Read on Twitter
Who feels like a sobering wake up call? Well here goes. This paper from @hsalje et al (inc @SCauchemez) examines what we can learn from France’s initial surge in the pandemic. And what it means for herd immunity
https://science.sciencemag.org/content/early/2020/05/12/science.abc3517">https://science.sciencemag.org/content/e... 1/n
https://science.sciencemag.org/content/early/2020/05/12/science.abc3517">https://science.sciencemag.org/content/e... 1/n
France has seen considerably more than 20,000 deaths so far. We are at a point where we can start to estimate the numbers of actual infections – including those not picked up in case counts. This paper does that... 2/n
The results are compatible with existing data from France on serology, even if we would like to have better data on serology all told. They point to an infection fatality ratio of 0.7% (95% CrI: 0.4–1.0). Ie between 1 in 100, and 1 in 200 chance of death following infection 3/n
Unsurprisingly given what we know the risk of death is much higher in some groups. The infection fatality ratio is different from the case fatality ratio. We often tend to identify cases on the bases of symptoms, so we miss the milder ones. The IFR account for this 4/n
This is interesting because the estimate it matches well with reports from New York. This was serology that passes the smell test, but might be overestimating the numbers infected If you want a more informed viewpoint @florian_krammer is probably the best place to start 5/n
There’s also a preprint that is telling us a very similar message. https://www.medrxiv.org/content/10.1101/2020.05.02.20088898v1">https://www.medrxiv.org/content/1... Over three weeks, this found an increase in seroprevalence (ie exposure) from 3.1%, to 6.1%, to 9.7% over three weeks in Geneva 6/n
importantly “no differences in seroprevalence between children and middle age adults” which is relevant to the question of opening schools. This yields a similar IFR to the French study, and the rough one from NYC 7/n
And here is a thread from Madrid also. Again in a place that suffered a severe initial pandemic surge, 11% seroprevalence was obtained. We need at least 60% for significant population immunity
https://twitter.com/_MiguelHernan/status/1260625031119409156?s=20">https://twitter.com/_MiguelHe... 8/n
https://twitter.com/_MiguelHernan/status/1260625031119409156?s=20">https://twitter.com/_MiguelHe... 8/n
Now to put this into a little context – it takes time for immunity to develop. And continued transmission in these places will have likely increased the proportion immune, as well as the numbers of those dead 9/n
I will leave you with the upshot from @hsalje et al that the measures put in place in France pushed the reproductive number down to around 0.6. Those measures are no longer in place. Here it comes again - in France and elsewhere 10/fin
