The fatality of 41 is surprisingly low. I suspect it is incomplete or something, and will probably catch up tomorrow.
I've been looking at the data, and it's showing something interesting. I've had to rely on using the number of deaths to track the epidemic, because I don't believe the testing in any European country is sufficiently reliable to give meaningful results. One issue with doing this is that deaths have a 2-3 week delay from infections, so this tells us what was happening two or more weeks ago.
First, let's compare what's happening in Italy, France and the UK for the data as of 25th March:
I've used the scientist's favourite - the log plot as it shows very clearly if the epidemic is increasing exponentially or not. A straight line means the epidemic is still growing rapidly. On the other hand, if the line curves to the right, the epidemic is slowing down. The data is plotted from the time that ten deaths were reported - this removes the noise present in the early stages so that the trends can be more clearly seen.
For the first two weeks, Italy showed a growth rate of 20% (doubling every 3.5 days). This has slowed to around 5%
(doubling every 2 weeks). It looks like Italy is turning the corner, at last: the growth is no longer exponential. I take marinyork's point that many deaths are likely not being recorded. There's no real way to account for that: the data is what we've got. I suspect that under reporting can't account for slow down, so the trend is likely genuine.
France looks grim. The number of deaths is doubling every three days - and there's no sign of this growth rate slackening off. I fear that France will turn out worse than Italy: if France continues on this trajectory, it'll overtake Italy in about a week.
The UK is just odd. A few days ago, I predicted that there would be over 1000 deaths by this weekend. I'm happy to say that I was wrong - it's going to be much less (and sad that I wasn't
more wrong).
There are two ways I can explain this. The first is that there was a major change in behaviour at the start of the month. This was when people were being exhorted into washing their hands and paying careful attention to hygiene. Could these measures actually be working to reduce the number of transmissions? That would imply that the hand-oral route is the major infection route rather than aerosol.
The other explanation is that the case fatality rate has dropped. This most likely would be due to improved treatment protocols, so that the critically ill are more likely to survive. This is something I've no knowledge about -
@Brompton Bruce, would the prof be able to shed some light on this?
The frustrating thing is that without a robust testing strategy in the community, we have no idea which is the more likely explanation. This lack of testing means we're essentially blind: it will take a long time before anything we do to reduce R0 will show up in the data (that's probably the reason behind BoJo saying that the restrictions will be reviewed in three weeks, as that's just about the time in which changes can be expected to show up).
There is one other sting in the tail: if the favourable decrease of death growth rate is down to improved treatment, this will rapidly unwind if the NHS becomes overwhelmed. A sharp increase could be expected at that point.
