Epidemiological semantics

The tweet embedded above is the first in a fairly lengthy thread about the use of the words “endemic” and “epidemic” (“pandemic” gets a bit part in the drama) in the discourse about COVID.

There’s quite a bit of bad-tempered bickering (well, it is Twitter after all) about whether or not COVID is or isn’t now endemic. Some of the disagreement appears to arise from the fact that some of the combatants in the discussion don’t seem to have twigged that the word endemic has two similar, but distinct, meanings. Both of them get used in discussions of disease and public health, but they don’t mean the same thing. There is also the issue of the word endemic being used in a contrastive way with the word epidemic to unpick. Let’s start with what endemic means.

There’s the “is natural here” meaning, as in: “Malaria is endemic across this entire region”.

And then, there’s the “has become a feature of” meaning, as in: “The problem of deliberate planting of misinformation has become endemic on some social media platforms”.

Something that is contagious, and invading a new region, can never be said to be endemic in that region in the first sense of the word while the invasion is going on; if it’s invasive it doesn’t belong and so it’s not endemic. On the other hand, if the invasion succeeds and the pathogen (or weed, or insect, or rodent, etc…) becomes established in the region, its population dynamics will typically settle into some kind of pattern in which it is no longer spreading. Once that happens, it would be appropriate to use the word endemic in the second sense to describe its dynamics.

Of course, with the passage of sufficient time, something invasive that becomes endemic in the second sense might appear to be endemic in the first sense, if any trace of the invasion has been lost in the mists of time and the thing appears to have always been there. In fact, given that everything was at some point in the past, co-located at/in the big bang, this temporal shift between the two meanings must be true (in a trivial sense) in every case.

So much for endemicity, what about epidemics? In common use, the word epidemic has a connotation of rapid increase and spread of something (usually disease). The question in the tweet is really asking, “when can we start thinking about this thing as having settled down to some kind of stable background level and stop worrying about it increasing/spreading?”. This is a contrast between the second meaning of endemic and the common meaning of epidemic.

In a technical sense an epidemic is any change in the intensity or extent of disease over time (including decreases); endemic diseases (in either sense of the word endemic) can have local, transient epidemics; colds and the flu are examples of this phenomenon. While there are circumstances where it matters to be really clear about what we mean when we used words like endemic and epidemic, most of the time the intended meaning is clear enough from the context and really doesn’t — at least to this epidemiologist — seem worth getting into an argument about.