BCG epidemiology supports its protection against COVID-19? A word of caution.

The COVID-19 pandemic, caused by type 2 Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV-2), puts all of us to the test. Epidemiologic observations could critically aid the development of protective measures to combat this devastating viral outbreak. Recent observations, linked nation based universal Bacillus Calmette-Guerin (BCG) vaccination to potential protection against morbidity and mortality from SARS-CoV-2, and received much attention in public media. We wished to validate the findings by examining the country based association between COVID-19 mortality per million population, or daily rates of COVID-19 case fatality (i.e. Death Per Case/Days of the endemic [dpc/d]) and the presence of universal BCG vaccination before 1980, or the year of the establishment of universal BCG vaccination. These associations were examined in multiple regression modeling based on publicly available databases on both April 3rd and May 15th of 2020. COVID-19 deaths per million negatively associated with universal BCG vaccination in a country before 1980 based on May 15th data, but this was not true for COVID-19 dpc/d on either of days of inquiry. We also demonstrate possible arbitrary selection bias in such analyses. Consequently, caution should be exercised amidst the publication surge on COVID-19, due to political/economical-, arbitrary selection-, and fear/anxiety related biases, which may obscure scientific rigor. We argue that global COVID-19 epidemiologic data is unreliable and therefore should be critically scrutinized before using it as a nidus for subsequent hypothesis driven scientific discovery.

Introduction

There is a current global crisis from the Coronavirus Disease of 2019 (COVID-19) pandemic [1]. COVID-19 is caused by type 2 Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV-2), which is a medium-sized, enveloped, positive-stranded RNA virus of the Coronaviridae family. SARS-CoV-2 is the pathogen of the third, large, severe respiratory syndrome outbreak caused by Coronaviruses (CoVs) (1: SARS [severe acute respiratory syndrome] which emerged in late 2002 and disappeared by 2004; 2: MERS [Middle East respiratory syndrome], which emerged in 2012 and remains in circulation in camels]) [2].

COVID-19 cases and associated deaths continue to rise [3], which naturally induces fear, anxiety and sadness in all of us. Time is of essence towards finding definite solutions for stopping the pandemic, and scientists are under significant pressure trying to balance speed with safety and precision [4]. Since fear and sadness can alter our cognitive control [5], there is valid concern about loosened scientific rigor in respect to the massive surge of publications amidst the time pressure on biomedical scientists racing for a cure. Even though the outbreak likely began in December of 2019 in Wuhan of Hubei Province in China [6], there are ongoing uncertainties about SARS-CoV-2 epidemiology. Rigorous studies (including multiple site and repeated nucleic acid based-, and also viral culture based testing) in 9 symptomatic patients with mild disease course have shown active viral replication in the upper airway, and high viral shedding in pharynx (but lower than in sputum) peaking at 4 days of symptoms [7]. SARS-CoV-2 virus was readily isolated from throat- and lung-derived samples, but not from stool, in spite of high virus RNA concentration in the fecal samples [7]. Blood and urine never yielded live virus [7]. On the contrary, among the first cases in Europe, at least RNA based viremia was detected in a severe form of the disease progressing to multi-organ failure [8]. Investigators from China found the highest SARS-CoV-2 viral load in throat swabs at the time of symptom onset, and inferred that infectiousness peaked on or before symptom onset [9]. These findings underscore the primary respiratory spread of SARS-CoV-2, at least during mild disease course, which has been traditionally considered to be large droplet/contact communicated based on findings with SARS and MERS. However, some work indicates the potential airborne spread of the virus [10], leading to debates amongst infectious disease experts between contact vs. airborne protection recommendations. Adding to the difficulties in making clear cut regulations for personal protective equipment (PPE) use is viral shedding before symptomatic presentation [9]. Importantly, a large proportion of infected people can be asymptomatic who can spread the virus [11]. Such viral spreading may be prominent in children and young adults, with over 50% being asymptomatic or having mild disease [12, 13]. Additionally, some patients present with gastrointestinal complaints, and never develop respiratory symptoms [14], for example. Furthermore, due to limitations in nucleic acid based analysis including quality of sample collection and variable methodology [15], a single nasopharyngeal swab can have as low as 32% sensitivity over the course of infection [16]. In the meantime, an asymptomatic patient may have similar viral loads as symptomatic ones, indicating the transmission potential from such cases [17]. These observations add to the tremendous difficulties in developing clear and consistent guidelines for PPE use during the COVID-19 pandemic, especially in the medical setting.

As already mentioned, similarly to MERS [18] and SARS [19], pediatric patients with COVID-19 run a much milder disease course than the elderly (especially above 60 years of age) [20]. The exact reason for this is unknown [13], but at least in non-human primates with experimental SARS-CoV infection, immune responses (most prominently CD8 T cell and B cell associated) were greatly reduced in the aged host compared to younger animals [21]. Consequently, a number of immunomodulatory treatments are being explored to help patients in fighting the infection. Amidst these explorations, the effect of prophylactic Bacillus Calmette-Guerin (BCG) vaccination on COVID-19 outcomes was being investigated in 14 registered clinical trials (on clinicaltrials.gov) as of June 16th 2020. Linking to these investigations is a not yet peer-reviewed paper [22], suggesting a connection between universal BCG vaccination policy and the peculiarly significant nation based variation in case frequency and death rates from COVID-19, based on data from March 21st, 2020. This manuscript received both public and scientific attention, supporting [23] and contradicting [24] it. A more recently published work found significant correlation between an arbitrarily designed, country specific BCG index and COVID-19 mortality, following attempts to control for confounding variables [25]. The epidemiology of the pandemic, however, is in ongoing flux. Therefore, we decided to examine this question with a modified definition of COVID-19 death rate based on both April 3rd, and May 15th 2020 data.

Methods

COVID-19 epidemiologic data was extracted from [3] on the afternoon of April 3rd, 2020 for the top 68 countries based on number of cases. Epidemiologic data was repeatedly extracted from [3] on May 15th of 2020. Further demographic data (i.e. median age, population density, percentage of urban population) was extracted from https://www.worlometers.info/population/ [26] on May 15th, 2020. Data on air passengers (a surrogate value of mobility)—registered carrier departures worldwide was extracted from https://data.worldbank.org/indicator/IS.AIR.DPRT. Day of country dependent “onset” was defined as first confirmed case reported and extracted from https://ourworldindata.org/coronavirus. Total number of days from day of onset to April 3rd, and May 15th 2020 was calculated for each country. Due to tremendous variation in population based testing (36/million in Indonesia to 74,416/million in Iceland on April 3rd 2020) and the importance of time between diagnosis and death, we arbitrarily defined death rate as Death Per Case (i.e. case fatality)/Days (dpc/d) for the endemic of each country.

Data on BCG vaccination was extracted from the BCG World Atlas [27] similarly as in [22], or from online searches for those few countries, which were not analyzed in the Atlas, one example being Iceland [28]. Modern "Colonial Era" countries to colonize America and Africa were defined as: Netherlands, Spain, United Kingdom, France, Belgium, Portugal, and Germany. We examined these countries as ones with ‘historic colonization status’ to highlight the significant potential for selection bias in the COVID-19 global epidemiologic data.

As opposed to Miller, et al. [22], we did not exclude countries with a population less than 1 million from our analyses, arguing that smaller countries may actually have better policies for universal testing (i.e. supporting rigorous epidemiologic analyses) than larger ones, Iceland being the prime example. Rather, we decided to study the top 68 countries for number of cases reported on April 3rd. This subjective cut-off was made for the Diamond Princess Cruise ship included in the list ranking at 68th (the ship’s data was excluded). Amongst these 68 countries, we could identify the initiation year of universal BCG vaccination in 40 (S1 File of S1 Table). Out of the countries examined, 9 did not have universal BCG vaccination before 1980 (S1 File of S1 Table), which date we arbitrarily selected as the cutoff for having BCG vaccination “introduced” in respect to COVID-19 (since that would have affected the population of a country 40 years old and above [i.e. the population with increased vulnerability towards the infection]).

As for the May 15th, 2020 dataset, we arbitrarily examined those countries which had more than 1,000 cases reported by that time.

We fitted two multiple regression models with the April 3, 2020 data-set using mortality rates (death/million OR dpc/days) as dependent variables for each model respectively, and BCG vaccination status before 1980, historic colonization status, median age, urban population percentage, population density, and air passengers as independent variables (predictors or confounders, depending on how one approaches the question). We repeated the analyses using the May 15, 2020 dataset including tests/million as additional independent variable (S1 File of S2 Table). The death/million and dpc/d variables were square-root transformed (SQRT), while population density, air passengers, and tests/million were log-transformed in order to achieve normality and decrease heteroscedasticity [29]. In order to compare the predictors’ relative importance on the mortality rates, we reported standardized regression coefficients. Spearman‘s rank-order correlations were used to assess the relationships between the year of the establishment of universal BCG vaccination and mortality rates (death/million or dpc/days) using April 3 and May 15 data sets. All statistical tests used in this study were two-tailed. Results were considered significant if p < 0.05. Analyses were performed using IBM SPSS 22 for Windows.

Results

There were no significant correlations between the year of the establishment of universal BCG vaccination and mortality rates: death/million or dpc/days using April 3, 2020 data (rs = -.216, p = .18, n = 40 and rs = -.052, p = .751, n = 40, respectively). There was a significant negative correlation between the year of the establishment of universal BCG vaccination and death/million for the May 15 data-set (rs = -.28, p = .035, n = 57). However, there was no significant correlation between the year of the establishment of universal BCG vaccination and the dpc/days (rs = -.20, p = .135, n = 57) for the same, more recent data set.

Following direct correlation analyses, we proceeded with multiple regression analyses. We first examined deaths/million as the dependent variable for COVID-19 mortality. For the April 3 data-set, multiple regression statistically significantly predicted deaths/million, F(6, 61) = 10.181, p < .001, R2 = .50, Adj. R2 = .451 (Table 1). Arbitrary ‘historic colonization status’, BCG vaccination status before 1980, and median age added statistically significantly to the prediction model (p < .001, p = .004 and p = .015, respectively), while urban percentage, population density, and air passengers were non-significant, p > .05 (Table 1).

Table 1

Parameter estimates for predictors of mortality (SQRT death/million) in models of April 3 (n = 68) and May 15 (n = 92).

Predictors	Standardized coefficients	t	p	Model
				Adj. R2
April 3				.451
Historic colonization status	.481***	4.795	< .001
BCG vaccination before 1980	-.280**	-3.012	.004
Median age	.243*	2.498	.015
Population density (LOG)	.024	.253	.801
Air passengers (LOG)	-.023	-.249	.804
Urban percentage	.020	.212	.833
May 15				.545
Historic colonization status	.502***	6.497	< .001
BCG vaccination before 1980	-.227**	-3.151	.002
Median age	.201*	2.039	.045
Tests/million (LOG)	.170	1.662	.100
Air passengers (LOG)	.097	1.252	.214
Population density (LOG)	-.072	-.993	.323
Urban percentage	-.019	-.220	.826

SQRT indicates square-root transformation, LOG indicates log10 transformation. Asterisks mark significant coefficients:

*p < 0.05

**p < 0.01

***p < 0.001.

For the May 15 data-set multiple regression model, F(7, 84) = 15.726, p < .001, R2 = .58, Adj. R2 = .545 (Table 1), significant predictors of COVID-19 deaths/million were historic colonization status (p < .001), BCG vaccination status before 1980 (p = .002) and median age (p = .045), while, urban percentage, population density, tests/million and air passengers were non-significant, p > .05 (Table 1).

We then examined dpc/d as the dependent variable for COVID-19 mortality by multiple regression analysis. The April 3 data model explained a relatively small amount of dpc/days variation F(6, 61) = 2.805, p = .018, R2 = .216, Adj. R2 = .139, with the historic colonization status being the only significant predictor (p = .004) (Table 2). For the May 15 data-set, however, the multiple regression model explained more variation: F(7, 84) = 9.206, p < .001, R2 = .434, Adj. R2 = .387. Significant predictors of May 15 dpc/days were test/million, median age, and historic colonization status (all p < .001), while BCG vaccination status before 1980, urban percentage, population density, and air passengers were non-significant, p > .05 (Table 2).

Table 2

Parameter estimates for predictors of mortality (SQRT dpc/days) in models of April 3 (n = 68) and May 15 (n = 92).

Predictors	Standardized coefficients	t	p	Model
				Adj. R2
April 3				.139
Historic colonization status	.374**	2.978	.004
Urban percentage	-.230	-1.925	.059
Median age	-.210	-1.723	.090
BCG vaccination before 1980	-.193	-1.655	.103
Air passengers (LOG)	-.079	-.668	.507
Population density (LOG)	-.037	-.317	.752
May 15				.387
Median age	.620***	5.426	< .001
Tests/million (LOG)	-.596***	-5.023	< .001
Historic colonization status	.348***	3.872	< .001
Population density (LOG)	-.155	-1.833	.07
BCG vaccination before 1980	-.132	-1.579	.118
Urban percentage	-.057	-.562	.575
Air passengers (LOG)	.043	.482	.631

SQRT indicates square-root transformation, LOG indicates log10 transformation, dpc indicates death per case (i.e. case mortality), days indicates days between first case reported and the date the analysis was performed (i.e. the reported length of the epidemic by country). Asterisks mark significant coefficients:

**p < 0.01

***p < 0.001.

Discussion

In this study, we found no significant association between universal BCG vaccination and country based COVID-19 mortality variation as defined by an arguably more precise death rate (i.e. dpc/d) definition than simply death/million, as examined by Miller, et al. [22], or by Escobar, et al. [25]. We underscore that both testing for-, and reporting of death from COVID-19 is highly influenced by nation specific political, cultural, and socioeconomic bias. Such bias is exemplified by no reported cases in North Korea, underreported cases and deaths in Yemen [30], very likely underreporting of deaths in various countries [31], and exploiting the pandemic for political gains [32], just to name a few. Consequently, the worldwide epidemiologic data is highly unreliable, which specifically pertains to COVID-19 deaths per million death rate, by our opinion. Such simple death rates are bound to have the most prominent political influence, and therefore are most vulnerable to bias. In the meantime, once a COVID-19 positive case is officially reported, a dependent obligation is created to provide outcomes for the case, less influenced by political, social, or cultural bias. Such bias is difficult to enumerate or account for, and current “evidence” for it in respect to COVID-19 relies on social media. Therefore, it is ‘unscientific’. Pretending for scientific scrutiny, however, by using country specific death rate or mortality as an “objective” COVID-19 outcome is misleading. In the meantime, such highly biased epidemiology based conclusions may be used as “evidence” for hypotheses, such as infantile/pediatric BCG vaccination providing lifelong protection against COVID-19 complications, in this case. For these reasons, we underscore that case mortality by length of the endemic for each country, or death per case per day (i.e. our dpc/d) measure is a more reliable measure of COVID-19 case severity than death rate alone. This conclusion is supported by our observation that tests/million (i.e. a rather objective parameter) was a significant predictor of dpc/d based COVID-19 severity for the more recent May 15th data set, but not of death/million severity outcome.

BCG vaccination is currently performed shortly after birth (newborns) in most of the countries, which universally vaccinate, as a protective measure against infantile/pediatric tuberculosis. There are some animal model and observational studies indicating that BCG vaccination can modulate host immunity (designated as ‘trained immunity’) and may protect against non-mycobacterial respiratory (and even other) infections as well (off target effect), especially in early childhood (reviewed in [33]). However, the effects of BCG vaccination fades with age, even against tuberculosis [34]. Hence, many countries have stopped their universal newborn vaccination programs or never even started that, since infantile/pediatric tuberculosis has become very rare in the economically advanced world. Consequently, there is no biologic evidence that newborn/baby age delivered BCG vaccination may have any protective effects against COVID-19, especially in adults and the elderly (who have received the BCG vaccine as an infant or child). Importantly, Hamiel, et al. did not find any difference in COVID-19 infection rates or case severity between similarly aged young adults who were either BGG immunized or not as infants in Israel [35].

The question whether BCG vaccination may acutely protect against COVID-19 infection and/or complications is very different form the one this paper addresses. We simply claim that this question cannot be answered through epidemiologic observations on historical (pediatric age delivered), country based BCG vaccination policy and associated COVID-19 endemic outcomes. Nevertheless, the most convincing support for the “acute BCG protection hypothesis” comes from a very recently published, double blind placebo controlled trial. Giamarellos-Bourboulis, et al. [36] enrolled recently hospitalized elderly (>65y old) patients to receive BCG vaccination (strain 1331; Intervax), or placebo (0.1.ml normal saline) by intradermal injection in a double blinded, randomized fashion. The primary outcome was the time interval to the first infection post hospital. Patients were followed for 12 months. Most importantly, significant protection against respiratory tract infections of probable viral origin (hazard ratio 0.21, p: 0.013) was observed in the BCG group. The investigators, however, did not clearly describe how blinding of patients and study personnel could be achieved in this case. Most commonly, a skin reaction occurs in 10–14 days at the site of the BCG injection and a permanent small scar develops in 3–6 months after the immunization in more than 85% of adults receiving the vaccine [37]. Therefore, it is biologically not possible to blind BCG with normal saline placebo. This oversight in one of the highest impact scientific journals (i.e. Cell) repeatedly emphasizes the significant bias that surrounds the COVID-19 pandemic at all levels of biomedicine from research to publication.

It is important to recognize that universal BCG vaccination establishment and current policy is a highly dependent variable, commonly inversely correlating with country specific economic status. Consequently, it is the economically most advanced countries, frequently with the most established democracies, which have abandoned or never established (such as the USA) universal BCG vaccination. We speculate that these countries are actually the ones where SARS-CoV-2 testing is most widespread and the reporting of cases and deaths is the most transparent. Historically, these countries were commonly those, which participated in colonizing other parts of the world since the late 1400s, designated as Modern Era Colonizers (‘historic colonization status’ predictor in the Tables). Therefore, to demonstrate the arbitrary selection bias (i.e. post hoc explanation for variable COVID-19 severity) in universal (pediatric age) BCG vaccination modulating COVID-19 death rates decades after its delivery, we included historic colonization status as an “independent” variable/predictor into our multiple regression models for COVID-19 outcomes (death/million vs. dpc/d). This arbitrary ‘historic colonization status’ variable actually turned out to be a much stronger predictor for both of the COVID-19 mortality measures than BCG vaccination status.

Speculative biologic explanation (similar to that to newborn BCG vaccination) for Colonizer countries having higher mortality rates from COVID-19 could be generated (long standing, transgenerational influence of improved prenatal nutrition [38] on postnatal immune responses and life expectancy [leading to increased vulnerability to COVID-19] in these richer countries compared to the rest of the world, for example). In the meantime, we rather conclude that both BCG vaccination and Colonizer status are dependent variables of country based socioeconomic and political status, which latter features are the strongest predictors for COVID-19 outcomes (especially death/million) due the highly politicized nature of the pandemic.

Our work highlights the difficulties in drawing reliable epidemiologic conclusions from the currently available worldwide data on the COVID-19 pandemic. We advise for extreme caution and self-reflective scrutiny to balance publication pressure, inherent drive for scientific discovery, and financial social and political gains, when examining COVID-19 related biomedical research, including epidemiology. This conclusion is in line with experts in the field, emphasizing that “the data is not from peer-reviewed research, but rather is almost real-time clinical data–which can be messy and come with many caveats” [39]. The experts also underscore that “the lack of widespread, systematic testing in most countries is the main source of discrepancies in death rates internationally” [39]. It is our responsibility, as the medical scientific community around the world, to promote consistent and reliable epidemiologic reporting, and de-politicizing the current COVID-19 pandemic in order to prepare for other expectable large-scale infectious outbreaks in the future.

(XLSX)

Click here for additional data file.

4 May 2020

PONE-D-20-10246

BCG protects against COVID-19? A word of caution

PLOS ONE

Dear Dr. Kellermayer,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Specifically, please take in account the limitation and advance highlighted by the two reviewers.

We would appreciate receiving your revised manuscript by Jun 18 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Pierre Roques, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following financial disclosure:

'NO - Include this sentence at the end of your statement: The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript'

At this time, please address the following queries:

Please clarify the sources of funding (financial or material support) for your study. List the grants or organizations that supported your study, including funding received from your institution.

State what role the funders took in the study. If the funders had no role in your study, please state: “The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

If any authors received a salary from any of your funders, please state which authors and which funders.

If you did not receive any funding for this study, please state: “The authors received no specific funding for this work.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The article from Richard Kellermayer’s group presents a critical analysis of the recent publication by Miller et al 2020 ref. [15] and proposes an original approach to establish a representation of the daily rates of COVID-19 case fatality (i.e. Death Per Case /Days of the endemic [dpc/d]). However, only strictly identical groups (age/sex proportion, living conditions, COVID diagnosis) could be strictly compared and would better answer the question of the effect of the BCG vaccine on current acute epidemic infection; could you address this question by adding an analysis of equivalent groups (for ex. in urban conditions)? Furthermore, to support the article statements, it is necessary to discuss whether the people who have benefited from BCG vaccination or not are really comparable (age/sex proportion, living conditions, COVID diagnosis), and otherwise to highlight the differences observed. Finally, he possible bias of the chosen representation is not discussed.

In the abstract and elsewhere after, it’s supported that “the use of personal protective equipment (PPE) are the only epidemiologic measures” while the facts exemplified may demonstrate opposite situations: please add also that PPE in our day-to-day lives are efficient when associated with frequent and proper hand washing.

There is too much citations from media and general press (e.g. BBC, YouTube…) in the R. Kellermayer’s article that could be replaced with scientific short communications or opinion views. A lot of references citations of key publications are lacking concerning the current epidemic of COVID-19 both in introduction and in discussion part (for example in the introduction: Lescure FX (https://doi.org/10.1016/S1473-3099(20)30237-1) in parallel to ref. [3] and He X https://www.nature.com/articles/s41591-020-0869-5 in parallel to ref. [4, 5]).

One point that particularly deserves to be developed in this article is the explanation of the interest especially in the BCG vaccination (even extension of use, see thereafter) in the context of COVID-19 epidemic: please explain the scientific rational.

Some evaluations in children of a BCG vaccine cross-protection against several types of microbial or fungal infection were published yet (doi:10.1016/j.cmi.2019.04.020 ; doi:10.3389/fimmu.2019.02806) and these studies are not mentioned or discussed. Likewise, there is no reference to the phase III clinical trials which have recently started (Holland (NCT04328441); Australia (NCT04327206)) to answer the question of the prophylactic effect of BCG vaccine on COVID-19, dedicated to health care, aiming at reducing the symptomatology and considered as an emergency measure and not a lasting solution.

In addition, many applications for prophylactic purposes currently take advantage of the anti-inflammatory response induced by this BCG vaccination, e.g. for type 1 diabetes, bladder cancer, or autoimmune diseases. This rational for use should also be discussed.

Reviewer #2: The authors explored the association between daily rates of COVID-19 case fatality and the presence of universal BCG vaccination.

At the time of submission of their manuscript, there was one preprint reporting a correlation between universal BCG vaccination policy and reduced morbidity and mortality for COVID-19, which has not published yet after peer-review (at least, it is not listed in pubmed as for today, the 2nd of May).

The current analysis defined BCG vaccination data based on the same database as the preprint they refer to. For the death rate data, they used the one reported on April 3rd (instead of March 31st) and they did not exclude countries with less than one million inhabitants.

They analyzed the association between the death rate or the Death Per Case (or case fatality rate)/Days from onset (dpc/d) to account for testing and time bias, and BCG vaccination. They also compared Netherlands, Spain, United Kingdom, France, Belgium, Portugal, and Germany (they called Modern "Colonial Era" countries to colonize America and Africa) to others.

They showed that there is significantly less death/million when BCG was introduced before 1980, but it is not the case when Death Per Case/Days of endemic (dpc/d) is considered. Finally, they reported that dpc/d is significantly higher in Modern Era Colonizer countries compared to others.

MAJOR RECOMMANDATIONS

The choice of 1980 has a cut off for having BCG was defined arbitrary. The authors state that “BCG would have affected the population of a country 40 years old and above [i.e. the population with increased vulnerability towards the SARS-CoV-2 infection]”. In studies reporting a non specific effect of live attenuated vaccines (such as BCG, oral polio, measles or smallpox), like the one published by Christine Benn and Peter Aaby, the heterologous protection against all cause morbidity and mortality is long-lasting, but as compared to the transient immediate defense mediated by the innate effector response. It is not as long-lasting as adaptive immune memory, and in no way it lasts 40 years. Moreover, the non potential beneficial non specific effect of BCG might be modulated with time, and other vaccines for instance. Thus, the authors could focus their analysis on a younger population, rather than the entire population of each of the Top 68 countries in terms of covid-19 induced death, or covid-19 severity.

The year of introduction of BCG vaccination can be one variable to study. However, the authors are certainly aware of the heterogeneity of the vaccine coverage depending on the country. Whether BCG vaccination is mandatory or not can result in different BCG vaccine coverage. In addition, the vaccine schedules and strains of BCG with different immunogenicity can differ between countries for instance. This should ideally be taken into account because it can be confounding, or at least be discussed by the authors.

MINOR RECOMMANDATIONS

Please update the introduction and discussion sections with the very recent literature about:

-The presence of viral RNA found in plasma and urine, and the successfully isolation of infectious viruses from these compartments (as for the presence of infectious particles in feces, there are inherent technical difficulties to isolate virus from feces because of the high risk of contamination of cell cultures with other microrganisms than the virus of interest)

-the transmission through aerosols

-the pathogenesis of SARS-CoV-2 (and not SARS-CoV) in non human primates, including the comparison between n=2 old and n=2 young animals

-there are currently several preprints on observationnal studies and the analysis of the association between BCG vaccination and covid-19, with positive, negative or no correlation

-there are now more than one trial to evaluate whether BCG could reduce the susceptibility to SARS-CoV2 infection and severity of covid-19

Please include a paragraph on the hypothesis as to how BCG could protect, or not, or even worsen the susceptibility to SARS-CoV2 infection and severity of covid-19, with few words on the concept of non specific effect of vaccines and on the new paradigm of trained immunity.

The quality of the figures is too low.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Anne-Sophie Beignon

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

26 Jul 2020

Response to the Reviewers

The time and expertise of the reviewers is greatly appreciated. We responded to the valuable recommendations point-by-point below. The corresponding changes in the manuscript are highlighted with Word ‘track changes’. The modifications have greatly improved the quality of our manuscript. In general, we would like to underscore that our work is intended to demonstrate that the currently available epidemiological data on COVID-19 can not be used to make any reliable clinical prediction in regards to prevention or treatment against SARS-CoV-2, especially if examined as death/million. We wished to deliver this message in the most scientific way possible in order to effectively reach the biomedical research community. Hence, the “word of caution” in our title. The title has also been modified to more precisely cover our findings and message, which pertains to BCG epidemiology, and does NOT intend to evaluate the biologic evidence for BCG vaccination as an acute (i.e. short lived) protectant against COVID-19. We feel that PLOS ONE is and outstanding means to communicate this message. We added Giedrius Trakimas to our author list and performed additional analyses to more strongly support our conclusions.

Reviewer #1: The article from Richard Kellermayer’s group presents a critical analysis of the recent publication by Miller et al 2020 ref. [15] and proposes an original approach to establish a representation of the daily rates of COVID-19 case fatality (i.e. Death Per Case /Days of the endemic [dpc/d]). However, only strictly identical groups (age/sex proportion, living conditions, COVID diagnosis) could be strictly compared and would better answer the question of the effect of the BCG vaccine on current acute epidemic infection; could you address this question by adding an analysis of equivalent groups (for ex. in urban conditions)? Furthermore, to support the article statements, it is necessary to discuss whether the people who have benefited from BCG vaccination or not are really comparable (age/sex proportion, living conditions, COVID diagnosis), and otherwise to highlight the differences observed. Finally, he possible bias of the chosen representation is not discussed.

Response: Our general response partly addresses this recommendation of the distinguished Reviewer. We absolutely agree that if reliable and sound epidemiologic data were available, stricter analyses incorporating as many variables as possible would be desired. In the meantime, we have now performed multiple regression modeling and incorporated additional variables such as the recommended ‘urban living’ as predictors into our models. Our main conclusion has not been significantly modified by these analyses. We do highlight the arbitrary selection bias not only in respect to BCG vaccination (as the essence of our critique manuscript), but in regards to our ‘historical colonization status’ variable.

In the abstract and elsewhere after, it’s supported that “the use of personal protective equipment (PPE) are the only epidemiologic measures” while the facts exemplified may demonstrate opposite situations: please add also that PPE in our day-to-day lives are efficient when associated with frequent and proper hand washing.

Response: We decided to omit the discussions about PPE in respect to COVID-19, since that is not the focus of our work.

There is too much citations from media and general press (e.g. BBC, YouTube…) in the R. Kellermayer’s article that could be replaced with scientific short communications or opinion views. A lot of references citations of key publications are lacking concerning the current epidemic of COVID-19 both in introduction and in discussion part (for example in the introduction: Lescure FX (https://doi.org/10.1016/S1473-3099(20)30237-1) in parallel to ref. [3] and He X https://www.nature.com/articles/s41591-020-0869-5 in parallel to ref. [4, 5]).

Response: Thank you for the recommendations, we have incorporated the references. In the meantime, we would like to emphasize that the focus of our work is the critique of worldwide epidemiology data, not a comprehensive review of COVID-19 pathogenesis.

One point that particularly deserves to be developed in this article is the explanation of the interest especially in the BCG vaccination (even extension of use, see thereafter) in the context of COVID-19 epidemic: please explain the scientific rational.

Some evaluations in children of a BCG vaccine cross-protection against several types of microbial or fungal infection were published yet (doi:10.1016/j.cmi.2019.04.020 ; doi:10.3389/fimmu.2019.02806) and these studies are not mentioned or discussed. Likewise, there is no reference to the phase III clinical trials which have recently started (Holland (NCT04328441); Australia (NCT04327206)) to answer the question of the prophylactic effect of BCG vaccine on COVID-19, dedicated to health care, aiming at reducing the symptomatology and considered as an emergency measure and not a lasting solution.

In addition, many applications for prophylactic purposes currently take advantage of the anti-inflammatory response induced by this BCG vaccination, e.g. for type 1 diabetes, bladder cancer, or autoimmune diseases. This rational for use should also be discussed.

Response: We have included a brief summary of the plausible biologic basis for BCG vaccination protection, and referenced a comprehensive and freely accessible web page based review on the topic in our discussion. We also included into our introduction the number of clinical trials registered in clinicaltrials.gov in June. However, our focus is to emphasize the highly biased and interdependent nature of arbitrary predictors of COVID-19 epidemiology, especially in review of the unreliable nature of country based case and death reporting.

Reviewer #2: The authors explored the association between daily rates of COVID-19 case fatality and the presence of universal BCG vaccination.

At the time of submission of their manuscript, there was one preprint reporting a correlation between universal BCG vaccination policy and reduced morbidity and mortality for COVID-19, which has not published yet after peer-review (at least, it is not listed in pubmed as for today, the 2nd of May).

The current analysis defined BCG vaccination data based on the same database as the preprint they refer to. For the death rate data, they used the one reported on April 3rd (instead of March 31st) and they did not exclude countries with less than one million inhabitants.

They analyzed the association between the death rate or the Death Per Case (or case fatality rate)/Days from onset (dpc/d) to account for testing and time bias, and BCG vaccination. They also compared Netherlands, Spain, United Kingdom, France, Belgium, Portugal, and Germany (they called Modern "Colonial Era" countries to colonize America and Africa) to others.

They showed that there is significantly less death/million when BCG was introduced before 1980, but it is not the case when Death Per Case/Days of endemic (dpc/d) is considered. Finally, they reported that dpc/d is significantly higher in Modern Era Colonizer countries compared to others.

MAJOR RECOMMANDATIONS

The choice of 1980 has a cut off for having BCG was defined arbitrary. The authors state that “BCG would have affected the population of a country 40 years old and above [i.e. the population with increased vulnerability towards the SARS-CoV-2 infection]”. In studies reporting a non specific effect of live attenuated vaccines (such as BCG, oral polio, measles or smallpox), like the one published by Christine Benn and Peter Aaby, the heterologous protection against all cause morbidity and mortality is long-lasting, but as compared to the transient immediate defense mediated by the innate effector response. It is not as long-lasting as adaptive immune memory, and in no way it lasts 40 years. Moreover, the non potential beneficial non specific effect of BCG might be modulated with time, and other vaccines for instance. Thus, the authors could focus their analysis on a younger population, rather than the entire population of each of the Top 68 countries in terms of covid-19 induced death, or covid-19 severity.

Response: Thank you for the insightful observations and recommendations. We absolutely agree with this Reviewer that there is no biologic basis for newborn/baby age BCG vaccination to protect against COVID-19 infection and/or morbidity in the elderly. We have now performed analyses on an additional (May 15th) dataset with multiple regression modeling, including median age as a predictor. We have also added a recent publication as reference (37) that contradicts newborn BCG effects on COVID-19, even in young adults around the age of 40. Our main conclusion is the unreliable nature of worldwide COVID-19 epidemiologic data and that researchers should consequently avoid, or highly critically perform detailed analyses of that.

The year of introduction of BCG vaccination can be one variable to study. However, the authors are certainly aware of the heterogeneity of the vaccine coverage depending on the country. Whether BCG vaccination is mandatory or not can result in different BCG vaccine coverage. In addition, the vaccine schedules and strains of BCG with different immunogenicity can differ between countries for instance. This should ideally be taken into account because it can be confounding, or at least be discussed by the authors.

Response: We would like to repeatedly emphasize the main point of our work, which underscores the unreliable nature of COVID-19 worldwide data. We trust that the added analyses and more clear discussion of those along with a stronger message to the epidemiologist community makes it understandable that we decided not to perform the recommended studies on the heterogeneity of the vaccine coverage, depending on the country.

MINOR RECOMMANDATIONS

Please update the introduction and discussion sections with the very recent literature about:

-The presence of viral RNA found in plasma and urine, and the successfully isolation of infectious viruses from these compartments (as for the presence of infectious particles in feces, there are inherent technical difficulties to isolate virus from feces because of the high risk of contamination of cell cultures with other microrganisms than the virus of interest)

-the transmission through aerosols

-the pathogenesis of SARS-CoV-2 (and not SARS-CoV) in non human primates, including the comparison between n=2 old and n=2 young animals

-there are currently several preprints on observationnal studies and the analysis of the association between BCG vaccination and covid-19, with positive, negative or no correlation

-there are now more than one trial to evaluate whether BCG could reduce the susceptibility to SARS-CoV2 infection and severity of covid-19

Please include a paragraph on the hypothesis as to how BCG could protect, or not, or even worsen the susceptibility to SARS-CoV2 infection and severity of covid-19, with few words on the concept of non specific effect of vaccines and on the new paradigm of trained immunity.

The quality of the figures is too low.

Response: We repeatedly hope that the added analyses, additional references and more clear discussion of those along with a stronger message to the epidemiologist community is acceptable to this Reviewer without us following all the minor recommendations. We wished to be as concise as possible and not elaborate more on viral transmission and animal modeling in respect to SARS-CoV-2 pathogenesis. We did include a paragraph referencing a rather recent comprehensive review on the plausible connections between BCG vaccination, and off target viral infections, including COVID-19. We also excluded our figures during the major revision of our work.

Submitted filename: ResponseToReviewers_BCG COVID-19 R1 Letterhead.doc

Click here for additional data file.

15 Sep 2020

PONE-D-20-10246R1

BCG epidemiology supports its protection against COVID-19? A word of caution

PLOS ONE

Dear Dr. Kellermayer,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

as suggested by the reviewer, it might be of good practice to indicate the last paper about BCG and respiratory infection he highlighted.

Please submit your revised manuscript by Oct 30 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Pierre Roques, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (if provided):

sorry for the delay but the end of shutdown and return to normal life is sometime quite complicated now;

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: I Don't Know

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: The authors have addressed the comments from both reviewers.

The goal of their work has been better defined (including by rephrasing the title of the paper and their conclusions). They highlight and alert that global COVID-19 epidemiological data (as well as global BCG epidemiological data?) are not reliable enough to base hypothesis-driven research.

After direct correlation analysis between the year of the establishment of universal BCG vaccination and the mortality rates, expressed as death/million or death per case/days, and using 2 datasets (from the beginning of April and mid of May), they now also used multiple regression analyses. They have included more variables, such as the urban population

percentage, population density, and air passengers, in addition to historic colonization status and median age.

The conclusions from the ongoing clinical trials testing whether BCG could protect against COVID-19 are not yet known, however a recent paper demonstrates that BCG can protect the elderly against respiratory infections (Giamarellos-Bourboulis et al., Cell, 31 Aug 2020). The authors might want to cite this reference as a stronger evidence to test the hypothesis rather than epidemiological data.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

20 Sep 2020

Reviewer #2: The conclusions from the ongoing clinical trials testing whether BCG could protect against COVID-19 are not yet known, however a recent paper demonstrates that BCG can protect the elderly against respiratory infections (Giamarellos-Bourboulis et al., Cell, 31 Aug 2020). The authors might want to cite this reference as a stronger evidence to test the hypothesis rather than epidemiological data.

Response: We agree, and have added an additional paragraph to the discussion, which critically reviews the recommended publication from our manuscript’s point of view. We also made minor modifications to the paper in order to more clearly emphasize that our work only pertains to the epidemiology of pediatric age delivered, universal BCG vaccination, and whether that may be protective against COVID-19 deaths in the adult/elderly population, who received the vaccine when they were very young (i.e. decades before a possible infection with COVID-19).

Submitted filename: ResponseToReviewers_BCG COVID-19 R2 Letterhead.doc

Click here for additional data file.

23 Sep 2020

BCG epidemiology supports its protection against COVID-19? A word of caution

PONE-D-20-10246R2

Dear Dr. Kellermayer,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Pierre Roques, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Thanks to have take in acount all the referee comments. I congratulate you for this nice paper.

Reviewers' comments:

25 Sep 2020

PONE-D-20-10246R2

BCG epidemiology supports its protection against COVID-19? A word of caution

Dear Dr. Kellermayer:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Pierre Roques

Academic Editor

PLOS ONE