Back

Association of BCG vaccination policy with prevalence and mortality of COVID-19

Sala, G.; Miyakawa, T.

2020-04-06 epidemiology
10.1101/2020.03.30.20048165 medRxiv
Show abstract

BackgroundEvidence suggests non-specific benefits of the tuberculosis vaccine bacillus Calmette-Guerin (BCG) against non-related infections. Recent studies propose such protection may extend to the novel COVID-19 as well. This is a contested hypothesis. MethodsOur ecological study confronts this hypothesis. We examine the effects of BCG vaccination on countries COVID-19 (a) cases and deaths (per million) and (b) exponential growth factors over specific periods of the pandemic. Since the BCG vaccine was derived from Mycobacterium bovis, a bacterium causing tuberculosis in cattle, having suffered from tuberculosis also may exert a non-specific protection against the COVID-19 as well. Along with BCG vaccination, we test the effect of the prevalence of tuberculosis. We employ multiple regression and principal component analysis (PCA) to control for potentially confounding variables (n = 16). ResultsBCG vaccination policy and incidence of tuberculosis is associated with a reduction in both COVID-19 cases and deaths, and the effects of these two variables are additive ({approx} 5% to 15% of total unique variance explained). The study of exponential growth factors in the initial stages of the pandemic further shows that BCG vaccination exerts a significant effect (up to 35% of unique variance explained). ConclusionsOverall, these findings corroborate the hypothesis that BCG vaccination and exposure to tuberculosis may induce a non-specific protection against the novel SARS-CoV-2 infection, even after accounting for a large number of confounding influences. However, given the potential public-health benefits, our results indicate that the hypothesis deserves further attention and should not be hastily dismissed.

Matching journals

The top 9 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.