Five years after its emergence, COVID-19 became endemic, like other respiratory viruses. In Brazil, it marked impact, reflected in high morbidity and mortality, highlighting structural, economic, and social vulnerabilities and inequalities.

Brazil adopted mobility restrictions, mask use, isolation of suspected or confirmed cases, and suspension of non-essential activities in a disorganized manner,1 becoming one of the global leaders in cases and deaths.

In March 2020, the Federal Government implemented the Emergency Aid Program (EAP) to support vulnerable populations. However, irregular payments and fluctuating amounts hindered financial planning.2 Moreover, the central government failed to leverage features of existing social protection programs, the Bolsa Família Program (BFP) and the Continuous Benefit Program,2, namely, the use of cash transfers to promote preventive behaviors and the provision of regular, predictable benefits.

Benefits granted in 2020–2021 varied in duration, value, coverage, and the regulatory mechanisms employed. In 2020, the EAP ran from April to December, with reductions in value and coverage starting in September. Peak coverage reached 64 million people, who received BRL 600 for the first five months; single mothers received double. In 2021, the EAP was paid from April to October, reaching a maximum of 39 million beneficiaries, who received BRL 150, BRL 250, or BRL 375 according to family characteristics, again with double payments for single mothers.2

During the EAP period, the BFP was paid only to individuals ineligible for the EAP or entitled to a lower benefit. From April 2020 onward, the BFP payroll decreased sharply, as 86%–95% of its beneficiaries shifted to EAP payments.2,3

Failure to use the BFP registration system led to major distribution errors. The Federal Court of Accounts estimated that 7.3 million people were improperly benefited, and 6.4 million women falsely declared themselves single mothers.4 These inclusion errors resulted in excess spending of BRL 54 billion, nearly one-sixth of total expenditures that year.4 The EAP ended in October 2021, coinciding with the implementation of the Auxílio Brasil Program, which replaced the BFP.

Despite inconsistencies, the EAP increased transfer amounts and expanded coverage. In 2020, spending of BRL 303.4 billion generated positive economic effects. Sanches et al.5 estimated that these transfers acted as a fiscal multiplier, mitigating the economic downturn. Without this investment, the 4.1% GDP decline in 2020 would have been substantially larger, ranging from 8.4% to 14.8% compared with 2019.5

Brazil began COVID-19 vaccination in January 2021, prioritizing biologically and socially vulnerable groups and expanding coverage as supply increased.6 Despite early delays, the country benefited from its strong National Immunization Program), with broad reach and experience in mass campaigns. Nonetheless, structural inequalities, political disputes, logistical barriers, and misinformation produced marked heterogeneity in vaccine uptake. Studies show lower coverage associated with income, education, race, housing conditions, and urban density, indicating delayed access among populations already more exposed and vulnerable to COVID-19.6–9

Given this context, it is important to assess how vaccination and income transfer programs interacted with disease spread. Specifically, evaluating whether large-scale financial support to vulnerable populations reduced disease burden, particularly alongside expanding vaccine coverage, may inform future social protection and public health strategies. This longitudinal ecological study analyzed whether the availability of the EAP and COVID-19 vaccination was associated with COVID-19 morbidity and mortality in Brazilian municipalities, controlling for relevant covariates, including demographic, health, socioeconomic, health service availability, and testing factors.

Figure 1 shows the municipal spatial distribution of hospitalization (left panels) and mortality rates (right panels) per 100,000 inhabitants, standardized by sex and age group, across three periods. Hospitalization maps reveal spatial clustering, with neighboring municipalities showing similar rates. Higher mortality generally coincides with areas of high hospitalization in all periods.

Figure 1.

Spatiotemporal patterns of COVID-19 hospitalization and mortality in Brazil. Spatial distribution of COVID-19 hospitalization (left) and mortality (right) rates per 100,000 inhabitants in Brazilian municipalities in July 2020, March 2021, and July 2021. July 2020 marks the first wave with higher rates in western regions, while March 2021 corresponds to the peak of the second wave, with widespread high hospitalization and mortality.

July 2020 (Fig. 1 A) corresponds to the first wave, with the highest numbers of hospitalizations and deaths, particularly in the western regions. In March 2021 (Fig. 1 B), hospitalization and mortality peaked during the second wave, with marked clustering of hospitalization rates exceeding 80 per 100,000 inhabitants in most municipalities, except for lower notifications in parts of the Northeast.

In July 2021 (Fig. 1 C), hospitalization and mortality rates were similar to those in March 2020, but with a different spatial pattern, concentrated mainly in central regions. Overall, the spatial distribution of hospitalization and mortality rates changed markedly over time, in magnitude and clustering, highlighting the relevance of spatiotemporal effects in pandemic analyses. Within municipalities, temporal trends of hospitalizations and deaths followed similar wave patterns (descriptive results not shown).

Figure 2 shows the spatial distribution of the EAP (left panels) and first-dose COVID-19 vaccination (right panels) across the same periods. During the first wave (July 2020), EAP coverage exceeded 15% of the population in most municipalities (Fig. 2 A). In March 2021, during peak incidence, coverage declined and was concentrated in the North and Northeast (Fig. 2 B), regions with consistently higher coverage, possibly reflecting prior BFP reach. In July 2021, EAP coverage increased again in the South, Southeast, and Midwest (Fig. 2 C). Vaccination coverage was still limited in March 2021 but expanded substantially nationwide by July 2021 (Fig. 2 C).

Figure 2.

Spatiotemporal distribution of the Emergency Aid Program (EAP) and COVID-19 vaccination coverage in Brazil. Spatial distribution of EAP coverage (left) and COVID-19 first-dose vaccination coverage (right) across Brazilian municipalities in July 2020, March 2021, and July 2021.

After Pearson correlation analysis, one level 2 and six level 3 control variables were excluded from the initial set of 29 (Table 1). Zero-inflated negative binomial GLMM showed the lowest AIC for both outcomes. Significant variables were retained for Bayesian spatiotemporal modeling (12 for hospitalization and 15 for mortality) (Table 1), all with VIF <5. The Zero-inflated negative binomial Bayesian model was selected for both outcomes, as it yielded the lowest DIC.

Results of the final Bayesian spatiotemporal model for COVID-19 hospitalizations are shown in Table 2. Chronic disease mortality (risk ratio [RR]: 1.048; 95%CI: 1.021-1.076), number of ICU beds (RR: 1.202; 95%CI: 1.194-1.212), and RT-PCR testing rate (RR: 1.956; 95%CI: 1.952-1.962) were positively associated with hospitalization risk, indicating higher hospitalization volumes in municipalities with greater bed availability and testing capacity, and in populations with higher chronic disease burden.

In contrast, EAP coverage (RR: 0.986; 95%CI: 0.984-0.988) and first-dose COVID-19 vaccine coverage (RR: 0.982; 95%CI: 0.979-0.985) were inversely associated with hospitalization risk. A 1-percentage-point increase in vaccine coverage was associated with a 1.8% (95%CI: 1.5-2.1) reduction in risk, while a similar increase in EAP coverage corresponded to a 1.4% (95%CI: 1.2-1.6) reduction. Although individual estimates are numerically small, their impact increases with larger coverage changes. Under the log-link negative binomial model, a 10% increase in vaccine and EAP coverage was associated with 16.6% (95%CI: 14.4-18.6) and 13.9% (95%CI: 11.7-14.1) reductions in hospitalization risk, respectively.

Table 3 presents the results for COVID-19 mortality. Positive associations with mortality risk were observed for chronic disease mortality (RR: 1.215; 95%CI: 1.167-1.265), ICU beds (RR: 1.206; 95%CI: 1.088-1.125), clinical beds (RR: 1.081; 95%CI: 1.062-1.102), RT-PCR testing rate (RR: 2.210; 95%CI: 2.122-2.273], tuberculosis incidence (RR: 1.150; 95%CI: 1.112-1.191), mortality due to unsafe sanitation (RR: 1.092; 95%CI: 1.061-1.124), and the proportion of households in favelas (RR: 1.219; 95%CI: 1.175-1.265). These findings indicate higher mortality in municipalities with greater hospital capacity and testing, and reinforce the role of comorbidity, poor sanitation, and socioeconomic disadvantage, including residence in urban agglomerations.

Conversely, EAP coverage (RR: 0.975; 95%CI: 0.973-0.977), first-dose COVID-19 vaccine coverage (RR: 0.983; 95%CI: 0.981-0.985), and routine vaccination coverage in children under five (RR: 0.996; 95%CI: 0.994-0.996) were inversely associated with mortality. A 10% increase in vaccine and EAP coverage was associated with estimated reductions of 21.9% (95%CI: 19.8-24.0) and 16.6% (95%CI: 14.8-18.3) in mortality risk, respectively.

Investigating regional inequalities in COVID-19 remains highly relevant to inform future pandemic policies.22 This study evaluated associations between socioeconomic and demographic factors, access to health services, and COVID-19 hospitalization and mortality in Brazilian municipalities from March 2020 to December 2021. First-dose COVID-19 vaccine and EAP coverage were associated with reduced relative risks for both outcomes after adjustment for socioeconomic and health conditions. For mortality, regular vaccination coverage in children under five showed a similar association, suggesting that municipalities with stronger routine immunization capacity may be better prepared to implement care protocols during pandemics.

The pandemic disproportionately affected historically marginalized populations with structural vulnerabilities.23,24 Poor living conditions and poverty are key social determinants of health, underscoring the need for public policies to improve living and health conditions.25 Most governments responded with mobility restrictions.1 While closures of restaurants, bars, informal trade, and industrial activities helped contain the pandemic, they increased underemployment and unemployment.26,27 Economists have warned that rising unemployment and inflation would increase poverty and reduce living standards after COVID-19,27 a trend reflected in growing food insecurity.28

Our findings suggest that pandemic impacts could have been further reduced if income support had been more stable and conditional on adherence to disease-prevention measures.2–4

Income transfers to the most vulnerable populations reduce poverty, albeit temporarily. However, broader social policies such as education, housing, sanitation, working conditions, and remuneration directly influence the impact of specific health policies and should be enforced. Despite limitations, our findings indicate that income transfer policies contributed to reducing COVID-19 morbidity and mortality in Brazil, after adjustment for vaccination and other factors.

The proportion of households in favelas was positively associated with mortality, suggesting that urbanization and population density facilitated virus transmission. Tuberculosis incidence and mortality due to unsafe sanitation, also markers of socioeconomic deprivation, were similarly associated with higher risk. These results align with previous evidence showing that socioeconomic conditions strongly influenced COVID-19 morbidity and mortality in Brazil.2

We found evidence that first-dose COVID-19 vaccine coverage reduced hospitalization and mortality in Brazil. Vaccine supply was limited for a period,6,29 and immunization initially prioritized biologically vulnerable and socially disadvantaged groups, such as homeless and incarcerated populations.6,9,29 This strategy did not fully address residents of densely populated and deprived areas, where mobility restrictions were less feasible.6

COVID-19 vaccination substantially influenced pandemic dynamics, reducing severe cases, hospitalizations, and deaths, and remains one of the most effective measures for prevention and mitigation.6–9

For both outcomes, the spatial component and state-level hierarchical effects explained most residual variability after adjustment for socioeconomic, sociodemographic, and health indicators, highlighting spatial dependence between neighboring municipalities and differences in state-level pandemic management. The superior fit of the zero-inflated negative binomial model is consistent with the data structure, given the high proportion of zero counts for hospitalizations (37.5%) and deaths (87.2%), as well as overdispersion and heterogeneity across municipalities.

Finally, despite substantial financial resources, Brazil experienced high mortality and weak social protection. This raises a critical question: what might have occurred if income transfers had been more regular and better coordinated with social distancing measures?

Some limitations should be acknowledged. Ecological studies are useful for hypothesis generation but have limited causal inference, mainly due to ecological fallacy, whereby group-level associations may not apply to individuals. Reliance on aggregated data prevents control for individual-level confounders, reduces exposure precision, and hampers assessment of temporal relationships; therefore, results should be interpreted cautiously. In addition, we did not account for changes in eligibility criteria or benefit values for the EAP, considering only coverage for both EAP and vaccination. We also did not model counterfactual scenarios to estimate potential effects under alternative policy designs. Finally, individual-level factors related to health needs and susceptibility to COVID-19 could not be directly controlled.

In conclusion, this longitudinal ecological analysis shows that EAP coverage and COVID-19 vaccination were independently associated with lower COVID-19 hospitalization and mortality in Brazilian municipalities from March 2020 to December 2021. After adjustment for socioeconomic conditions, health service availability, comorbidities, and spatiotemporal effects, both income transfer and vaccination remained protective, with complementary roles in mitigating severe outcomes. These findings underscore the importance of coordinated social protection and public health policies in reducing the impact of health emergencies, especially in settings of structural inequality.

Alberto Lana.

The corresponding author, on behalf of the other authors guarantee the accuracy, transparency and honesty of the data and information contained in the study, that no relevant information has been omitted and that all discrepancies between authors have been adequately resolved and described.

M. Silva Monteiro de Castro: conceptualization, study design, data interpretation, writing, review and editing. A. Andrade de Souza: conceptualization, study design, data interpretation, writing, review and editing. G. Lopes de Oliveira: study design, data collection, data analysis, figures, writing, review and editing. G. Drummond Marques da Silva: study design, data collection, data analysis, figures, writing, review and editing. L. Lemos Jardim: literature search, writing, review and editing. W. Debôrtoli Miranda: literature search and editing. R. Paes de Sousa: conceptualization, project administration, writing, review and editing.

None.

None.