One of the principles of the ‘less wrong’ community, which I find very attractive, is a commitment to open thinking, in other words, to be clear about what evidence is being relied on to make what judgement. In addition – and possibly the most important of all – is a commitment to be clear when a view is changing as a result of finding new evidence.
So I’m glad to have written what I did earlier, as it provoked some good conversations and lines of investigation that have changed how I am seeing this. This represents progress, and is my new drug of choice (actually, that’s a bit flippant, but I’ll make a more serious point on that topic tomorrow). So this post is to explain a shift over the last 24 hours.
The issue that had engaged me was a perceived discrepancy between rocketing infection rates and an unchanging bed occupancy rate. This was the source for the point about bed occupancy:
Up-to-date situation report on the state of hospital bed occupancy in England. pic.twitter.com/LMRMl5UMI5
— Joel Smalley (@RealJoelSmalley) January 2, 2021
I didn’t just go from the tweet; I did go to the NHS site to see if the raw figures back up what was shown in the graph, which they did.
However, in the light of the explosion of infections – why was the bed occupancy rate not changing? Perhaps it was because the argument offered by people like Mike Yeadon was true, ie that the testing regime is compromised by, amongst other things, a very high false-positive rate.
Some tweets from John Bye made me reconsider that possible answer:
It’s sad we still need to say this, but false positives are NOT driving rising cases.
In the summer the False Positive Rate (FPR) at lots of labs was below 0.1%.
— John Bye (@_johnbye) December 28, 2020
In other words, if the false positive rate was madly high then it would show up in other places. That seemed plausible to me (although there were a couple of tweets in that thread that made me go ‘hmmmm’).
So there needed to be an alternative explanation for the discrepancy between the bed occupancy rate and the infection rate (ie why did one not reflect the other; same issue with the death rate, of course, but I thought that had alternative explanations along the lines indicated in my earlier post). I have now had a good conversation with a nurse on the front-line, who unpacked the seeming contradiction, and made a further essential point.
The reason why the bed occupancy rate isn’t changing is because it cannot change – there are only so many bed spaces available. What is happening is that the people who would otherwise be occupying those beds (in a ‘normal’ winter) are now not in hospital at all, displaced in favour of Covid+ patients. In other words, simply looking at bed occupancy rate is not a sufficient guide for measuring the impact of the virus.
The further essential point, though, is that the NHS has still not recovered from the earlier peak in April, most especially in terms of the availability of trained personnel, and that this has a huge impact on what is being demanded of nurses and doctors and support staff now. My friendly nurse was extremely concerned at the capacity of the hospitals to cope with the surge coming down the line due to the Christmas break. The NHS is really up against it.
Lockdown it is then.