Impact of Routine Infant BCG Vaccination on COVID-19

Abstract

Objectives: In Japan, the first case of coronavirus disease 2019 (COVID-19) was diagnosed on January 15, 2020 and subsequent infections rapidly increased. The Bacillus Calmette-Guérin (BCG) vaccination program is the principal element of tuberculosis control in Japan. We investigated the impact of routine infant BCG vaccination on prevention of local COVID-19 spread.

Methods: Data on the prevalence of SARS-CoV-2 infection, annual routine infant BCG vaccine coverage (represented by the number of BCG vaccinations per live births), and other candidate factors in each prefecture were obtained from the official notifications database in Japan. We analysed the association of vaccine coverage with the prevalence of SARS-CoV-2 infection.

Results: The BCG vaccine coverage in 1999-2002, 2004, and 2012 in five prefectures with no COVID-19 infections was significantly higher than that in five prefectures with a high prevalence of infections (Mann-Whitney U test, p<0.05). The prevalence of SARS-CoV-2 infection was significantly negatively correlated with BCG vaccine coverage in 2004 and was significantly positively correlated with age groups 20-34 and 40-54 years (Spearman's rank correlation, p<0.01).

Conclusions: Our findings suggest that routine infant BCG vaccination coverage in young generation had a significant impact on prevention of local COVID-19 spread in Japan.

Keywords: BCG; Bacillus Calmette-Guérin vaccination; COVID-19; Coronavirus disease 2019; Prevention.

Fig. 1

Annual change in routine infant BCG vaccine coverage. Top: Change in annual number of routine infant BCG vaccines (grey squares) and live births (black diamonds). The number of live births has gradually decreased over the last 20 years. The number of BCG vaccines has fluctuated according to changes in the routine vaccination system. Bottom: Annual change in routine infant BCG vaccine coverage, represented by the number of BCG vaccines per number of live births. *p<0.05, Mann–Whitney U test. ^aCOVID-19 prevalence on March 29 (patients per million).

Fig. 2

Correlation between COVID-19 prevalence and related factors. Scatter plot of COVID-19 prevalence (ordinate) against related factors (abscissa). Prevalence of (a) patients with COVID-19 on March 29, (b) SARS-CoV-2 PCR+ individuals on April 24, and (c) deaths on April 24 showed significant negative correlation with BCG vaccine coverage in 2004, the population age 65 years or more, and workers in the primary sector of industry; and significant positive correlation with days from the first patient diagnosis with COVID-19 up to March 29. Dotted squares indicate significant correlation (Spearman's rank correlation, p<0.01). Regression lines are shown for reference.

Fig. 3

Correlation between prevalence of patients with COVID-19 on March 29 and related factors. Scatter plot of the prevalence of patients with COVID-19 (ordinate) against related factors (abscissa). (A) Significant positive correlation of age groups 20–24, 25–29, 30–34, 40–44, 45–49 and 50–54 years with COVID-19 prevalence. (B) Workers in the tertiary sector of industry were significantly correlated with SARS-CoV-2 infection in a positive manner. Dotted squares indicate significant correlation (Spearman's rank correlation, p<0.01). Regression lines are shown for reference.

Fig. 4

Number of PCR+ individuals (top), seriously ill patients and deaths owing to COVID-19 (middle), and their proportions (bottom) in each age group on May 1. A drastic increase in case fatality rate was observed with age, especially among individuals age 70 years or older. Shaded squares show age groups born before commencement of routine BCG vaccination for infants using qualified freeze-dried BCG.