ArticlePDF Available

Abstract and Figures

►► Accepted manuscript, pre-copyedit version of the article: https://www.researchgate.net/publication/345941695_The_Pandemic_of_Poverty_Vulnerability_and_COVID-19_Evidence_from_a_Fuzzy_Multidimensional_Analysis_of_Deprivations_in_Brazil ►► Preliminary Briefing Abstract: Using the Brazilian Consumer Expenditure Survey (POF-IBGE) for 2017-18, the briefing aims to show how much and in which way people in Brazil are deprived in living standards indicators directly related to the capacity to prevent and heal COVID-19. We use the Alkire-Foster (AF) method to measure multidimensional vulnerability selecting eight interlinked indicators in the dimensions of hygiene, shelter, physical distance, and health recovery capacity. This approach, applied to the analysis of vulnerability to COVID-19 in Brazil, revealed considerable inequality among regions and ethnicities. These pieces of evidence reinforce the urgent need to act to avoid the virus to achieve the most vulnerable groups and be prepared in advance in case of necessity.
Content may be subject to copyright.
* Department of Economics and Statistics, University of Siena, Siena, Italy; email: fern.florestavares@student.unisi.it.
** Department of Economics and Statistics, University of Siena, Siena, Italy; email: gianni.betti@unisi.it.
The Pandemic of Poverty, Vulnerability, and COVID-19: Evidence from
a Fuzzy Multidimensional Analysis of Deprivations in Brazil
Fernando Flores Tavares*, Gianni Betti**
This is an accepted manuscript, pre-copyedit version of the article published in World Development.
The final version is available online at: https://doi.org/10.1016/j.worlddev.2020.105307.
Abstract
This paper aims to show how much and in which way people in Brazil are deprived in terms of
indicators directly related to the capacity to prevent and recover from infection with COVID-19. We use
the Alkire-Foster (AF) method and a fuzzy-set approach as complements to measure multidimensional
poverty within the context of the coronavirus pandemic. We propose two pandemic-specific indexes to
account for the vulnerability related to the capacity to prevent infection with and to recover from the
disease. The outcomes reveal structural deprivations in the country and considerable inequality among
regions and ethnic groups. Rank correlation analyses suggest that the proposed indexes can trace the
trends in increasing infection and a higher mortality rate in vulnerable regions. Compared to headcount
ratio results, the fuzzy measures have more precise outcomes and are better able to capture the evolution
in mortality patterns. Our empirical evidence offers an additional warning that the pandemic responses
need to prioritize the most vulnerable groups and reinforces the need for coordinated national action.
Keywords: COVID-19 ∙ Multidimensional Poverty ∙ Fuzzy-set approach ∙ Alkire-Foster (AF) method
Latin America ∙ Brazil
1. Introduction
The COVID-19 outbreak has exposed the inequality and interlinked socioeconomic deprivation
affecting developing countries to a greater extent than before. The fact that some of the population has
these problems not only is related to the pandemic but mainly reveals historical gaps that are exacerbated
by the virus. In Brazil, minority groups are at a disadvantage in terms of economic, social, and health
deprivations (Hoffman, 2018; Fernandes, 2017; Raupp et al., 2017). Planning an efficient response to the
pandemic requires an understanding of the increased risk of exposure, especially among those living in
unsafe conditions. In this sense, interest in analyzing the vulnerability to infection with COVID-19
among subgroups has grown (Pareek, 2020; Khalatbari-Soltani et al., 2020). By examining how much
and in what ways people in Brazil are deprived in terms of indicators directly related to the ability to
prevent infection with and recover from COVID-19, this paper joins others on this topic.
The first confirmed case in Brazil was diagnosed on February 25, 2020. By May 10, the country had
162,699 confirmed cases in a pattern of rapid infection (DATASUS, 2020). Even though Brazil climbed
to second in the worldwide number of confirmed COVID-19 cases on May 22 and in the number of
confirmed deaths from COVID-10 on June 12, its national government is still struggling to recognize the
problem and promote coordinated action (see Lancet, 2020). The pandemic is worsening the quality of
life in entire communities, and the lack of effective policies poses an additional threat to the population.
Families experiencing multidimensional poverty face at least two sets of additional risk factors.
First, people living in poverty might not be able to follow the recommendations for prevention (see
WHO, 2020a, 2020b). Sheltering at home might be infeasible if their housing is inadequate for keeping
them safe and comfortable during a quarantine. It is not always possible to wash hands, clean and
disinfect the home properly if one has inadequate access to clean water and sanitation conditions are
poor. Keeping a safe distance from others is not practicable in an overcrowded residence. Furthermore,
transmission of the virus might be enhanced in high-density communities (Lusignan et al., 2020; Rubin
et al., 2020) and in places with insufficient social distancing (Rubin et al., 2020, Chu et al., 2020); and
the spread of COVID-19 can be mitigated where the mobility control measures are stricter (Kraemer et
al., 2010).
Second, poor living standards and insufficient health services reduce the ability to recover from
COVID-19. Drinking unsafe water and being exposed to improper sewage disposal is highly correlated
with the contraction of preventable diseases (WHO, 2019a, 2019b), which can compromise the immune
system. Families who use highly polluting fuels for cooking might be a risk group as indoor air pollution
is associated with respiratory diseases (WHO, 2018a). Because the schools are closed, food security is
now under threat for families with schoolchildren who depend on schools for daily free meals. The lack
of physicians and intensive-care beds in hospitals is critical for people in need of treatment. The distance
from hospitals is an additional factor in vulnerability, particularly for several Indigenous communities
that live far from urban areas.
1
The literature on infectious disease outcomes for subgroups suggests that risks are higher among
minority groups and in more deprived regions. For instance, Zhao et al. (2016) show that, during the
2009/2010 influenza A(H1N1) pandemic, the risk of mortality in England was higher for non-White
populations than White populations and for people living in the most deprived areas compared with those
in less deprived areas. Lusignan et al. (2020) estimate that, within the Oxford Royal College of General
Practitioners Research and Surveillance Centre primary care network, Black people have higher risk
factors for testing positive for COVID-19 than White people and so do individuals in more deprived
areas.
Studies about racial and ethnic disparities in the United States in terms of infection with and mortality
from COVID-19 also show that minorities are the hardest hit. Laurencin and McClinton (2020)
demonstrate that in Connecticut, the Black population had a proportion of infection and death that
exceeded its share of the population even at the beginning of April. Yancy (2020) shows that this
disproportion is also present in Chicago, Louisiana, Michigan, and New York City. In Chicago, for
example, Black people make up 30% of the population but more than 50% of the confirmed cases and
almost 70% of the deaths. Millett et al. (2020) and Holtgrave et al. (2020) confirm these discrepancies
among racial and ethnic minorities, and they conclude that social characteristics, structural racism, less
access to health care, and other factors might be driving these results.
Research on the impact of the COVID-19 pandemic on minority groups and in different regions in
Latin America is still thin, but it confirms the same outcomes there. A pioneering study by Baqui et al.
(2020) uses the SIVEP-Gripe (Sistema de Informação de Vigilância Epidemiológica da Gripe) dataset
to analyze COVID-19 hospital mortality in Brazil. The analysis selects only observations that account
for ethnicity to assess the relation between health risk, ethnicity, and regional differences. The authors
find that Black and Brown people are at the highest risk of a hospital death. They also show that people
at hospitals in the northern region had comorbidities more often and a higher risk of mortality than people
in most of the central-south region.
To contribute to the pandemic literature on Brazil, we use the Alkire-Foster (AF) method and the
fuzzy-set approach as complementary measures of multidimensional poverty in the context of COVID-
19 (Alkire and Foster, 2011; Betti and Verma, 2008). Because families have multiple difficulties at the
same time, unidimensional poverty measureswhich usually focus only on monetary povertyare
insufficient to account for the reality for these people. Therefore, the methods proposed in this work are
appropriate for collecting clear evidence of overlapping kinds of deprivation. The latter, also seen as the
intersection of multidimensional aspects of poverty, are considered high-risk factors in any
multidimensional approach (Lemmi and Betti, 2006), and this is particularly evident when poverty and
deprivation are analyzed at the regional or subnational level (Betti et al., 2012).
This paper is inspired by the policy briefing on multidimensional poverty and COVID-19 risk factors
written by Alkire et al. (2020). They show that the Global Multidimensional Poverty Index (GMPI)
(OPHI and UNDP, 2019) provides information that is useful for identifying risks and vulnerabilities
related to COVID-19. They estimated that 472 million people in the world face simultaneous deprivation
in terms of water, nutrition, and indoor air pollution.
This paper innovates in at least three lines of research, both theoretical and applied:
1. it proposes two COVID-19-specific multidimensional indices: the COVID-19 prevention index
and the COVID-19 recovery index;
2. it proposes a rank correlation analysis to determine how the vulnerability indexes can capture
the mortality patterns in vulnerable regions;
3. it introduces a fuzzy counterpart to these indices.
To achieve these original contributions, we have moved step by step; the first step was to adapt the
GMPI in the context of COVID-19 in Brazil, creating a multidimensional vulnerability index (MVI). We
selected eight interlinked vulnerability indicators in the dimensions of sanitation, home shelter, physical
distance, and recovery from illness. Five of those indicators are also among the ten GMPI
2
indicators. To
better account for groups and regional disparities, we took a further step in building an appropriate
multidimensional index. The fact that the variables previously selected for the MVI are all interlinked
makes it difficult to observe the immediate relation to COVID-19. Therefore, we propose two
multidimensional poverty indexes related to the COVID-19 pandemic in terms of prevention and the
ability to recover as the first contribution of this paper. In this way, we can obtain a more comprehensive
and detailed picture of deprivation in these two aspects. The indexes reveal considerable inequality
between regions and ethnic groups, confirming the existing evidence that minority groups and vulnerable
regions have more exposure to the virus.
The second contribution of the paper is our estimation of rank correlations to clarify whether the
states with the highest vulnerability are also those with highest death rates. In Brazil, the first cases
emerged in the wealthiest states in the southeast and gradually spread to some of the poorest states in the
north and northeast. By calculating the evolution of the correlations, our indexes identify this path,
showing that the virus is progressively hitting harder the most vulnerable regions. This trend is observed
in the two COVID-19 multidimensional poverty indexes and the unidimensional monetary poverty index.
Interestingly, the index of monetary poverty shows the highest correlations in almost all of the
epidemiological weeks, which suggests that a lack of money is an immediate factor of vulnerability when
people face unexpected shocks and reinforces the importance of using both monetary and nonmonetary
indexes as complementary tools in a multidimensional poverty analysis.
Our third contribution is in using the fuzzy approach to overcome the limitation of standard poverty
measures, which treat poverty as a binary phenomenon (poor/non-poor). Using this approach enriches
the other two contributions. Fuzzy measures are more suitable for analyses at the subnational level and
for subgroups because they have smaller standard errors in the estimation of poverty and are better at
capturing mortality trends by showing higher rank correlations in most of the results.
The paper is organized as follows. Section 2 presents the empirical strategy, as well as the description
and sources of the data, and the scope of the indexes. Section 3 presents and discusses the results, and
Section 4 concludes.
2. Methodology and Data
2.1 Empirical Strategy
2.1.1 The Alkire-Foster Method
The most traditional measure of poverty is the headcount ratio (HCR), also known as the incidence
of poverty or poverty rate, which shows the percentage of people identified as poor. In this approach, by
defining a poverty line, the result is a dichotomic measure that splits the population into the poor and the
non-poor.
The Alkire-Foster methodology (AF), developed by Alkire and Foster (2009, 2011), goes beyond the
traditional approach by measuring multidimensional poverty based on its incidence (HCR) and intensity
(A). The latter is the average share of deprivation across individuals who are identified as poor. The
adjusted headcount ratio (M0), or multidimensional poverty index (MPI), is defined as the product of
incidence and intensity, M0 = HCR*A.
The identification of multidimensional poverty is calculated using the two-cutoff approach. The first
is the deprivation cutoff set for each variable. In this way, individuals can be identified as being deprived
in terms of a specific indicator, which means that we must define a deprivation threshold for each of the
variables. We apply the second cutoff by calculating the weighted sum of deprivation and classifying an
individual as poor if the resulting score is above the chosen poverty cutoff. Because the estimation of M0
is particularly well-suited to ordinal/binary data (Alkire and Foster, 2009), when applying the AF method,
we use our variables as ordinal indicators and transform the continuous variables into binary indicators.
2.1.2 The Fuzzy-Set Approach
Both the traditional monetary approach (HCR) and the MPI approach are based on deprivation cutoffs
(poverty lines), which treat poverty indicators as binary (poor/non-poor); instead, the fuzzy-set approach
treats poverty and multidimensional deprivation as matters of degree, determined in terms of the
individual’s position in the distribution of the monetary variable concerned (either income or
consumption expenditure) and other aspects of living conditions (Betti and Verma, 2008). The state of
deprivation is thus seen in the form of fuzzy sets, to which all members of the population belong in
varying degrees. In particular, within a determined poverty range, the approach uses membership
functions to identify the degree of certainty of individual poverty in a specific dimension (Alkire et al.,
2015).
The fuzzy-set approach was first proposed by Cerioli and Zani (1990) and developed by Cheli and
Lemmi (1995) in the so-called totally fuzzy and relative approach. Later, Betti et al. (2006) proposed the
integrated, fuzzy, and relative (IFR) approach, in which the membership function used for the fuzzy
monetary (FM) measure is defined as:
 


where is the cumulative distribution function for consumption expenditure, L is the corresponding
Lorenz curve, is the ranked individual sample weight, is individual consumption expenditure, and
is a parameter. The definition of the membership function is based on the monetary variable, in which
the alpha parameter is chosen such that the mean is “anchored” to the headcount ratio. The FM measure,
as defined previously, can also be applied in terms of the generalized Gini measures when we define α =
1.
In a multidimensional context, is an individual composite index, in which the weights of the single
indicators are not predetermined but, rather, follow the prevalence-correlations principles proposed by
Betti and Verma (2008). If the prevalence of an indicator is high, then its weight is low, and if correlations
with other vulnerable variables are high, then its weight is low. In this way, we determine appropriate
weights without the necessity of recurrence in potential arbitrary weight choices.
Another important advantage of fuzzy measures is that they are more informative and have smaller
standard errors (Betti et al., 2018). Therefore, fuzzy measures are more useful for subnational poverty
measures (Betti et al., 2012), which means that we can obtain poverty estimations for areas with relatively
small samples that are more statistically significant than those yielded by other measures.
The fuzzy approach and the AF method are complementary measures. The latter has the advantage
of providing intuitive measures that can be decomposed by population groups. In contrast, the former
has the advantage of overcoming the poor/non-poor dichotomy and enables more precise measures for
subnational regions.
2.2 Data
To construct the multidimensional indexes, we combine different publicly available sources. In this
section, we describe the data sources and the indicators.
2.2.1 Household Expenditure Survey
The primary source of data is the Brazilian Consumer Expenditure Survey (POF) for 2017-18, the
most recent round, released by the Brazilian Institute of Geography and Statistics (IBGE) on May 3,
2020. The POF is a high-quality household survey conducted to investigate the profile of consumption
expenditure, income, and living standards of Brazilian households. The data are widely used in poverty
and inequality research and have particular national importance because they are used to construct
consumption baskets in order to calculate official consumer price indexes.
The sample design of the POF is structured to cover the entire territory of the country; it is
representative in terms of the country, major regions, capitals, metropolitan regions, other parts of the
states, and urban or rural areas. The survey sample in 2017-18 totals 69,660 households, providing
information at the household and individual level.
The variables derived from the POF are Drinking water, an indicator that accounts for the
household’s frequency of supply, whether the household has running water, and the system of
distribution; Sanitation, which represents whether the household has at least one indoor bathroom with
shower and toilet, whether it is shared with other households, and whether it is connected to the public
sanitation system; Electricity, which represents whether the household has access to electricity and the
frequency of this access; Housing, which assesses the materials used in the household’s flooring, walls,
and roof; School meals, which, for households that have children who used to have daily free meals at
school, calculates how many children had access to this service, and how many meals per day; Share of
food consumption expenditure, as a proxy for a household’s food security; Overcrowded housing,
calculated as the number of residents per permanent bedroom in the household; Older adults per resident,
calculated as the number of people age 60 or more per number of members of the household who are
younger; Commuting time, which represents the number of members of the household who spend more
than an hour to get to work; Indoor air pollution by cooking fuel, which refers to the kind of fuel used by
the household for cooking; and Private insurance, which shows whether the individual has private health
insurance. The scores are presented in Table 2.
The remaining variables (described below) come from other sources and were merged with the most
possible disaggregated subnational level in the POF (state, capital, metropolitan region, or other parts of
the state). These variables are uniform across the population at the corresponding merged level.
Studies on COVID-19 stress that demographic and social variables matter when it comes to the
consequences of the pandemic (Oke and Heneghan, 2020; Souza et al., 2020). Table 1 shows
demographic and social characteristics estimated from the POF 2017-2018 dataset for the Brazilian
population
3
. Age and gender are two important factors in COVID-19 risk. According to the estimations
of Oke and Heneghan (2020), the case fatality rate is 66% higher for males older than 30 than for women
and 4.47 times higher for people age 80-89 than for those age 60-69. The median age in Brazil is 33 (the
mean is 35.26), while the median age across Europe Union members in 2018 range from 37.3 years in
Ireland to 46.3 years in Italy (Eurostat, 2019). This difference could reflect lower risk in terms of age in
Brazil, but other factors that affect risk remain to be proved.
For instance, as mentioned in the Introduction, few analyses are available about the impacts on ethnic
minorities. In Brazil, this is particularly important because ethnic minorities are at a relative disadvantage
in terms of the risk of infection. As we discuss in the next section, Indigenous people, who make up 0.4%
of Brazil’s population, predominantly live in regions with higher vulnerability to infection with COVID-
19 (22.7% live in the northern region) and have the highest vulnerability scores.
Table 1. Mean of Demographic and Social Variables
Variable
Mean
Gender
Women
51.61%
Men
48.39%
Color/ethnicity
White
44.00%
Black
10.22%
Asian
0.68%
Brown
44.42%
Indigenous
0.38%
Not identified
0.30%
Area type
Urban
85.26%
Rural
14.74%
Age in Years
35.26
Literacy ratio (>14 years)
92.41%
Years of education (>14 years)
9.37
Number of observations
178,431
2.2.2 Data on Access to Health Care and Risk Ratio by Age and Gender
The survey Area of Influence of Cities (REGIC) conducted by the IBGE in 2018 is used to provide
information on the distance that people need to travel from their city to other cities to access intensive-
care health services. To calculate the distance, we used geographic coordinates to measure the length in
kilometers of the shortest path between two cities. The final indicator is the mean for each POF
subnational level (capitals and other parts of the states) of the distance, weighted by the frequency of the
corresponding destination.
The data on the number of physicians and intensive-care hospital beds in the public health system is
available at the city level on the National Registry of Health Facilities (CNES) website. We used the
CNES data processed by the IBGE at the municipal level for December 2019, calculating the mean for
each POF subnational level. Both indicators are calculated per 1,000 people.
The risk ratio by age and gender was built based on the estimation of Oke and Heneghan (2020),
which use Italian data from the Italian National Institute of Health (ISS). The indicator sets the risk
reference score to the age between 60-69 and increases/decreases if the age is above/below this range.
2.2.3 Legal Measures of Social Distancing and Mobility Indexes
In Brazil, to date there has been no coordinated social distancing policy implemented at the national
level. The federal states and municipalities started to adopt measures to contain the spread of COVID-19
regardless of the decisions of the national government. However, these policies were implemented at
different times and in different ways. To capture the differences in the level of each state’s strictness, we
used the index of legal measures for social distancing developed by Moraes (2020a, 2020b). Moraes
considered all the decrees by state legislatures adopted April 6-24, 2020, to construct the index. The
measure considers the suspension or restriction of six types of activities: cultural, athletic, and religious;
bars and restaurants; non-essential services and business; non-essential industries; schools and
universities; and transportation. In this paper, the score was adapted to range from 0 (strict restrictions)
to 10 (no restrictions) (see Table 2).
For the mobility index, we used the Google Community Mobility Report from March 11, 2020, which
is the day on which the World Health Organization (WHO) declared COVID-19 a pandemic, to April
30, 2020. The Google indicator provides population-wide information on the relative change in mobility
in each state and in the following categories: retail stores and recreation, grocery stores and pharmacies,
parks, transit stations, workplaces, and residents. The change in mobility is the percentage change from
a baseline day before the pandemic. We use the mean of the changes in mobility in retail and recreation,
parks, transit stations, and workplace categories as a proxy for changes in behavior regarding daily
activities.
2.2.4 COVID-19 Indicators
The data on confirmed cases of COVID-19 and deaths from it are available on a daily basis on the
coronavirus website of the Ministry of Health (https://covid.saude.gov.br). The first confirmed case was
identified on February 25 and the first confirmed death on March 17, 2020. We collected statistics for
the states and capitals using official data from the Brasil.io (2020) website,
https://brasil.io/dataset/covid19/caso_full/. Based on the number of deaths confirmed as being due to
COVID-19 and the population estimated by the POF, we calculated the number of confirmed deaths per
one million people. It is important to stress that the official number of confirmed deaths from COVID-
19 underestimates the actual number, mostly due to limited testing.
2.2.5 Descriptive Statistics
Table 2 shows the score range and descriptive statistics for all the indicators used in the COVID-19
multidimensional poverty indexes. Figure 1 presents the correlations between each pair of indicators
calculated as Pearson coefficients. The heatmap is colored using a range from -1 (blue) to +1 (red). The
deprivations that are commonly explored in research on multidimensional povertysuch as having clean
drinking water, sanitation, electricity, housing, housing density, and indoor pollutionare all positively
correlated. The correlations of these variables with the share of expenditure on food, the distance from a
hospital, monetary poverty (measured by the household consumption expenditure per capita, $3.20 a day,
in 2018 purchasing power parity [PPP]), and COVID-19 deaths per million people are also positive.
However, they have a negative correlation with population density and indicators related to health-care
resources, such as private health insurance, physicians per 1,000 people, and intensive-care beds per
1,000 people. The correlations with the remaining variables are negative or near zero.
Table 2. Score range and descriptive statistics for the variables used in the COVID-19 multidimensional
poverty indexes
Variable
Mean
Standard Error
Min
Max
Drinking water
0.605
0.003
0
6
Sanitation
0.494
0.002
0
4
Electricity
0.056
0.001
0
4
Housing
1.027
0.003
0
9
School meals
0.476
0.003
0
16
Share of expenditure on
food
0.177
0.0003
0
1
Overcrowded housing
(residents per permanent
bedroom)
1.905
0.002
0.333
13
Older adults per
household
0.193
0.001
0
4
Commuting time
0.1501
0.001
0
4
Population density
(inhabitants per km2)
1261.859
5.827
0.673
8435.358
Index of legal measures of
social distancing
3.269
0.004
0.8
6.7
Mobility index (%
reduction from a baseline
day before the pandemic)
51.294
0.011
39
61.392
Risk ratio by age and
gender (1 is the risk
reference score set for age
60-69)
0.496
0.003
0
8.018
Indoor pollution due to
cooking fuel
0.010
0.0002
0
2
Private insurance
0.260
0.001
0
1
Distance from hospital (in
km)
30.503
0.101
0
606.544
Physicians per 1,000
people
1.174
0.002
0.365
4.695
Intensive-care hospital
beds per 1,000 people
0.441
0.001
0
3.01
Note: For ordinal variables, the score ranges are from no deprivation to total deprivation. The variable for private
insurance is the only binary variable, in which 0 means no insurance, and 1 means the person has insurance. The
continuous variables are identified as such.
(1) Drinking water
(2) Sanitation
(3) Electricity
(4) Housing
(5) School meals
(6) Share of food expenditure
(7) Overcrowded housing
(8) Older adults per resident
(9) Commuting time
(10) Population Density
(11) Ind. of legal meas. of social dist.
(12) Mobility Index
(13) Risk ratio by age and gender
(14) Indoor pollution
(15) Private insurance
(16) Distance from hospital
(17) Physicians per 1,000 p.
(18) Int. care hosp. beds per 1,000 p.
(19) Monetary poverty*
(20) COVID-19 deaths per mill.
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
-0.3
0
0.5
1
Figure 1. Pearson correlation matrix of the variables used in the COVID-19 multidimensional poverty
indexes, monetary poverty, and COVID-19 deaths per million people
*Monetary poverty is measured by the household consumption expenditure per capita ($3.20 a day, 2018 purchasing power
parity).
2.3 Multidimensional Poverty Indexes: Defining the Scope
This section proposes in detail the two COVID-19-related multidimensional poverty indexes
(CMPIs), comprising variables directly related to the capacity for preventing infection with COVID-19
and for recovery from it. Table 3 shows the dimensions and variables of each index, the definition of the
deprivation cutoffs and weights used in the AF method, and the resulting prevalence-correlation weight
scores for the fuzzy analysis.
Most of the cutoffs in the AF analysis were adapted from the United Nations sustainable development
goals and consider the Brazilian context and data availability. In the fuzzy application, to avoid
dichotomization of the variables and to obtain more information, we regard the variables as ordinal or
continuous indicators when possible. Only the indicator for private insurance is binary in both
approaches. With respect to the weights in the AF method, for simplicity, we follow the GMPI by
assuming that the dimensions are of equal weight. Also following the GMPI standard, for each
multidimensional index, we consider people vulnerable
4
(VN) to infection with COVID-19 if they are
deprived of at least one-third of the weighted indicators and consider people at severe risk (SR) of
infection with COVID-19 if they are deprived of at least half the weighted indicators.
Table 3. Structure of the COVID-19 multidimensional poverty indexes, variable cutoffs, and weights
Index
Dimension
Variable
AF Deprivation Cutoff
AF
Weight
Prevalence
correlation
weight
COVID-19 prevention
Hygiene
Drinking water
The household does not have daily access to
water, or does not have indoor running water, or
the water does not come from a public water
system.
0.100
0.103
Sanitation
The household does not have indoor bathroom
with shower and toilet, or the bathroom is shared
with other households, or the disposal of human
waste is not connected to a public sewage
system.
0.100
0.097
Staying at home
Electricity
The household has no access to electricity.
0.100
0.103
Housing
The households housing materials for at least
one floor, wall, and roof are inadequate.
0.100
0.097
Food security
School meals
One or more children in the household have
breakfast, lunch, snacks, or dinner free at school
every day.
0.100
0.094
Share of food
consumption
expenditure
Food represents 75% or more of the total
consumption expenditure of the household.
0.100
0.106
Household
density
Overcrowded
housing
There are three or more residents per permanent
bedrooms in the household.
0.100
0.098
Older adults per
household
Two or more older adults per members of a
household.
0.100
0.102
Public social
distancing
Commuting time
At least one individual in the household spends
more than an hour to get to work.
0.050
0.063
Population density
The household is in a region* where the pop.
density is higher than the mean of the Brazilian
capitals (> 2,700/km2)
0.050
0.062
Index of legal
measures of social
distancing
The household is in a state where the index is
higher than 2 (out of 10, which is the least
restrictive)
0.050
0.040
Mobility index
The household is in a state where the index is
less than 60% of the relative reduction in
mobility
0.050
0.035
Index
Dimension
Variable
AF Deprivation Cutoff
AF
Weight
Prevalence
correlation
weight
COVID-19 health recovery
Living standards
Electricity
The household has no access to electricity.
0.083
0.087
Housing
The household housing materials for at least one
of floor, wall, and roof are inadequate.
0.083
0.082
Overcrowded
housing
There are three or more residents per permanent
bedrooms in the household.
0.083
0.081
Risk groups
Risk ratio by age
and gender
The indicator is 1 or more;1 is the risk reference
score at age 60-69 (Oke and Heneghan, 2020). It
is an individual-level indicator.
0.125
0.120
Indoor air pollution
due to cooking fuel
The household’s cooking fuel is wood, oil,
kerosene, or another liquid fuel.
0.125
0.130
Healthy immune
system
Drinking water
The household does not have daily access to
water, or does not have indoor running water, or
the water does not come from a public water
system.
0.062
0.057
Sanitation
The household does not have indoor bathroom
with shower and toilet, or the bathroom is shared
with other households, or the disposal of human
waste is not connected to a public sewage
system.
0.062
0.060
School meals
One or more children in the household have
breakfast, lunch, snacks, or dinner free at school
every day.
0.062
0.063
Share of food
consumption
expenditure
Food represents 75% or more of the total
consumption expenditure of the household.
0.062
0.071
Access to health
care
Private insurance
The individual has no private health insurance.
0.062
0.054
Distance from
hospital
The household is in a region* where the
weighted mean distance from a hospital is more
than 100 km.
0.062
0.083
Physicians per
1,000 people
The household is in a region where the mean of
the indicator is less than 1 physician per 1,000
people.
0.062
0.063
Intensive-care
hospital beds per
1,000 people
The household is in a region* where the mean of
the indicator is less than 1 bed per 1,000 people.
0.062
0.050
Notes: AF deprivation cutoff refers to the description of the cutoff in the AF method. A cutoff definition is not necessary in
the fuzzy approach, because it does not treat the variables as binary measures. AF weights are the values of the weights used
in the AF method. Prevalence-correlation weights are the weights calculated in the fuzzy approach analysis.
* Capital, metropolitan region, or other parts of the state.
In Table 4, we present the number and percentage of deprived people in terms of all the variables
used in the two CMPIs. The indicator with the highest percentage of deprived people is the index that
measures mobility reduction, with 95.6% of the population deprived. This means that most of the
Brazilian population lives in states in which the mean reduction in daily mobility was less than 60%
(from March 11 to April 30). The lack of national coordination in social distancing measures, as the result
for the index of legal measures suggest (81.7% of people deprived), is one possible factor in the small
reduction in mobility. Moreover, the participation in protests opposing coronavirus lockdowns and the
continuous calls by the president to end social distancing is another possible factor that demotivated
people to decrease mobility.
By looking at the data, it is possible to observe that, independent of the pandemic context, a large
proportion of Brazilians do not have access to basic public services. The lack of public health-care
infrastructure is widespread. More than 83% of the population is deprived in terms of intensive-care
hospital beds, and the mean is 0.44 beds per 1,000 people (see Table 2 and Table 4). Moreover, 51% of
the population has access to less than one physician per 1,000 population. In some states, the private
health sector offers proportionately more physicians and hospital beds. In any case, although only 26%
of the population has private insurance, it does not mean that they will have access to all the private
health infrastructure, only to the hospitals and services specified by their contract
5
. It is also evident that
for several people the access to sanitation and drinking water is inadequate: 30.2% and 39.1%,
respectively. Clean water is crucial for preventing infection with COVID-19 because it is needed for
frequent and sufficient hand washing (WHO, 2020b) and is essential for human health and well-being
(WHO, 2019a). Moreover, improper sanitation is a major cause of infectious disease (WHO, 2018b,
2020b) and can compromise the immune system, with a possible impact on recovery from COVID-19.
Table 4. Number and percentage of deprived people per indicator
Variable
AF Pop. Deprived
AF % Deprived
Drinking water
62,511,394
30.18%
Sanitation
80,970,486
39.09%
Electricity
457,742
0.22%
Housing
22,155,805
10.69%
School meals
47,170,894
22.78%
Share of food consumption expenditure
283,702
0.14%
Overcrowded housing (residents per
permanent bedroom)
26,313,116
12.71%
Older adults per household
6,666,591
3.22%
Commuting time
26,390,282
12.74 %
Population density (people per km2)
34,501,439
16.66%
Index of legal measures of social distancing
169,164,425
81.68%
Mobility index (% reduction from a baseline
day before the pandemic)
198,056,358
95.63%
Risk ratio by age and gender (1 is the risk
reference score set for age 60-69)
31,702,592
15.31%
Indoor pollution due to cooking fuel
2,094,513
1.01%
Private insurance
153,306,719
74.02%
Distance from hospital (in km)
8,187,121
3.95%
Physicians per 1,000 people
105,612,592
51.00%
Intensive-care hospital beds per 1,000 people
172,871,751
83.47%
Total population
207,103,790
Note: AF Pop. Deprived and AF % Deprived refers, respectively, to the number and incidence of deprived people using the
cutoff defined in the Alkire-Foster (AF) model.
3. Results
3.1 Multidimensional Poverty Analysis
This section presents the results for both the AF and fuzzy approaches. For the AF method, Table 5
shows the outcomes for the multidimensional headcount ratio (HCR), the poverty intensity (A), and the
adjusted headcount ratio (M0). In addition, the results for each CMPI are shown for VN and SR. For
group decompositions, we use only the HCR indicator, as it is more intuitive and the comparison with
the fuzzy measure is more appropriate.
According to the AF and fuzzy results, between 16.2% and 15.6% of the population is vulnerable to
infection with COVID-19 as measured by the prevention index. This implies that between 32.4 million
and 33.5 million people cannot implement proper prevention measures related to at least one-third of the
weighted indicators. Severe risk, which represents deprivation in half the weighted indicators, is 3.4% in
the AF and 4.1% in the fuzzy results, respectively. In the health recovery index, the two approaches
diverge more. The AF estimates 19.8% of people are vulnerable (41 million people), while the fuzzy
estimate is 13.7% (28.4 million people). In terms of severe risk, the results are 2.2% for AF and 2.6% for
fuzzy.
Table 5. COVID-19 multidimensional poverty indexes per approach and indicators
AF
Fuzzy
Index
M0
VN
HCR
VN
A
VN
M0
SR
HCR
SR
A
SR
VN
SR
Prevention
0.068
16.17%
0.419
0.018
3.43%
0.528
15.64%
4.11%
Health recovery
0.081
19.81%
0.409
0.012
2.15%
0.555
13.72 %
2.60 %
Table 6 presents the results for each state, showing the number of confirmed COVID-19 deaths per
million people and the share of the monetarily poor people measured by household consumption
expenditure per capita ($3.20 a day, 2018 PPP). Alternatively, Figure 2 illustrates in maps the distribution
of confirmed deaths per million, FM poverty, and fuzzy vulnerability for the health recovery index by
state. The outcomes demonstrate the vast regional inequality in Brazil. The northern and northeastern
regions have the highest proportion of vulnerability and severe risk of infection with COVID-19. For
instance, Amazonas state (AM) has the most deaths per million people and among the highest risk: the
incidence of vulnerable people according to the health recovery index is 50.5%. By comparison, in São
Paulo (SP), the state with the most infections in absolute terms, it is 3.9%.
Table 6. COVID-19 death indicator and estimation results for the COVID-19 multidimensional poverty
indexes and unidimensional monetary poverty by state
State
COVID-
19
Prevention Index
Health Recovery Index
Monetary Poverty
Death
per
million
HCR
VN
HCR
SR
Fuzzy
VN
Fuzzy
SR
HCR
VN
HCR
SR
Fuzzy
VN
Fuzzy
SR
HCR
Fuzzy
RO
39.70
36.74%
7.28%
45.72%
2.47%
35.10%
1.43%
16.04%
1.55%
14.28%
14.01%
AC
69.61
52.05%
24.79%
48.19%
22.99%
58.45%
24.64%
44.02%
14.90%
16.42%
16.06%
AM
353.17
45.14%
17.48%
27.34%
11.43%
50.45%
16.18%
27.52%
7.52%
25.48%
23.60%
RR
96.74
26.81%
8.43%
16.17%
2.78%
48.12%
7.72%
19.75%
3.26%
21.00%
20.15%
PA
144.91
41.55%
14.10%
48.98%
17.19%
44.20%
6.22%
32.34%
7.78%
22.13%
20.16%
AP
132.70
41.71%
12.94%
24.69%
6.31%
51.63%
9.02%
25.81%
4.51%
11.77%
12.13%
TO
17.59
20.35%
5.70%
34.35%
4.67%
34.65%
2.47%
13.64%
2.65%
21.57%
19.74%
MA
78.61
44.37%
14.78%
36.19%
18.26%
50.13%
9.14%
35.86%
13.28%
22.34%
21.45%
PI
22.08
29.03%
8.58%
18.84%
9.61%
51.83%
13.79%
23.05%
7.92%
14.68%
15.46%
CE
178.39
8.13%
0.92%
11.89%
4.45%
29.17%
3.56%
21.28%
5.43%
22.81%
20.36%
RN
39.23
29.09%
5.54%
21.27%
5.69%
31.51%
2.74%
22.68%
3.97%
10.81%
10.86%
PB
46.16
23.93%
3.76%
28.85%
9.59%
27.57%
1.92%
25.80%
5.58%
23.89%
21.39%
PE
155.20
20.73%
1.19%
17.51%
5.46%
25.59%
2.10%
22.91%
5.45%
17.93%
16.68%
AL
60.12
22.93%
1.78%
21.34%
5.31%
31.26%
2.44%
25.14%
3.48%
31.07%
27.45%
SE
23.36
15.39%
1.71%
9.12%
3.82%
22.80%
2.19%
13.59%
2.59%
10.02%
10.24%
BA
19.37
19.56%
4.32%
28.92%
8.05%
26.71%
2.34%
20.78%
5.21%
15.71%
14.88%
MG
7.15
8.32%
1.38%
8.64%
1.17%
13.97%
0.88%
9.82%
0.85%
5.70%
6.23%
ES
68.65
12.22%
1.74%
10.15%
0.98%
15.23%
0.31%
8.31%
0.43%
7.68%
8.38%
RJ
152.77
15.79%
2.45%
8.57%
1.36%
9.50%
0.04%
7.77%
0.39%
8.25%
8.86%
SP
103.37
8.17%
1.47%
6.67%
1.20%
3.89%
0.01%
6.19%
0.28%
2.56%
3.65%
PR
11.15
9.26%
1.00%
16.74%
2.22%
12.78%
0.37%
7.61%
0.97%
7.12%
7.40%
SC
11.54
8.04%
0.08%
2.71%
0.38%
16.50%
0.55%
7.92%
0.70%
2.81%
3.42%
RS
12.23
10.22%
1.50%
5.69%
0.51%
16.29%
0.79%
7.38%
0.63%
2.37%
3.22%
MS
5.63
14.47%
2.26%
37.39%
4.00%
21.29%
0.59%
7.69%
0.96%
5.23%
5.64%
MT
8.01
23.32%
5.63%
36.31%
4.54%
48.58%
6.58%
13.79%
0.93%
7.13%
7.51%
GO
10.03
10.18%
0.59%
6.39%
0.40%
19.27%
0.54%
7.79%
0.32%
4.97%
5.76%
DF
18.96
13.66%
1.47%
5.50%
0.83%
7.62%
0.03%
3.85%
0.28%
3.74%
3.66%
Total
75.79
16.17%
3.43%
15.64%
4.11%
19.81%
2.15%
13.72%
2.60%
9.96%
9.96%
Notes: Covid-19 deaths per million as of May 10, 2020.
Northern region: RO = Rondônia; AC = Acre; AM = Amazonas; RR = Roraima; PA = Pará; AP = Amapá; TO = Tocantins.
Northeastern region: MA = Maranhão; PI = Piauí; CE = Ceará; RN = Rio Grande do Norte; PB = Paraíba; PE = Pernambuco;
AL = Alagoas; SE = Sergipe; BA = Bahia.
Southeastern region: MG = Minas Gerais; ES = Espírito Santo; RJ = Rio de Janeiro; SP = São Paulo.
Southern region: PR = Paraná; SC = Santa Catarina; RS = Rio Grande do Sul.
Central-western region: MS = Mato Grosso do Sul; MT = Mato Grosso; GO = Goiás; DF = Distrito Federal.
The risk of infection with COVID-19 also differs among ethnic groups. Table 7 shows the indexes
for each group. Overall, the picture is not favorable for the Indigenous population, which has the worst
conditions in all indicators. Brown and Black groups also are at a disadvantage to the White and Asian
groups and the total population.
The fact that the virus is spreading toward the northern region is an additional concern, as it is the
region with the highest proportion of the population that is vulnerable and at severe risk of infection with
Figure 2. Distribution of COVID-19 confirmed deaths per million population, fuzzy monetary poverty,
and fuzzy vulnerability for the COVID-19 health recovery index by state
Note: For abbreviations, see note to Table 6.
COVID-19, and, as mentioned previously, where 22.7% of the Brazilian Indigenous population is
concentrated.
Table 7. Estimation results for the COVID-19 multidimensional poverty indexes and unidimensional
monetary poverty by color/ethnicity
Color/Ethnicity
Prevention Index
Health Recovery Index
Monetary Poverty
HCR
VN
HCR
SR
Fuzzy
VN
Fuzzy
SR
HCR
VN
HCR
SR
Fuzzy VN
Fuzzy
SR
HCR
Fuzzy
White
10.57%
1.69%
9.80%
1.84%
13.49%
0.94%
9.53%
1.24%
5.25%
5.79%
Black
19.54%
4.23%
19.12%
5.01%
21.52%
2.35%
14.95%
3.11%
13.06%
12.31%
Asian
6.71%
0.89%
5.72%
0.73%
7.31%
0.47%
6.98%
0.84%
4.19%
4.61%
Brown
20.96%
4.96%
20.71%
6.20%
25.88%
3.28%
17.67%
3.85%
13.98%
13.62%
Indigenous
34.98%
9.46%
28.64%
7.43%
28.58%
8.88%
23.33%
6.61%
14.97%
13.72%
Not identified
10.60%
0.99%
8.87%
0.39%
8.33%
0.03%
5.68%
0.10%
8.21%
9.40%
Total
16.17%
3.43%
15.64%
4.11%
19.81%
2.15%
13.72%
2.60%
9.96%
9.96%
3.2 The Link between the Multidimensional Poverty Indexes and COVID-19 Deaths
The first confirmed cases were in São Paulo, the richest state and one of the least vulnerable to
infection with COVID-19 according to the CMPI. At the beginning of March, it was estimated that 85.3%
of the transmission came from outside the country, with 54.8% probably coming from travelers infected
in Italy, 9.3% in China, and 8.3% in France (Candido et al., 2020). This suggests that in Brazil the initial
infection was concentrated among the middle and upper classes (who could afford to fly outside the
country). Later data show the spread of the virus to the most vulnerable regions (see Table 6 and Figure
2).
In this section, we propose an innovative analysis, measuring rank correlations at the state level for
the unidimensional and multidimensional poverty indexes and the number of COVID-19 deaths in Brazil
per million people for many consecutive weeks. We used Spearman and Kendall rank correlation
coefficients; for each subgroup, the latter calculates the average of the square difference between the two
ranks, and the former is based on the difference in the number of pairs that do and do not match. The
coefficients are between -1 and +1; an extreme value means that the rankings are perfectly associated
either negatively or positively. Even if the analysis does not imply causality, our indexes can be tested
by determining the evolution in the correlation throughout over the epidemiological weeks.
The results are given for week 12 (April 5) to week 20 (May 10). Figure 3 and Figure 4 illustrate the
results for the HCRs of vulnerability and severe risk (blue and orange lines) and the fuzzy indicators of
vulnerability and severe risk (gray and yellow lines). Figure 5 plots outcomes for the HCR of monetary
poverty (green line) and FM poverty (light blue line).
a. Spearman b. Kendall
Figure 3. The evolution per week in the rank correlations between the prevention index and deaths per
million people at the state level
Overall, the outcomes confirm that our indexes capture the trend in infection from the richest regions,
which are less vulnerable, to the more vulnerable regions. Initially, all the measures show a negative
relation between the vulnerability indicators and deaths per million people. Beginning in week 14, the
correlations increase, meaning that states with the highest vulnerability and severe risk scores have
increasing numbers of deaths per million people.
-0,8
-0,6
-0,4
-0,2
0
0,2
0,4
0,6
week
12 week
13 week
14 week
15 week
16 week
17 week
18 week
19 week
20
Correlation
HCR VN SR VN
Fuzzy VN Fuzzy SR
-0,5
-0,4
-0,3
-0,2
-0,1
0
0,1
0,2
0,3
0,4
week
12 week
13 week
14 week
15 week
16 week
17 week
18 week
19 week
20
Correlation
HCR VN SR VN
Fuzzy VN Fuzzy SR
The health recovery index has greater correlations than the prevention index. The highest Spearman
coefficients in the prevention index are 0.41 and 0.44 (for HCR vulnerability and fuzzy severe risk),
whereas for the health recovery index they are 0.46 and 0.53 (for fuzzy severe risk and fuzzy
vulnerability). In addition, most of the results suggest that fuzzy measures are more appropriate for
explaining the link with deaths because they show the highest correlations in Figure 4 and Figure 5.
a. Spearman b. Kendall
Figure 4. The evolution per week in rank correlations between the health recovery indexes and deaths
per million people at the state level
Finally, beginning in week 14, the correlations are predominantly steeper and higher for
unidimensional monetary poverty than for the other two indexes. One possible interpretation is that
monetary poverty is an immediate factor of vulnerability and risk to shocks. In times of difficulty, money
seems to be the first thing that plays an essential role in addressing the increasing threat from the
pandemic. Because vulnerability is a multidimensional phenomenon, these results reinforce that
nonmonetary and monetary variables are complementary indicators.
-0,6
-0,4
-0,2
0
0,2
0,4
0,6
week
12 week
13 week
14 week
15 week
16 week
17 week
18 week
19 week
20
Correlation
HCR VN SR VN
Fuzzy VN Fuzzy SR
-0,5
-0,4
-0,3
-0,2
-0,1
0
0,1
0,2
0,3
0,4
week
12 week
13 week
14 week
15 week
16 week
17 week
18 week
19 week
20
Correlation
HCR VN SR VN
Fuzzy VN Fuzzy SR
a. Spearman b. Kendall
Figure 5. The evolution per week in rank correlations between the unidimensional monetary poverty
indicators and deaths per million people at the state level
4. Concluding Remarks
This paper contributes to the literature on the potential social impacts of the COVID-19 pandemic.
We use the AF method and fuzzy-set approach as complements to measure multidimensional poverty
within the context of the coronavirus pandemic in Brazil. We propose two multidimensional poverty
indexes to account for the vulnerability related to the ability to prevent the spread of infection and to
recover from infection with the virus.
The data reveal structural deprivations in the country due to the fact that a big part of the population
cannot fully implement the recommended preventive measures and because the social conditions and the
health-care system do not meet the basic requirements for avoiding preventable deaths. Moreover, the
estimations of the indexes illustrate the considerable inequality among regions and ethnic groups. This
is in line with the extensive literature on inequality in Brazil.
Two of the innovations in this paper are presenting pandemic-specific indexes and proposing a rank
correlation analysis that can trace the increasing spread of infection and higher mortality rate in
-0,4
-0,3
-0,2
-0,1
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
week
12 week
13 week
14 week
15 week
16 week
17 week
18 week
19 week
20
Correlation
HCR Monetary Fuzzy Monetary
-0,3
-0,2
-0,1
0
0,1
0,2
0,3
0,4
0,5
week
12 week
13 week
14 week
15 week
16 week
17 week
18 week
19 week
20
Correlation
HCR Monetary Fuzzy Monetary
vulnerable regions. Most of the correlations increase weekly, which means that most states with the
highest vulnerability and severe risk outcomes also have the largest increase in death rates. The monetary
poverty indicator has the highest correlation when compared with the two CPMIs for almost all the
epidemiological weeks. This indicates that money is very important in battling the threat of the pandemic
and that nonmonetary and monetary indexes are complementary variables, rather than competing
variables, in multidimensional poverty analysis.
Another innovation is the application of fuzzy measures, which are more appropriate for the
characteristics of the vulnerability variables because they avoid a binary split between deprivation and
non-deprivation, have more precise measures in subnational analysis, and have higher rank correlation.
Despite the limitations of the data on confirmed deaths from COVID-19, our empirical evidence
offers an additional warning that responses to the pandemic need to prioritize the most vulnerable groups,
and our analysis reinforces the need for coordinated national action. In the short run, rapid measures are
needed to stop the virus from spreading, to ensure that the entire population follows the recommendations
for prevention as well as they can, and to guarantee universal coverage by public health services. In the
medium and long run, this analysis reinforces the urgent necessity of public policies that promote health,
adequate housing, and sanitation.
Acknowledgments
We are grateful to Francesca Bettio, Antonella D'Agostino, Ugo Pagano, Ana Cecília Kreter, Isabel
Sofia Fanciulli, Thiago D. Oliveira, Kamaiaji de S. Castor, Guilherme S. Morlin, Chandni Dwarkasing,
Marwil J. Dávila-Fernández, Henrique C. de Mesquita, Mario A. M. P. de Almeida, the Associate Editor
Christelle Dumas, and two anonymous referees for their helpful suggestions and comments.
References
Alkire, S., & Foster, J. (2011). Counting and multidimensional poverty measurement. Journal of Public
Economics, 95(7-8), 476-487.
Alkire, S., & Foster, J. (2009). Counting and multidimensional poverty measurement (OPHI Working
Paper No. 32). Oxford Poverty and Human Development Initiative, University of Oxford.
Alkire, S., Roche, J. M., Ballon, P., Foster, J., Santos, M. E., & Seth, S. (2015). Multidimensional poverty
measurement and analysis. New York: Oxford University Press.
Alkire, S., Dirksen, J., Nogales, R., & Oldiges, C. (2020). Multidimensional poverty and COVID-19 risk
factors: A rapid overview of interlinked deprivations across 5.7 billion people, OPHI Briefing 53,
Oxford Poverty and Human Development Initiative, University of Oxford.
Baqui, P., Bica, I., Marra, V., Ercole, A., & Schaar, M. V. (2020). Ethnic and regional variations in
hospital mortality from COVID-19 in Brazil: A cross-sectional observational study. Lancet Global
Health, 8(8).
Betti, G., Cheli, B., Lemmi, A., & Verma, V. (2006). Multidimensional and longitudinal poverty: An
integrated fuzzy approach, in Lemmi, A., & Betti, G. (eds.), Fuzzy Set Approach to
Multidimensional Poverty Measurement. New York: Springer, 111-137.
Betti, G., Gagliardi, F., & Verma, V. (2018). Simplified Jackknife variance estimates for fuzzy measures
of multidimensional poverty. International Statistical Review, 86(1), 68-86.
Betti, G., Gagliardi, F., Lemmi, A., & Verma, V. (2012). Sub-national indicators of poverty and
deprivation in Europe: methodology and applications. Cambridge Journal of Regions, Economy
and Society, 5, 149-162.
Betti, G., & Verma, V. (2008). Fuzzy measures of the incidence of relative poverty and deprivation: a
multi-dimensional perspective. Statistical Methods and Applications, 17(2), 225-250.
Brasil.io. (2020). COVID-19: Boletins informativos e casos do coronavírus por município por dia.
Retrieved August 2020, from https://brasil.io/dataset/covid19/caso_full/.
Candido, D. D. S., Watts, A., Abade, L., Kraemer, M. U., Pybus, O. G., Croda, J., ... & Faria, N. R.
(2020). Routes for COVID-19 importation in Brazil. Journal of Travel Medicine, 27(3), taaa042.
Cerioli, A., & Zani, S. (1990). A fuzzy approach to the measurement of poverty, in Dagum, C., & Zenga,
M. (eds.), Income and Wealth Distribution, Inequality and Poverty, Studies in Contemporary
Economics. Berlin: Springer, 272-284.
Cheli, B., & Lemmi, A. (1995). A totally fuzzy and relative approach to the multidimensional analysis
of poverty, Economic Notes, 24, 115-134.
Chu, D. K., Akl, E. A., Duda, S., Solo, K., Yaacoub, S., Schünemann, H. J., . . . Reinap, M. (2020).
Physical distancing, face masks, and eye protection to prevent person-to-person transmission of
SARS-CoV-2 and COVID-19: A systematic review and meta-analysis. The Lancet, 395(10242),
1973-1987.
DATASUS. (2020). Coronavírus Brasil. Retrieved August 2020, from https://covid.saude.gov.br.
Eurostat. (2019). Median age over 43 years in the EU. Retrieved August, 2020, from
https://ec.europa.eu/eurostat/web/products-eurostat-news/-/DDN-20191105-1
Fernandes, G. A. D. A. L. (2017). Is the Brazilian tale of peaceful racial coexistence true? Some evidence
from school segregation and the huge racial gap in the largest Brazilian city. World Development,
98, 179-194.
Hoffman, R. (2018). Changes in income distribution in Brazil. In Amann, E., Azzoni, C., & Baer W.
(Authors), The Oxford Handbook of the Brazilian Economy (pp. 467-488). New York: Oxford
University Press.
Holtgrave, D. R., Barranco, M. A., Tesoriero, J. M., Blog, D. S., & Rosenberg, E. S. (2020). Assessing
racial and ethnic disparities using a COVID-19 outcomes continuum for New York State. Annals
of Epidemiology, 48, 9-14.
InfoAmazon. (2020). Distantes de UTIs e respiradores, indígenas da Amazônia tentam se blindar do
vírus. Retrieved May 20, 2020, from https://infoamazonia.org/pt/2020/05/distantes-de-utis-e-
respiradores-indigenas-da-amazonia-tentam-se-blindar-do-virus/
Khalatbari-Soltani, S., Cumming, R. G., Delpierre, C., & Kelly-Irving, M. (2020). Importance of
collecting data on socioeconomic determinants from the early stage of the COVID-19 outbreak
onwards. Journal of Epidemiology and Community Health, 74, 620-623.
Kraemer, M. U., Yang, C. H., Gutierrez, B., Wu, C. H., Klein, B., Pigott, D. M., ... & Brownstein, J. S.
(2020). The effect of human mobility and control measures on the COVID-19 epidemic in China.
Science, 368(6490), 493-497.
Lancet (2020). COVID-19 in Brazil: So what? (Editorial). The Lancet, 395 (10235), 1461.
Laurencin, C. T., & Mcclinton, A. (2020). The COVID-19 pandemic: A call to action to identify and
address racial and ethnic disparities. Journal of Racial and Ethnic Health Disparities, 7(3), 398-
402.
Lemmi, A., & Betti, G., eds. (2006). Fuzzy set approach to multidimensional poverty measurement. New
York: Springer.
Lusignan, S. D., Dorward, J., Correa, A., Jones, N., Akinyemi, O., Amirthalingam, G., . . . Hobbs, F. D.
(2020). Risk factors for SARS-CoV-2 among patients in the Oxford Royal College of General
Practitioners Research and Surveillance Centre primary care network: A cross-sectional study.
Lancet Infectious Diseases.
Massuda, A., Hone, T., Leles, F. A., Castro, M. C., & Atun, R. (2018). The Brazilian health system at
crossroads: Progress, crisis and resilience. BMJ Global Health, 3(4), e000829.
Millett, G. A., Jones, A. T., Benkeser, D., Baral, S., Mercer, L., Beyrer, C., . . . Sullivan, P. S. (2020).
Assessing differential impacts of COVID-19 on black communities. Annals of Epidemiology, 47,
37-44.
Moraes, R.F.D. (2020a). Medidas legais de incentivo ao distanciamento social: comparação das políticas
de governos estaduais e prefeituras das capitais no Brasil. Nota Técnica n.16, Instituto de Pesquisa
Econômica Aplicada (IPEA).
Moraes, R.F.D. (2020b). COVID-19 e medidas legais de distanciamento social: tipologia de políticas
estaduais e análise do período de 13 a 26 de abril de 2020. Nota Técnica n.18, Instituto de Pesquisa
Econômica Aplicada (IPEA).
Oke, J., & Heneghan, C. (2020). Global COVID-19 case fatality rates. Centre for Evidence-Based
Medicine. Retrieved April 2020, from https://www.cebm.net/covid-19/global-covid-19-case-
fatality-rates/.
OPHI & UNDP (2019). Global multidimensional poverty index 2019: illuminating inequalities. United
Nations Development Programme and Oxford Poverty and Human Development Initiative.
Retrieved April 2020, from https://ophi.org.uk/wp-content/uploads/G-
MPI_Report_2019_PDF.pdf.
Paim, J., Travassos, C., Almeida, C., Bahia, L., & Macinko, J. (2011). The Brazilian health system:
History, advances, and challenges. The Lancet, 377(9779), 1778-1797.
Pareek, M., Bangash, M. N., Pareek, N., Pan, D., Sze, S., Minhas, J. S., . . . Khunti, K. (2020). Ethnicity
and COVID-19: An urgent public health research priority. The Lancet, 395(10234), 1421-1422.
Raupp, L., Fávaro, T. R., Cunha, G. M., & Santos, R. V. (2017). Condições de saneamento e
desigualdades de cor/raça no Brasil urbano: Uma análise com foco na população indígena com
base no Censo Demográfico de 2010. Revista Brasileira De Epidemiologia, 20(1), 1-15.
Rubin, D., Huang, J., Fisher, B. T., Gasparrini, A., Tam, V., Song, L., . . . Tasian, G. (2020). Association
of Social Distancing, Population Density, and Temperature with the Instantaneous Reproduction
Number of SARS-CoV-2 in Counties Across the United States. JAMA Network Open, 3(7),
e2016099.
Souza, W. M., Buss, L. F., Candido, D. D., Carrera, J., Li, S., Zarebski, A. E., . . . Faria, N. R. (2020).
Epidemiological and clinical characteristics of the COVID-19 epidemic in Brazil. Nature Human
Behaviour.
WHO (2018a). Household air pollution and health (World Health Organization Fact sheets). Retrieved
April, 2020, from https://www.who.int/news-room/fact-sheets/detail/household-air-pollution-and-
health/.
WHO (2018b). WHO Housing and health guidelines. Geneva: World Health Organization; 2018.
Licence: CC BY-NC-SA 3.0 IGO.
WHO (2019a). Water, sanitation, hygiene and health: a primer for health professionals. Geneva: World
Health Organization, WHO reference number: (WHO/CED/PHE/WSH/19.149).
WHO (2019b). Drinking-water (World Health Organization Fact sheets). Retrieved April 2020, from
https://www.who.int/news-room/fact-sheets/detail/drinking-water/.
WHO (2020a). Coronavirus disease (COVID-19) advice for the public. Retrieved May 26, 2020, from
https://www.who.int/emergencies/diseases/novel-coronavirus-2019/advice-for-public/.
WHO (2020b). Water, sanitation, hygiene, and waste management for the COVID-19 virus. World
Health Organization, WHO reference number: WHO/2019-nCoV/IPC_WASH/2020.3. Retrieved
May, 2020, from https://www.who.int/publications-detail/water-sanitation-hygiene-and-waste-
management-for-the-covid-19-virus-interim-guidance/.
Yancy, C. W. (2020). COVID-19 and African Americans. JAMA, 323(19), 1891.
Zhao, H., Harris, R. J., Ellis, J., & Pebody, R. G. (2015). Ethnicity, deprivation and mortality due to 2009
pandemic influenza A(H1N1) in England during the 2009/2010 pandemic and the first post-
pandemic season. Epidemiology and Infection, 143(16), 3375-3383.
Notes
1
InfoAmazonia (2020) estimates that, in the Amazon forest region, the Indigenous tribes live on average about 315 km away
from public hospitals equipped with intensive-care departments.
2
The GMPI indicators are nutrition, child mortality, years of schooling, school attendance, cooking fuel, sanitation, drinking
water, electricity, housing, and assets.
3
The color/ethnicity classification follows the POF/IBGE, in which the individuals in the survey declared their race identity
without any influence from the interviewer. The categories are White, Black, Yellow (people that claimed to have Asian
origin), Brown (people that claimed to be parda, mulata, cabocla, cafuza, mameluca, or black mixed-race), indigenous, and
not identified (not declared). For an ethnic background, a discussion about race as a social construction, and segregation in
Brazil, see Fernandes (2017).
4
In GMPI deprivation, the cutoff for being considered poor is deprivation of one-third of the ten indicators. In the GMPI, the
term “vulnerable” is used differently: a person is considered vulnerable to poverty if he/she is deprived of between one-fifth
and one-third of the indicators.
5
Brazil’s health-care is provided by both public and private sectors, and people can use the two sectors depending on
accessibility and ability to afford costs. The public health system, through the Unified Health System (SUS), aims to offer
universal, free of charge, health service provision. The private sector offers services mainly through health plans, insurance
premiums, out-of-pocket payments, and provides services for the SUS. For more details, see Paim et al. (2011) and Massuda
et al. (2018).
... Além disso, Palmas-TO teve o maior número de casos de COVID-19 na faixa etária de 20 a 59 anos, pois é um grupo com maior número de pessoas que frequentam bares, shoppings e podem ter maiores compromissos presenciais de trabalho, em conformidade com o estudo realizado pela Fiocruz, que, ao comparar dados da semana epidemiológica 1 (3 a 9 de janeiro) e da semana epidemiológica 20 (16 a 22 de maio), mostrou um aumento de 172,22% no número de casos de COVID-19 entre pessoas com idade entre 20 e 29 anos 17 . Ainda, em relação à dengue, a realidade do município de Palmas também reflete a realidade do Brasil quando se diz respeito à faixa etária com maior incidência de dengue, uma vez que, de acordo com o Ministério da Saúde, ela concentrava-se entre 20-59 anos, sendo predominante entre 20-29 anos, entre as semanas epidemiológicas 1 e 53 17,18 . ...
... Além disso, Palmas-TO teve o maior número de casos de COVID-19 na faixa etária de 20 a 59 anos, pois é um grupo com maior número de pessoas que frequentam bares, shoppings e podem ter maiores compromissos presenciais de trabalho, em conformidade com o estudo realizado pela Fiocruz, que, ao comparar dados da semana epidemiológica 1 (3 a 9 de janeiro) e da semana epidemiológica 20 (16 a 22 de maio), mostrou um aumento de 172,22% no número de casos de COVID-19 entre pessoas com idade entre 20 e 29 anos 17 . Ainda, em relação à dengue, a realidade do município de Palmas também reflete a realidade do Brasil quando se diz respeito à faixa etária com maior incidência de dengue, uma vez que, de acordo com o Ministério da Saúde, ela concentrava-se entre 20-59 anos, sendo predominante entre 20-29 anos, entre as semanas epidemiológicas 1 e 53 17,18 . ...
Article
Full-text available
Introdução: A COVID-19 é causada pelo novo coronavírus SARS-CoV-2. A dengue é uma arbovirose com sintomas semelhantes à COVID-19 e, portanto, são doenças que geram confusão e podem ser negligenciadas. Objetivos: analisar casos de infecções pelos vírus da Dengue e de COVID-19, em Palmas-TO. Metodologia: Realizou-se um estudo quantitativo descritivo com dados secundários de Dengue e COVID-19, em Palmas-TO, em 2020 e 2021. Resultado: A incidência de COVID-19 foi de 130 casos/1000 habitantes e a de dengue foi de 3,28 casos/1000 habitantes, maior na faixa etária de 20 a 59 anos. A mortalidade por COVID-19 foi de 1,6 óbitos/1000 habitantes e, a letalidade, 0,98%. Discussão: Os maiores números de casos confirmados de COVID-19 em Palmas- TO ocorreram em agosto de 2020 e em março de 2021, mesmo com seis meses de diferença e políticas adotadas para contenção da pandemia. No período de maior incidência de COVID-19, a dengue teve registro de menor número de casos, provavelmente por subnotificação de casos da dengue, que corrobora com estudos em outros municípios. A análise de casos concomitantes das duas doenças é importante para entender a correlação entre elas. Considerações Finais: Os achados do estudo permitem avaliar o perfil endêmico de dengue em meio à pandemia por COVID-19, bem como analisar a circulação dos dois vírus, diagnósticos e notificações.
... Palmas-TO had the highest number of COVID-19 cases in the 20-59 age group, as it is a group with the largest number of people who frequent bars, malls and may have greater face-to-face work commitments, in accordance with the study carried out by FIOCruz, which, when comparing data from epidemiological week 1 (January 3 to 9) and epidemiological week 20 (May 16 to 22), showed an increase of 172.22% in the number of COVID cases -19 among people aged between 20 and 29 years 17 . Still, in relation to dengue, the reality of the municipality of Palmas also reflects the reality of Brazil when it comes to the age group and sex with the highest incidence of dengue, since, according to the Ministry of Health, it is concentrated in 20-59 years group, being predominant between 20-29 years, between epidemiological weeks 1 and 53 17,18 . ...
... Palmas-TO had the highest number of COVID-19 cases in the 20-59 age group, as it is a group with the largest number of people who frequent bars, malls and may have greater face-to-face work commitments, in accordance with the study carried out by FIOCruz, which, when comparing data from epidemiological week 1 (January 3 to 9) and epidemiological week 20 (May 16 to 22), showed an increase of 172.22% in the number of COVID cases -19 among people aged between 20 and 29 years 17 . Still, in relation to dengue, the reality of the municipality of Palmas also reflects the reality of Brazil when it comes to the age group and sex with the highest incidence of dengue, since, according to the Ministry of Health, it is concentrated in 20-59 years group, being predominant between 20-29 years, between epidemiological weeks 1 and 53 17,18 . ...
... Due to multiple deprivations on health, education, and living standards (e.g., sanitation), those who live in poverty are at a particularly higher risk for COVID-19 infection and adverse consequences (Alkire, Nogales, & Oldiges, 2020;Evans & Kovesdi, 2020). As people who live in poverty do not have wealth to live without working, and do not have an adequate house structure to keep them safe and comfortable during stay-at-home orders, they are unlikely to comply with prevention recommendations, particularly those who have multiple marginalized identities (Tavares & Betti, 2020;The Lancet, 2020). The health disparities of COVID-19 are particularly stark among people of color, as a disproportionate number of African Americans have been infected and died from the novel coronavirus (Lahut, 2020). ...
... Efforts should focus not only on providing online mental health services, but on diminishing the stressors stemming from economic insecurity. Economic Insecurity and COVID-19 People who live in poverty and experience economic insecurity are at higher risk of infection and adverse consequences from COVID-19 (Alkire et al., 2020), partially due to the inability to comply with community mitigation procedures because of financial needs (Evans & Kovesdi, 2020;Tavares & Betti, 2020). People of color in the U.S. are at particular risk for COVID-19 (Lahut, 2020), as gentrification, residential segregation, and racist practices prevent them from accessing healthy food and from having physical access to supermarkets (Alkon et al., 2020). ...
Article
The novel coronavirus disease 2019 (COVID-19) is currently spreading at a rapid rate worldwide. The current pandemic may have several adverse effects on overall psychological functioning and health behaviors. Economic insecurity, operationalized as financial strain and employment uncertainty, can be a significant risk factor for both psychological outcomes and compliance with shelter-in-place recommendations (i.e., health behaviors). One hundred and twenty four participants answered survey data on economic security, fear of COVID-19, health care system distrust, anxiety, well-being, and compliance with CDC recommendations to curb the spread of COVID-19 (i.e., health behaviors; CDC, 2020). Economic security was significantly associated with well-being, health behaviors, and fear of COVID-19, beyond health care system distrust. Economic insecurity appears to be a risk factor during the COVID-19 pandemic, as it appears to deter people from engaging in social distancing and shelter-in-place recommendations. More robust public policies geared toward alleviating economic distress among vulnerable populations are needed, as they may inadvertently help curb the rapid spread of COVID-19.
... Diante dessa situação, a qual se somam os retrocessos imprimidos pelo avanço neoliberal sobre as políticas sociais como um todo, e com as investidas específicas sobre os direitos indígenas, estes se encontram numa situação de maior fragilidade, pois, além de não fruírem de políticas sociais culturalmente sensíveis, vivenciam um atraso na resposta emergencial em forma de política pública contra os impactos da covid-19, o que pode levar ao maior número de contaminações e de mortes entre os povos indígenas. Com o avanço da doença, somado à extensão continental do território brasileiro, às disparidades sociais de sua população e ao multiculturalismo dos diversos segmentos populacionais, os povos indígenas se encontram em condições de maior vulnerabilidade, medida em termos de acesso às necessidades básicas 1 , frente à pandemia de covid-19, o que resulta em maiores chances de contaminação e reduzida capacidade de recuperação da doença (Tavares & Betti, 2020). Neste ponto, é importante salientar que as populações indígenas são historicamente mais vulneráveis às ameaças provenientes de infecções virais, em especial às ameaças respiratórias, que dizimaram grande número de indígenas que viviam no território brasileiro ( Isso implica que os casos e óbitos por covid-19 de indígenas em ambientes urbanos não têm sido notificados pelo seu caráter étnico. ...
... Neste ponto, é importante salientar que as populações indígenas são historicamente mais vulneráveis às ameaças provenientes de infecções virais, em especial às ameaças respiratórias, que dizimaram grande número de indígenas que viviam no território brasileiro ( Isso implica que os casos e óbitos por covid-19 de indígenas em ambientes urbanos não têm sido notificados pelo seu caráter étnico. Considerando tal subnotificação como um ato de racismo institucional, a Articulação dos Povos Indígenas do Brasil ([APIB], 2020) e outros movimentos indígenas passaram a apurar casos confirmados e mortes por covid-19 paralelamente, de modo que, até 6 de junho de 2020, dados de 1 São considerados vulneráveis os indivíduos que apresentam privações de pelo menos 2 de 8 indicadores, a saber, acesso à água potável, saneamento básico, eletricidade, moradia adequada, combustíveis de cozinha não poluentes, densidade morador-cômodo e alimentação na escola (Tavares & Betti, 2020) 3 organizações indígenas regionais, organizados pela APIB e Mobilização Nacional 1988, Ferreira, 2013. No entanto, a diferenciação da política pública indigenista de saúde se encontra limitada, como evidenciado pela subnotificação de casos e óbitos pela SESAI. ...
Article
Full-text available
As condições de vulnerabilidade socioeconômica e biológica dos povos indígenas frente à pandemia de coronavírus impõem a questão sobre quais são as limitações do amparo promovido pelas políticas de seguridade social. A fim de responder a esse questionamento, esta pesquisa buscou conceituar os direitos sociais na imbricação com a diferenciação cultural indígena, assim como buscou revisar a mais recente trajetória das políticas de saúde, assistência social e previdência social para esse público. Para tanto, foi realizada uma revisão bibliográfica e documental sobre os termos envolventes ao objetivo. Com isso, verificou-se que, até então, somente a política de saúde apresenta uma alternativa de diferenciação institucionalizada, enquanto as políticas de assistência social e de previdência social, embora possam ter boa cobertura dos usuários indígenas em alguns aspectos, se apresentam distantes das demandas desses povos. Diante dessa situação, a qual se somam os retrocessos imprimidos pelo avanço neoliberal sobre as políticas sociais como um todo, e com as investidas específicas sobre os direitos indígenas, estes se encontram numa situação de maior fragilidade, pois, além de não fruírem de políticas sociais culturalmente sensíveis, vivenciam um atraso na resposta emergencial em forma de política pública contra os impactos da covid-19, o que pode levar ao maior número de contaminações e de mortes entre os povos indígenas.
... Diante da relevância do tema, o efeito da pandemia da Covid-19 sobre o nível de pobreza multidimensional é um tema debatido na literatura. Nesse ínterim, alguns autores indicam que a pandemia poderia aumentar a pobreza multidimensional e as privações da população mais vulnerável (Alkire et al., 2021;Ba, 2020;Flores Tavares & Betti, 2020;Millan-Guerrero et al., 2021). Outros autores destacaram a proteção social como medida para garantir o consumo e o bem-estar dos cidadãos e desse modo reduzir o efeito da pandemia sobre o aumento da pobreza multidimensional (Bárcena et al., 2021;Bonfiglio & Robles, 2021;Carvalho, 2021 breza multidimensional mais elevados, particularmente para a parcela da população que sofre privações na dimensão de saúde, seguida pelo desenvolvimento industrial, emprego e renda. ...
Article
Full-text available
pobreza multidimensional está associada à teoriada pobreza e desenvolvimento humano de Amartya Sen.Uma das metodologias mais utilizadas para estimá-la, bemcomo determinar suas múltiplas privações, é o Índice dePobreza Multidimensional (IPM), desenvolvido por SabinaAlkire e James Foster. O objetivo deste estudo é analisar oefeito da pandemia da Covid-19 sobre a evolução da pobrezamultidimensional no Paraguai para o período 2016-2020. Osresultados demonstram que os programas sociais ajudaram areduzir os efeitos negativos da Covid-19, diminuindo o IPM,particularmente no caso das privações relacionadas a práticasinadequadas ou falta de serviços de eliminação de resíduos,além de pessoas doentes ou acidentadas sem acesso a cuidadosde saúde. Por outro lado, as privações correspondentes asuperlotação e desemprego aumentaram com o advento dapandemia e se mostraram pouco sensíveis aos efeitos dosprogramas sociais implementados.
... In support of these results are the studies conducted by DeCaprio et al. (2020) and Hägg et al. (2020) stating that age, frailty, and comorbidity were considered predictive factors for COVID-19 effects. Other studies demonstrated that COVID-19 affected differently communities subject to racial inequality and social exclusion (Kim & Bostwick, 2020), homelessness (Banerjee & Bhattacharya, 2020), low and middle income (Cénat, 2020), or poverty (Tavares & Betti, 2020). ...
Article
Full-text available
The assessment and identification of risk/vulnerable groups and risk factors are vital elements that can help quantify the pandemic potential of the SARS‐CoV‐2 virus in order to plan prevention and treatment measures. The aim of the study is to identify a methodological approach of population vulnerability to the SARS‐CoV‐2 virus infection. The study identifies reliable data sources and sets up a unitary database with statistical variables, quantitative and qualitative indicators with potential for being updated and improved. The analysis takes into account a number of variables/indicators (e.g., elderly persons, population without physician care, number of people suffering from cardiovascular diseases, number of people suffering from respiratory diseases, dwellings not connected to the public water supply network, no. of medical staff, number of COVID‐19 hospitals, PCR testing laboratories, number of vaccinated persons) grouped into the key vulnerability components: exposure, sensitivity, coping capacity and adaptive capacity. They allowed the computation of the final Index of Population Vulnerability to the SARS‐CoV‐2 virus infection and the mapping of different dimensions of vulnerability. The study was performed using the statistical data available at NUTS3/County level provided by different institutions (e.g., the Ministry of Health, the National Institute of Public Health, the Strategic Communication Group, and the National Institute of Statistics). The mapping of the different degrees of vulnerability could solve a problem of visibility for possible areas with vulnerable population, but also a problem of communication between different institutional health and administrative levels, as well as between all of them and the local communities and/or professionals.
... It is hoped that government initiatives such as guaranteed basic income and the provision of essential supplies may indeed reach people in vulnerable conditions. In addition, a unified and agile information system, expanded diagnostic testing capacity, and strengthened preventive actions in primary health services could help fight the pandemic Tavares and Betti, 2020). ...
Article
Full-text available
The pandemic caused by SARS-CoV-2 that leads to COVID-19 has had an incalculable impact on society, services, and health professionals, including physiotherapists. The field of physiotherapy is currently facing ongoing challenges and pandemic-related repercussions, with Brazilian physiotherapists being especially affected. This article describes the current scenario in Brazil which has exposed the shortage of professionals and resources and reinforces the need for professional recognition, especially of those working in intensive care units. Barriers and challenges faced by Brazilian physiotherapists during the COVID-19 pandemic are described, followed by innovative solutions. Remote physiotherapy prior to the pandemic was merely one of many patient care options, but now has become a necessity, and several international organizations have provided guidelines for physiotherapists to follow, including guidelines on offering digital physiotherapy. Related demands, challenges, and perspectives of physiotherapy practice are likely to emerge amid the current uncertainties of the COVID-19 crisis, as well as during the post-pandemic period. The lessons learned by the Brazilian physiotherapy community may be of assistance to physiotherapists in other countries particularly those with comparable social and demographic characteristics.
Article
Full-text available
This paper aims at identifying the Covid-19 infection and mortality risk factors in Brazil during the pandemic's first wave. Three groups of variables are considered: socioeconomic and health vulnerabilities, factors related to the virus transmission channels (mobility and density) and the effects of the policy responses. The analysis at the level of all 5,570 municipalities, drawing on a matching of different statistical and administrative databases, returns three main results. First, structurally vulnerable populations are hardest hit-non-white, poor, in poor health, favela residents and informal workers-showing the impact of socioeconomic inequalities. Second, we highlight some policy repercussions. The Auxilio Emergencial (emergency cash transfer) has had a mitigating effect in communities with relatively more informal workers. Finally, Covid-19 has hit hardest in municipalities that are more pro-Bolsonaro. The president's rhetoric and attitudes may have prompted his supporters to adopt more risky behaviour, suffer the consequences and infect others. Supplementary information: The online version contains supplementary material available at 10.1057/s41287-021-00487-w.
Article
Full-text available
Objective To study the profile of hospitalizations due to COVID-19 in the Unified Health System (SUS) in Brazil and to identify factors associated with in-hospital mortality related to the disease. Methods Cross-sectional study, based on secondary data on COVID-19 hospitalizations that occurred in the SUS between late February through June. Patients aged 18 years or older with primary or secondary diagnoses indicative of COVID-19 were included. Bivariate analyses were performed and generalized linear mixed models (GLMM) were estimated with random effects intercept. The modeling followed three steps, including: attributes of the patients; elements of the care process; and characteristics of the hospital and place of hospitalization. Results 89,405 hospitalizations were observed, of which 24.4% resulted in death. COVID-19 patients hospitalized in the SUS were predominantly male (56.5%) with a mean age of 58.9 years. The length of stay ranged from less than 24 hours to 114 days, with a mean of 6.9 (±6.5) days. Of the total number of hospitalizations, 22.6% reported ICU use. The odds on in-hospital death were 16.8% higher among men than among women and increased with age. Black individuals had a higher likelihood of death. The behavior of the Charlson and Elixhauser indices was consistent with the hypothesis of a higher risk of death among patients with comorbidities, and obesity had an independent effect on increasing this risk. Some states, such as Amazonas and Rio de Janeiro, had a higher risk of in-hospital death from COVID-19. The odds on in-hospital death were 72.1% higher in municipalities with at least 100,000 inhabitants, though being hospitalized in the municipality of residence was a protective factor. Conclusion There was broad variation in COVID-19 in-hospital mortality in the SUS, associated with demographic and clinical factors, social inequality, and differences in the structure of services and quality of health care.
Article
Full-text available
The first case of COVID-19 was detected in Brazil on 25 February 2020. We report and contextualize epidemiological, demographic and clinical findings for COVID-19 cases during the first 3 months of the epidemic. By 31 May 2020, 514,200 COVID-19 cases, including 29,314 deaths, had been reported in 75.3% (4,196 of 5,570) of municipalities across all five administrative regions of Brazil. The R0 value for Brazil was estimated at 3.1 (95% Bayesian credible interval = 2.4–5.5), with a higher median but overlapping credible intervals compared with some other seriously affected countries. A positive association between higher per-capita income and COVID-19 diagnosis was identified. Furthermore, the severe acute respiratory infection cases with unknown aetiology were associated with lower per-capita income. Co-circulation of six respiratory viruses was detected but at very low levels. These findings provide a comprehensive description of the ongoing COVID-19 epidemic in Brazil and may help to guide subsequent measures to control virus transmission. Brazil has one of the fastest-growing COVID-19 epidemics in the world. De Souza et al. report epidemiological, demographic and clinical findings for COVID-19 cases in the country during the first 3 months of the epidemic.
Article
Full-text available
Importance Local variation in the transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) across the United States has not been well studied. Objective To examine the association of county-level factors with variation in the SARS-CoV-2 reproduction number over time. Design, Setting, and Participants This cohort study included 211 counties, representing state capitals and cities with at least 100 000 residents and including 178 892 208 US residents, in 46 states and the District of Columbia between February 25, 2020, and April 23, 2020. Exposures Social distancing, measured by percentage change in visits to nonessential businesses; population density; and daily wet-bulb temperatures. Main Outcomes and Measures Instantaneous reproduction number (Rt), or cases generated by each incident case at a given time, estimated from daily case incidence data. Results The 211 counties contained 178 892 208 of 326 289 971 US residents (54.8%). Median (interquartile range) population density was 1022.7 (471.2-1846.0) people per square mile. The mean (SD) peak reduction in visits to nonessential business between April 6 and April 19, as the country was sheltering in place, was 68.7% (7.9%). Median (interquartile range) daily wet-bulb temperatures were 7.5 (3.8-12.8) °C. Median (interquartile range) case incidence and fatality rates per 100 000 people were approximately 10 times higher for the top decile of densely populated counties (1185.2 [313.2-1891.2] cases; 43.7 [10.4-106.7] deaths) than for counties in the lowest density quartile (121.4 [87.8-175.4] cases; 4.2 [1.9-8.0] deaths). Mean (SD) Rt in the first 2 weeks was 5.7 (2.5) in the top decile compared with 3.1 (1.2) in the lowest quartile. In multivariable analysis, a 50% decrease in visits to nonessential businesses was associated with a 45% decrease in Rt (95% CI, 43%-49%). From a relative Rt at 0 °C of 2.13 (95% CI, 1.89-2.40), relative Rt decreased to a minimum as temperatures warmed to 11 °C, increased between 11 and 20 °C (1.61; 95% CI, 1.42-1.84) and then declined again at temperatures greater than 20 °C. With a 70% reduction in visits to nonessential business, 202 counties (95.7%) were estimated to fall below a threshold Rt of 1.0, including 17 of 21 counties (81.0%) in the top density decile and 52 of 53 counties (98.1%) in the lowest density quartile.² Conclusions and Relevance In this cohort study, social distancing, lower population density, and temperate weather were associated with a decreased Rt for SARS-CoV-2 in counties across the United States. These associations could inform selective public policy planning in communities during the coronavirus disease 2019 pandemic.
Article
Full-text available
Background Brazil ranks second worldwide in total number of COVID-19 cases and deaths. Understanding the possible socioeconomic and ethnic health inequities is particularly important given the diverse population and fragile political and economic situation. We aimed to characterise the COVID-19 pandemic in Brazil and assess variations in mortality according to region, ethnicity, comorbidities, and symptoms. Methods We conducted a cross-sectional observational study of COVID-19 hospital mortality using data from the SIVEP-Gripe (Sistema de Informação de Vigilância Epidemiológica da Gripe) dataset to characterise the COVID-19 pandemic in Brazil. In the study, we included hospitalised patients who had a positive RT-PCR test for severe acute respiratory syndrome coronavirus 2 and who had ethnicity information in the dataset. Ethnicity of participants was classified according to the five categories used by the Brazilian Institute of Geography and Statistics: Branco (White), Preto (Black), Amarelo (East Asian), Indígeno (Indigenous), or Pardo (mixed ethnicity). We assessed regional variations in patients with COVID-19 admitted to hospital by state and by two socioeconomically grouped regions (north and central-south). We used mixed-effects Cox regression survival analysis to estimate the effects of ethnicity and comorbidity at an individual level in the context of regional variation. Findings Of 99 557 patients in the SIVEP-Gripe dataset, we included 11 321 patients in our study. 9278 (82·0%) of these patients were from the central-south region, and 2043 (18·0%) were from the north region. Compared with White Brazilians, Pardo and Black Brazilians with COVID-19 who were admitted to hospital had significantly higher risk of mortality (hazard ratio [HR] 1·45, 95% CI 1·33–1·58 for Pardo Brazilians; 1·32, 1·15–1·52 for Black Brazilians). Pardo ethnicity was the second most important risk factor (after age) for death. Comorbidities were more common in Brazilians admitted to hospital in the north region than in the central-south, with similar proportions between the various ethnic groups. States in the north had higher HRs compared with those of the central-south, except for Rio de Janeiro, which had a much higher HR than that of the other central-south states. Interpretation We found evidence of two distinct but associated effects: increased mortality in the north region (regional effect) and in the Pardo and Black populations (ethnicity effect). We speculate that the regional effect is driven by increasing comorbidity burden in regions with lower levels of socioeconomic development. The ethnicity effect might be related to differences in susceptibility to COVID-19 and access to health care (including intensive care) across ethnicities. Our analysis supports an urgent effort on the part of Brazilian authorities to consider how the national response to COVID-19 can better protect Pardo and Black Brazilians, as well as the population of poorer states, from their higher risk of dying of COVID-19. Funding None.
Article
Full-text available
Background Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes COVID-19 and is spread person-to-person through close contact. We aimed to investigate the effects of physical distance, face masks, and eye protection on virus transmission in health-care and non-health-care (eg, community) settings. Methods We did a systematic review and meta-analysis to investigate the optimum distance for avoiding person-to-person virus transmission and to assess the use of face masks and eye protection to prevent transmission of viruses. We obtained data for SARS-CoV-2 and the betacoronaviruses that cause severe acute respiratory syndrome, and Middle East respiratory syndrome from 21 standard WHO-specific and COVID-19-specific sources. We searched these data sources from database inception to May 3, 2020, with no restriction by language, for comparative studies and for contextual factors of acceptability, feasibility, resource use, and equity. We screened records, extracted data, and assessed risk of bias in duplicate. We did frequentist and Bayesian meta-analyses and random-effects meta-regressions. We rated the certainty of evidence according to Cochrane methods and the GRADE approach. This study is registered with PROSPERO, CRD42020177047. Findings Our search identified 172 observational studies across 16 countries and six continents, with no randomised controlled trials and 44 relevant comparative studies in health-care and non-health-care settings (n=25 697 patients). Transmission of viruses was lower with physical distancing of 1 m or more, compared with a distance of less than 1 m (n=10 736, pooled adjusted odds ratio [aOR] 0·18, 95% CI 0·09 to 0·38; risk difference [RD] −10·2%, 95% CI −11·5 to −7·5; moderate certainty); protection was increased as distance was lengthened (change in relative risk [RR] 2·02 per m; pinteraction=0·041; moderate certainty). Face mask use could result in a large reduction in risk of infection (n=2647; aOR 0·15, 95% CI 0·07 to 0·34, RD −14·3%, −15·9 to −10·7; low certainty), with stronger associations with N95 or similar respirators compared with disposable surgical masks or similar (eg, reusable 12–16-layer cotton masks; pinteraction=0·090; posterior probability >95%, low certainty). Eye protection also was associated with less infection (n=3713; aOR 0·22, 95% CI 0·12 to 0·39, RD −10·6%, 95% CI −12·5 to −7·7; low certainty). Unadjusted studies and subgroup and sensitivity analyses showed similar findings. Interpretation The findings of this systematic review and meta-analysis support physical distancing of 1 m or more and provide quantitative estimates for models and contact tracing to inform policy. Optimum use of face masks, respirators, and eye protection in public and health-care settings should be informed by these findings and contextual factors. Robust randomised trials are needed to better inform the evidence for these interventions, but this systematic appraisal of currently best available evidence might inform interim guidance. Funding World Health Organization.
Article
Full-text available
Background There are few primary care studies of the COVID-19 pandemic. We aimed to identify demographic and clinical risk factors for testing positive for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) within the Oxford Royal College of General Practitioners (RCGP) Research and Surveillance Centre primary care network. Methods We analysed routinely collected, pseudonymised data for patients in the RCGP Research and Surveillance Centre primary care sentinel network who were tested for SARS-CoV-2 between Jan 28 and April 4, 2020. We used multivariable logistic regression models with multiple imputation to identify risk factors for positive SARS-CoV-2 tests within this surveillance network. Findings We identified 3802 SARS-CoV-2 test results, of which 587 were positive. In multivariable analysis, male sex was independently associated with testing positive for SARS-CoV-2 (296 [18·4%] of 1612 men vs 291 [13·3%] of 2190 women; adjusted odds ratio [OR] 1·55, 95% CI 1·27–1·89). Adults were at increased risk of testing positive for SARS-CoV-2 compared with children, and people aged 40–64 years were at greatest risk in the multivariable model (243 [18·5%] of 1316 adults aged 40–64 years vs 23 [4·6%] of 499 children; adjusted OR 5·36, 95% CI 3·28–8·76). Compared with white people, the adjusted odds of a positive test were greater in black people (388 [15·5%] of 2497 white people vs 36 [62·1%] of 58 black people; adjusted OR 4·75, 95% CI 2·65–8·51). People living in urban areas versus rural areas (476 [26·2%] of 1816 in urban areas vs 111 [5·6%] of 1986 in rural areas; adjusted OR 4·59, 95% CI 3·57–5·90) and in more deprived areas (197 [29·5%] of 668 in most deprived vs 143 [7·7%] of 1855 in least deprived; adjusted OR 2·03, 95% CI 1·51–2·71) were more likely to test positive. People with chronic kidney disease were more likely to test positive in the adjusted analysis (68 [32·9%] of 207 with chronic kidney disease vs 519 [14·4%] of 3595 without; adjusted OR 1·91, 95% CI 1·31–2·78), but there was no significant association with other chronic conditions in that analysis. We found increased odds of a positive test among people who are obese (142 [20·9%] of 680 people with obesity vs 171 [13·2%] of 1296 normal-weight people; adjusted OR 1·41, 95% CI 1·04–1·91). Notably, active smoking was linked with decreased odds of a positive test result (47 [11·4%] of 413 active smokers vs 201 [17·9%] of 1125 non-smokers; adjusted OR 0·49, 95% CI 0·34–0·71). Interpretation A positive SARS-CoV-2 test result in this primary care cohort was associated with similar risk factors as observed for severe outcomes of COVID-19 in hospital settings, except for smoking. We provide evidence of potential sociodemographic factors associated with a positive test, including deprivation, population density, ethnicity, and chronic kidney disease.
Article
Full-text available
Purpose Given incomplete data reporting by race, we used data on COVID-19 cases and deaths in US counties to describe racial disparities in COVID-19 disease and death and associated determinants. Methods Using publicly available data (accessed April 13, 2020), predictors of COVID-19 cases and deaths were compared between disproportionately (>13%) black and all other (<13% black) counties. Rate ratios were calculated and population attributable fractions (PAF) were estimated using COVID-19 cases and deaths via zero-inflated negative binomial regression model. National maps with county-level data and an interactive scatterplot of COVID-19 cases were generated. Results Nearly ninety-seven percent of disproportionately black counties (656/677) reported a case and 49% (330/677) reported a death versus 81% (1987/2,465) and 28% (684/ 2465), respectively, for all other counties. Counties with higher proportions of black people have higher prevalence of comorbidities and greater air pollution. Counties with higher proportions of black residents had more COVID-19 diagnoses (RR 1.24, 95% CI 1.17-1.33) and deaths (RR 1.18, 95% CI 1.00-1.40), after adjusting for county-level characteristics such as age, poverty, comorbidities, and epidemic duration. COVID-19 deaths were higher in disproportionally black rural and small metro counties. The PAF of COVID-19 diagnosis due to lack of health insurance was 3.3% for counties with <13% black residents and 4.2% for counties with >13% black residents. Conclusions Nearly twenty-two percent of US counties are disproportionately black and they accounted for 52% of COVID-19 diagnoses and 58% of COVID-19 deaths nationally. County-level comparisons can both inform COVID-19 responses and identify epidemic hot spots. Social conditions, structural racism, and other factors elevate risk for COVID-19 diagnoses and deaths in black communities.
Article
Full-text available
Disadvantaged socioeconomic position (SEP) is widely associated with disease and mortality, and there is no reason to think this will not be the case for the newly emerged coronavirus disease 2019 (COVID-19) that has reached a pandemic level. Individuals with a more disadvantaged SEP are more likely to be affected by most of the known risk factors of COVID-19. SEP has been previously established as a potential determinant of infectious diseases in general. We hypothesise that SEP plays an important role in the COVID-19 pandemic either directly or indirectly via occupation, living conditions, health-related behaviours, presence of comorbidities and immune functioning. However, the influence of socioeconomic factors on COVID-19 transmission, severity and outcomes is not yet known and is subject to scrutiny and investigation. Here we briefly review the extent to which SEP has been considered as one of the potential risk factors of COVID-19. From 29 eligible studies that reported the characteristics of patients with COVID-19 and their potential risk factors, only one study reported the occupational position of patients with mild or severe disease. This brief overview of the literature highlights that important socioeconomic characteristics are being overlooked when data are collected. As COVID-19 spreads worldwide, it is crucial to collect and report data on socioeconomic determinants as well as race/ethnicity to identify high-risk populations. A systematic recording of socioeconomic characteristics of patients with COVID-19 will be beneficial to identify most vulnerable groups, to identify how SEP relates to COVID-19 and to develop equitable public health prevention measures, guidelines and interventions.
Article
Full-text available
The Coronavirus disease 2019 (COVID-19) pandemic has significantly impacted and devastated the world. As the infection spreads, the projected mortality and economic devastation are unprecedented. In particular, racial and ethnic minorities may be at a particular disadvantage as many already assume the status of a marginalized group. Black Americans have a long-standing history of disadvantage and are in a vulnerable position to experience the impact of this crisis and the myth of Black immunity to COVID-19 is detrimental to promoting and maintaining preventative measures. We are the first to present the earliest available data in the peer-reviewed literature on the racial and ethnic distribution of COVID-19-confirmed cases and fatalities in the state of Connecticut. We also seek to explode the myth of Black immunity to the virus. Finally, we call for a National Commission on COVID-19 Racial and Ethnic Health Disparities to further explore and respond to the unique challenges that the crisis presents for Black and Brown communities.
Article
Purpose Heightened COVID-19 mortality among Black non-Hispanic and Hispanic communities (relative to white non-Hispanic) is well established. This study aims to estimate the relative contributions to fatality disparities in terms of differences in SARS-CoV-2 infections, diagnoses, and disease severity. Methods We constructed COVID-19 outcome continua (similar to the HIV care continuum) for white non-Hispanic, Black non-Hispanic, and Hispanic adults in New York State. For each stage in the COVID-19 outcome continua (population, infection experience, diagnosis, hospitalization, fatality), we synthesized the most recent publicly-available data. We described each continuum using overall percentages, fatality rates, and relative changes between stages, with comparisons between race and ethnicity using risk ratios. Results Estimated per-population COVID-19 fatality rates were 0.03%, 0.18%, and 0.12% for white non-Hispanic, Black non-Hispanic, and Hispanic adults. The 3.48-fold disparity for Hispanic, relative to white, communities was explained by differences in infection-experience, whereas the 5.38-fold disparity for non-Hispanic Black, relative to white, communities was primarily driven by differences in both infection-experience and in the need for hospitalization, given infection. Conclusions These findings suggest the most impactful stages upon which to intervene with programs and policies to build COVID-19 health equity.
Article
As the coronavirus disease 2019 (COVID-19) pandemic continues advancing globally, reporting of clinical outcomes and risk factors for intensive care unit admission and mortality are emerging. Early Chinese and Italian reports associated increasing age, male sex, smoking, and cardiometabolic comorbidity with adverse outcomes. Striking differences between Chinese and Italian mortality indicate ethnicity might affect disease outcome, but there is little to no data to support or refute this.