pública de saúde no contexto da transição demográfica brasileira: exigências atuais e futuras. Cad Saude Publica. 2012 May;28(5):955-64. https://doi.org/10.1590/S0102-311X2012000500014
https://doi.org/10.1590/S0102-311X201200... .

In 2013, the Brazilian Institute of Geography and Statistics (IBGE) in association with the Ministry of Health (MS) carried out the first Pesquisa Nacional de Saúde (PNS – National Health Survey). The research had many objectives, which included evaluating the national health system in terms of access to and use of its services, as well as measuring access to medical care at different levels of care¹⁸18. Szwarcwald CL, Malta DC, Pereira CA, Vieira ML, Conde WL, Souza Júnior PR, et al. Pesquisa Nacional de Saúde no Brasil: concepção e metodologia de aplicação. Cien Saude Colet.2014 Feb;19(2):333-42. https://doi.org/10.1590/1413-81232014192.14072012
https://doi.org/10.1590/1413-81232014192... . The survey revealed that 53.4% of households were registered in Family Health Units, which correspond to Primary Care in Brazil¹⁹19. Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional de Saúde 2013. Acesso e utilização dos serviços de saúde, acidentes e violências Brasil, Grandes Regiões e Unidades da Federação. Rio de Janeiro: IBGE; 2015 [cited 2020 Jul 10]. Available from: https://biblioteca.ibge.gov.br/visualizacao/livros/liv94074.pdf
https://biblioteca.ibge.gov.br/visualiza... . Later, in 2019, the second edition of the survey was carried out, with changes in some questions and an expansion of the questionnaire²⁰20. Stopa SR, Szwarcwald CL, Oliveira MM, Gouvea EC, Vieira ML, Freitas MP, et al. Pesquisa Nacional de Saúde 2019: histórico, métodos e perspectivas. Epidemiol Serv Saude.2020 Oct;29(5):e2020315. https://doi.org/10.1590/s1679-49742020000500004
https://doi.org/10.1590/s1679-4974202000... . Both editions of the PNS presented information on hospitalizations and, in this second edition, an increase in the registration of households in Primary Care was observed to 60.0%²¹21. Instituto Brasileiro de Geografia e Estatística. PNS 2019: sete em cada dez pessoas que procuram o mesmo serviço de saúde vão à rede pública. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 4 set 2020 [cited 2020 Dec 3]. Available from: https://agenciadenoticias.ibge.gov.br/agencia-sala-de-imprensa/2013-agencia-de-noticias/releases/28793-pns-2019-sete-em-cada-dez-pessoas-que-procuram-o-mesmo-servico-de-saude-vao-a-rede-publica
https://agenciadenoticias.ibge.gov.br/ag... .

According to PNS 2013 data, 6.0% of individuals had been hospitalized for 24 hours or more in the last 12 months¹⁹19. Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional de Saúde 2013. Acesso e utilização dos serviços de saúde, acidentes e violências Brasil, Grandes Regiões e Unidades da Federação. Rio de Janeiro: IBGE; 2015 [cited 2020 Jul 10]. Available from: https://biblioteca.ibge.gov.br/visualizacao/livros/liv94074.pdf
https://biblioteca.ibge.gov.br/visualiza... , a prevalence that increased to 6.6% in PNS 2019²¹21. Instituto Brasileiro de Geografia e Estatística. PNS 2019: sete em cada dez pessoas que procuram o mesmo serviço de saúde vão à rede pública. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 4 set 2020 [cited 2020 Dec 3]. Available from: https://agenciadenoticias.ibge.gov.br/agencia-sala-de-imprensa/2013-agencia-de-noticias/releases/28793-pns-2019-sete-em-cada-dez-pessoas-que-procuram-o-mesmo-servico-de-saude-vao-a-rede-publica
https://agenciadenoticias.ibge.gov.br/ag... . From 1999 to 2016, the rate of hospitalizations financed by the SUS had been gradually decreasing¹⁵15. Souza AS, Braga JU. Trends in the use of health services and their relationship with multimorbidity in Brazil, 1998-2013. BMC Health Serv Res. 2020 Nov;20(1):1080. https://doi.org/10.1186/s12913-020-05938-4
https://doi.org/10.1186/s12913-020-05938... ,¹⁷17. Mendes AC, Sá DA, Miranda GM, Lyra TM, Tavares RA. Assistência pública de saúde no contexto da transição demográfica brasileira: exigências atuais e futuras. Cad Saude Publica. 2012 May;28(5):955-64. https://doi.org/10.1590/S0102-311X2012000500014
https://doi.org/10.1590/S0102-311X201200... ,²²22. Fundação Osvaldo Cruz. Instituto de Comunicação e Informação Científica e Tecnológica em Saúde. Boletim informativo: monitoramento da assistência hospitalar no Brasil (2009-2017). Rio de Janeiro: Fundação Osvaldo Cruz; 2019 [citado 3 jun 2021]. Projeto Avaliação do Desempenho do Sistema de Saúde, Boletim Informativo, 4). Available from: https://www.proadess.icict.fiocruz.br/Boletim_4_PROADESS_Monitoramento%20da%20assistencia%20hospitalar_errata_1403.pdf
https://www.proadess.icict.fiocruz.br/Bo... . After this sustained reduction, in 2017 an increase in this rate was identified as compared to the previous year²²22. Fundação Osvaldo Cruz. Instituto de Comunicação e Informação Científica e Tecnológica em Saúde. Boletim informativo: monitoramento da assistência hospitalar no Brasil (2009-2017). Rio de Janeiro: Fundação Osvaldo Cruz; 2019 [citado 3 jun 2021]. Projeto Avaliação do Desempenho do Sistema de Saúde, Boletim Informativo, 4). Available from: https://www.proadess.icict.fiocruz.br/Boletim_4_PROADESS_Monitoramento%20da%20assistencia%20hospitalar_errata_1403.pdf
https://www.proadess.icict.fiocruz.br/Bo... , and this increase was also noticed in the last edition of the PNS, in 2019. It is still unclear whether these facts mean points outside the curve or a new growth trend of hospitalizations in Brazil. Our study intended to clarify this issue. Based on the 2013 and 2019 PNS estimates, we aimed to compare the prevalence of hospitalizations according to geographic, socioeconomic and health conditions and to identify any changes in the hospitalizations profile in Brazil.

METHODS

This work is a serial cross-sectional study that used data from the PNS 2013 and PNS 2019, which took place respectively between August 2013 and February 2014²³23. Damacena GN, Szwarcwald CL, Malta DC, Souza Júnior PR, Vieira ML, Pereira CA, et al.O processo de desenvolvimento da Pesquisa Nacional de Saúde no Brasil, 2013. Epidemiol Serv Saude. 2015 Jun;24(2):197-206. https://doi.org/10.5123/S1679-49742015000200002
https://doi.org/10.5123/S1679-4974201500... , and between August 2019 and March 2020²⁰20. Stopa SR, Szwarcwald CL, Oliveira MM, Gouvea EC, Vieira ML, Freitas MP, et al. Pesquisa Nacional de Saúde 2019: histórico, métodos e perspectivas. Epidemiol Serv Saude.2020 Oct;29(5):e2020315. https://doi.org/10.1590/s1679-49742020000500004
https://doi.org/10.1590/s1679-4974202000... , whose object of study was the Brazilian population.

The 2013 PNS interviews were conducted in 64,348 households, in approximately 1,600 municipalities throughout Brazil²⁴24. Pinto LF, Freitas MP, Figueiredo AW. Sistemas Nacionais de Informação e levantamentos populacionais: algumas contribuições do Ministério da Saúde e do IBGE para a análise das capitais brasileiras nos últimos 30 anos. Cien Saude Colet. 2018 Jun;23(6):1859-70. https://doi.org/10.1590/1413-81232018236.05072018
https://doi.org/10.1590/1413-81232018236... , with information on 205,546 residents, referring to a population projection of 199,551,444 inhabitants. The PNS 2019 interviews were conducted in 94,114 households, referring to 279,382 residents and a population projection of 209,589,607 inhabitants. The two surveys followed a similar methodology. The items were divided into 3 groups. The first provided information about the household and visits by family health teams or endemic agents. The second investigated the general characteristics of each resident, income, health insurance coverage, use of health services, and health status of individuals. Questions in the third group were outside the scope of this study and were not used. The questionnaire, the interviewer’s manual and the microdata are available from the IBGE website²⁵25. Instituto Brasileiro de Geografia e Estatística. Coordenação de Trabalho e Rendimento. Pesquisa Nacional de Saúde: 2013: manual de entrevista. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2013 [citado 8 dez 2020]. Available from: https://biblioteca.ibge.gov.br/visualizacao/instrumentos_de_coleta/doc2963.pdf
https://biblioteca.ibge.gov.br/visualiza... .

In this study, we used the following condition as the outcome: having been hospitalized for 24 hours or more in the last 12 months. As a health need is the main determinant of hospital use, we describe the participation of chronic illness and perception of poor health status in the Brazilian population and in total hospitalizations. The age group was also included in the description. From the microdata from the two PNS, we estimated the proportion of the population in different categories of age, presence or absence of chronic diseases and perception of health status, and the number of hospitalizations related to each category. With the estimated mean total numbers, and their respective confidence intervals, we calculated how much each category of these variables represented, proportionally, in the general total of hospitalizations, in eachPNS edition.

To analyze changes in the prevalence of hospitalization, we used the factors described in Andersen’s behavioral model of health service use³⁰30. Andersen RM. National health surveys and the behavioral model of health services use. Med Care. 2008 Jul;46(7):647-53. https://doi.org/10.1097/MLR.0b013e31817a835d
https://doi.org/10.1097/MLR.0b013e31817a... . We used the following variables: as predisposing factors – sex, age group, skin color, and level of education; as facilitating factors – urban or rural housing, usual place of care, presence or absence of medical health insurance, household income per capita, and region of the federation; and as health need – perception of health and presence or absence of chronic disease. In the age group variable, ages were grouped into 6 categories: 0–5 years, 6–17 years, 18–29 years, 30–39 years, 40–59 years, and 60 years of age or older. All variables are categorical. The prevalence of hospitalization and respective 95% confidence intervals were calculated according to the mentioned factors. In the prevalence calculations, hospitalizations for childbirth were excluded.

The estimates found in the two editions of the PNS were compared using Student’s t-test for independent samples. We considered significant differences when p was less than 0.01³¹31. Li C, Balluz LS, Ford ES, Okoro CA, Zhao G, Pierannunzi C. A comparison of prevalence estimates for selected health indicators and chronic diseases or conditions from the Behavioral Risk Factor Surveillance System, the National Health Interview Survey, and the National Health and Nutrition Examination Survey, 2007-2008. Prev Med. 2012 Jun;54(6):381-7. https://doi.org/10.1016/j.ypmed.2012.04.003
https://doi.org/10.1016/j.ypmed.2012.04.... .

The consistency of survey data was assessed by comparing administrative data with estimates from the two editions. Hospitalizations estimated by the PNS were broken down into hospitalizations covered by the SUS, hospitalizations covered by private health insurance, and obstetric hospitalizations. Information on births in hospitals in 2013 and 2019 was obtained from the website IBGE – Civil Registry Statistics, through the IBGE Automatic Recovery System (SIDRA)³²32. Instituto Brasileiro de Geografia e Estatística. Sistema IBGE de Recuperação Automática – SIDRA. Tabela 2680: nascidos vivos, ocorridos no ano, por mês do nascimento, sexo, local de nascimento, número de nascidos por parto, idade da mãe na ocasião do parto e lugar do registro; Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2020 [cited 2020 Jul 22]. Available from: https://sidra.ibge.gov.br/tabela/2680
https://sidra.ibge.gov.br/tabela/2680... . Hospitalizations through the SUS in 2013 and 2019 were obtained from the SUS Hospital Information System (SIH/SUS)³³33. Ministério da Saúde (BR). Sistema de Informações Hospitalares do SUS. Informações de Saúde. Brasília, DF: Ministério da Saúde [cited 2020 Jun 5]. Available from: http://tabnet.datasus.gov.br/cgi/deftohtm.exe?sih/cnv/qiuf.def
http://tabnet.datasus.gov.br/cgi/deftoht... . Information on hospitalizations covered partially or fully by private health insurance was obtained from the website of the Agência Nacional de Saúde Suplementar (National Agency of Supplementary Health – ANS)³⁴34. Ministério da Saúde (BR). Agência Nacional de Saúde Suplementar. Dados e Indicadores do Setor. Brasília, DF: Ministério da Saúde; 2020 [cited 2020 Jun 22]. Available from: http://www.ans.gov.br/perfil-do-setor/dados-e-indicadores-do-setor
http://www.ans.gov.br/perfil-do-setor/da... , and referred to the years 2014 and 2019. Data referring to the year 2013 were not found.

As this is a study that used complex sampling, we used the Stata® statistical software, version SE 15.1 for data analysis, which, through its survey module, takes into account the effect of the sampling plan. In the adjustment, strata, primary sampling units and sample weights were incorporated.

PNS 2013 was approved by opinion no. 328.159 of the National Research Ethics Committee. PNS 2019 was approved by opinion no. 3,529,376 of the National Research Ethics Commission. All participants signed, in both editions, an informed consent form, assuring them anonymity and the possibility of withdrawing from the study at any time.

RESULTS

According to the two surveys’ estimates, there were changes in the demographic and epidemiological profile of the Brazilian population between 2013 and 2019. A significant change was observed among people aged 6 to 29 years and among people aged 40 years or older. The percentage of people with chronic diseases increased significantly between the two PNS editions, and the perception of health status changed in 4 of the 5 categories, with an increase in the very good, fair, and very poor categories. This information is detailed in Table 1.

Changes in the hospitalization profile were also observed. While hospitalizations by people without chronic diseases predominated in the 2013 PNS, patients with chronic diseases predominated in the 2019 PNS, as shown in Table 2.

Excluding hospitalizations for childbirth, the prevalence of hospitalization increased significantly from 5.32% in the 2013 PNS to 5.81% in the 2019 PNS. There was a statistically significant increase in prevalence between 2013 and 2019 in the Southeast Region in both sexes, in people with white skin color, in people with no schooling, in people with complete high school education, in people living in urban areas, in people with health insurance, in people who always go the same place when they need health care, and for people with a household income per capita greater than 3 and up to 5 minimum wages. On the other hand, significant reductions in prevalence were observed in people with chronic disease and in people without chronic disease, as can be seen in Table 3.

When analyzing admissions for childbirth, estimated in each survey, with administrative data, referring to live births in equivalents years, we noticed different values. Similarly, we noticed a difference in non-obstetric hospitalizations, both in those financed by the SUS and in those financed by health insurance. The values are shown in Table 4.

DISCUSSION

Our study shows an increase in the estimated prevalence of hospitalizations between the 2013 and 2019 PNS editions. This growth was accompanied by important changes in the population’s epidemiological and demographic profile, especially in the doubling of the percentage of people with chronic diseases, along with a decrease in the percentage of those aged 6 to 29 and an increase among those aged 40 and over.

Similarly, other studies, with data from the 2013 and 2019 PNS, show that the prevalence of chronic diseases has increased in Brazil³⁵35. Simões TC, Meira KC, Santos JD, Câmara DC. Prevalências de doenças crônicas e acesso aos serviços de saúde no Brasil: evidências de três inquéritos domiciliares. Cien Saude Colet.2021 Set;26(9):3991-4006. https://doi.org/10.1590/1413-81232021269.02982021
https://doi.org/10.1590/1413-81232021269... ,³⁶36. Szwarcwald CL, Stopa SR, Damacena GN, Almeida WD, Souza Júnior PR, Vieira ML, et al. Mudanças no padrão de utilização de serviços de saúde no Brasil entre 2013 e 2019. Cien Saude Colet. 2021 Jun;26 suppl 1:2515-28. https://doi.org/10.1590/1413-81232021266.1.43482020
https://doi.org/10.1590/1413-81232021266... . The concomitant increase observed in the proportion of people with chronic diseases and people in middle age explains an apparently paradoxical finding. As can be seen in our results, there was a reduction in the prevalence of hospitalizations in people with chronic diseases. With a greater number of people in the middle age group, there are more individuals with chronic diseases, but who have not yet had complications. Consequently, there has been an increase in the proportion of people with chronic diseases who do not require hospitalization.

Although the prevalence of hospitalization has decreased among people with chronic diseases, as they have proportionally increased, and hospitalization among them remains very high, there has been an increase in the general prevalence of the population. The participation of patients with chronic diseases in total hospitalizations increased from 36.76% to 57.61% between the two PNS editions. These findings demonstrate the influence of the epidemiological profile and demography on hospitalization rates in Brazil.

In Brazil, the epidemiological transition is characterized by a triple burden of disease. Non-transmissible chronic diseases (NTCDs) coexist with a high incidence and prevalence of infectious and parasitic diseases (IPD) and external causes³⁷37. Martins TC F. Silva JHCM, Máximo GC, Guimarães RM. Transição da morbimortalidade no Brasil: um desafio aos 30 anos de SUS. Cienc Saúde Coletiva. 2021 out;26(10):4483-96. https://doi.org/10.1590/1413-812320212610.10852021
https://doi.org/10.1590/1413-81232021261... . In 2019, proportionally, the burden of IPD was higher in early childhood, the burden of external causes was higher in young male adults, and the burden of NTCDs increased with age³⁷37. Martins TC F. Silva JHCM, Máximo GC, Guimarães RM. Transição da morbimortalidade no Brasil: um desafio aos 30 anos de SUS. Cienc Saúde Coletiva. 2021 out;26(10):4483-96. https://doi.org/10.1590/1413-812320212610.10852021
https://doi.org/10.1590/1413-81232021261... . According to PNS 2019 estimates, in more than 40% of hospitalizations, there was no diagnosis of chronic disease. If hospitalizations reflect the burden of disease, we will have a predominance of hospitalizations due to IPD, in early childhood, due to external causes, in young males, and due to obstetric causes in young females.

Emerging situations also influence hospitalizations. In 2020, due to the covid-19 pandemic, there was a 15% decrease in the hospital admission rate, accompanied by a 9% increase in in-hospital lethality of patients with cardiovascular diseases admitted through the SUS³⁸38. Normando PG, Araujo-Filho JA, Fonseca GA, Rodrigues RE, Oliveira VA, Hajjar LA, et al. Redução na hospitalização e aumento na mortalidade por doenças cardiovasculares durante a pandemia da COVID-19 no Brasil. Arq Bras Cardiol. 2021 Mar;116(3):371-80. https://doi.org/10.36660/abc.20200821
https://doi.org/10.36660/abc.20200821... .

A study on the profile of hospital admissions by the SUS between 2013 and 2017 showed that in 24.4% of hospitalizations, the patient was 60 years old or over³⁹39. Gomes HG, Dias SM, Gomes MS, Medeiros JSM, Ferraz LP, Pontes FL, et al.Perfil das internações hospitalares no Brasil no período de 2013 a 2017. Rev Interd.2017 dez;10(4):96-104.. We estimate that people aged 60 or over accounted for 25.56% of all hospitalizations in the 2013 PNS and 28.35% in the 2019 PNS, but without a statistically significant difference between the two surveys. Brazil has been increasing life expectancy due to reductions in child mortality and cardiovascular disease mortality, despite the negative influence of mortality from external causes, which occurs mainly in young male adults⁴⁰40. BORGES. G. M. A transição da saúde no Brasil: variações regionais e divergência/convergência na mortalidade. Cad Saude Publica. 2017;33(8):e00080316. https://doi.org/10.1590/0102-311x00080316
https://doi.org/10.1590/0102-311x0008031... . The increase in life expectancy is due, in part, to the improvement in health care. The challenge is not only to increase life expectancy, but mainly years with quality of life⁴¹41. World Health Organization. Global action plan for the prevention and control of noncommunicable disease 2013-2020. Geneva: World Health Organization; 2013..

Similarly, we found that people with poor or very poor health status correspond to less than 5% of the population, but account for almost 20% of all hospitalizations. These people probably have worse health, and we already know that people in this situation are the ones who need hospitals the most¹⁴14. Castro MS, Travassos C, Carvalho MS. Fatores associados às internações hospitalares no Brasil. Cien Saude Colet. 2002;7(4):795-811. https://doi.org/10.1590/S1413-81232002000400014
https://doi.org/10.1590/S1413-8123200200... . Adequate primary care and an integrated health care network can reduce hospitalizations, reduce these people’s hospital stay, and also reduce readmissions⁴²42. Buttigieg SC, Abela L, Pace A. Variables affecting hospital length of stay: a scoping review.J Health Organ Manag. 2018 May;32(3):463-93. https://doi.org/10.1108/JHOM-10-2017-0275
https://doi.org/10.1108/JHOM-10-2017-027... .

In Brazil, between 2013 and 2019, the proportion of individuals who consulted a doctor in the last year increased. This increase was greater among SUS users than among health insurance holders³⁶36. Szwarcwald CL, Stopa SR, Damacena GN, Almeida WD, Souza Júnior PR, Vieira ML, et al. Mudanças no padrão de utilização de serviços de saúde no Brasil entre 2013 e 2019. Cien Saude Colet. 2021 Jun;26 suppl 1:2515-28. https://doi.org/10.1590/1413-81232021266.1.43482020
https://doi.org/10.1590/1413-81232021266... . This is probably a reflection of the primary care expansion policy, with the Family Health Strategy.

An adequate supply of physicians, associated with long-term relationships between physicians and their patients, can reduce hospitalizations due to chronic diseases⁴³43. Loenen T, Berg MJ, Westert GP, Faber MJ. Organizational aspects of primary care related to avoidable hospitalization: a systematic review. Fam Pract. 2014 Oct;31(5):502-16. https://doi.org/10.1093/fampra/cmu053
https://doi.org/10.1093/fampra/cmu053... . A study carried out in 2018 demonstrated a reduction in hospitalizations for conditions sensitive to primary care⁴⁴44. Pinto LF, Giovanella L. Do programa à Estratégia Saúde da Família: expansão do acesso e redução das internações por condições sensíveis à atenção básica (ICSAB). Cien Saude Colet. 2018 Jun;23(6):1903-14. https://doi.org/10.1590/1413-81232018236.05592018
https://doi.org/10.1590/1413-81232018236... , associated with the advance of the Family Health Strategy’s coverage in Brazil. On the other hand, we found a significant increase in the prevalence of hospitalizations among people with private health insurance and among those who always seek the same place when they need care, which probably represent the individuals with better access to health services⁴⁵45. Stopa SR, Malta DC, Monteiro CN, Szwarcwald CL, Goldbaum M, Cesar CL. Acesso e uso de serviços de saúde pela população brasileira, Pesquisa Nacional de Saúde 2013. Rev Saude Publica. 2017;51(suppl 1):3s.. Unlike the progress observed in primary care offered by the SUS, initiatives to invest in this level of care by the private sector, especially for people with chronic diseases, are still incipient in the country, which may have influenced this increase in hospitalizations. Complementarily, it is expected that there has been an increase in the diagnosis of conditions that occasionally require hospitalization, especially in people with poor health¹²12. Castro MS, Travassos C, Carvalho MS. Efeito da oferta de serviços de saúde no uso de internações hospitalares no Brasil. Rev Saude Publica. 2005 Apr;39(2):277-84. https://doi.org/10.1590/S0034-89102005000200020
https://doi.org/10.1590/S0034-8910200500... ,²²22. Fundação Osvaldo Cruz. Instituto de Comunicação e Informação Científica e Tecnológica em Saúde. Boletim informativo: monitoramento da assistência hospitalar no Brasil (2009-2017). Rio de Janeiro: Fundação Osvaldo Cruz; 2019 [citado 3 jun 2021]. Projeto Avaliação do Desempenho do Sistema de Saúde, Boletim Informativo, 4). Available from: https://www.proadess.icict.fiocruz.br/Boletim_4_PROADESS_Monitoramento%20da%20assistencia%20hospitalar_errata_1403.pdf
https://www.proadess.icict.fiocruz.br/Bo... ,⁴⁶46. Macinko J, Dourado I, Aquino R, Bonolo PF, Lima-Costa MF, Medina MG, et al. Major expansion of primary care in Brazil linked to decline in unnecessary hospitalization. Health Aff (Millwood). 2010 Dec;29(12):2149-60. https://doi.org/10.1377/hlthaff.2010.0251
https://doi.org/10.1377/hlthaff.2010.025... ,⁴⁷47. Cavalcante DF, Brizon VS, Probst LF, Meneghim MC, Pereira AC, Ambrosano GM. Did the Family Health Strategy have an impact on indicators of hospitalizations for stroke and heart failure? Longitudinal study in Brazil: 1998-2013. PLoS One. 2018 Jun;13(6):e0198428. https://doi.org/10.1371/journal.pone.0198428
https://doi.org/10.1371/journal.pone.019... .

Our study found variations in the prevalence of hospitalizations across Brazilian regions. In both PNS editions, the South and Midwest regions have the highest prevalence. The Southeast region, on the other hand, had the highest prevalence increase across all regions, while the North and Northeast regions remained below the national average. These findings reflect the large historical regional differences observed in Brazil, in relation to socioeconomic characteristics and social and health investments. The 2018 IBGE Household Budget Survey showed, for example, a huge regional discrepancy in health expenditures per capita, ranging from R$4.70 in the North region to R$70.04 in the Southeast region⁴⁸48. Instituto Brasileiro de Geografia e Estatística. Coordenação de Trabalho e Rendimento. Pesquisa de orçamentos familiares: 2017-2018: perfil das despesas no Brasil: indicadores selecionados. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2020 [cited 2020 Dec 11]. Available from: https://biblioteca.ibge.gov.br/visualizacao/livros/liv101761.pdf
https://biblioteca.ibge.gov.br/visualiza... . In addition, there are also differences in relation to age structures, HDI, provision of health services, and epidemiological profiles²²22. Fundação Osvaldo Cruz. Instituto de Comunicação e Informação Científica e Tecnológica em Saúde. Boletim informativo: monitoramento da assistência hospitalar no Brasil (2009-2017). Rio de Janeiro: Fundação Osvaldo Cruz; 2019 [citado 3 jun 2021]. Projeto Avaliação do Desempenho do Sistema de Saúde, Boletim Informativo, 4). Available from: https://www.proadess.icict.fiocruz.br/Boletim_4_PROADESS_Monitoramento%20da%20assistencia%20hospitalar_errata_1403.pdf
https://www.proadess.icict.fiocruz.br/Bo... ,⁴⁹49. Monteiro Neto A, Castro CN, Brandão CA. Desenvolvimento regional no Brasil: políticas, estratégias e perspectivas. Rio de Janeiro: IPEA; 2017..

In evaluating the estimates of hospitalizations from the two surveys, with the official number of hospitalizations, we found that the number of obstetric hospitalizations, in the public and supplementary health systems, was underestimated, as well as the number of non-obstetric hospitalizations covered by health insurance. Non-obstetric hospitalizations financed by the SUS were overestimated as well. As the information on hospital admissions referred to the last 12 months, there may have been a memory bias among the participants. In addition, the PNS items analyzed in this study refer to all of the household residents, but were answered by only one resident, which can lead to greater inaccuracy in the information. One of the limitations of household surveys is the bias associated with the use of secondary informants⁵⁰50. Viacava F, Dachs N, Travassos C. Os inquéritos domiciliares e o Sistema Nacional de Informações em Saúde. Cien Saude Colet. Dez 2006;11(4):863-9. https://doi.org/10.1590/s1413-81232006000400002
https://doi.org/10.1590/s1413-8123200600... . However, the discrepancy between the actual number of live births and the estimates of births in the two surveys may indicate the need for some methodological adjustment.

With the aim of ensuring greater homogeneity in gender distribution in the different age groups, in the calculations of the prevalence of hospitalizations, we chose not to include obstetric hospitalizations. Thus, the prevalence figures we found are lower than those previously published. This methodological decision has been observed in previous studies¹³13. Pazó RG, Frauches DD, Maria DC, Cade NV. Modelagem hierárquica de determinantes associados a internações por condições sensíveis à atenção primária no Espírito Santo, Brasil. Cad Saude Publica. 2014 Set;30(9):1891-902. https://doi.org/10.1590/0102-311x00099913
https://doi.org/10.1590/0102-311x0009991... ,¹⁴14. Castro MS, Travassos C, Carvalho MS. Fatores associados às internações hospitalares no Brasil. Cien Saude Colet. 2002;7(4):795-811. https://doi.org/10.1590/S1413-81232002000400014
https://doi.org/10.1590/S1413-8123200200... ,⁵¹51. Mullachery P, Silver D, Macinko J. Changes in health care inequity in Brazil between 2008 and 2013. Int J Equity Health. 2016 Nov;15(1):140. https://doi.org/10.1186/s12939-016-0431-8
https://doi.org/10.1186/s12939-016-0431-... . Nevertheless, it should be noted that this methodological decision may lead to losses in mean estimates and consequent reduction in statistical power, making it difficult to identify differences between editions. Our focus, however, is on information in which significant differences were observed.

Our study had as a limitation the absence of administrative data regarding hospitalizations by health insurance in 2013 and we used, as an alternative, data from 2014 in the comparison with the 2013 PNS, which may generate bias in the comparison. Likewise, the divergences between the administrative data and the estimates of the two PNS may represent a greater amplitude in the degree of uncertainty of our findings, resulting from under or overestimates observed in the two PNS editions. Another issue is that, due to the large sample size in both PNS editions, minimal differences can be considered statistically significant and the differences found in this work must be evaluated not only in statistical terms, but in their epidemiological significance.

The 2019 PNS identified a predominance of hospitalizations by people with chronic diseases. This population subgroup should be better studied, identifying risk factors for hospitalization. Special attention should be given to heavy users, especially in the determinants of re-hospitalization. It is also necessary to evaluate the huge number of small hospitals that exist in Brazil. The effectiveness of the hospital care network should be investigated, bringing information to support new public policies, for a more rational use of financial resources.

CONCLUSIONS

There was an increase in overall hospitalization rates in the period between the 2013 and 2019 PNS, especially among people with better access to health services. In addition, the hospitalization profile itself has changed. While in the 2013 edition, hospitalizations of people without chronic diseases predominated in the total number of hospitalizations, in the PNS 2019 more than half of the total number of hospitalizations were of people with chronic diseases.

It was evident that changes in the demographic and epidemiological profile of the Brazilian population are already impacting hospitalization rates. The number of chronically ill people in the Brazilian population has doubled, but the prevalence of hospitalization among them has decreased, probably because they have not yet developed complications. In this context, accessible and effective Primary Health Care becomes even more relevant, otherwise more hospitalizations and higher costs can be expected in the medium term. In addition, the hospital care policy must be continuously improved, seeking efficiency and cost reduction. Without these measures, there is a serious risk of collapse of Brazil’s health systems.

REFERENCES

Publication Dates

History