On Herd Immunity

BACKGROUND

The recent emergence of the SARS-CoV-2 virus, and the COVID-19 disease which it causes in humans, has brought the old term “herd immunity” into prominence again.

In particular, there have been credible claims that the UK government’s initial pandemic response plan (as of Mar 12th 2020) relied heavily upon the herd immunity effect, as apparently does the still-current (as of Apr 18th 2020) strategy of Sweden.

Many people have been very upset by this, and have demanded the abandonment of these policies for a variety of reasons, apparently including:

1. people don’t like being thought of as part of a “herd”; especially not when it is done by the hated “shepherds” of the Tory Party
2. various misunderstandings that this somehow represents a callous “survival of the fittest” do-nothing Social Darwinist approach, as perhaps dreamed-up by Dominic Cummings or one of his “weirdos”
3. sober-sounding professional assessments that it is very dangerous and irresponsible to rely on such risky “mitigation” effects instead of strong suppression measures.

However, in this as-brief-as-I-could-make-it essay, I will try to show that taking advantage of the herd immunity effect has the potential (at least in theory) to significantly reduce the overall death toll of a highly contagious infection for which a vaccine is not yet available, and that it is therefore not a fundamentally bad idea in principle. Rather, the merits (and demerits) must be considered on a case by case basis.

DEFINITION

So, what is “herd immunity”? Well, it is perhaps a rather unfortunate term, dating from the 1920s, but popularised in the 1970s, for what we might otherwise call “Immunity In The Community”. It describes the observed, and now well-understood, effect whereby partial immunity to an infection within a group can sometimes provide indirect protection even to the un-immunised (and otherwise vulnerable) members of that group.

It works like this: if one case of an infection generally spreads to 3 other people, then the total number of infections will obviously grow substantially over time, at least to begin with. However, if more than two out of three people in a mixed community develop immunity to such an infection (e.g. either by recovering from it, or from being vaccinated against it), the disease will actually go into decline, because there will be too few susceptible people to maintain the necessary chains of transmission. Better still, a further large outbreak will then be unlikely for as long as the immunity persists, even if some newly infected people turn up from outside — as long as they don’t pull the numbers below the critical two-thirds “herd immunity threshold” limit.

As such, if an infection has severe effects on some people, but mild effects on others, it might be possible to reduce the total impact of the resulting disease by selectively isolating the more vulnerable people while allowing a protective shield of immunity to be developed among those who are more robust, and for whom the effects are milder.

NAIVE EXAMPLE

To illustrate this idea, allow me to create a simple hypothetical. Obviously, this is not a realistic scenario, but I’m trying to illustrate a principle:

Imagine there is a disease which is fatal to people with red/ginger hair, but which only causes mild cold-like symptoms in “normal” people. Let’s also assume that we don’t have a vaccine for this disease yet either, and that it spreads quickly and is hard to contain (in part because the symptoms are not always obvious in mild cases). What can we do to prevent the precious redheads from dying in huge numbers?

One approach would be to temporarily LOCK-UP THE GINGERS in some kind of protected environment where they are adequately fed and watered, and then simply allow the infection to naturally pass through the remaining People Of Boring Hair Colours, who will mostly get an annoying cold but not much worse.

At the end of this process, when sufficient immunity has been achieved, the infection will die out naturally and it should be possible to re-introduce the gingers to polite society, without killing them. It is important to note that they will still not be immune, but they will nevertheless be significantly (but not perfectly) protected from new outbreaks of the disease by the immunity of their fellows, for as long as that immunity lasts.

Now, whether this “herd immunity” approach actually works in reality depends on the numbers in any particular case: technical stuff like the initial “R Value”, the frequency of redheads, the practical effectiveness of the isolation policy, the degree and duration of immunity, and so on. But my point is that it can potentially work, at least in theory, and there are examples of it having worked in practice (e.g. with measles and smallpox, historically).

LESS NAIVE EXAMPLE

A real disease is unlikely to be as simple as our contrived hypothetical. But what if there was a disease that was highly fatal to the elderly, but essentially harmless to the young? Similar arguments would apply here as with the redheads.

Now, what if, rather than being “essentially harmless” to the young, it did have a mortality risk for them too… but it was 100 times smaller than the risk to the elderly. People would still die under a hypothetical “lock-up the old; immunise the young” strategy, but they would do so in far smaller numbers than under a “do nothing” approach. And this would still be true even if the relative mortality risk was not 100, but merely a closer-to-reality factor of 10.

Combined with general “social distancing” measures and/or other approaches for reducing load on the health services etc., this kind of heightened protective isolation of the most vulnerable might provide a reasonable management strategy during a disease outbreak.

COVID-19

The real question of the hour is: is COVID-19 anything like these hypotheticals? Well, the answer appears to be that it does indeed have a very varied mortality rate across the age spectrum: varied enough that the UK initially pushed heavily for an “isolate the over 70s” response, while being less severe in the measures proposed for younger groups.

The actual numbers that we have for COVID-19 are quite distressing. The fatality rate among the infected may be something around 1% (estimates vary). If two thirds of the UK eventually becomes infected, that would lead to around 440,000 deaths. However, the mortality rate for those under 70 is much lower than this general rate. So, if people above that age could somehow be effectively isolated for the duration, it might be possible to achieve a herd immunity “exit strategy” with a final death toll of maybe 150,000 (optimistically). This is still a very large number, of course — but let me draw your attention to one critically important point: IT IS A MUCH SMALLER LARGE NUMBER.

However, it is certainly true that the overall impact of COVID-19 may indeed be higher than we currently realise (e.g. if it does long-term damage even to the “survivors”), and it’s not guaranteed that sufficient herd immunity would even develop, although most experts seem to suspect that it will. These are very difficult judgements to make — but, no matter what is done, we will be forced to make highly significant decisions in the presence of far less information and far more uncertainty than we would wish. And, yet, decide we must…

(Incidentally, if immunity does not develop naturally, then it will presumably be very difficult to create a vaccine too.. because the point of a vaccine is to stimulate such “natural” immunity, but without risking the full negative effects of the disease).

I should also note that a widely reported model by researchers at Imperial College London suggests an “intermittent suppression” approach which they claim could perhaps keep the UK fatalities below 20,000 over two years, pending the development of a vaccine, although I think it’s fair to say that there is ongoing debate about the viability and correctness of their proposal, and this 20,000 figure doesn’t account for the potential mortality costs of their measures themselves, which I will get to later.

ALTERNATIVE(S?)

Perhaps we might prefer a safer-seeming, containment-centric, “extensive testing” approach instead? But what if the necessary tests simply cannot be made available in time, or they aren’t reliable enough to be effective, or they don’t detect the infection soon enough in the disease course to be helpful, or they cannot be made available at sufficient scale, and administered with enough speed and accuracy, or that containment SIMPLY DOESN’T WORK in the long run (because the disease is ultimately too contagious)?

Even so, it may still be that the mortality (and the long-term morbidity) costs of COVID-19, about which we still know relatively little, are potentially too high for the risk of a “herd immunity” approach to be responsibly taken in practice. Similarly, there might be worries about super-spreaders, overshoot effects, creating excessive viral load, or mutation risks.

If so, we might have to adopt a fairly extreme “LOCK EVERYONE UP” strategy instead, and simply bear the costs which come along with that. But what are these costs? Well, no-one knows for sure, but it’s widely felt that a significant economic recession would ensue, and that lives would be lost from that too in various ways e.g. by general erosion of much of the civic infrastructure (including the normal supplies of food and medicine etc.) that otherwise keeps us healthy.

As a way of providing a ballpark reference, it has been claimed (although the details have been disputed, with justification I feel) that, in the aftermath of the 2008 economic crash, reductions in public spending arising from economic “austerity” measures in the UK led to over 100,000 excess deaths over a period of several years. To the extent that you accept that kind of reasoning, you must presumably accept that a “lockdown recession” would potentially have very significant mortal consequences too. It may, in fact, have negative effects which are greater than those of the disease!

Incidentally, it’s also worth noting that any countries which are taking the Herd Immunity approach would be expected to have higher death tolls in the early phases of their local epidemics, but should hopefully have less severe subsequent waves of infection, and less collateral damage if they manage to avoid other aggressive control measures which are themselves not without cost, as I have indicated above.

THE QUESTION

The ultimate question here is surely this:

Of the options available, which one seems to be the Least Bad at the time when we have to make the decision e.g. what appears to be the realistically attainable path-of-least-death (as evaluated over the duration of the crisis, which is perhaps a year or more until a good vaccine is hopefully available)?

That’s a very hard question to answer, but it’s essentially a question that needs to be investigated with data and estimates and “best guesses” and “models” from informed experts (while also being adapted to political realities and developments “on the ground”), rather than being something that can be decided up-front by any automatic principled commitment for or against any one strategy or other.