Thread by @ewanbirney

An explainer thread (often I feel I am pitching these to journalists as much as anyone else) on COVID this month.

Context: I am an expert in one area (human genetics) with battlescars in complex data flow+analysis; I know experts in most other areas and aim to be curious about their viewpoints; I have a clear conflict of interest in that I am consultant to ONT, which makes a new COVID test

Again, worth reminding people of the overall situation; SARS_CoV_2 is an infectious virus which causes a nasty, often lethal, disease in a subset of people. It is now across the world.

We know a lot about the virus since January. We know its sequence and can design tests; we know many aspects of its infection dynamics (it is not the best virus; it infects in a bursty way); we know aspects of what makes people at risk of the disease (age, sex, obesity).

If we left the virus to move unchecked through the population a large number of people would die in a short period of time, many of them with plenty of natural life ahead of them + many people would get a long term disease (we know less about this follow on disease, "LongCOVID")

Further this rate of disease burden would swamp modern healthcare systems; the modern healthcare systems would be forced to make horrific choices of who lives or dies. Even if one was rational about it, it would be horrible.

Like all Coronaviruses, it looks like this virus' infectivity has risen in the autumn and winter in Europe. Many factors could drive this - straightforward temperature, or the increased clustering of people indoors, out of rain etc. All of Europe has seen a strong growth in cases

(because of the deeper testing across Europe, we can see this far more clearly than the equivalent time of infection rise in Feb. One important thing to remember is not to compare *case* numbers Spring to Autumn; comparing hospitalisations or deaths is valid).

Because of our understanding, testing, mask wearing, avoidance of risky situations, isolation and tracing the rise in infections is certainly less than in Spring - however, it is still steadily rising.

In the UK context this is best shown in the unbiased population samples of ONS and REACT; similar assessments happen in other European countries; hospitalisation levels is another unbiased marker for this.

This slow rate of rise is good, but frustratingly we need to have in shrink + be suppressed (* this is broadly a tactic not a strategy - see below). We've improved many things, but not enough. This is why more restrictions - non pharmaceutical interventions - NPIs are happening

Now a little sidebar on NPIs. I think having a crude on/off view of "lockdown" vs "no lockdown" is mistaking a huge amount of complexity. For example, it is clear that schools should be kept open. Because of the way broad NPIs effect lives, life and livelihoods this is complex

NPIs are not for free in any of these aspects, and one does need to think this through. One criticism of the current epidemic is we view this too much as a tension between "Health" and "Wealth" and not a joint "overall wealth".

We have the frameworks to make this weighing - Welfare Economics from the economics side; QALYs including "societal" or "life" QALYs from the healthside. We should use these frameworks more.

However, no matter the framework we have to impose more NPIs in many countries because once there is a high level of infections two main things happen and then there is one additional consideration

(1) we really will fill up healthcare capacity at a predictable rate. Just waiting for that point of filling up is crazy - (a) it is virtually impossible to "tune" epidemic to have some constant rate. Epidemics either grow or shrink. (b) you will have to make this decision.

(2) Certain things, most notably "trace" work more effectively at low infection levels. R is the growth parameter, not the infection level. So an R around 1 but low infection levels is reasonable. R around 1 with high levels is very much not reasonable.

The other additional consideration is that one has to think about things getting worse. Because of the lag in the system of making a NPI change to that impact hospitalisation rates (its at least a month) one has to have *additional* capacity present in the system for worse case.

Now - as I mentioned (and I can feel some of my colleagues metaphorically shooting daggers at me in this thread) suppression is not a strategy by itself - we need an exit.

There are two levels of exit. One level of exit is this upping of NPIs might well be able to be relaxed when more efficiencies come into TTI. Remember one efficiency is natural (lower infection levels, better Trace efficiency)

More can be done. This is all operational stuff and really about lots and lots of details. Beware anyone thinking this is simple; also beware the statement that TTI has failed in the UK - parts of it have worked well, parts less well and it is a hard task to suppress

One external way of viewing this is that one of the best in class in Europe TTI - Germany - has also had its capacity broached. We are, across Europe, going to have to up our collective game.

Can more be done - yes. Will it be enough? In many ways we will trade off broad NPIs for better TTI in the mid-term. We need TTI the best it can be.

Even this is not the full story - that goes to vaccines. There is a veritable wall of vaccine Phase III trials, some that will almost certainly report this year. Keep your fingers crossed (I do - I am on one of these trials!)

Early successful vaccines will not be a fairy tale ending. They might not work so well; there is a depressing and bleak possibility that *none* of them work (but there are an awful lot of them). They might work well but even then just logistics will make it complex.

My view is that early vaccines will be much like broad NPIs, but ones one can administer in a single shot, and, when appropriate target to risk groups - they are sort of "super" risk adjusted NPIs

As such, again, we can relax other NPIs, keep our TTI levels working as high as vaccines come in. As more diverse vaccines come in, we can use combinations (either disjointly across the populations or jointly with extreme care on safety)

Hopefully then the next thing on the horizon is Spring, and genuinely coronavirus infection levels dropping, and then we have to prepare for autumn 2021 with even more tools (vaccines and hopefully more therapeutics).

The final strategy is therefore this becomes likely another circulating coronavirus; perhaps one we will all need vaccination from regularly in older age.