| Description: |
Background: Bacille Calmette-Guérin (BCG) vaccination has off-target (non-specific) effects that are associated with protection against unrelated infections and decreased all-cause mortality in infants. We aimed to determine whether BCG vaccination prevents febrile and respiratory infections in adults. Methods: This randomised controlled phase 3 trial was done in 36 healthcare centres in Australia, Brazil, the Netherlands, Spain, and the United Kingdom. Healthcare workers were randomised to receive BCG-Denmark (single 0.1 ml intradermal injection) or no BCG in a 1:1 ratio using a web-based procedure, stratified by stage, site, age, and presence of co-morbidity. The difference in occurrence of febrile or respiratory illness were measured over 12 months (prespecified secondary outcome) using the intention-to-treat (ITT) population. This trial is registered with ClinicalTrials.gov, NCT04327206. Findings: Between March 30, 2020, and April 1, 2021, 6828 healthcare workers were randomised to BCG-Denmark (n = 3417) or control (n = 3411; no intervention or placebo) groups. The 12-month adjusted estimated risk of ≥1 episode of febrile or respiratory illness was 66.8% in the BCG group (95% CI 65.3%–68.2%), compared with 63.4% in the control group (95% CI 61.8%–65.0%), a difference of +3.4 percentage points (95% CI +1.3% to +5.5%; p 0.002). The adjusted estimated risk of a severe episode (defined as being incapacitated for ≥3 consecutive days or hospitalised) was 19.4% in the BCG group (95% CI 18.0%–20.7%), compared with 18.8% in the control group (95% CI 17.4%–20.2%) a difference of +0.6 percentage points (95% CI −1.3% to +2.5%; p 0.6). Both groups had a similar number of episodes of illness, pneumonia, and hospitalisation. There were three deaths, all in the control group. There were no safety concerns following BCG vaccination. Interpretation: In contrast to the beneficial off-target effects reported following neonatal BCG in infants, a small increased risk of symptomatic febrile or respiratory illness was observed in ... |