The serology hype has passed, which is good since it was driven by unscientific hopes that silent infections were >10x higher than all previous data suggested, so that we could end distancing and shutdown. And now we are seeing good serology studies (1/n) mscbs.gob.es/ciudadanos/ene…
It's important to do serological surveys because we want to know what % has indeed been infected and may be immune, at least temporarily. It will also tell us the IFR, how effective shutdown measures were, and what dangers lie in store for us until vaccines are developed. (2/n)
This study validated an antibody test made by Zhejiang Orient Gene Biotech, finding specificity of 98% for IgM and 100% for IgG. Specificity is crucial when surveying low-frequency events, because your false-positive rate must be lower than the actual event frequency. (3/n) 
With specificity of 100% for IgG then you can assume any positive IgG result in the surveyed population would be a true positive (so it would be good to see the primary data to know the # of negative samples here, and know how confident we are in that 100%). (4/n)
The conclusion from surveying 60,000 residents in different areas of Spain was that about 5% of the population, i.e. 2.3M, had been exposed. With 27k deaths, that's an IFR of 1.1%, which is completely in line with previous estimates of 0.7 to 1.5% from data from China (5/n).
So yes COVID-19 is > 10x more lethal than flu. Yes shutdowns have worked to limit spread. Yes continued distancing and face masks are required. Yes 20x more people can still die. Here's a good article with further links if you are interested (6/n)
the-scientist.com/news-opinion/r…
the-scientist.com/news-opinion/r…
It upsets me how earlier poorly done serology studies misled people, and how the press spread that misinformation, suggesting we were nearing herd immunity or we could use antibodies to allow us to return to work soon - ideas that were way premature and thereby harmful. (7/n)
People tend to stick with the first thing they hear, not noticing or ignoring later facts. People are reluctant to take time to weigh new evidence to overturn established beliefs. Thus the press has a huge responsibility in an epidemic to get the facts right the first time (8/n)
I expect second-tier press outlets to publish extreme headlines to attract views, but the disappointment has been the prestige organizations have also been skipping expert fact-checking in their rush to publish. I'm hoping with the hype level dropping, they can do better now.
