Income inequality and risk of infection and death by COVID-19 in Brazil

ABSTRACT:

Objective:

To assess, through space-time analyses, whether the income inequality of the Federative Units (FUs) in Brazil can be associated with the risk of infection and death by COVID-19.

Methods:

This was an ecological study, based on secondary data on incidence and mortality rates for COVID-19. Data were analyzed at the state level, having the Gini coefficient as the main independent variable. Records of twelve days were used, spaced one week each, between April 21^th and June 7^th, 2020. The weekly variation in the rates was calculated through Prais-Winsten regression, aiming at measuring the evolution of the pandemic in each FU. Spearman’s correlation test was used to assess correlation between the rates and their weekly evolution and the independent variables. Lastly, a spatial dependence diagnosis was conducted, and a Spatial Regression lag model was used when applicable.

Results:

Incidence and mortality rates of COVID-19 increased in all Brazilian FUs, being more pronounced among those with greater economic inequality. Association between Gini coefficient and COVID-19 incidence and mortality rates remained even when demographic and spatial aspects were taken into account.

Conclusion:

Income inequality can play an important role in the impact of COVID-19 on the Brazilian territory, through absolute and contextual effects. Structural policies to reduce inequality are essential to face this and future health crises in Brazil.

Keywords:
COVID-19; Pandemic; Health status disparities; Social determinants of health; Gini coefficient

RESUMO:

Objetivo:

Avaliar, por meio de análise espaçotemporal, se a desigualdade econômica das Unidades Federativas (UF) do Brasil pode estar associada com o risco de infecção e morte por COVID-19.

Métodos:

Trata-se de um estudo ecológico, baseado em dados secundários das taxas de incidência e mortalidade para COVID-19. Os dados foram analisados em nível estadual, tendo como principal variável independente o coeficiente de Gini. Foram utilizados os registros de 12 dias, espaçados em uma semana cada, entre 21 de abril e 7 de julho de 2020. A variação semanal das taxas foi calculada pela regressão de Prais-Winsten, com o objetivo de medir a evolução da pandemia em cada UF. O teste de correlação de Spearman foi empregado para avaliar a correlação entre as taxas e suas evoluções semanais e as variáveis independentes. Por fim, realizou-se diagnóstico de dependência espacial dos dados e usou-se o modelo de defasagem da regressão espacial, quando aplicável.

Resultados:

As taxas de incidência e mortalidade por COVID-19 foram crescentes em todas as UF brasileiras, tendo sido mais acentuada entre aquelas com maior desigualdade econômica. A associação entre coeficiente de Gini e incidência e mortalidade por COVID-19 manteve-se mesmo quando levados em consideração aspectos demográficos e espaciais.

Conclusão:

A desigualdade econômica pode exercer papel importante no impacto da COVID-19 em território brasileiro, por meio de efeitos absolutos e contextuais. Políticas estruturais para a redução da desigualdade são fundamentais para o enfrentamento dessa e de futuras crises sanitárias no Brasil.

Palavras-chave:
COVID-19; Pandemia; Desigualdade em saúde; Determinantes sociais da saúde; Coeficiente de Gini

INTRODUCTION

Humanity faces one of the greatest public health challenges in contemporary history when dealing with the disease caused by a new type of coronavirus, called coronavirus disease 2019 - COVID-19. The disease emerged in China, rapidly spreading across the globe and causing the World Health Organization (WHO) to recognize it as a pandemic^1,2. It is a highly transmissible disease and of high clinical severity³. On July 7, 2020, 15 million cases and 500 thousand deaths due to COVID-19 were recorded around the world². In Brazil, official data point to 1.7 million cases of infection and 66 thousand deaths⁴, and the numbers keep increasing.

Since treatments and vaccines for COVID-19 have not yet been developed, strategies to contain the spread of the virus have been implemented such as encouraging social distancing and, in the most affected regions, mandatory population confinement⁵. The main objective of adopting such measures is limiting the number of infected people to a threshold at which the healthcare system and services are able to meet the demand, distributing the total number of cases over time, a phenomenon that has been popularly called flattening the curve⁶. Countries, such as Italy and Spain, which reached a high number of cases and deaths from this disease very quickly, managed to control the situation by adopting these strategies².

In addition to the impact on the health of populations worldwide, health crises involving infectious agents often have an even more damaging impact: they unevenly affect population subgroups⁷. Based on the assumption that the COVID-19 pandemic is a global phenomenon, people are believed to be equally likely of being infected; however, pandemics occur in a local context, with different impacts among socially different populations⁷. Failure to recognize this aspect has already led to various plans and policies for coping with health crises to exacerbate preexisting biological, social, and economic disadvantages⁸.

In the last decades, Brazil has achieved impressive advances in its health indicators as a result of its development and the project to universalize the access to health in the country^9,10. However, since rankings on income concentration of nations have been released, Brazil remains among the 10 countries with the greatest inequality in the world¹¹. Advances in health are also uneven, with less progress among the subgroups in the worst socioeconomic position^12,13,14.

Considering both the Brazilian experience with dengue¹⁵, tuberculosis¹⁶, and HIV/AIDS¹⁷ and the international experience with the H1N1, SARS, and Ebola pandemics⁷, it is assumed that there will be differences in the incidence and mortality rates due to COVID-19 related to income inequality in Brazil. Therefore, the objective of this study was to evaluate, based on space-time analyses, whether there is a relationship between income inequality and infection and death by COVID-19 in the Federative Units (FUs) of Brazil.

METHODS

This is an ecological study carried out with secondary data obtained from publicly-accessible databases in Brazil, which used Brazilian FUs as a unit of analysis. The evaluated outcomes were the incidence and mortality rates of COVID-19, with income concentration as the main independent variable. In addition, taking into account the possibility that the rates of infection and death by COVID-19 may be correlated with the population density of each FU, this variable was included in the model to control the confounding effect.

The outcomes were obtained from data on the number of infected people and deaths by COVID-19 and the total population of each FU, which were extracted from Painel Coronavírus portal⁴ on July 7, 2020. With this information, the incidence (cases ÷ population) and mortality (deaths ÷ population) rates were calculated, which were expressed in the proportion of cases/deaths per 1 million inhabitants. For this study, data recorded on twelve different days were used, with an interval of one week between each, namely: April 21^st and 28^th; May 5^th, 12^th, 19^th, and 26^th; June 2^nd, 9^th, 16^th, 23^rd, and 30^th; and July 7^th.

The Gini coefficient, a measure adopted for estimating the degree of income concentration of the population of each FU, was also extracted from data from the Brazilian Institute for Geography and Statistics (IBGE)¹⁸. This coefficient can vary between 0 and 1, and the higher the value, the greater the concentration of income. Population density was calculated by dividing the total population and the territorial area (in km²) of each FU (data from IBGE¹⁹).

The weekly temporal evolution of the incidence and mortality rates due to COVID-19 was evaluated using the Prais-Winsten regression²⁰, based on the Durbin-Watson test, with the STATA 13.1 software. From this analysis, it was possible to extract the magnitude of the average weekly growth of these rates for each FU in the selected period. In addition, an analysis of the association between incidence and mortality recorded on each of the 12 dates included in the respective weekly evolutions (Prais-Winsten regression coefficient) was performed with the Gini coefficient; and the population density of each FU, by the Spearman’s correlation test.

Finally, the hypothesis of spatial autocorrelation between the outcomes (incidence and mortality due to COVID-19) and the independent variables (Gini coefficient and population density) was considered. For this purpose, a diagnosis of spatial autocorrelation of two regression models was conducted, one using the weekly variation in incidence as the outcome and the other, the weekly variation in mortality. This procedure was conducted using the GeoDa software, employing the mesh of Brazilian states as a weighting matrix, with queen-type contiguity²⁰. Moran’s and Lagrange Multiplier tests (of the types of lag and errors, including robust tests for testing the spatial autocorrelation) were performed^21,22. All statistical tests were performed considering a 5% significance level for two-tailed tests.

RESULTS

Brazil is a country characterized by wide social, geographical, and economic diversity. According to data obtained from the accessed databases, São Paulo is the most populous state (45,919,049 inhabitants), whereas Rio de Janeiro accounts for the highest population density (394.62 inhabitants/km²). On the other hand, Roraima is the state with the lowest number of inhabitants (605,761) and Amazonas has the lowest population density (2.65 inhabitants/km²). The FU with the highest income inequality, according to the Gini coefficient, is Amazonas (0.6664), followed by Roraima (0.6398) and Acre (0.6394), whereas the one with the lowest concentration is Santa Catarina (0.4942), followed by Paraná (0.5416) and Rio Grande do Sul (0.5472).

Table 1 shows the incidence and mortality rates recorded at the beginning (April 21^th) and at the end (July 7^th) of the studied period as well as the weekly variation in this indicator obtained by the Prais-Winsten regression. In the first week, recorded on April 21^th, the states with the highest incidence rates were Amazonas (547.7/1 million inhabitants), Amapá (540.36/1 million inhabitants), and Roraima (407.75/1 million inhabitants); conversely, Tocantins (23.52/1 million inhabitants), Sergipe (40.02/1 million inhabitants), and Mato Grosso (51.94/1 million inhabitants) had the lowest rates on that day. In the last week included, recorded on July 7^th, the FUs with the highest incidence rates were Amapá (35,819.90/1 million inhabitants), Roraima (31,510.78/1 million inhabitants), and the Federal District (20,792.18/1 million inhabitants). On the other hand, the states with the lowest incidence on that date were Minas Gerais (2,876.73/1 million inhabitants), Rio Grande do Sul (2,970.84/1 million inhabitants), and Paraná (3,000.54/1 million inhabitants). In all Brazilian FUs, an increase in the incidence rate was verified, which was more prominent in Amapá (+3,280.05/1 million inhabitants per week), Roraima (+2,807.82/1 million inhabitants per week), and the Federal District (+1,857.97/1 million inhabitants per week), and less remarkable in Minas Gerais (+254.27/1 million inhabitants per week), Paraná (+261.13/1 million inhabitants per week), and Rio Grande do Sul (+262.33/1 million inhabitants per week).

Table 1.
Characteristics of Federative Units (FUs) and their respective incidence and mortality rates due to COVID-19^#. Brazil, 2020.

	FU	Population density†	Gini	Incidence 04/21	Incidence 07/07	Incidence Variation$	Mortality 04/21	Mortality 07/07	Mortality Variation$
North	AC	5.37	0.6394	221.10	16,941.16	+1,539.90**	9.07	452.41	+40.66**
	AM	2.65	0.6664	547.70	19,101.25	+1,727.84**	46.57	712.25	+61.08**
	AP	5.94	0.6157	540.36	35,819.90	+3,280.05**	15.37	538.00	+48.48**
	PA	6.91	0.6260	119.26	13,501.55	+1,222.89**	4.42	596.08	+56.38**
	RO	7.47	0.5686	111.97	13,821.55	+1,247.48**	2.25	324.66	+29.42**
	RR	2.71	0.6398	407.75	31,510.78	+2,807.82*	4.95	620.71	+55.74**
	TO	5.67	0.6099	23.52	8,267.71	+750.64**	0.64	144.96	+13.25**
Northeast	AL	119.86	0.6343	62.92	12,442.18	+1,127.55**	5.69	357.17	+32.10**
	BA	26.34	0.6278	100.11	6,182.59	+549.76*	3.16	148.99	+13.25**
	CE	61.33	0.6193	406.92	13,682.76	+1,211.94**	23.54	717.91	+64.54**
	MA	21.46	0.6291	197.31	13,015.64	+1,176.63**	8.48	323.10	+28.68**
	PB	71.16	0.6139	65.45	13,638.69	+1,236.05**	8.21	284.96	+25.17**
	PE	97.45	0.6366	304.28	6,921.68	+609.44**	27.20	547.66	+48.29**
	PI	13.00	0.6193	56.82	8,405.77	+755.18*	4.28	254.79	+22.73*
	RN	66.40	0.6074	173.37	10,214.29	+909.56*	7.98	368.14	+32.64*
	SE	104.84	0.6288	40.02	13,764.33	+1,245.36**	2.18	370.21	+33.39*
Midwest	DF	523.48	0.6370	292.18	20,792.18	+1,857.97*	7.96	254.37	+22.34*
	GO	206.30	0.5588	59.99	4,460.88	+398.50*	2.71	99.88	+8.69*
	MS	7.78	0.5650	62.25	3,845.65	+342.24*	2.16	46.06	+3.85*
	MT	3.86	0.5652	51.94	6,430.25	+574.63*	1.72	242.79	+21.72*
Southeast	ES	87.22	0.5723	301.59	14,108.72	+1,254.20**	8.46	467.82	+41.85**
	MG	36.09	0.5634	58.10	2,876.73	+254.27*	2.08	60.56	+5.28*
	RJ	394.62	0.6116	307.33	7,187.17	+631.68**	26.70	630.24	+55.94**
	SP	184.99	0.5768	335.05	7,245.53	+627.96**	23.80	358.78	+30.48**
South	PR	57.37	0.5416	89.56	3,000.54	+261.13*	4.46	74.43	+6.32*
	RS	40.39	0.5472	79.46	2,970.84	+262.33**	2.37	69.52	+6.07**
	SC	74.84	0.4942	148.36	4,932.73	+430.87*	4.89	58.48	+4.85**

• ^#Incidence and mortality rates per 1 million inhabitants; ^$weekly variations extracted by using the Prais-Winsten regression (+ = increasing variation); ^†population density in people/km²; *p < 0.05; **p < 0.001.
• Source: Brasil⁴, Brazilian Institute for Geography and Statistics - (IBGE)18^,19.

Regarding mortality rates, in the first week, registered on April 21^th, the states with the highest mortality rates were Amazonas (46.57/1 million inhabitants), Pernambuco (27.2/1 million inhabitants), and Rio de Janeiro (26.7/1 million inhabitants), whereas Tocantins (0.64/1 million inhabitants), Mato Grosso (1.72/1 million inhabitants), and Minas Gerais (2.08/1 million inhabitants) accounted for the lowest rates on that day. In the last week included in the analysis, July 7^th, the FUs with the highest mortality rates were Ceará (717.91/1 million inhabitants), Amazonas (712.25/1 million inhabitants), and Rio de Janeiro (630.24/1 million inhabitants), whereas those with the lowest rates were Mato Grosso do Sul (46.06/1 million inhabitants), Santa Catarina (58.48/1 million inhabitants), and Minas Gerais (60.56/1 million of inhabitants). There was an increasing mortality rate in all Brazilian FUs, with a greater increase in Ceará (+64.54/1 million inhabitants per week), Amazonas (+61.08/1 million inhabitants per week), and Pará (+56.38/1 million inhabitants per week); and with a modest increase in Mato Grosso do Sul (+3.85/1 million inhabitants per week), Santa Catarina (+4.85/1 million inhabitants per week), and Minas Gerais (+5.28/1 million inhabitants per week).

Table 2 shows the results of the correlation analyses. Population density was not correlated with any of the rates. The Gini coefficient, in its turn, was correlated with both rates in all recorded periods. As for the incidence rate, the association evolved from a weak positive correlation on April 21^th (rho = +0.4115, p = 0.033), reaching strong correlations (greater than +0.7) in May and June, to a moderate positive (though almost strong) correlation on July 7^th (rho = +0.6906, p < 0.001). The mortality rate was also correlated with the Gini coefficient, evolving from a weak positive correlation on April 21^th (rho = +0.4760, p = 0.012) to a moderate positive correlation on July 7^th (rho = +0, 6564, p < 0.001). The visual dispersion between the weekly variation in the incidence and mortality rates of each state, according to the respective Gini coefficients, can be seen in Figure 1.

Table 2.
Correlation between incidence and mortality rates due to COVID-19 and their weekly variation in the observed period (April 21^th to July 7^th), with population density and Gini coefficient^# of the Federative Units (FUs) of Brazil, 2020.

Period	Incidence		Mortality
	Population density	Gini coefficient	Population density	Gini coefficient
04/21	+0.0617	+0.4115*	+0.2821	+0.4760*
04/28	+0.1233	+0.5892*	+0.2088	+0.5199*
05/05	+0.0299	+0.6582*	+0.1758	+0.5846*
05/12	-0.0342	+0.6955**	+0.1306	+0.6164**
05/19	-0.0745	+0.7474**	+0.0360	+0.6384**
05/26	-0.1306	+0.7834**	+0.0195	+0.6350**
06/02	-0.1441	+0.7816**	+0.0049	+0.6613**
06/09	-0.1886	+0.7608**	-0.0360	+0.6640**
06/16	-0.1917	+0.7639**	-0.0409	+0.6652**
06/23	-0.1947	+0.7538**	-0.0427	+0.6616**
06/30	-0.2265	+0.7226**	-0.0110	+0.6475**
07/07	-0.1636	+0.6906**	-0.0140	+0.6558**
Weekly variation	-0.1685	+0.6924**	-0.0330	+0.6564**

• ^#Coefficients were extracted by using Spearman’s correlation test; *p < 0.05; **p < 0.001.

Figure 1.
Scatter plots between COVID-19 and Gini Coefficient of Federative Units (FUs) in Brazil, 2020: (A) Dispersion of the correlation between the weekly variation in incidence rates by COVID-19 and the Gini coefficient of FUs; (B) Dispersion of the correlation between the weekly variation in mortality rates due to COVID-19 and the Gini coefficient of FUs.

By assessing the spatial dependence diagnosis conducted, it was observed that the model for the weekly variation in the incidence rate did not show spatial autocorrelation (Moran’s I = 0.724, p = 0.469; Lagrange Multiplier (lag) = 0.605, p = 0.437; Lagrange Multplier (error) = 0.02, p = 0.875). In turn, the model for the weekly variation in the mortality rate showed spatial autocorrelation (Moran’s I = 2.051, p = 0.040; Lagrange Multiplier (lag) = 3.288, p = 0.069, Lagrange Multipler (error) = 1,792, p = 0.180). Based on these results, the spatial regression with a lag model was chosen to be used, and the results are shown in Table 3. In both models (linear and spatial), even with the control for population density of the FUs, the Gini coefficient remained associated with both the evolution of incidence and mortality due to COVID-19.

Table 3.
Analysis of simple and spatial linear regression (lag model) of the outcomes weekly variation in incidence and in mortality rates due to COVID-19 in the observed period (April 21^th to July 7^th)*. Brazil, 2020.

Outcome	Variables	Simple Linear Regression			Spatial Regression (lag model)
		Coefficient	SE	p-value	Coefficient	SE	p-value
Weekly variation in incidence	Constant	-5,167.34	1,875.32	0.011	-4,401.26	1,810.33	0.015
	Gini coefficient	10,406.80	3,121.57	0.003	8,648.91	3,160.19	0.006
	Population density	-0.42	1.01	0.682	0.03	0.95	0.974
	Aut. Coef.	-	-	-	0.289	0.196	0.140
Weekly variation in mortality	Constant	-161.33	43.84	0.001	-116.12	39.77	0.004
	Gini coefficient	319.122	72.87	< 0.001	218.65	69.91	0.002
	Population density	-0.01	0.02	0.916	0.02	0.02	0.276
	Aut. Coef.	-	-	-	0.448	0.166	0.006

• *Weekly variations in incidence and mortality rates extracted from time series data using Prais-Winsten regression; SE: standard error; Aut. Coef.: Autoregressive coefficient of the lag model.

DISCUSSION

The results of this study indicate a possible negative reflection of income inequality on facing the COVID-19 pandemic in Brazil. This result seems to be consistent because, in addition to the correlation observed in 12 different weeks, there was an increase in the strength of the correlation during the evaluated period, considering that the associations evolved from weak positive correlations to practically strong ones²³. Furthermore, among the most unequal states, the progression in the incidence and mortality rates due to COVID-19 was more prominent, whereas among the less unequal states there were modest increases. Finally, even considering demographic (population density) and spatial (spatial autocorrelation) aspects, it can be stated that the Gini coefficient was associated with an increase in the incidence and mortality rates of this disease.

Thus, the findings regarding the association between economic inequality and infection and death by COVID-19 seem to be true, and provide a plausible explanation for the differences between Brazilian states concerning the COVID-19 pandemic. Economic inequality can have a significant impact on the health of populations, in addition to the effect of poverty itself. In the case of COVID-19, this seems to occur due to at least two distinct effects: the absolute and the contextual²⁴.

The absolute effect concerns the direct impact of income distribution on health outcomes. Small changes in the income of poorest individuals produce significant changes in health outcomes, whereas among the wealthier individuals the same changes in income (higher or lower) do not produce a major change in health standards²⁴. For example, according to estimates based on data from 1.3 million New Zealanders of working age monitored for three years, if it was possible to transfer the income of 10% of the population to meet the average (which would correspond to a 10% reduction in the Gini coefficient), mortality rates for the entire population would decrease by at least 4% (over 1,100 deaths avoided per year)²⁵. It may seem a slight reduction, but it is equivalent to three times the number of deaths per year from traffic accidents in that country²⁵, which also has less economic inequality than Brazil.

Within the Brazilian context, among the 20% poorest individuals of the population, 94.4% have no health insurance and 10.9% rate their health as fair, poor, or very poor; conversely, among the 20% richest individuals, only 35.7% do not have health insurance and 2.2% evaluate their health likewise²⁶. Moreover, the availability of beds in intensive care units (ICU) for users of the Brazilian Unified Health System is almost five times lower than for those who have access to the private health network²⁶. International data indicate that, in more unequal regions, the proportion of individuals with impaired health is greater²⁷, including chronic diseases that are currently recognized as risk factors for COVID-19. Therefore, in more unequal FUs, the burden of morbidity is likely to be greater, making them structurally more vulnerable to the COVID-19 pandemic.

The contextual effect, in turn, demonstrates that people (regardless of socioeconomic status) who live in unequal societies end up paying a health fee. It is like air pollution: it is difficult for individuals to completely escape from the negative effects of air pollution in the place where they live. In uneven locations, the public structures of health, safety, sanitation, and urban planning are worse. Such conditions degrade everyone’s quality of life, but the poorest people are more severely affected within the social structure²⁴.

The unequal distribution of opportunities can allocate individuals in different socioeconomic positions according to their social group, sex, gender, and ethnicity, creating cascading difficulties in the access to education, work, and income²⁸ . People at greater socioeconomic disadvantage tend to have differential exposure to the virus (because they have poor housing conditions, live with a larger number of people in smaller residences, use public transport with greater agglomeration, and have job insecurity, which makes social distancing difficult); differential susceptibility (due to food insecurity and poorer nutritional quality, increased psychological stress, and difficulty of access to healthcare professionals); and differential consequence (less social capital and reduced options of primary prevention and treatment)^{7,28,29,30,31,32}. Altogether, differential exposure, susceptibility, and consequence can produce higher rates of illness and death in these subgroups. Such an effect has already been observed in the National Household Sample Survey (PNAD) to assess the impact of COVID-19, which showed that black and mixed-race, poor and uneducated people, in addition to being more likely to be infected, were also more severely affected by the pandemic in terms of economy³³.

It is estimated that the risk of dying from COVID-19 may be up to 10 times higher among individuals living in the most vulnerable neighborhoods in the same city, and that black people are 62% more likely to be victims of the virus³⁴. These contextual effects, which impact everyone by the degradation of the public structure, but affect poorest individuals in a more severe way, can be a plausible explanation for the greater increase in incidence and mortality rates in more unequal states in the evaluated period (and in less unequal states, this increase was smaller or almost stable).

However, the results should be carefully interpreted considering the limitations of this study. Taking into account that these are secondary data, it is not possible to state that the most economically vulnerable people are the most affected ones, but rather that in more unequal locations the impact of the pandemic is likely to be more severe for the entire population. Secondly, considering that Brazil does not carry out mass testing, the underreporting of cases and deaths by COVID-19 may influence the results of this study. Thirdly, considering the speed at which the disease has been changing in Brazil, the results may change. For instance, the results of temporal analysis, as well as the correlations, indicate stronger associations in mid-June, with a slight reduction from July onward. This finding may reflect the fact that FUs that experienced a more prominent advance of the disease (those with greater economic inequality) managed to reduce the speed of infections and deaths over time. Finally, association does not mean causality, in such a way that investigations are required to identify the causal paths of these results. Multilevel studies that simultaneously evaluate variables at the individual level (for example, skin color/ethinicity, occupation, presence of risk factors for COVID-19 etc.) and at the contextual level (such as social, demographic, and economic aspects of the place of residence in regional, municipal, or state levels) will be able to identify the specific contributions of individuals’ socioeconomic position and the inequality of the context in which they live to being affected by COVID-19.

Hence, it is concluded that income inequality can play an important role in the impact of COVID-19 on the Brazilian population, either by the unequal distribution of opportunities, which has negative cascading impacts on those at greater socioeconomic disadvantage, or by contextual effects that hinder the capacity of a locality to adequately respond to this health crisis. This finding highlights the urgency of formulating intersectoral policies aimed at reducing economic inequality. In the pandemic context, the emergency financial assistance for the most vulnerable people³⁵ seems to have been a positive short-term measure. Nevertheless, long-term structural measures are paramount for this and future health crises to have a reduced impact on the Brazilian population.

REFERÊNCIAS

• ¹
  Lai CC, Shih TP, Ko WC, Tang HJ, Hsueh PR. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and coronavirus disease-2019 (COVID-19): The epidemic and the challenges. Int J Antimicrob Agents 2020; 55(3): 105924. https://doi.org/10.1016/j.ijantimicag.2020.105924
  » https://doi.org/https://doi.org/10.1016/j.ijantimicag.2020.105924
• ²
  World Health Organization. Coronavirus disease (COVID-19) situation dashboard [Internet]. Genebra: World Health Organization; 2020 [acessado em 7 jul. 2020]. Disponível em: Disponível em: https://experience.arcgis.com/experience/685d0ace521648f8a5beeeee1b9125cd
  » https://experience.arcgis.com/experience/685d0ace521648f8a5beeeee1b9125cd
• ³
  Freitas ARR, Napimoga M, Donalisio MR. Análise da gravidade da pandemia de Covid-19. Epidemiol Serv Saúde 2020; 29(2): e2020119. https://doi.org/10.5123/S1679-49742020000200008
  » https://doi.org/https://doi.org/10.5123/S1679-49742020000200008
• ⁴
  Brasil. Ministério da Saúde. Painel coronavírus [Internet]. 2020 [acessado em 7 jul. 2020]. Disponível em: Disponível em: https://covid.saude.gov.br/
  » https://covid.saude.gov.br/
• ⁵
  Brooks SK, Webster RK, Smith LE, Woodland L, Wessely S, Greenberg N, et al. The psychological impact of quarantine and how to reduce it: Rapid review of the evidence. Lancet 2020; 395(10227): 912-20. https://doi.org/10.1016/S0140-6736(20)30460-8
  » https://doi.org/https://doi.org/10.1016/S0140-6736(20)30460-8
• ⁶
  Kenyon C. Flattening-the-curve associated with reduced COVID-19 case fatality rates- an ecological analysis of 65 countries. J Infect 2020; 81(1): E98-E99. https://doi.org/10.1016/j.jinf.2020.04.007
  » https://doi.org/https://doi.org/10.1016/j.jinf.2020.04.007
• ⁷
  Quinn SC, Kumar S. Health inequalities and infectious disease epidemics: a challenge for global health security. Biosecur Bioterror 2014; 12(5): 263-73. https://doi.org/10.1089/bsp.2014.0032
  » https://doi.org/https://doi.org/10.1089/bsp.2014.0032
• ⁸
  Garoon JP, Duggan PS. Discourses of disease, discourses of disadvantage: A critical analysis of National Pandemic Influenza Preparedness Plans. Soc Sci Med 2008; 67(7): 1133-42. https://doi.org/10.1016/j.socscimed.2008.06.020
  » https://doi.org/https://doi.org/10.1016/j.socscimed.2008.06.020
• ⁹
  Marmot M. Brazil: Rapid progress and the challenge of inequality. Int J Equity Health 2016; 15(1): 177. https://doi.org/10.1186/s12939-016-0465-y
  » https://doi.org/https://doi.org/10.1186/s12939-016-0465-y
• ¹⁰
  Paim J, Travassos C, Almeida C, Bahia L, Macinko J. The Brazilian health system: History, advances, and challenges. Lancet 2011; 377(9779): 1778-97. https://doi.org/10.1016/S0140-6736(11)60054-8
  » https://doi.org/https://doi.org/10.1016/S0140-6736(11)60054-8
• ¹¹
  Victora CG. Socioeconomic inequalities in Health: Reflections on the academic production from Brazil. Int J Equity Health 2016; 15(1): 164. https://doi.org/10.1186/s12939-016-0456-z
  » https://doi.org/https://doi.org/10.1186/s12939-016-0456-z
• ¹²
  Barreto ML, Teixeira MG, Bastos FI, Ximenes RAA, Barata RB, Rodrigues LC. Successes and failures in the control of infectious diseases in Brazil: Social and environmental context, policies, interventions, and research needs. Lancet 2011; 377(9780): 1877-89. https://doi.org/10.1016/S0140-6736(11)60202-X
  » https://doi.org/https://doi.org/10.1016/S0140-6736(11)60202-X
• ¹³
  Chor D. Desigualdades em saúde no Brasil: É preciso ter raça. Cad Saúde Pública 2013; 29(7): 1272-5. https://doi.org/10.1590/S0102-311X2013000700002
  » https://doi.org/https://doi.org/10.1590/S0102-311X2013000700002
• ¹⁴
  Victora CG, Aquino EML, Leal MC, Monteiro CA, Barros FC, Szwarcwald CL. Maternal and child health in Brazil: progress and challenges. Lancet 2011; 377(9780): 1863-76. https://doi.org/10.1016/S0140-6736(11)60138-4
  » https://doi.org/https://doi.org/10.1016/S0140-6736(11)60138-4
• ¹⁵
  Johansen IC, Carmo RL, Alves LC, Bueno MCD. Environmental and demographic determinants of dengue incidence in Brazil. Rev Salud Pública 2018; 20(3): 346-51. https://doi.org/10.15446/rsap.v20n3.54315
  » https://doi.org/https://doi.org/10.15446/rsap.v20n3.54315
• ¹⁶
  Harling G, Castro MC. A spatial analysis of social and economic determinants of tuberculosis in Brazil. Health Place 2014; 25: 56-67. https://doi.org/10.1016/j.healthplace.2013.10.008
  » https://doi.org/https://doi.org/10.1016/j.healthplace.2013.10.008
• ¹⁷
  Grangeiro A, Escuder MML, Castilho EA. A epidemia de AIDS no Brasil e as desigualdades regionais e de oferta de serviço. Cad Saúde Pública 2010; 26(12): 2355-67. https://doi.org/10.1590/S0102-311X2010001200014
  » https://doi.org/https://doi.org/10.1590/S0102-311X2010001200014
• ¹⁸
  Instituto Brasileiro de Geografia e Estatística. Índice de Gini da renda domiciliar per capita - Brasil [Internet]. Brasil: Instituto Brasileiro de Geografia e Estatística; 2010 [acessado em 3 jun. 2020]. Disponível em: Disponível em: http://tabnet.datasus.gov.br/cgi/ibge/censo/cnv/giniuf.def
  » http://tabnet.datasus.gov.br/cgi/ibge/censo/cnv/giniuf.def
• ¹⁹
  Instituto Brasileiro de Geografia e Estatística. Áreas Territoriais [Internet]. Brasil: Instituto Brasileiro de Geografia e Estatística ; 2019 [acessado em 3 jun. 2020]. Disponível em: Disponível em: https://www.ibge.gov.br/geociencias/organizacao-do-territorio/estrutura-territorial/15761-areas-dos-municipios.html?t=acesso-ao-produto&c=17
  » https://www.ibge.gov.br/geociencias/organizacao-do-territorio/estrutura-territorial/15761-areas-dos-municipios.html?t=acesso-ao-produto&c=17
• ²⁰
  Antunes JLF, Cardoso MRA. Uso da análise de séries temporais em estudos epidemiológicos. Epidemiol Serv Saúde 2015; 24(3): 565-76. https://doi.org/10.5123/S1679-49742015000300024
  » https://doi.org/https://doi.org/10.5123/S1679-49742015000300024
• ²¹
  Anselin L. Exploring spatial data with GeoDa: A workbook. Urbana: Center for Spatially Integrated Social Science; 2005.
• ²²
  Anselin L, Syabri I, Kho Y. GeoDa: An introduction to spatial data analysis. In: Fischer MM, Getis A, editores. Handbook of applied spatial analysis. Berlim: Springer; 2010. p. 78-89.
• ²³
  Mukaka MM. Statistics corner: A guide to appropriate use of correlation coefficient in medical research. Malawi Med J 2012; 24(3): 69-71.
• ²⁴
  Kawachi I, Subramanian SV. Income inequality. In: Berkman LF, Kawachi I, Glymour MM, editores. Social Epidemiology. 2ª ed. Nova York: Oxford University Press; 2014. p. 126-52.
• ²⁵
  Blakely T, Wilson N. Shifting dollars, saving lives: What might happen to mortality rates, and socio-economic inequalities in mortality rates, if income was redistributed? Soc Sci Med 2006; 62(8): 2024-34. https://doi.org/10.1016/j.socscimed.2005.08.059
  » https://doi.org/https://doi.org/10.1016/j.socscimed.2005.08.059
• ²⁶
  Pires LN, Carvalho L, Xavier LL. COVID-19 e desigualdade: a distribuição dos fatores de risco no Brasil. Experiment Findings 2020. https://doi.org/10.13140/RG.2.2.27014.73282
  » https://doi.org/https://doi.org/10.13140/RG.2.2.27014.73282
• ²⁷
  Dewan P, Rorth R, Jhund PS, Ferreira JP, Zannad F, Shen L, et al. Income Inequality and Outcomes in Heart Failure: A Global Between-Country Analysis. JACC Heart Fail 2019; 7(4): 336-46. https://doi.org/10.1016/j.jchf.2018.11.005
  » https://doi.org/https://doi.org/10.1016/j.jchf.2018.11.005
• ²⁸
  Solar O, Irwin A. A conceptual framework for action on the social determinants of health. Social determinants of health discussion paper 2 (Policy and Practice). Genebra: World Health Organization ; 2010.
• ²⁹
  Barata RB. Como e por que as desigualdades sociais fazem mal à saúde. Rio de Janeiro: Editora FIOCRUZ; 2009. 120 p.
• ³⁰
  Buss PM, Pellegrini Filho A. A saúde e seus determinantes sociais. Physis 2007; 17(1): 77-93. https://doi.org/10.1590/S0103-73312007000100006
  » https://doi.org/https://doi.org/10.1590/S0103-73312007000100006
• ³¹
  Diderichsen F, Evans T, Whitehead M. The social basis of disparities in health. In: Evans T, editor. Challenging inequities in health. Nova York: Oxford University Press ; 2001. p. 12-23.
• ³²
  Shim RS, Compton MT, Manseau MW, Koplan C, Langheim FJP, Powers RA. Overview of the social determinants of mental health. In: Compton MT, Shim RS, editores. The social determinants of mental health. Washington, D.C.: American Psychiatric Publishing; 2015. p. 1-21.
• ³³
  Instituto Brasileiro de Geografia e Estatística. Pesquisa Nacional por Amostra de Domicílios: PNAD COVID19 (maio/2020) [Internet]. Brasil: Instituto Brasileiro de Geografia e Estatística ; 2020 [acessado em 25 jun. 2020]. Disponível em: Disponível em: https://biblioteca.ibge.gov.br/index.php/biblioteca-catalogo?view=detalhes&id=2101727
  » https://biblioteca.ibge.gov.br/index.php/biblioteca-catalogo?view=detalhes&id=2101727
• ³⁴
  Mello D. Risco de morrer por coronavírus varia até 10 vezes entre bairros de SP [Internet]. Agência Brasil; 2020 [acessado em 3 jun. 2020]. Disponível em: Disponível em: https://agenciabrasil.ebc.com.br/saude/noticia/2020-05/risco-de-morrer-por-coronavirus-varia-ate-10-vezes-entre-bairros-de-sp
  » https://agenciabrasil.ebc.com.br/saude/noticia/2020-05/risco-de-morrer-por-coronavirus-varia-ate-10-vezes-entre-bairros-de-sp
• ³⁵
  Agência Brasil. Auxílio emergencial [Internet]. Agência Brasil; 2020 [acessado em 26 jul. 2020]. Disponível em: Disponível em: https://agenciabrasil.ebc.com.br/tags/auxilio-emergencial
  » https://agenciabrasil.ebc.com.br/tags/auxilio-emergencial

• Financial support: none

Publication Dates

• Publication in this collection
  05 Oct 2020
• Date of issue
  2020

History

• Received
  26 July 2020
• Accepted
  17 Aug 2020