UK (Sky News)

Coronavirus: What the COVID-19 statistics tell us – and what they don’t

By April 24, 2020 No Comments

There are a few things we know about COVID-19 but they are far outweighed by what we don’t know.

We don’t know how to cure the disease. We don’t know how to make a vaccine. We don’t know precisely where or how it started. We don’t know how many people have COVID-19 here or indeed around the world.

Even in other countries where there is more widespread testing, we don’t know how close the case fatality rate (which is to say the number of positive cases versus deaths) is to the real fatality rate – though we are pretty clear it’s some way off.

We don’t know how many people are true asymptomatic carriers. We don’t know precisely how long people can remain infectious after their symptoms have diminished. While we know it tends to kill more elderly people than young, we don’t really know why the disease affects some but is far more dangerous for others, why some people’s immune systems react the way they do, with terrible consequences.

Human beings crave certainty. Our world is built on a bedrock of data, scientific findings and probabilities. And central to everything is the notion that there is nothing we cannot quantify, model or manage, given enough expertise or money.

But COVID-19 throws most of that out of the window. In time we will have a better understanding of the disease and be able to answer most of the questions above. But for the time being there is an astonishing amount we simply don’t know.

Even a simple question – how many people have died of COVID-19 – is actually far more difficult to answer than you might expect. We get figures every day from the Department for Health and Social Care about how many people have died thus far. But many of these deaths actually happened some days or indeed weeks before.

We know as much because last week the NHS started publishing details of when those COVID-19 fatalities actually happened, and it turns out that for much of the past fortnight or so those DHSC numbers have dramatically understated (by 20% or more) the number of people who have actually died each day.

Hylton Murray Philipson
Coronavirus survivor Hylton Murray Philipson is full of praise for the NHS

It is to be expected that the on-the-day statistics are not comprehensive – it takes time to diagnose the disease, carry out post-mortems, seek familial consent and go through the administration necessary to report these deaths, even before you consider that we are in nationwide lockdown.

There are other difficulties too: did someone really die of COVID-19 or did something else kill them even though they tested positive for the virus? We don’t know the answer and for the time being the statistics don’t provide much guidance.

We know that the DHSC numbers only include hospital deaths, so they are certainly undercounting the true total, but by how much? Some people have talked about a terrifying COVID19 “iceberg” where the majority of the death toll is hidden beneath the surface.

Then again, numbers produced by the Office for National Statistics today suggest the opposite – that 93% of all COVID19 deaths happened in hospital – so we are aware of nine-tenths of the iceberg instead of only a tenth.

Those ONS numbers are supposed to be more definitive, but even they aren’t necessarily an accurate reflection of where the country is at any given point.

Death certificates take time to process and comparing this period to previous years – as we tend to get a sense of whether the country is in the midst of a period of “excess deaths” – is even more difficult than normal this time of year because it so happens the date of Easter plays havoc with those yardsticks of what constitutes a “normal” week in a “normal” year.

Pic: Beretta/Sims/Shutterstock
How are people reacting to Prime Minister Boris Johnson’s admission to intensive care?

There are other wrinkles too: in England and Wales the total number of deaths from all causes rose sharply above the average in the week to 27 March – in other words well into “excess death” territory. Yet in Northern Ireland that same week, weekly deaths were well below the average.

Or consider this: in a “normal” year, around 19.6% of all deaths registered in the week to 27 March have influenza or pneumonia. Yet in that week this year the proportion mentioning influenza, pneumonia or indeed COVID-19 was actually lower at 18.8%. A promising sign or simply statistical noise?

For all its apparent certainty, data is often clouded by doubt. And this is especially the case with COVID-19.

But having got those provisos out of the way, many of the statistics suggest that the UK might well be starting to turn a corner. Hospital admissions are no longer rising at scary rates, nor are confirmed cases (though let’s not get into the wormhole about how reliable those numbers are).

Even if the disease is no longer being spread so quickly, it will take some time before the growth rate of deaths drops significantly. In other words we may well still be heading for around 10,000 deaths by the weekend. This, note, is not a forecast but an extrapolation based on the growth rate from the past week.

But bear in mind that UK deaths been growing as fast as they were last week and we would likely have hit that milestone by the middle of this week, in other words more or less now.

Of course, that gets us back to the data problem.

The deaths we’re hearing about today may have happened last week. Those people may have caught the virus two, three, maybe more weeks before that. We are looking through a series of rear view mirrors, each more distorted than the last.

The upshot is that even as we get the worst news of all – deaths rising into five figures – the reality might be both worse and better than the numbers suggest. Worse because the daily deaths numbers are understating the reality on a given day. Better because those deaths reflect contagion from before the lockdown took effect.

These are tragic statistics. Each “datapoint” is in reality a family losing a loved one, it is friends and colleagues who will never bump into an old mate. It is heart-rending.

But by focusing on the statistics we can glean whether the UK lockdown is working or not. And there are tentative signs in the data – very tentative but there nonetheless – that the curve is flattening and that the UK might be starting to bring the pandemic under control.

But the reality is we may not know we have passed the worst until the peak has already passed. There is so much we do not know about this disease: and that includes the extent to which we are vanquishing it.

0

Leave a Reply