#Covid19UK
Modelling UK data tells us: at an assumed fatality rate of 1%, there are currently about 1M active infections in the UK (as of 10/05). At a fatality rate of 0.5%, it's ~2M. Is this the time to send people back to work? The UK's strategy has been deeply flawed... 1/

- The UK's COVID-19 strategy has meant wide spread of infection with lots of deaths (and probably without much collateral benefit of herd immunity).
- The large pool means infections will take a long time to subside to levels which can be managed with careful monitoring. 2/

Many things, not least failure to protect the vulnerable, and failure to provide adequate PPE to key workers, contributed to the bad shape of the UK at the moment. Here are some other observations from modelling... 3/

There was rapid initial growth of infections - comparable to several other European countries and the US, but faster than, say, India. The slow time course of infection means it takes time to see deaths, and by this point there are already tens of thousands of people infected. 4/

Low testing meant only a small fraction of infections were being picked up - testing, tracing and isolation does not seem to have been a serious part of any control strategy (compare, say, Germany). Also, low testing meant it was hard to see what was going on. 5/

People got cautious in early March, but serious mitigation only began after mid-March. Modelling suggests that by 18/03 at least half a million people were infected. Wide geographical spread meant restrictions slowed the spread but couldn't really limit COVID19 geographically. 6/

Where infection had already arrived - a family, a care home... it spread. The lockdown was "leaky" - infection continued to enter new settings, slower. This is normal, but modelling suggests the scale of leak was higher than, say, in Germany. 7/

Right now: "Back to work" with a large number of active infections still in the community could be the UK's next mistake. Infection levels can easily pick up. By the start of June, even if active infections have halved, there will still be hundreds of thousands out there. 8/

Any silver lining? Depending on the fatality rate about 6-12% may have been infected. This would only slightly slow a future outbreak. (Even a relatively small level of immunity slows an outbreak a little.) There is likely more substantial immunity in some badly hit areas. 9/

To conclude: time is needed for infections to subside to levels manageable by monitoring, testing and tracing. And for this to work there needs to be a really good level of testing - it is essential for people to be able to find out quickly if they have, or have had, COVID-19.