COVID-19 Related Mortality and The BCG Vaccine

Introduction

The coronavirus disease 2019 (COVID-19) is the result of infection by the novel coronavirus SARS-CoV-2; previously 2019-nCoV. First detected in late December 2019 in Wuhan, China; it has since spread to 6 continents and is estimated to have infected over 2.1 million people and caused over 145,000 deaths worldwide as of April 16, 2020.^{1, 2, 3} The rapid spread of the disease around the world has become a worldwide emergency. The high mortality of the disease in certain populations (i.e. the elderly), along with the lack of vaccinations or effective antiviral treatment have resulted in significant efforts in trying to find interventions that would improve outcomes.^{4, 5} Early studies have shown an inverse relationship between the bacille Calmette-Guérin (BCG) vaccination policy of countries and COVID-19 related mortality. Countries with a current universal BCG vaccination policy seem to have significantly lower mortality rates when compared to countries that do not implement BCG in their childhood vaccination schedules.^{6, 7, 8} The immunostimulant properties of the vaccine could in theory improve outcomes through heterologous immunity, which refers to the cross-protection the immune system confers between two unrelated pathogens. ⁹

Past studies support the notion of BCG induced heterologous immunity as they have found a BCG-associated reduction in all cause mortality within the first years of life.^10,11 This reduction in mortality has been found to be more noticeable in regions with high childhood mortality and infectious disease burden such as West Africa.^12,13 Similarly, the rates of infection by non mycobacterial pathogens, were found to be lower in children vaccinated with BCG.¹⁴ In a study conducted by Hollm et al. in 2014, more than 150,000 children in 33 different countries were found to have a risk reduction of 17% to 37% for the development of acute lower respiratory infections.¹⁵

All these findings seem to suggest that the BCG might indeed be associated to improved outcomes of COVID-19. However, some studies debate the efficacy of the heterologous immunity induced by the BCG vaccine. The Danish Calmette Study, conducted by Lone Graff et al., found no differences in the rates of infection or hospitalization in children up to the age of 15 months; partially explained by the lower infection disease burden in the region.¹⁶ Moreover, the protective effect conferred by the BCG seems to wane with time and a previous study conducted by Yu et al. in 2007 found no cross protection between BCG and the Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV), which belongs to the same family as SARS-CoV-2.^{14, 17} Regarding the negative association between universal BCG vaccination and COVID-19 related mortality, several confounders were identified by the authors themselves. Under-reporting and reduced SARS-CoV-2 capability testing in low income countries, along with comparing mortality rates at different times in each countries’ epidemiological curve, could all serve as potential confounders.^6,7,8

This paper tries to account for these potential confounders for the first time. This study would then serve in answering the question as to whether a statistically significant difference in mortality rates exists between countries with differing BCG vaccination policies. We gathered data from publicly available sources and compared the number of COVID-19 related deaths per 1 million inhabitants as well as the number of deaths at the time the countries hit the 1000^th COVID-19 case. To avoid any possible bias on each country’s capability for testing, reporting and their resources available for healthcare, only high income countries were compared.¹⁸ Our findings suggest no statistical difference exists between countries with different vaccination programs.

Material and Methods

The study conducted is an observational ecological study, comparing the mortality rates of countries with differing BCG vaccination programs. The data retrieved included: total population, total number of COVID-19 cases, total number of COVID-19 attributable deaths, COVID-19 attributable deaths per 1 million population, COVID-19 attributable deaths at the 1000^th case, percentage of the population over 65 years old, and the year of introduction and cessation of the BCG vaccine where applicable.

The number of total COVID-19 cases, total COVID-19 attributable deaths, deaths per 1 million population and deaths at the 1000^th case were retrieved from the Worldometer website.³ Data on the total population, percentage of the population over 65 years old and the GDP income was retrieved from The World Bank Data Website.¹⁹ Data on BCG vaccination policies was retrieved from The BCG World Atlas.²⁰ The IBM SPSS program was then used to analyze statistical data.

The inclusion criteria were countries with:

• At least 1,000 COVID-19 cases per the Worldometer website.

• At least 1 million inhabitants per The World Bank Data Website.

• A Gross National Income (GNI) per capita higher than 12,375 US dollars, per The World Bank Data Website.

Countries with no data available were excluded from the study.

Countries that met all criteria were divided in three groups:

1. Countries with no history of universal BCG vaccination policy.

2. Countries with an actual vaccination policy.

3. Countries with a prior universal BCG vaccination policy.

For the third group, countries that had a universal BCG vaccination policy introduced in 1955 or earlier that lasted until at least 1975 were included (as to ensure that the elderly population would have been vaccinated).

For each group, the mean and standard deviation of COVID-19 attributable deaths per 1 million population and number of deaths at the 1000^th case were calculated from the combined data of each country in the respective group.

First compared was mean and standard deviation of the number of COVID-19 attributable deaths per 1 million population. Countries with no history of universal vaccination programs were compared to the combined data of countries with a current or prior vaccination policy using the t-test. Then, the three groups were compared separately using the ANOVA test. Afterwards, the number of COVID-19 attributable deaths at the 1000^th was compared following the same method, presuming a normal distribution data where pertinent.

Finally, adjustments were made to account for differences in demographics. The mean and standard deviation was calculated for the percentage of the population over 65 years old for all the countries included in the study. The previous comparisons were repeated, this time excluding countries that strayed more than 1.5 standard deviations from the mean.

Results

After analysis, 20 countries met the inclusion criteria for the study. Out of these, 5 high income countries never had a universal BCG vaccination policy (table 1), 5 currently have a universal BCG vaccination policy and 10 had a previous BCG vaccination policy introduced in 1955 or earlier; with its respective withdrawal no earlier than 1975 (table 2).

The mean number of deaths per 1 million population for countries with no universal BCG vaccination (223.2 ± 166.1) was not statistically significant from countries with current or previous BCG vaccination programs (55 ± 82.5) [95% Confidence Intervals (CI) (−34.8, 371); p = 0.85]. When comparing the means of the 3 groups separately, a statistically significant difference was noted between countries with a current BCG vaccination program (7.8 ± 4.8) and countries that never implemented a universal BCG vaccination policy (223.2 ± 166.1) [95% CI (−386.5, −44.2); p = .013].

The mean number of deaths at the 1000th case for countries with current or previous BCG vaccination programs and countries that never implemented BCG vaccination in their schedules was 15.7 ± 17.5 and 20.2 ± 9.2 respectively. No statistically significant difference was noted between these two groups [95% CI (−12.9, 21.8); p value = .597]. When comparing the mean number of deaths of the three different groups, a statistically significant difference was noted between the means of countries with an actual BCG program (29.2 ± 23.1) and countries with previous BCG programs (9 ± 9.4) [95% CI (0.66, 39.7); p value = .042]. (Graph 1)

• Download figure
• Open in new tab

Graph 1 Number of SARS-CoV-2 Related Deaths at the 1000th Case

Data on BCG vaccination policies was retrieved from The BCG World Atlas. The number of COVID-19 cases and COVID-19 attributable mortality were retrieved from the Worldometer website. No statistically significant difference was found when comparing mortality at the 1000th case. (p = > 0.05).

It is important to note that some countries have a different constitution in the age of their population. The mean percentage and standard deviation of adults aged >65 in the 20 countries was 18.7 ± 3.6; Japan (28.2%), Israel (11.3%) and Chile (11.8%) were more than 1.5 SD’s away from the mean. When comparing the number of deaths at the 1000th case and the total number of deaths per 1 million population of the three groups, no statistically significant difference was found when these countries were excluded from the analysis (p = > 0.05).

Discussion

The results found in this study seem to suggest no statistically significant difference in mortality arises from BCG vaccination. This is in contrast to ealy studies that found an inverse relationship between mortality rates and BCG vaccination status.^6,7,8 A striking difference in deaths per 1 million population was noted between current BCG using countries and countries that never implemented a universal vaccination program, which is similar to the results found in previous studies. However, just as the WHO noted, this doesn’t reflect demographic differences, disease burden and the different stages of the pandemic in these countries.²¹

When trying to account for confounders by comparing the number of deaths at a particular point in time and considering demographic differences, no statistically significant difference was found. Furthermore, the number of deaths at the 1000th case were actually lower in countries that had a prior BCG vaccination program (9 ± 9.4) when compared to countries with an actual universal vaccination policy (29.2 ± 23.1). Nevertheless, this finding was brought into question as no difference was found when accounting for demographic differences in these countries.

These results seem to fit in line with previous findings regarding BCG-induced heterologous immunity. Findings seem to suggest an increase in immune response, however this is highly dependent on the time the research subjects received the BCG vaccine. The population that has been previously shown to benefit from BCG-induced heterologous immunity is restricted to the pediatric population.^{10-15, 22-23} Furthermore, this benefit seems to only be apparent when the regions involved already have high childhood mortality, as evidenced by the Danish Calmette Study.¹⁶ The pediatric population can contract COVID-19 and even serve as potential carriers because of their mostly asymptomatic course.^{24, 25} However, this fact alone is not enough evidence to suggest a change in the vaccination program of countries that do not implement BCG in their schedules. This if further emphasized by the apparent lack of cross protection the BCG vaccine confers against coronaviruses. This was evidenced by a study conducted by Yu et al. in 2007 when he looked at the potential protection childhood vaccines conferred against SARS-CoV.¹⁷

Although childhood vaccination seems to have no effect on mortality rates of the adult population, the BCG vaccine could still prove useful. The early genome sequencing of SARS-CoV-2 is helping current efforts in developing a vaccine against it. The use of recombinant technology could help with the early development of a vaccine by using vectors and embedding them with SARS-CoV-2 antigens.²⁶ Recombinant BGC vaccines could help in this field as they induce a considerable immune response and have been used before as vectors to non-mycobacterial related antigens.^27-29 Also of interest is the potential effect revaccination might have on the elderly population. The potential of BCG-induced heterologous immunity after revaccination has shown promise, as demonstrated by Warhana et al.³⁰ Two current clinical trials are looking into this by vaccinating healthcare workers with BCG to see if there is any potential benefit.^{31, 32}

There are some limitations to be considered in this study. Although great care was taken to try to avoid as many confounders as possible, some still remain. The difference in mortality seems to be higher in the male population than in the female population, which could affect the overall mortality in countries with differing gender compositions.³³ Also, some populations are more genetically susceptible to certain diseases and complications than others (e.g. Latin Americans and diabetes).³⁴ Some researchers are looking into this by analyzing genetic differences in the expression of the ACE2 receptor or the CCL2 and MBL genes, which could affect susceptibility to SARS-CoV-2.^{6, 35, 36} Finally, each country’s response to the outbreak affects subsequent policies taken by other countries and although the coverage of countries currently using BCG was over 90% no data was available for countries with prior vaccination programs.²⁰

Conclusion

The status of BCG vaccination policy does not seem to have an effect in COVID-19 related mortality. Although data is difficult to analyze due to the presence of multiple confounders, statistically significant differences seem to disappear once much of the confounders are taken into consideration. Nevertheless, the immunostimulant properties of the BCG vaccine could prove useful in future developments of a vaccine using recombinant technology. Likewise, revaccination with BCG has shown promise in the past for the induction of heterologous immunity. Current medical trials are underway looking at the potential benefits of vaccinating healthcare workers.

Conflicts of interest

No conflicts of interest to disclose.