Thread by @VikiLovesFACS

#HerdImmunity is a talking point again. So let's talk about it! What conditions do we need to fulfil for a herd immunity strategy to work? And is it possible to do this for #COVID19 #SARSCoV2? 🧵

The idea behind herd immunity is simple. We know infectious diseases spread in a population that is susceptible to them. We also know people can become immune. If enough people are immune, the virus doesn't have enough people to spread to. Then small outbreaks will die out. 2/21

Instinctively, we can feel that the more contagious a disease is, the more people need to be immune to prevent it from spreading. 3/21

You might remember R0 - the number of people each infected person passes a disease on to, if everyone is susceptible. For #COVID19 this number is around 3. For measles, it's more like 13. 4/21

The number of people who need to be immune to achieve herd immunity is given by 1 - 1/R0. So for #COVID19, that's about 66% (2/3) of the population. For measles, it's more like 92% of the population. 5/21

This really tells us all we need to know to decide if a herd immunity strategy will work: can we achieve a situation where that percentage of the population is immune? 6/21

Let's think first about measles, where a herd immunity strategy famously can work. 7/21

There are two ways we can get immunity. One is if the virus induces long-lived immunity in people who have recovered. Measles does this. People who had the virus >60 years ago are still immune! 8/21

tinyurl.com/r3qya2p

But wait! Anyone who is old enough to remember the pre-vaccine days will tell you that wasn't enough. Measles epidemics would still happen every couple of years. 9/21

This is because susceptible individuals were joining the population: babies. Although "herd immunity" from natural infection prevented outbreaks of measles for a bit, after a couple of years, there were enough susceptible people that the proportion immune was < 92%. 10/21

And then another outbreak would occur. 11/21

The way around this is the second way we can induce immunity in a population: vaccination. For measles, this works great. The vaccine works really well and so long as almost all children get it, we shouldn't see measles outbreaks. 12/21

So even for measles, which is famous for inducing long-lasting immunity, a herd immunity strategy was possible only once we had a vaccine. 13/21

Now let's think about #COVID19. We don't have a vaccine yet, but can we achieve herd immunity with natural infections? To do that we have to get to 66% of the population immune. 14/21

As a first pass, we might say that 66% of the population therefore needs to catch the virus. But we're seeing that not everyone makes a good response, so maybe it would have to be more like 80% of people catching it for 66% to be immune. 15/21

In Britain, this would be about 54 million people catching #COVID19, and we might expect something like half a million of them to die. 16/21

Now we *can* have a debate about whether it's worth sacrificing half a million people to achieve this aim. But I think that's the wrong discussion to have, because this strategy assumes that #COVID19, like measles, gives long-lasting immunity. 17/21

And we know that most coronaviruses *don't*. Studies on the coronaviruses that cause colds suggest that immunity tends to wear off after about 18 months (most of us know this instinctively, from our experience with colds!). 18/21

Furthermore, we're now seeing some people who are getting reinfected five or so months after their first infection, so #COVID19 also seems to follow this pattern of not giving long-lasting immunity. 19/21

If immunity wears off, this will make that 66% immunity even harder - I would probably say impossible, though it depends exactly how quickly the immunity wears off - to achieve. 20/21

So to summarise: even for diseases where natural infection induces long-lasting immunity (eg measles) #HerdImmunity strategies rely on vaccination. #COVID19 probably does *not* induce long-lasting immunity, so to for a herd immunity strategy, we definitely need a vaccine. \thread

Afterthought: R0 depends on population mixing factors, as well as viral factors. So one thing I suppose we could do to make a "herd immunity" strategy more viable is to reduce R0 by reducing social interactions, mandatory mask wearing etc...

This would reduce the proportion of people who would have to be immune. BUT my feeling is that people who want to pursue a "herd immunity" strategy want to do so precisely to avoid have to reduce social interactions, mandatory masks etc. So I'm not sure this would help.