After talking to the ONS (by Twitter and e-mail) I’m somewhat clearer now about what they’re doing, and while it’s confusing, it’s not quite as bad as I thought and at least it hasn’t made the respiratory death figures useless.
ONS now has a separate category for Covid-19 deaths, which includes all deaths where Covid-19 is mentioned on the death certificate, even if it’s not put down as the main cause of death. Thing is, Covid-19 is a ‘notifiable’ disease, which means that anyone who fills out a death certificate has to note down if the person had Covid-19, or if they even just suspect the person had Covid-19, whatever the person died of. So there will quite a few people who have ‘Covid-19’ written on their death certificate even if they didn’t die of it, and they will go into the Covid-19 death numbers. This means that this category is misleading, and basically useless. Unfortunately, it’s the category that the media is using.
From what I can gather the same thing is happening in the USA, and Italy too.
(I should note that things are different with the flu. It has long been noted by flu researchers that a lot of elderly people die with the flu, but this is not noted on their death certificate, because it’s not considered to be the main cause of death, even though in some of these cases it may have been what finished someone off. So Covid-19 figures aren’t directly comparable to flu death figures, because the former will be inflated in a way that the latter are not.)
‘Covid-19 deaths’ is, then, a most unusual category, it’s not like the other ones. There will be some deaths that are included in this category and in another category, eg. if someone died of a heart attack and they had Covid-19 then they’d be in both lists. I’m assuming that ONS’s other categories aren’t like this.
(You might then worry that ONS is double-counting the overall deaths. They won’t be, though. This Covid-19 category will not be used for that. They don’t just add up all the subsets including this category to get the total death figure, so that’s not something to worry about. There are other things to worry about, but not this.)
Another issue I asked about is, how does the ‘Covid-19 death’ category relate to the respiratory deaths category? The situation is this. It’s neither a subset of respiratory deaths, not additional to it. Basically, it partly overlaps the respiratory deaths. Those Covid-19 deaths which are respiratory deaths get included in the respiratory death stats, the others don’t. The good news about this is that it means that the respiratory death figures are still reliable and useful.
It’s very poor that the ONS does not provide a breakdown of the Covid-19 figures, into, for instance, those who were listed as dying of Covid-19 and those who died with it. (The issues here are actually more complicated than they sound, but still, they could at least do this.) They could also list the number of people in the Covid-19 category who were also in the respiratory deaths category, and the number of respiratory deaths that had Covid-19 as the main cause of death.
It’s very, very poor in general that the ONS has not provided a more detailed and clear explanation of what is going on with the Covid-19 figures (‘disgrace’ might be a more apt word). And of course it’s poor that the national statistics service of an advanced nation cannot manage to do some graphs like my readers have managed to do in their spare time. They’re hard to get useful information out of, and I’m still not even 100% certain that my new understanding is correct. If anyone more knowledgeable than me, or anyone who has managed to have longer discussions with the ONS, wants to weigh in, please comment. I will try them again soon to get further clarification.
Update: I also note that Oxford’s Centre for Evidence-Based Medicine is also complaining about how hard it is to get clarity on these sorts of issues (in general, not just from the ONS). If even they’re struggling, it’s no surprise everyone else, including the media, also is:
Readers of the blog will be aware that we have little trust in the current reported COVID data. Everywhere we look we cannot get a handle on the essential facts or at times we get 2 completely different answers to the same question. The military historian Sir Basil Liddle Hart would have called this “the fog of a pandemic” or perhaps the “fog of information overload”.