SARS CoV-2 seroprevalence and diagnostic accuracy during a COVID-19 outbreak in a major penitentiary complex in Brazil, June to July 2020.

Purpose: This study aims to estimate the overall SARS-CoV-2 seroprevalence and evaluate the accuracy of an antibody rapid test compared to a reference serological assay during a COVID-19 outbreak in a prison complex housing over 13,000 prisoners in Brasília.

Design/methodology/approach: The authors obtained a randomized, stratified representative sample of each prison unit and conducted a repeated serosurvey among prisoners between June and July 2020, using a lateral-flow immunochromatographic assay (LFIA). Samples were also retested using a chemiluminescence enzyme immunoassay (CLIA) to compare SARS-CoV-2 seroprevalence and 21-days incidence, as well as to estimate the overall infection fatality rate (IFR) and determine the diagnostic accuracy of the LFIA test.

Findings: This study identified 485 eligible individuals and enrolled 460 participants. Baseline and 21-days follow-up seroprevalence were estimated at 52.0% (95% CI 44.9-59.0) and 56.7% (95% CI 48.2-65.3) with LFIA; and 80.7% (95% CI 74.1-87.3) and 81.1% (95% CI 74.4-87.8) with CLIA, with an overall IFR of 0.02%. There were 78.2% (95% CI 66.7-89.7) symptomatic individuals among the positive cases. Sensitivity and specificity of LFIA were estimated at 43.4% and 83.3% for IgM; 46.5% and 91.5% for IgG; and 59.1% and 77.3% for combined tests.

Originality/value: The authors found high seroprevalence of anti-SARS-CoV-2 antibodies within the prison complex. The occurrence of asymptomatic infection highlights the importance of periodic mass testing in addition to case-finding of symptomatic individuals; however, the field performance of LFIA tests should be validated. This study recommends that vaccination strategies consider the inclusion of prisoners and prison staff in priority groups.