Lessons from the COVID-19 pandemic: the unequal burden of COVID-19 on vulnerable populations in the Brazilian Central-West

Abstract: This study aimed to estimate the prevalence and identify social factors and preventive strategies associated with the coronavirus disease 2019 (COVID-19) in socio and economically vulnerable people (recyclable waste pikers, immigrants/refugees, and homeless people) in Goiânia, Goiás State, Central-Western Brazil. A cross-sectional study was conducted from July 2020 to October 2020. COVID-19 positivity was defined as a positive total anti-SARS-COV-2 antibody test and/or RNA test for SARS-COV-2. Univariable and multiple regression analyses were performed to identify the variables associated with COVID-19. Of the 594 participants, 47.3% were recyclable waste pickers, 29.6% were immigrants/refugees, and 23.1% were homeless people. The positivity for SARS-CoV-2 RNA was 14.1%, whereas for anti-SARS-CoV-2 a total of 30.8% were positive, and 39.4% were positive for at least one COVID-19 marker. Among the 541 individuals, being immigrants/refugees, not wearing a surgical mask, and having three or more people sleeping in the same room were associated with SARS-CoV-2 infection, while using TV news as the main source of information about the pandemic was a protective predictor of COVID-19. This study revealed ethnic and socioeconomic inequalities in the prevalence of COVID-19 among impoverished people in Brazil. Additionally, a high prevalence of COVID-19 was detected in all three groups. Developing new strategies to combat and prevent communicable diseases affecting this population is essential for mitigating future and ongoing pandemics.


Introduction
The coronavirus disease 2019 (COVID-19) pandemic represents the most significant public health crisis of the 21st century 1 .It is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which is transmitted primarily by droplets and aerosols generated during coughing or sneezing od infected individuals.Additionally, indirect transmission can occur via contact with objects or surfaces contaminated by respiratory secretions of infected individuals 2,3 .Generally, SARS-CoV-2 infection starts with mild respiratory symptoms, but it can quickly progress to a severe respiratory syndrome.Available evidence indicates that severe symptoms with worse prognoses are more frequent in men, older adults, individuals with comorbidities, and immunocompromised individuals 4 .In Brazil, SARS-CoV-2 infection initially affected high-income people returning from overseas.However, it quickly spread to other groups, overwhelmingly affecting poorer populations 5 .
According to the Human Development Report 2021/2022 6 , at least 1.3 billion people worldwide live in multidimensional poverty, facing health, education, and food deprivation, among others.The pandemic amplified this scenario, increasing the number of people living in extreme poverty (surviving on less than USD 1.90/day) for the first time in over two decades 6 .
Economic and political crises in Brazil have resulted in high unemployment rates, consequently increasing the figures of informal workers, such as recyclable waste pickers and homeless people.Moreover, there has been a large influx of immigrants in recent years, with more than 50,000 people, mainly from Venezuela, requesting refugee status in 2022 7 .Most of these emergent populations live in extreme poverty, in hazardous conditions and crowded housing facilities, socially isolated, and without basic sanitation.Additionally, they are often exposed to high-stress jobs with low wages and have difficulty accessing public healthcare services, making them vulnerable to infectious diseases such as COVID-19 8,9 .Therefore, COVID-19 has been recognized as a syndemic, in which the complex interaction between SARS-CoV-2 infection and existing social, economic, and health disparities can increase the chances of higher positivity rates and worse outcomes 10 .
The increasing global environmental changes, including deforestation and urbanization, contribute indirectly to the emergence of new viral agents 11 .Therefore, it is important to elucidate the extent to which SARS-CoV-2 affected the marginalized groups of developing countries, such as Brazil, and the measures undertaken to face the pandemic.Understanding the epidemiology of this pandemic in low-income and vulnerable populations will support appropriate handling of similar outbreaks affecting the above-mentioned groups in the future.In this study, we aimed to estimate the prevalence of COVID-19 and identify social factors and preventive strategies adopted against COVID-19 in socially and economically vulnerable people (immigrants/refugees, recyclable waste pickers, and homeless individuals) in Goiânia, Goiás, State, Brazilian Central-West, during the first and second waves of COVID-19.

Methodology Study design
A cross-sectional study was conducted from August 2020 to April 2021 among three socially and economically vulnerable populations: immigrants/refugees, recyclable waste pickers, and homeless people.

Population and location
In Brazil, COVID-19 testing availability was limited.Therefore, the Federal University of Goiás developed a community project for no-cost SARS-CoV-2 screening.To support this effort, a strict biosafety infrastructure was created on the university premises.All safety protocols were followed, including reducing the crowding of participants.Thus, the research team successfully reached vulnerable populations in the community and conducted COVID-19 screening.
Cad. Saúde Pública 2024; 40(8):e00199623 The population consisted of immigrants/refugees (80% Venezuelans), recyclable waste pickers, and homeless people from Goiânia, assisted by nongovernmental organizations (NGOs).During the data collection period, there were no data on COVID-19 prevalence in Brazil, and no vaccines were available.Therefore, we determined the cohort size as a minimum of 329 individuals, considering a hypothetical 5% prevalence, a 95% significance level (α < 0.05), a 3% precision, and a 2% effect design.
We included socially vulnerable individuals from Goiânia who were referred by NGOs, leaders, and government entities partnering in this project.We excluded individuals who were visibly under the influence of alcohol and/or illicit drugs and exhibited behaviors that jeopardized the team's safety, or faced difficulties in answering to questions.

Recruitment
In July 2020, the research team contacted NGO partners who were providing social assistance to immigrants/refugees, recyclable waste pickers, and homeless individuals in Goiânia, and disclosed the study purpose.All agreed to collaborate and were eager to participate in COVID-19 screening efforts.The representatives of the NGO then committed to invite eligible individuals to participate in the study, scheduling their data and sample collection, and providing their transportation to the location where data and samples were collected.

Ethical aspects
This study was approved by the Ethics Committee of the Federal University of Goiás (n.4,249,851) according to Resolution n. 466/12.

Data and samples collection
All participants were interviewed using a questionnaire regarding demographic and social characteristics (sex, age, marital status, years of education, employment, and monthly household income), flulike signs and symptoms in themselves or family since the beginning of the pandemic, and pandemic coping measures (related to their knowledge about the pandemic, hand hygiene, and use of mask).For immigrants/refugees, interviews were conducted with individuals who were fluent in Spanish and trained to adhere to all safety and research protocols.Interviews were conducted after explaining the project and obtaining written informed consent.
In this study, COVID-19 diagnosis was the dependent variable.Participants were classified as positive for COVID-19 if they tested positive for SARS-CoV-2 RNA and/or total anti-SARS-CoV-2 antibodies (Figure 1).
The independent variables included demographic and social characteristics, such as sex, age, marital status, years of education, employment, number of residents in the household, monthly household income, pandemic coping measures, COVID-19 symptoms in the family, knowledge of the pandemic, hand hygiene, and use of personal protective equipment.

Data analysis
The interview data and laboratory results were analyzed using the SPSS version 20.0 software (https://www.ibm.com/).Prevalence was estimated using 95% confidence interval (95%CI).For each group, prevalence ratio (PR) and 95%CI were used to measure the strength of the association between COVID-19 diagnosis (outcome) and the independent variables.Variables associated with the outcome, with p < 0.20, were included in Poisson regression with a robust variance model 12 .The significance level was set at 0.05.Pearson's chi-squared goodness-of-fit test was used to assess the model adequacy.

Discussion
Herein, we describe the prevalence of SARS-CoV-2 infection and the associated variables among immigrants/refugees, recyclable waste pickers, and homeless individuals in Central-Western Brazil.Notably, this study was conducted before the COVID-19 vaccination program.Therefore, the epidemiological data on COVID-19 in these vulnerable target groups are important to establish better strategies for preventive actions in future outbreaks.
At the onset of the pandemic in Brazil, Hallal et al. 13 conducted two successive nationwide serological household surveys in May and June 2020.They tested participants for total antibodies against SARS-CoV-2 using a lateral-flow point-of-care test and found that the pooled prevalence increased from 1.9% to 3.1% among cities, with 200 or more individuals tested in both surveys, highlighting accelerated viral dissemination.Furthermore, both surveys showed that the prevalence in the poorest people was twice as high as that of the richest people.Similarly, this study included people from the lowest layer of the population; 14.1% were already infected as determined by SARS-CoV-2 RNA positivity, and 30.8% had antibodies against SARS-CoV-2.These findings reinforce the uncontrolled viral dissemination in Brazil, particularly among the poorest , as well as the urgent need to promptly include socially and economically vulnerable groups in the target population.
We studied three groups of impoverished people living with limited social protection: immigrants/ refugees, recyclable waste pickers, and homeless people.A high frequency of SARS-CoV-2 exposure was found in all of them.However, the disproportional viral dissemination among immigrants/    refugees is noteworthy, as 67.5% had been exposed to the new coronavirus.These findings align with the epidemiological bulletin of the state of Roraima (Brazil-Venezuela border), which reported that over 80% of deaths caused by COVID-19 occurred among Venezuelan immigrants 14 .Investigations conducted with immigrants in other countries have reinforced their higher social vulnerability.For example, in the United Kingdom, hospital death rates are higher among immigrants, accounting for 35% of all patients with COVID-19 in intensive care units 15 .Even in the United States and Europe, hospitalization and death rates are higher among immigrant populations 16,17,18 .
Cad. Saúde Pública 2024; 40(8):e00199623 Most impoverished people live in overcrowded housing conditions with shared bedrooms or even beds, making it difficult to designate a separate room for everyone.Thus, infected, and noninfected individuals live in the same environment, increasing the chances of viral transmission and acquisition 19,20,21 .In this study, individuals who reported sleeping with three or more individuals in the same bedroom were more likely to have a positive test for COVID-19.
A low prevalence of COVID-19 was observed among individuals who relayed on TV news as the main source of information about the pandemic.TV is the mode of communication with most publicity and participation of government representatives, including public health managers.In addition, television information is provided in an accessible and clear language, without the need for doublechecking.Hence, traditional media is considered a reliable source of information and contributes to improving prevention behaviors 22,23 .
During the study period, COVID-19 vaccination and treatment were not available.Therefore, non-pharmacological interventions, such as physical distancing, face masks, and hand hygiene were the primary measures to control and prevent the new coronavirus outbreak 21,24 .Most participants reported using masks; however, most of them used cloth masks.Those who reported the use of surgical masks showed a lower frequency contracting COVID-19.Therefore, during epidemics, public managers must distribute free surgical masks for impoverished people as a strategy to mitigate the ongoing communicable disease.
This study had some limitations.We used a convenience sampling method and considered the three groups as a whole.Therefore, our data may not represent all the studied groups.Furthermore, the cross-sectional design provided limited evidence to support causality, although reverse causality was unlikely in this study.Despite these limitations, our data show epidemiological coherence and plausibility.

Conclusion
We showed a high prevalence of COVID-19 among the three studied groups, with the most affected being immigrants/refugees, followed by the Brazilian recyclable waste pickers, and homeless people.In addition, not wearing a surgical mask and having three or more people sleeping in the same room were associated with SARS-CoV-2 infection, while relying on TV news as the main source of information about the pandemic was a protective predictor of COVID-19.Finally, these findings highlight the need for stakeholders to consider the specificities of socially vulnerable individuals to control and prevent emergent infections.

Table 1
Sociodemographic characteristics of the 594 vulnerable people tested for SARS-CoV-2 markers during the first and second waves of the COVID-19 in Goiânia, Goiás State, Central-Western Brazil.

Table 2
Univariable analysis of sociodemographics and coping strategies according to COVID-19 markers in 541 vulnerable participants in Goiânia, Goiás State, Central-Western Brazil.

Table 3
Multiple Poisson regression analysis of variables associated with COVID-19 among vulnerable populations in Goiânia,Goiás State, Central-Western Brazil.* Adjusted by population, age, marital status, worked during the pandemic, surgical mask, number of people sleeping in the same room, COVID symptoms in family, TV news as the main source of information about the pandemic.