I have read claims that the Omicron variant of SARS-CoV-2 is:

Which one, if any, is right, and does it depend whether eradicated, or almost eradicated, diseases like smallpox are included? Or on other factors, maybe?

1 Answer 1


There is no one measure for "contagiousness", so ranking contagious diseases is just like ranking "best sports player": there are enough different ways to do it that there are entire TV networks devoted to arguing about those sorts of questions 24 hours a day. It's difficult enough to determine the best player in just one sport right now, let alone in one sport over history (lots of systemic changes in the sport change what the meaning of 'good' is; how do you account for teammates in a team sport? medical advances? professionalization?) or even worse, completely different sports. How do you compare a chess champion to a soccer/football goalie? They'll never be done arguing because there isn't an answer they can eventually come to consensus on, so it can be very entertaining but ultimately isn't very meaningful.

Basic reproduction number or "R0" is one possible measure, though it can be difficult to estimate (so different estimates vary) and does not come close to telling the entire story. This is a parameter relevant to epidemiological modeling that is most easily interpreted as "the number of people that one infected person will infect in an entirely susceptible population" - that is, in a population with no existing immunity. In practice, an "entirely susceptible population" rarely exists (even for a completely new virus), and the number is estimated from models with many other free parameters and assumptions that may or may not hold. It also lacks information about the time course of spread.

There is also a lot of variation in spread, with some viruses demonstrating "superspreader" behavior, where average rates of spread are very misleading at a case-by-case level. Instead, particular individuals tend to cause most subsequent infections, and others may cause none at all, so R0 estimates may only really work on a large population scale and not describe behavior in smaller groups.

One of the articles you cite seems to use R0, but doesn't seem to entirely understand the dynamics. The other one doesn't explain how they made the conclusion at all, so we can't really approach the claim. In your "the most contagious disease in history" link they write:

...compared to the highly contagious measles virus, where one person can infect 15 other people in 12 days

I take this to imply here that measles has an R0 = 15 (which would be in line with reported estimates, though again, those estimates vary). Then in the next sentence:

Omicron doubles every 2-3 days, with one case spreading to six people in four days, 36 people in eight days, and a hefty 216 people after 12 days.

So here, "one case spreading to six people" seems to imply R0 = 6, much less than measles; this doesn't really support their suggestion that it's the most contagious disease. Instead, they seem to be basing it on "216 people after 12 days" which is based on 6 in four days, then those 6 to 36 in the next four days, etc: they're counting just one "round" of spread with measles, but 3 "rounds" of spread with Omicron. There isn't really enough information to make that judgment, though, because epidemiology doesn't work that way where you can just sequentially multiply the spread. It's not like all the people that person A infects wait until person A is done infecting to start spreading themselves.

It's also highly unlikely that in an epidemiological model, all the people exposed are "entirely susceptible". Consider spread in a 7-person household: one infected family member may infect all 6 other people in their household, but after that, many of the closest contacts of those family members have already been infected, so aren't available to infect again. So it's not quite as simple as just taking R0b, where b is the number of infection cycles in some unit time. The rate of spread does matter, too, when considering the pressure on medical systems to keep up, but simple back-of-envelope math just gets you simple answers. They might be useful anyways, but not definitive in ranking the "most".

I'd recommend instead of saying "most contagious", it would be more rigorous to take some particular measurable quantity and report that instead. You might choose R0; Wikipedia does have a list of some viruses by R0 where it seems that the most contagious COVID-19 strains come in just behind viruses considered very contagious like measles, mumps, and chickenpox, while being quite similar to other viruses like smallpox. Some estimates for Omicron are as high as 10; your article used 6. I'd caution against multiplying based on "effective reproduction number" relative to another variant like Delta (for example, this paper estimated 3.19), because those are not really estimates of R0 but rather Reffective, and those estimates are made well into the Delta wave where a lot of the population already had immunity to Delta (including from vaccination and exposure to previous strains).

As far as ranking "number", also consider that there is no fair way to "count" virus strains. We give labels because they are useful, but it isn't really fair to consider "measles" just one thing, and clump every possible variety of measles together as one virus, while also separating out COVID variants and counting them all separately. It's also not true that "Omicron" is all identical, there will be many mutations in any viral lineage, some useful/important and some not, but there won't be any definitive way to come up with a "top 5" or "top 10" when there's no definitive way to even decide how many players there are in the game.

  • Isn't the R-number also depending on the population density (or rather the average number of contact persons), which can vary between cultures and also between time periods? May 10, 2022 at 0:15
  • @PaŭloEbermann Yep; so is pretty much every aspect of infectious disease transmission. Host behavior and physical susceptibility are at least as important as any property of a virus itself.
    – Bryan Krause
    May 10, 2022 at 0:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.