Brazil was severely affected by the COVID-19 pandemic, and cases and mortality data had problems of timeliness and correctness, hindering risk analysis. We have implemented a mixed-methods approach providing data interoperability to understand the relationship between risk and inequality. We implement a GIS-based risk score and focus groups in São Paulo. The risk score includes age structure, social vulnerability, and behavior. The COVID-19 risk score results identified more frequent high-risk districts among peripheral and vulnerable areas influenced by social vulnerability. Risk was also strongly influenced by behavior, hence varying over time. We found that the average risk score may overestimate the risk for privileged population groups and underestimate it for vulnerable groups. To verify this issue, we extended our analysis with qualitative data comparing COVID-19 impact between focus groups at opposing ends of the vulnerability spectrum: in the city center and a vulnerable, peripheral, and informal community. The focus groups reveal further inequality in the pandemic impact, coupling health issues with social, economic, and mental health problems. Our findings highlight the importance of providing data-interoperability and mixing methods at the local scale to account for contextual and societal risk factors regarding social vulnerability.