"External immunity effect" is the key to understand COVID-19, design the clinical trials, and calculate R0.

Introducing the word/concept of "external immunity effect" and visualizing its size

In this age group analysis post, I showed the group immunity effect was larger than the individual immunity effect and I hope someone quantifies them. But, so far it seems no one has done it. Let me emphasize this point in this post and how this affects the clinical trial design and R0 calculation.

Also, let me change my wording of "herd immunity" as it is a very misleading word now. Instead, I would like to propose to use "external immunity effect"

I studied economics a long time ago when I was in college. In economics, we call "external economies / external economy effect" when something benefits someone who doesn't bear its cost (if I remember correctly). I think this is happening for vaccinations and immunity in general. If your family or close friends can catch a cold often, you are likely to catch a cold often and vice versa.

For individual immunity effect, I would like to keep using the "individual immunity effect".

I did the same analysis for Italy, Spain, and Portugal using the most recent data and visualize how large the external immunity effect and individual immunity effect are. And here is the summary of BCG vaccination of these countries.

  • Italy: never mandatory
  • Spain: from 39 to 78 years old has been vaccinated
  • Portugal: from 3 to 78 old have been vaccinated

You can see from the charts that there are individual immunity effect but external immunity effect seems much larger than individual immunity effect.

This explains why you can't see a big difference in those who were BCG vaccinated and those who were not in the same community/country. Please check the articles below for my counter-arguments for the papers which deny the BCG hypothesis. They all try to deny it without thinking of the fact that external immunity effect is much larger than individual effect.

High-quality analysis of the former East/West Germany testing the BCG hypothesis... if the herd immunity effect is added...

The BCG hypothesis is validated by the Israel COVID-19 death tolls.


The article below about Hungary doesn't think of external immunity effect, either.

Newborn BCG vaccination does not protect the elderly from coronavirus


There comes another preprint denying the BCG effect by ignoring external immunity effect.

The BCG vaccine does not protect against COVID-19. Evidence from a natural experiment in Sweden


Clinical trials should be designed considering "external immunity effect"

This is not a minor point as this affects the design of clinical trials.

All of the clinical trials of BCG and COVID-19 are designed with the standard clinical trial method; randomized and mixed. These may prove the individual immunity effect but may conclude that its individual immunity effect is not significant. That is right in a sense. BCG vaccine is not a specific vaccine against SARS-CoV-2. A specific vaccine may make us 100% immune to SARS-CoV-2 but BCG vaccine may make us 30% less susceptible to SARS-CoV-2. 30% less may not significant as an individual effect but it is significant and creates a strong group immunity if its external immunity effect is included. 

If I were to design a clinical trial, I would pick up three villages; BCG vaccination 100% for one village, 50% for another village, 0% for the last village. This clinical trial will tell us how large external immunity effect is.

A slight decrease in susceptibility can decrease R0 significantly in a group

In infectious diseases, there are jargons of R0, Rt, susceptibility, and transmission. Simply put, R0 is determined by the disease irrelevant of people's susceptibility. And Rt is a function of susceptibility and transmission is a function or equivalent of Rt. These are my guess, please correct me if I am wrong.

R0s of COVID-19 seems to vary SIGNIFICANTLY by countries because people's susceptibility seems to vary SLIGHTLY. Yes, you are right. I write significantly and slightly. How can a slight change of susceptibility result in a significant change of R0?  

Here is my hypothesis/understanding. If a Japanese is 30% less susceptible to COVID-19 than an Italian, R0 in Japan would not be only 30% less than R0 in Italy because the transmission will decrease accordingly such as 0.7 * 0.7 = 0.49. I think R0 in Japan can be less than half of R0 in Italy because of this mechanism. 

I hope some experts pick this post and rethink the clinical trial design/evaluation and realize that R0 of COVID-19 is not universal but should be calculated for each country.

(Added on 15 June) Interesting simulation regarding influenza vaccination

I found an interesting paper about the influenza vaccine simulation.

Optimizing the impact of low-efficacy influenza vaccines

The elderly people have a much higher mortality rate from influenza than the young people as with COVID-19. And the elderly people are more willing to get a flu vaccine and young people are less interested in getting a flu vaccine. However, if the U.S. government can dose the same coverage(43%) of flu vaccination and decide which generation should get, the computer simulation tells us the school-age children and the young adults have higher priority than the elderly themselves because they are the groups to which most transmission is attributable. This vaccine distribution optimization under the same total coverage(43%) can reduce death by one-third. And the optimal uptake can differ by vaccine efficacy. If the vaccine efficacy is less than 30%, vaccination to the elderly will have a higher priority.

And this article uses the words "direct protection" and "indirect protection". These words correspond to my wording of "individual immunity effect" and "external immunity effect". Indirect protection/external immunity effect can be much larger than direct protection/individual immunity effect.

"As a consequence, protection via herd immunity is more effective than direct protection when vaccine efficacy is above 30%."

The current ongoing randomized controlled trials only see direct individual protection/immunity and miss the much larger part, indirect/external immunity effect.

"Randomized controlled trials are the gold standard for measuring vaccine efficacy in terms of the direct individual-level protection conferred. Mathematical modeling provides a complementary approach through which transmission dynamics can be simulated and population-level effectiveness of vaccination programs quantified in terms of key outcome measures."

Sources of this analysis; accessed on 10 Jun.



  1. J. Soto,
    I have been trying find your email address. Could you email me russelllamontagne@gmail.com. I help organize a BCG working group, www.bcgandautoimmunity.org that includes many of the researchers you have referenced. would be interested in chatting with you about your analysis of BCG/COVID-19. One of our member, D. Faustman is starting a COVID-19 study using the Tokyo strain and has a type 1 diabetes trial underway. www.faustmanlab.org.

    1. Thanks Russell,

      The readers of this blog will be very glad to hear that Dr. Faustman is starting a COVID-19 study using the BCG Tokyo strain. I think clinical trials should use the BCG Tokyo strain or at least include it.

      Kind regards,
      Jun Sato

  2. Did you see this?


Post a Comment

Popular posts from this blog

(COMMENTS reached the max) If I were North American/West European/Australian, I would take BCG vaccination now against the novel coronavirus pandemic.

Please leave comments here regarding the BCG hypothesis, the correlation between COVID-19 and BCG vaccination