Artigos

Conteúdo (article):

Trends in inequalities and inequities in the utilization of health services by older adults in Rio Grande do Sul, Brazil: Evidence for the period 1998-2019

ABSTRACT
Objective: To assess inequalities and inequities in the utilization of health services by older people (50 years and over) in Rio Grande do Sul during the period 1998-2019. Methods: Using data from the National Household Sample Survey and the National Health Survey, four outcomes (medical consultations, dental consultations, hospital admissions and mammograms) were analyzed using concentration and horizontal indices. Results: The results show a decrease in inequities for medical consultations and mammograms and an increase for dental consultations, showing that the benefits were greater for the richest segment of the population. Conclusion: Public policies should focus on increasing access and reducing disparities, especially in access to mammograms and dental consultations, which exhibit significant inequities.
Keywords: Health services; measuring health inequalities; older people.

INTRODUCTION
Access to health services is a fundamental aspect of health and well-being. However, despite its importance, evidence in the literature shows that access is generally uneven and that certain specific population groups end up benefiting more than others1–4. The assessment of access to healthcare in different population groups consists of carrying out studies of health care inequity and inequality. These studies have the potential to provide valuable information that can inform public health policies and are particularly important in countries with marked social and regional inequalities, such as Brazil5.
Brazil’s public health system, o Sistema Único de Saúde (SUS) or Unified Health System, aims to ensure the provision of universal free public health services. However, evidence in the literature suggests that access to services is greater for more developed regions and among groups with higher socioeconomic status3,6–8.
A group that warrants special attention, and has been little explored in the literature, is the older population9. This group has gained increasing relevance due to the challenges epidemiological and demographic changes in the country pose to public administration, placing pressure on the social security system and increasing demand for health services10,11. Population aging is likely to further exacerbate inequalities, as the elderly population is particularly vulnerable due to higher incidence of chronic diseases and the need for a wider range of health services in this group11–13.
It is estimated that around 19% of Brazil’s population will be older people by 2030, reaching around one-third of the population by 2060 (32.2%). In the state of Rio Grande do Sul, the pace of population aging is even faster, with the proportion of older people in the population set to reach 24% in 2030 and around 36% in 206014. According to data from the Brazilian Institute of Geography and Statistics (IBGE), Rio Grande do Sul is the state with the highest proportion of older people. It is estimated that around 18% of the state population is aged over 60, in comparison to the national rate of 14%. Although there are some important contributions in the literature, little attention has been given to measuring inequities in access to and utilization of health services among this population group. We found only three pieces of evidence in the literature regarding this age group in Brazil4,13,15.
The aim of this study was to analyze inequalities and inequities in the utilization of health services by older adults (aged 50 years and over) in Rio Grande do Sul during the period 1998-2019. In addition to being at a more advanced stage of the demographic transition than other states in Brazil, Rio Grande do Sul is also experiencing financial difficulties, which makes it an interesting case study, as population aging tends to put pressure on the health system.
METHODS

Data

We conducted an exploratory quantitative empirical study using indicators and statistical/econometric analysis, making comparisons and discussing existing literature on the subject.
The data source was the National Household Sample Survey (PNAD) and National Health Survey (PNS), both conducted by the IBGE. The PNAD used the following complex three-stage probability sampling design: primary units (municipalities); secondary units (census tracts); and tertiary units (households). The PNS also used a three-stage sampling design: primary units (census tracts); secondary units (households); and tertiary units (randomly selected adults from each household, including individuals aged 18 years and over in 2013 and 15 years and over in 2019). Despite sample design differences, the PNS was designed to enable comparisons with the PNAD health component, ensuring continuity in the monitoring of indicators of access to and utilization of health services16. The indicators of the selected adults were not used to ensure greater data compatibility. All of the data are secondary; however, the surveys were approved by the National Research Ethics Committee.
The study therefore focused on the five years where data are available: PNAD - 1998, 2003, 2008; and PNS - 2013 and 2019, except for the variable mammogram, which was only included from the 2003 PNAD onwards and in the 2019 PNS. This indicator was only included in the questionnaire for the third stage of the sample (the selected adult), meaning it was not possible to make comparisons with the other three years. Thus, for this variable we analyzed the period 2003-2013.
The sample consisted of people aged 50 years and over, except for the mammogram variable, which is limited to women aged between 50 and 69 years, following the recommendation of the Ministry of Health on breast cancer screening17.
Four types of indicators were used: outcome variables (health services - medical consultations, dental consultations, hospital admissions and mammograms); socioeconomic status variable (per capita household income); health need variables (age, gender, activity restriction and self-reported health); and other socioeconomic and demographic variables (race, education, health insurance and place of residence). The variables are described in Box 1. Income was chosen as a socioeconomic status indicator for two main reasons: first, because most of the literature providing evidence in Brazil uses this indicator5,7,8,15,18–20; and second, because the indicator is more compatible with the overall study period and respective databases.

Statistical analysis

The concentration index (CI) was used to measure inequality, and the horizontal inequity index (HI) was adopted to measure inequity. While the terms inequality and inequity are often used interchangeably, they have distinct meanings. Inequality in healthcare is when individuals with different levels of socioeconomic status (SES) have access to different amounts of healthcare (i.e., inequality related to SES), while with inequity (horizontal equity) it is necessary to distinguish between factors that should affect access (need variables) and those that should not (non-need variables)21. From this perspective, there is inequity when the utilization of health services is affected by non-need factors, meaning that the amount of health care accessed by individuals with the same health service needs varies with SES22.
The assessment of inequality and inequity is based on the distribution of an outcome in relation to a measure of SES, such as income, education or wealth index. For example, outcome variable hi, utilization of health services, and socioeconomic status variable xi, income. The concentration curve shows the cumulative proportion of individuals ranked in ascending order of income (x-axis) according to the cumulative proportion of utilization of health services (y-axis). When the concentration curve coincides with the line of equality (a 45-degree line), there is perfect equality; however, when the former lies below (above) the line of equality, then there is inequality in favor of the richest (poorest).
The CI is twice the area between the concentration curve and the line of equality, measuring in this study the degree of income-related inequality in the utilization of health services by older people. Simply put, this index helps understand whether the utilization of health services is concentrated among the richest or the poorest. When the index is positive, there is inequality favoring the richest (pro-rich), meaning utilization of health services is concentrated among individuals with higher levels of income. The opposite is the case for a negative index—the utilization of health services is concentrated among the poorest (pro-poor). The index is bounded between –1 and 1, where the closer to one, the greater the magnitude of inequality, with zero indicating no inequality21.
A cross-section concentration index (CIt) can be calculated by the covariance method (cov), using the following equation:

〖CI〗^t=2/h ̅_t cov(h_it,R_i^t) (1)
Where:
h ̅_t is the mean of the variable utilization of health services during a given period (t).
R_i^t is the order (rank) of the individual i (i=1,2, ..., N) in the distribution of income.

However, when the variable of interest is dichotomous, Wagstaff23 has demonstrated that these bounds are h ̅-1 and 1- h ̅ (where h ̅ is the mean of the outcome). The author therefore proposes correction by dividing the CI by 1-h ̅. Another approach is correction based on Erreygers24, where the CI is multiplied by 4h ̅. The second approach was used in the present study because it is the most widely used method in the literature.
However, simply measuring concentration is not enough. It is necessary to ensure that we are comparing people with the same health needs. To this end, this study used the horizontal inequity index (HI) as an indicator of inequity measured for each PNAD and PNS. The HI can be calculated using the concentration index of the estimated outcome variable standardized by needs (h ̂_it^IS). Considering the indirect standardization approach, h ̂_it^IS is calculated using equation 2. When the HI is positive (negative), there is horizontal inequality favoring the richest (poorest).

h ̂_it^IS=h_it-h ̂_it^N+h ̅_t (2)

Where i represents the individual, t is each PNAD or PNS, h ̂_it^IS is (for example) the variable utilization of medical consultations standardized by needs, h_it is the observed sample value of this variable, h ̂_it^N is expected predicted utilization and h ̅_t is the measure of utilization of medical consultations for PNAD t (1998, 2003 and 2008) or the PNS (2013 and 2019). These predictions (h ̂_it^N) can be obtained using the regression models represented by equation 3.

h ̂_it^N=G(∑_N▒〖β ̂_N x_it^N+〗 ∑_NN▒〖β ̂_NN x_it^NN 〗)+e_it (3)

Where “G” is a non-linear functional form. The variables included in these models are health need (x_it^N) and “non-need” (x_it^NN, hereafter called socioeconomic variables). The coefficients β ̂_N and β ̂_NN are the model estimators.
The decomposition of the CI is then measured for each year t using equation 4, adopting the approach proposed by Wagstaff et al.25, but incorporating Erreygers correction24.

〖CI〗_t=4*[∑_k▒(β_kt 〖 x ̅〗_kt ) 〖CI〗_(x_kt )+GCI(e_kt )] (4)

As can be seen, a linear model is applied, where the index is written as weighted CIs (by the marginal effect and mean of the associated factor) of the explanatory variables (∑_k▒(β_kt 〖 x ̅〗_kt ) 〖IC〗_(x_kt )), plus the contribution of an error term (GCI(e_kt )).

Elasticity is calculated based on the literature, using the linear approximation for nonlinear models. Thus β_kt is estimated using the mean marginal effect of a probit econometric model for dichotomous variables. The decomposition analysis shows the main drivers of inequality in the utilization of health services during each period analyzed. All analyses were performed using Stata 15 and accounted for the sample design of the databases using the svy syntax command, which considers weights, strata and primary sampling units.

RESULTS
Table 1 shows the proportions (percentage) for the categorical variables and means of the continuous variables used for the five periods of analysis. The findings reveal improvements in access to medical consultations and dental consultations over the period, reaching around 85% and 46%, respectively, in 2019. Hospital admissions dropped over the period, from 13% in 1998 to 9% in 2019. Trends in mammogram were unstable throughout the period, showing an increase between 2003 and 2013.
Table 2 presents the results for CI and HI. The findings of the inequality analyses show there is pro-rich inequality in access to medical consultations, dental consultations and mammograms. For hospital admissions, statistically significant rates were only observed for the years 1998 and 2008, revealing the opposite - a higher concentration in the poorest group. This greater degree of access among the most vulnerable group may be due to differences in needs, with the HI showing inequity. The results show that the HI for hospital admission was statistically significant only in 2003, with a positive index, demonstrating pro-rich inequities. The same pattern can be observed for the other outcomes, showing that the utilization of health services is in favor of the richest in the state of Rio Grande do Sul.
The outcomes with the highest levels of inequity were mammograms and dental consultations. The magnitude of statistically significant inequity in hospital admissions was small. While there was an overall reduction in inequity in medical consultations and mammograms over the study period, there was an increase in equity in these outcomes between 2008 and 2013. In contrast, although there was an overall pro-rich increase in inequity in dental consultations over the study period, reductions were observed between 2013 and 2019.
Table 3 shows the contributions of the health need and non-need variables in the first and last years of the analysis.
The findings reveal that income and health insurance make an important contribution to inequalities for all three outcomes. The results of the decomposition analysis show that, in both years, health insurance, income and self-rated health were the variables that contributed most to inequalities in medical consultations, with the first two variables contributing to a pro-rich pattern and the latter contributing to a pro-poor pattern.
With regard to dental consultations, besides health insurance and income, the variable education also contributed to a pro-rich pattern. The pattern was stable in both periods, and the variable that contributed most was income, followed by education and health insurance. The outcome mammogram showed the same pattern as dental consultations, with income contributing the most, followed by health insurance and education (Table 3).

DISCUSSION
The objective of this study was to assess income-related inequalities and inequities in the utilization of health services by older adults in Rio Grande do Sul. The study used data from five points in time (1998, 2003, 2008, 2013 and 2019), allowing for an analysis of temporal trends in the indices. Finally, we performed a decomposition analysis to identify the factors that contributed to the observed inequalities.
The results demonstrated that income-related inequality and inequity in the utilization of health services in the state of Rio Grande do Sul has a pro-rich pattern, with access to medical consultations, dental consultations and mammograms being concentrated among the richest segment of the population. The magnitude of the indices found in Rio Grande do Sul in the outcomes medical consultations and mammograms are lower than those found in a national study15. In contrast, for the dental consultations outcome, the indices in Rio Grande do Sul are higher than those found at national level. Our findings did not show inequality and inequity in hospital admissions in Rio Grande do Sul, which contrasts to the results found by Santos et al.15, who reported a pro-rich pattern throughout the period. Despite these minor differences, Rio Grande do Sul showed the same overall pattern as Brazil over time, with a reduction in inequity for medical consultations and mammograms and worsening of inequity for dental consultations.
The results for medical consultations are also similar to the evidence found by previous studies with the Brazilian population3,6,8,9,20. Almeida et al.6 and Macinko and Lima-Costa8 found a reduction in inequality in access to medical consultations between 1998 and 2008, while Mullachery et al.3 reported a pro-rich increase in inequity, a pattern similar to that observed for older adults in the state of Rio Grande do Sul.
The findings for dental consultations in Rio Grande do Sul differ slightly to those found at national level. The literature points to a reduction in inequalities in access to dental care between 1998 and 2008, followed by an increase between 2008 and 2013. In contrast, we found a continuous increase between 1998 and 2013, with a drop in the final year of the analysis. Previous research shows inequalities and inequities in the utilization of dental services, with studies at national level and in the municipalities of Pelotas and Porto Alegre showing a strong association between SES and pattern of utilization of dental services9,26–29.
Our results for mammograms are similar to those in the national population4,20,30 and to those reported by studies in municipalities in Rio Grande do Sul31–33. At national level, Coube et al.20 found that inequality dropped over the period 2003-2019, with a reduction in differences in utilization between rich and poor. In a study in São Leopoldo, Dias-da-Costa et al.31 found that SES was a determining factor for having mammograms, showing that women with lower SES were less likely to be examined. This continued to be the case when the analysis was limited to women who sought health services. The same result was found in the municipality of Pelotas32,33.
Despite the progress in reducing inequity in Rio Grande do Sul during this period, actions to promote equity need to be promoted, focusing efforts on disseminating information about the importance the mammogram as a preventive exam for the early detection of breast neoplasms34.
The results of the decomposition analysis of factors associated with inequality showed that income and health insurance were the main factors contributing to all outcomes where we found inequality. The variable good self-reported health also stands out for access to medical consultations, while education is an important factor for explaining inequality in dental consultations and mammograms.
These results are consistent with the literature, which highlights that health insurance plays a fundamental role in favoring richest individuals in the country5,6,8,15,19. Although Brazil has a public universal healthcare system, the private sector accounts for the largest share of health expenditure in the country, with the public sector being responsible for less than half of total healthcare spending 15. This is confirmed by the evidence in the literature, with individuals with higher SES (assessed by income) finding it easier to access health services1,35. Finally, it is worth highlighting the understanding of the different type of contributions the socioeconomic variables income and education make to access to healthcare in the literature. Higher income means that individuals can afford services, thus acting as an enabling factor for access to healthcare, allowing people to take out health insurance and pay for private consultations. Education on the other hand promotes greater understanding of information, both regarding forms of access and the importance of health services for illness prevention27. Thus, actions aimed at reducing the disparities found by this study should be integrated in order to increase the provision of public services and raise awareness among the population by disseminating information highlighting the importance of utilizing health services such as dental consultations and mammograms for disease prevention.
This study has some limitations. First, we used cross-sectional data, meaning that it was not possible to control for unobserved individual heterogeneity, potentially leading to the endogeneity problem. Second, it was not possible to include pre-existing health conditions that may lead individuals to seek the health services in the models. Another limitation related to database restrictions is the fact that aspects demonstrating difficulty or quality of access to health services were not considered, nor was it possible to distinguish between specialized medical consultations.
Despite these limitations, the study has a number of strengths, including the fact that it used two national household surveys with representative probability samples and covered a long time horizon (1998–2019), allowing for the analysis of trends. Furthermore, the study assessed four health service utilization outcomes, including indicators that are widely used in the literature, thus providing a comprehensive insight into health service utilization and enabling comparison with results of international studies.
In summary, inequities in the utilization of health services in Rio Grande do Sul by individuals aged 50 years and over showed different trends, decreasing for medical consultations and mammograms, but increasing for dental consultations, favoring the richest. These results show that advances in coverage alone do not guarantee greater equity. Public policies therefore need to go beyond the expansion of general access, adopting explicitly redistributive actions to ensure that the benefits reach the most economically vulnerable groups and reduce persistent disparities.
AUTHORS\' CONTRIBUTIONS
Triaca LM contributed to study conception and design, data analysis and interpretation of the results, and to the writing and critical revision of the manuscript. Santos AMA contributed to data analysis and interpretation and to the writing and critical revision of the manuscript. Tejada CAO contributed to study design and the writing and the critical revision of the manuscript. All authors approved the final version of the manuscript to be published and are responsible for all aspects of the work, including accuracy and integrity.

REFERENCES

1. Travassos C, de Oliveira EXG, Viacava F. Desigualdades geográficas e sociais no acesso aos serviços de saúde no Brasil: 1998 e 2003. Cien Saude Colet 2006; 11: 975–986.
2. Arruda NM, Maia AG, Alves LC. Inequality in access to health services between urban and rural areas in Brazil: a disaggregation of factors from 1998 to 2008. Cad Saude Publica; 34.
3. Mullachery P, Silver D, Macinko J. Changes in health care inequity in Brazil between 2008 and 2013. Int J Equity Health 2016; 15: 1–12.
4. Mullachery P, Macinko J, Silver D. Have Health Reforms in Brazil Reduced Inequities in Access to Cancer Screenings for Women? J Ambul Care Manage 2020; 43: 257–266.
5. Cambota JN, Rocha FF. Determinantes das desigualdades na utilização de serviços de saúde: análise para o Brasil e regiões.
6. Almeida G, Sarti FM, Ferreira FF, et al. Analysis of the evolution and determinants of income-related inequalities in the Brazilian health system, 1998-2008. Revista Panamericana de Salud Pública 2013; 33: 90–97.
7. Andrade MV, Noronha KVM de S, Menezes R de M, et al. Desigualdade socioeconômica no acesso aos serviços de saúde no Brasil: um estudo comparativo entre as regiões brasileiras em 1998 e 2008. Economia Aplicada 2013; 17: 623–645.
8. Macinko J, Lima-Costa MF. Horizontal equity in health care utilization in Brazil, 1998–2008. Int J Equity Health 2012; 11: 1–8.
9. Almeida APSC, Nunes BP, Duro SMS, et al. Determinantes socioeconômicos do acesso a serviços de saúde em idosos: revisão sistemática. Rev Saude Publica 2017; 51: 50.
10. Beard JR, Bloom DE. Towards a comprehensive public health response to population ageing. The Lancet 2015; 385: 658–661.
11. Martinez-Lacoba R, Pardo-Garcia I, Escribano-Sotos F. Aging, dependence, and long-term care: a systematic review of employment creation. INQUIRY: The Journal of Health Care Organization, Provision, and Financing 2021; 58: 00469580211062426.
12. Brody JA. Changing health needs of the ageing population. In: Ciba Foundation Symposium 134?Research and the Ageing Population: Research and the Ageing Population: Ciba Foundation Symposium 134. Wiley Online Library, 2007, pp. 208–220.
13. Channon AA, Andrade MV, Noronha K, et al. Inpatient care of the elderly in Brazil and India: Assessing social inequalities. Soc Sci Med 2012; 75: 2394–2402.
14. IBGE - Instituto Brasileiro de Geografia e Estatísitica. Pesquisa Nacional da Saúde 2019.
15. dos Santos AMA, Triaca LM, Tejada CAO. Evolution of inequalities in health care use among older people in Brazil: Evidence for the period 1998–2019. The Journal of the Economics of Ageing 2021; 20: 100347.
16. Szwarcwald CL, Malta DC, Pereira CA, et al. National Health Survey in Brazil: design and methodology of application. Cien Saude Colet 2014; 19: 333.
17. IBGE - Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional da Saúde 2013.
18. Almeida G, Sarti FM, Ferreira FF, et al. Analysis of the evolution and determinants of income-related inequalities in the Brazilian health system, 1998-2008. Revista Panamericana de Salud Pública 2013; 33: 90–97.
19. Politi R. Desigualdade na utilização de serviços de saúde entre adultos: uma análise dos fatores de concentração da demanda. Economia Aplicada 2014; 18: 117–137.
20. Coube M, Nikoloski Z, Mrejen M, et al. Persistent inequalities in health care services utilisation in Brazil (1998–2019). Int J Equity Health 2023; 22: 25.
21. O’Donnell O, Van Doorslaer E. Analyzing Health Equity Using Household Survey Data,. Washington. Epub ahead of print 2008. DOI: 10.2471/BLT.08.052357.
22. Morris S, Sutton M, Gravelle H. Inequity and inequality in the use of health care in England: an empirical investigation. Soc Sci Med 2005; 60: 1251–1266.
23. Wagstaff A. The bounds of the concentration index when the variable of interest is binary, with an application to immunization inequality. Health Econ 2005; 14: 429–432.
24. Erreygers G. Correcting the concentration index. J Health Econ 2009; 28: 504–515.
25. Wagstaff A, Van Doorslaer E, Watanabe N. On decomposing the causes of health sector inequalities with an application to malnutrition inequalities in Vietnam. J Econom 2003; 112: 207–223.
26. Araújo CS de, Lima R da C, Peres MA, et al. Utilização de serviços odontológicos e fatores associados: um estudo de base populacional no Sul do Brasil. Cad Saude Publica 2009; 25: 1063–1072.
27. Camargo MBJ, Dumith SC, Barros AJD. Uso regular de serviços odontológicos entre adultos: padrões de utilização e tipos de serviços. Cad Saude Publica 2009; 25: 1894–1906.
28. Machado LP, Camargo MBJ, Jeronymo JCM, et al. Uso regular de serviços odontológicos entre adultos e idosos em região vulnerável no sul do Brasil. Rev Saude Publica 2012; 46: 526–533.
29. Fagundes MLB, Bastos LF, Amaral Júnior OL do, et al. Socioeconomic inequalities in the use of dental services in Brazil: an analysis of the 2019 National Health Survey. Revista Brasileira de Epidemiologia 2021; 24: e210004.
30. Theme Filha MM, Leal M do C, Oliveira EFV de, et al. Regional and social inequalities in the performance of Pap test and screening mammography and their correlation with lifestyle: Brazilian national health survey, 2013. Int J Equity Health 2016; 15: 1–8.
31. Dias-da-Costa JS, Olinto MTA, Bassani D, et al. Inequalities in clinical breast examination in São Leopoldo, Rio Grande do Sul, Brazil. Cad Saude Publica 2007; 23: 1603–1612.
32. Costa JSD da, Olinto MTA, Gigante DP, et al. Cobertura do exame físico de mama: estudo de base populacional em Pelotas, RS. Revista Brasileira de Epidemiologia 2003; 6: 39–48.
33. Sclowitz ML, Menezes AMB, Gigante DP, et al. Condutas na prevenção secundária do câncer de mama e fatores associados. Rev Saude Publica 2005; 39: 340–349.
34. Schneider IJC, Giehl MWC, Boing AF, et al. Rastreamento mamográfico do câncer de mama no Sul do Brasil e fatores associados: estudo de base populacional. Cad Saude Publica 2014; 30: 1987–1997.
35. Galvão MHR, Roncalli AG. Does the implementation of a national oral health policy reduce inequalities in oral health services utilization? The Brazilian experience. BMC Public Health 2021; 21: 1–8.