Background

The Coronavirus disease 2019 (COVID-19) caused a global pandemic that brought the whole world to a halt. National governments instituted sweeping lockdowns, rigorous physical distancing, and aggressive contact tracing in an effort to stop the spread of the virus. Although most cases arise from asymptomatic individuals, severe COVID-19 proves very fatal because it has long term effects on the lungs, heart, and even the central nervous system [1]. As of January 2022, COVID-19 has now reached 379.09 million cases worldwide with 5.70 million deaths [2]. As with previous health crises such as polio and measles, one of the most effective ways to curb their spread is through vaccination. Vaccines have been shown to be the safest method to prevent COVID-19 by directly conferring immune protection to inoculated individuals as well as indirectly through herd immunity [3,4,5]. Herd immunity is defined as the indirect protection conferred by a large proportion of immune individuals to susceptible individuals against infection [4, 6, 7]. The herd immunity threshold for the initial SARS-CoV-2 virus, or the proportion of immune individuals needed to observe a decline in incidence of infection, has been estimated to be 67% of every country’s population [3, 4, 6, 7]. Due to evolving variants, the virus increased its transmissibility by more than 50%, giving rise to the need to increase the herd immunity threshold to up to 80% of the population [8].

In the hopes of permanently putting a stop to the pandemic, a multinational race began to produce vaccines capable of building herd immunity. Countries such as the United States, Russia, Germany, and China, have put forward their own vaccines which have shown varying effectiveness in preventing COVID-19 and reducing its severity and mortality. Prominent vaccines currently administered are Pfizer-BioNtech, Moderna, AstraZeneca, Gamaleya, and Sinovac with each having been granted emergency use authorization [9].

Despite this progress, there are still significant barriers that must be faced before reaching herd immunity. As of January 2022, there are 4.16 billion fully vaccinated individuals, which only comprise 52.5% of the world population even with 31.92 million vaccines administered each day [2]. These data attest that achieving herd immunity may still be far ahead. One of the most significant challenges with vaccination is vaccine inequity. New COVID-19 cases and deaths continue to rise in places with high community transmission and low vaccination coverage, which may also lead to eventual vaccine resistance [10]. Despite the high rollout of vaccines in high- and upper-middle income countries, low- and middle-income countries (LMIC) lag far behind in terms of vaccine distribution. Only 41% of people in LMICs have received at least one dose as of January 2022, compared to the 74.8% of upper-middle income countries [2, 11]. In addition to vaccine inequity, there are still issues on vaccine hesitancy and questions regarding the vaccine efficacy of current vaccines against emerging COVID-19 variants [12]. These factors may further prolong the road to herd immunity, especially in LMICs. Although COVID-19 vaccines present the most viable way to end the pandemic, other strategies have to be explored as possible adjunct measures to bolster the current vaccination effort.

One such strategy that has been heavily studied is the possible role of Bacille Calmette-Guerin vaccine, otherwise known as BCG, against COVID-19. BCG is a live attenuated vaccine derived from Mycobacterium bovis and is the only vaccine approved for use in the prevention of pulmonary and extra-pulmonary tuberculosis [13]. Aside from tuberculosis, studies have proposed that BCG provides non-specific immunity against non-tuberculous infections and some immunotherapeutic benefit to malignancies such as non-muscle invasive bladder cancer [13, 14].

This could potentially explain the delayed COVID-19 spread in low-income countries, where BCG vaccination is currently mandated [15]. Some properties of BCG vaccines can also support COVID-19 vaccine design, such as its established safety profile, easy accessibility, and better public acceptance [16,17,18]. Based on these factors, it could be used as a model for future vaccine development [13, 14, 16,17,18] The aim of this study, therefore, is to evaluate the current evidence on the effectiveness of national BCG vaccination policies in reducing infection and mortality of COVID-19 and discuss the continued importance of BCG vaccines.

Methods

Study design

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA-P) as outlined by Shamseer et al. [19]. This protocol was published previously [20] and was registered with the International Prospective Register for Systematic Reviews (PROSPERO ID: CRD40221244060, Additional file 1). Ethics approval was not required due to the nature of the study.

Study setting

This study reviewed literature that reported on BCG vaccination policies and its association with COVID-19 severity and mortality. Literature was classified according to BCG vaccination policies of countries according to the BCG World Atlas, Our World in Data, and the World Health Organization (WHO).

Study period

Studies included in the review were those published between January 2020, which was when COVID-19 was initially reported to the WHO, until July 2021. Literature search was conducted from April to August 2021.

Search strategy

Search terms were generated using the Population/Intervention/Comparison/Outcomes (PICO) approach, guided by the research question “What is the effectiveness of national BCG vaccination policies in reducing infection and severity of COVID-19 in their native population?” Databases included PubMed, Cochrane, HERDIN, Web of Science, EBSCO, and Western Pacific Region Index Medicus (WPRIM). The search terms used are indicated in Table 1. Snowball search was employed by manual review of citations and reference sections of the studies for additional relevant articles [21]. Figure 1 shows the study selection flow chart following the PRISMA-P guidelines.

Table 1 Detailed search and selection strategy
Full size table
Fig. 1
figure 1

PRISMA flow diagram for searches of databases, registers, and other sources

Full size image

Inclusion and exclusion criteria

Articles included studies classified under Level 1A to 2C of the Oxford Center for Evidence-Based Medicine (2009) [22]. This included longitudinal studies, randomized controlled trials, ecological studies, and systematic reviews. Descriptive studies, commentaries, editorials, ongoing research, unpublished studies, and other working papers were excluded from the study. Articles not written in English, those published prior to January 2020, or those unavailable in full text were excluded. Search results were tabulated in Google Sheets for duplicate removal. JCO, MS, CE, JT, and MJMP individually screened the studies for eligibility. Any disagreements were resolved via consensus or discussion (Additional file 2).

Studies that met the inclusion criteria were assessed by two authors regarding methodological quality using the Joanna Briggs Institute critical appraisal instruments appropriate for the study design. A quality assessment checklist adapted from Betran et al. was used to appraise ecological studies (Additional file 3) [23]. The researchers agreed on a cut-off score of 70% of the total items of the quality assessment tools to be included in the review [24]. Disagreements between two reviewers were resolved by having a third reviewer assess the study. Further disagreements were resolved through discussion and consensus (Additional file 4).

Data extraction and synthesis

Extracted data from the selected studies were tabulated via Google sheets. Data included the authors, type of study, month and year of publication, population (countries included for ecological studies), source of BCG vaccination policy status, source of COVID-19 data, outcomes and confounding factors included, and the results of the study. JCO and CTE checked the accuracy of the recorded data.

Data synthesis consisted of comparing and contrasting the results of the outcomes and was interpreted using descriptive and inferential statistics applicable in the studies. Results of confounders were also discussed to better understand if the association between BCG vaccination policies and COVID-19 infection and severity was affected by other factors.

Results

A total of 220 studies were identified on initial literature search. After screening for duplicates and eligibility, 38 studies were sought for retrieval. Upon quality assessment, 13 studies were included in the final review. Figure 1 shows the PRISMA flow diagram depicting the literature search. Of the 13 studies, 11 were ecological and 2 were cohorts. Table 2 shows the characteristics of the studies included in the review. Studies were further categorized into those that used statistical correlation tests (e.g., Spearman, Pearson, etc.) and those that used regression analysis.

Table 2 Characteristics of reviewed studies
Full size table

Correlation between BCG vaccination and COVID-19 morbidity and mortality

Studies that used statistical correlation methods were synthesized together to assess the correlation between BCG policies and COVID-19 outcomes (Table 3). Four studies were included in this assessment. Statistical tests used by reviewed studies were either Spearman rho [28, 29] or Pearson [30, 37]. Two of the studies showed no significant correlation between mandatory BCG vaccination and COVID-19 outcomes while two studies showed strong correlation between BCG vaccination and decreased COVID-19 outcomes.

Table 3 Effects of BCG Vaccination on morbidity and mortality based on correlation
Full size table

Association between BCG vaccination and COVID-19 morbidity and mortality

To assess for overall cause-and-effect, studies that used regression analysis were grouped together and compared. Table 4 presents the findings of the studies included. Of the 12 studies that used regression analysis, only 3 studies did not find a significant association between BCG vaccination and COVID-19 outcomes [26, 27, 35]. One study found a significant association between BCG vaccination and COVID-19 deaths per million but found no significant association between BCG vaccination and COVID-19 deaths per case [29]. Figures 2 and 3 show the magnitude of association between BCG vaccination policies and COVID-19 morbidity and mortality. These studies indicate the large magnitude of association between vaccination policies and COVID-19 outcomes, which may demonstrate the protective effect of such policies.

Table 4 Effects of BCG vaccination on morbidity and mortality based on association and confounding variables
Full size table
Fig. 2
figure 2

Heat map of the magnitude of association between presence of national BCG Vaccination Policies and COVID-19 morbidity

Full size image
Fig. 3
figure 3

Heat map of the magnitude of association between presence of national BCG Vaccination Policies and COVID-19 mortality

Full size image

Effects of confounding variables

Even after controlling for confounding variables, most studies still found a significant association between BCG vaccination policies and decreased COVID-19 outcomes. The list of confounding variables identified per study is presented in Table 3. Positive associations were consistently observed between demographic variables, such as higher median age [29,30,31, 37] proportion of population > 65 years [25, 27, 28, 34], population density [25, 30, 31, 34], and COVID-19 outcomes. Evidence of a country’s higher economic status, namely high gross domestic product (GDP), high gross national income (GNI), and high World Bank classification, was consistently associated with higher COVID-19 incidence and mortality [28,29,30, 34, 35]. ‘Historic colonization status,’ an arbitrary variable classifying countries as to their history of participation in colonization efforts in the late 1400 s, was also used as a marker for advanced economies and was strongly associated with COVID-19 outcomes [29].

Some studies also observed relationships between other mandated vaccination programs and COVID-19 outcomes. Diphtheria, tetanus, and pertussis (DTP) coverage was not seen to be significant in decreasing COVID-19 outcomes [30, 35]. In two studies, measles vaccination coverage either showed a significant albeit weaker association to decreased COVID-19 outcomes compared to BCG or no association at all [32, 34, 35]. Pneumococcal, meningococcal, and influenza vaccinations were also assessed and were shown to have no significant effect on COVID-19 outcomes [36]. Lastly, oral polio vaccination showed no significant association to COVID-19 outcomes [34].

Discussion

Summary of the findings

Overall, most studies indicate an inverse association between BCG vaccination policies and COVID-19 incidence and mortality. However, this review presents a different finding from the study done by Ricco et al. early on in the pandemic, which previously concluded that there is no sound evidence to recommend BCG vaccination for the prevention of COVID-19 [38]. Findings may result from the inclusion of more recent studies on the topic, including those from early to mid-2021 [30, 32, 34, 35]. For one, Ricco et al. report that only two out of the 14 studies reviewed controlled for confounding variables. Meanwhile, most studies included in this review controlled for some or all confounding variables reported. In addition, Ricco et al. point out that COVID-19 outcomes highly depend on the reliability of reporting COVID-19 cases and deaths. Although consistent with other studies [29, 39], this may have been a more significant factor in the early phases of the pandemic because of its more erratic and uncertain course at the time [40]. As the pandemic continued, more robust testing and reporting strategies were implemented [30, 41], thus increasing the raw data quality.

Nonspecific protective effect of BCG vaccination

Numerous studies have proposed the nonspecific immune effect of BCG on unrelated infections as a possible mechanism to explain its protective effect against COVID-19. BCG has long been reported to grant "trained immunity," a nonspecific innate immune response resulting from epigenetic reprogramming. Through epigenetic reprogramming, BCG modifies human monocytes leading to increased cytokine production. This epigenetic programming explains BCG's ability to confer protection against unrelated infections [42, 43]. In addition, possible long-term effects are brought about by heterologous T helper 1 (Th1) and Th17 immune responses [43] and more rapid seroconversion from enhanced leukocyte response and robust cytokine response to unrelated pathogens. This long-term effect may explain the protection BCG confers on individuals years after vaccination, which is significant since policies usually mandate its inoculation immediately after birth.

Studies included in this review have demonstrated this nonspecific protective effect. Rivas et al. explored the serologic effects of previous BCG vaccination on COVID-19 [36]. The study demonstrated that a history of BCG vaccination in healthcare workers lowers the incidence of clinically or laboratory-confirmed diagnosis of COVID-19. The study also demonstrated that prior BCG vaccination was associated with lower seroconversion and anti-SARS-CoV-2 IgG index even when controlled for age and sex [36]. These lower serological features are still consistent even in the presence of co-morbidities, even though these conditions increase the likelihood of acquiring COVID-19. This study, therefore, gives credence to the supposed nonspecific protection afforded by prior inoculation with BCG.

Compared to other vaccines, only BCG has consistently shown a robust significant association in decreasing COVID-19 outcomes. Studies reported on DTP, meningococcal, pneumococcal, measles, influenza, and oral polio against COVID-19 outcomes have failed to produce the same protective effect compared to BCG [30, 32, 34,35,36]. Results may further illustrate that BCG vaccination alone could confer nonspecific immunity, thereby reducing incidence and mortality related to COVID-19. This stresses the importance of BCG vaccines, not only for its potential against COVID-19, but also against other infectious diseases and non-communicable diseases such as malignancies [13]. Additional studies on BCG must be done to augment these findings, not only to help eradicate tuberculosis but also to increase its potential as a model for future vaccine design [13, 17, 18].

Confounding variables and lower COVID-19 adverse outcomes

This study is the first to synthesize the effects of the common confounding variables reported among studies on BCG and COVID-19 adverse outcomes. Although most of the studies controlled for these variables, their effect on a higher incidence of COVID-19 may still have played a role in the perceived effectiveness of BCG policies against COVID-19.

Markers of the high economic status of a country (e.g., GDP, GNI, World Bank classification) have consistently been associated with increased COVID-19 incidence and mortality [28,29,30, 34, 35]. Those higher-income countries are also likely to have phased out their mandated BCG policies [27, 28, 35]. However, it is essential to note that these higher-income countries are likely to be more developed and may have better healthcare infrastructure, thus enabling adequate testing and contact tracing measures [29, 30, 32]. Consequently, this increases the recorded COVID-19 adverse outcomes in these countries compared to those of lower-income countries, which may suffer from underreporting of cases. This may have contributed to the relatively decreased incidence in countries with a current mandated BCG policy [29, 30] and may therefore underscore the protective effect of BCG against COVID-19 outcomes.

Studies have also reported positive associations between specific population demographics and COVID-19 adverse outcomes. For example, countries with higher median age and population over 65 have shown an increase in COVID-19 incidence and mortality [25, 27,28,29,30,31, 34, 36]. These two factors show that a higher proportion of older residents in these countries may have increased COVID-19 adverse outcomes resulting in more susceptibility to acquiring COVID-19 and exhibiting severe manifestations leading to death [44]. Furthermore, in relation to this study, an older-aged population also reflects these countries' higher economic status, which may have also contributed to the increased incidence and mortality to COVID-19 in these high-income countries [28, 30].

Significance of the findings

This review was able to collate results from various papers about the effectiveness of BCG vaccine policies with COVID-19. Unlike the previous review by Ricco et al., the current study was able to find evidence that supports the protective effect of such policies against COVID-19. The study was also able to account for the influence of various confounding variables in the association. However, due to most studies being ecological or observational, actual cause-and-effect between BCG and COVID-19 incidence and mortality cannot be ultimately concluded. The studies reviewed also point out the possibility that the supposed protective effect may have merely been due to the effect of other variables.

This study, therefore, highlights the importance of conducting clinical trials to determine the effectiveness of an intervention adequately. Short papers and case studies on the topic have shown conflicting results, demonstrating either no protective effect [45, 46] or some effect such as lower hospitalization rates [47]. Studies on animal models have also been performed which may provide physiologic evidence for the perceived protective effect [48]. Recently completed randomized controlled trials are now attesting to the protective effect of BCG, supporting the findings of this study [49,50,51]. In contrast, some early clinical trials show no effect of BCG vaccination on COVID-19 infection [52, 53]. Because of the conflicting nature of the results, these data should be interpreted with utmost prudence [54]. It is therefore still recommended that further studies be continued to map the true effect of BCG vaccination against COVID-19, especially with the rise of variants and the possibility of waning immunity with the current vaccines [55]. Studies have now even shifted to explore the possibility of using BCG as a booster for current COVID-19 vaccines [56]. Further studies to be performed at the molecular or cellular level may also provide further basis for this protective effect. These can help in conclusively establishing a link between BCG vaccination and lower COVID-19 adverse outcomes and may also aid in the production of future vaccines.

Limitations and recommendations

This study has several limitations. For one, our database was only limited to open access journals and those that were published in English. Moreover, our study data gathering was only limited until August 2021. With the COVID-19 situation continuously evolving and with even more papers being published regarding this disease, recent breakthroughs published after the said date might not have been included in this study. We recommend future researchers to include recent data and information not covered in this study.

Conclusion

Even though the lower incidence and mortality due to COVID-19 may not merely be attributable to BCG vaccination policies, there is still a large amount of evidence that shows their possible protective effect. Only actual clinical trials, which are just recently being completed, may truly be able to elucidate the actual effect of BCG vaccination against COVID-19. Therefore, before recommendations can be made regarding the use of BCG vaccination as an adjunct solution to the pandemic, further studies must be done to adequately report its efficacy, including further review of these clinical trials. Lastly, further research to improve BCG vaccines must be continued to strengthen its effectiveness in not only eliminating tuberculosis but also in aid of reducing the morbidity and mortality of other communicable and non-communicable diseases.