Background: Socially vulnerable populations were vastly affected by the COVID-19 pandemic. The pandemic significantly impacted Brazil, pressuring its healthcare system for several months, with high mortality rates, even among the youngest population. Cohort studies combining disease surveillance are essential for understanding virus circulation in the community, surrogates of protection, vaccine effectiveness, and demand for health resources. Methods: Here, we present the protocol for a community-based prospective cohort study in the largest complex of favelas (slums) in Rio de Janeiro, Brazil (Complexo da Maré). The study participants are residents initially recruited during a massive vaccination campaign in the community. Five waves of data collection at approximately six-month intervals were planned. The first two waves have been completed at the time of writing this study protocol, and the third is underway. The protocol comprises interviews, blood sampling, and records linkage with secondary data to enrich the profiles of cohort participants and community information. We will describe COVID-19 seroprevalence, socio-demographic characteristics, and the burden of COVID-19, followed by estimating the association of socioeconomic factors and the burden of disease with seroprevalence. Discussion: The primary aims of the study are to assess COVID-19 clinical, epidemiological and genomic profiles and outcomes in residents from Maré, including vaccine effectiveness, surrogates of immune protection, virus transmission in households, and the overall burden of the pandemic.

IntroductionFew community-based interventions addressing the transmission control and clinical management of COVID-19 cases have been reported, especially in poor urban communities from low-income and middle-income countries. Here, we analyse the impact of a multicomponent intervention that combines community engagement, mobile surveillance, massive testing and telehealth on COVID-19 cases detection and mortality rates in a large vulnerable community (Complexo da Maré) in Rio de Janeiro, Brazil.MethodsWe performed a difference-in-differences (DID) analysis to estimate the impact of the multicomponent intervention inMaré,before (March–August 2020) and after the intervention (September 2020 to April 2021), compared with equivalent local vulnerable communities. We applied a negative binomial regression model to estimate the intervention effect in weekly cases and mortality rates inMaré.ResultsBefore the intervention,Marépresented lower rates of reported COVID-19 cases compared with the control group (1373 vs 1579 cases/100 000 population), comparable mortality rates (309 vs 287 deaths/100 000 population) and higher case fatality rates (13.7% vs 12.2%). After the intervention,Marédisplayed a 154% (95% CI 138.6% to 170.4%) relative increase in reported case rates. Relative changes in reported death rates were −60% (95% CI −69.0% to −47.9%) in Maré and −28% (95% CI −42.0% to −9.8%) in the control group. The case fatality rate was reduced by 77% (95% CI −93.1% to −21.1%) inMaréand 52% (95% CI −81.8% to −29.4%) in the control group. The DID showed a reduction of 46% (95% CI 17% to 65%) of weekly reported deaths and an increased 23% (95% CI 5% to 44%) of reported cases inMaréafter intervention onset.ConclusionAn integrated intervention combining communication, surveillance and telehealth, with a strong community engagement component, could reduce COVID-19 mortality and increase case detection in a large vulnerable community in Rio de Janeiro. These findings show that investment in community-based interventions may reduce mortality and improve pandemic control in poor communities from low-income and middle-income countries.