Statistics work. Back on 17th April, I wrote an article for Nation.Cymru in which I said the peak of deaths had passed. It was early days; figures were only just beginning to suggest a fall.
But I had no doubt about it; we’d locked down on 23rd March, and we knew that time from infection to death was around 18 days to three weeks. We also knew what had happened following lockdowns in other places, such as Italy.
I still remember lying on the bed in casualty in Bronglais hospital, Aberystwyth, where I had found myself at 3am on the morning of Good Friday, 10th April, suffering with what I thought might be Covid. I was there because of breathing difficulties, yet because I was getting enough oxygen they did not admit me. My breathing issues were neurological – my mind was not signalling to my body properly to breathe. I didn’t realise what was happening to me at the time, and nor did the doctors; the neurological aspect was not understood. Yet it turns out people have died from that; I was lucky.
Anyhow, the doctors and nurses were clearly concerned about the rise in cases. I said to the doctor, quite insistently, not to worry, and that the Easter weekend would see the peak. I explained the figures to him. I don’t think he knew how to take this. He nor, seemingly, anyone else at the time had any idea; a prominent health official had predicted, without any good reason as far as I could tell, that the local peak wouldn’t be until June!
We can now use statistics to predict what will happen for the forthcoming lockdown, and it is more important than ever that people understand this. If we do not, we will be vulnerable to the suggestion that “the lockdown isn’t working”, or even “see, it was going to go down anyway”.
So, what’s happening at the moment? Well, cases are rising; deaths are rising. Reported cases are at an all-time high, but we must not let that fact confuse us. Most people who got Covid back in March and April were not tested. Because I wasn’t admitted, they threw away the swab they’d already taken; there wasn’t enough testing capacity to waste it on me! But now, because (more or less) anyone who needs a test gets it, testing data are useful for spotting trends; that’s quite different to previously, when testing data gave very little useful information.
Figures over the week vary a lot. Those of us watching the figures on the PHW web site know, for example, that data released on Sunday and Monday are consistently lower than other days, and Tuesday tends to be a catch-up day from the weekend. So we need to look at 7-day averages to be able to spot the trends.
We also know that the average time from infection for symptoms to develop is somewhere around five days, but that is hugely variable.
When will deaths and cases fall?
For deaths, we have reasonably consistent data from March and April. I predicted it correctly at the time, so am reasonably confident in making another prediction; it will be around 18 days to 3 weeks after lockdown.
We will only really be able to judge in retrospect, because figures are noisy (especially when the numbers are relatively low, as they are now compared to April), but expect the peak deaths somewhere between 10th and 13th November. That’s after the lockdown is set to end.
What about cases? Well, they will peak earlier. It’s less certain exactly when, because there are more unknowns; we don’t have reliable data from the past to go on, so I am about to make some educated guesses. There’s still uncertainty about exactly how long the incubation period (time to first symptoms) lasts.
The length of time chosen for lockdown of 16 days was not accidental; it is likely that some can remain asymptomatic for this long. There’s also the question of when people tend to get tested, but if we assume it’s an average of two days after the onset of symptoms, we’re probably looking at around seven days from infection. Give another three days for the result to be formally reported, and we’re looking at a total of 10 days.
This would suggest a peak in cases around Monday 2nd November. I think that’s probably a minimum, and I wouldn’t be surprised to see it later. Especially because, when dealing with averages and a rising trend, around half will be above the average; so whilst the rate of increase might stall, the increase might continue on through the first week of November.
So when people point to figures and say “It’s not working”, try to bear in mind that it’s not expected to show in the figures until at least then.
Is the lockdown the right approach?
Yes. The figures are rising, and we simply have to address that. Local lockdowns have not succeeded – at least, not sufficiently (though some seem to have been more successful than others).
My concern however over the forthcoming lockdown is that primary schools, and some other pupils, are going to be allowed to return after the half-term break. A number of reports have shown that primary school children are one of the greatest sources of transmisson, with some of these putting them at the top.
For the sake of four days (Monday 2nd being an inset day – at least, around here), we are, as I made the analogy recently, patching all the holes in the bucket except for the largest one. I hope I’m wrong, but I fear that could prove to be a mistake.
And my Covid experience? It still goes on; the neurological symptoms I experienced in April were the first signs of nerve problems – be that damage, immune system interference, or whatever; no-one seems quite sure. I’m reasonably confident that it will go away eventually, because, although it’s been a bumpy road, it’s heading the right way.
But after my Long-Covid experience, my appeal to everyone is: treat this thing seriously. You might think you won’t die if you get it, but there are other serious consequences for many. And even if you don’t get it bad, the person you end up passing it on to might. The lockdown will only work if we cooperate.
Follow me on @AngharadHafod@toot.wales for daily Covid updates.
Angharad Shaw is a candidate for the Plaid Cymru list for Mid & West Wales in the Senedd elections in 2021.