COVID19 update, April 23, 2020: community immunity testing results from Belgium; non-COVID19 hospital care; India

(1) DE STANDAARD (in Dutch) reports on a new immunity study in Belgium. Researchers from the University of Antwerp collected residual blood samples of 3,686 patients that had originally been taken for other purposes (e.g., to check for anemia) and checked those for COVID19 antibodies. The samples were collected on March 30.

The Antverpian team found that about 3% of the samples had antibodies — if their sample were truly representative, that would imply about 300,000 people had antibodies for COVID19 around March 30. 

Let’s work with this a bit, shall we? According to worldometers, on that date (March 30) Belgium had 11,899 documented cases. This implies a Dunkelziffer/“dark number” (De Standaard uses this English term) to documented cases ratio of about 25:1.

As of April 22, Belgium had 41,889 documented cases — if we (dubiously) assume that the “dark number” ratio is constant, then about 10.6% of the population may have antibodies at present. 

How much would you need for herd immunity? The herd immunity threshold %HI depends in a very simple way on the effective reproductive number R of the virus: 

%HI = 100% * (1 – 1/R)

If R≤1 then the epidemic will die out anyway and %HI is zero. For R=1.1 just 9% would already be enough, while for R0=1.5 you’d need 33%, for R=2 you’d need 50%, and for R=2.5 you’d need 60%.  (Corollary: if Belgium does have about 10-11% with antibodies, it doesn’t need to keep R below 1.0 with social distancing measures, but can let things slide a little higher. As the percentage of immune residents grows, further relaxation is possible.)

(2) At Sarah Hoyt’s blog, a guest post by “Scarlett Doc” called “Healthcare Charlie Foxtrot” about the current situation in US hospitals for non-COVID19 care. These are the fruits of rigid edicts by domineering, not-too-bright bureaucrats: entire hospitals sitting on their hands waiting for the COVID19storm to hit (which is largely confined to NYC and a few other hotspots), while myriad non-COVID19 patients go untended. There are even hospitals furloughing most of their medical staff. The article is aptly illustrated with a picture of a dumpster fire. Read and weep.

It gets bad enough even without meddlesome middlebrow bureaucrats with Messiah complexes. German hospitals by and large continued normal operations. Yet DIE WELT (in German) reports on how internal medicine wards in German hospitals see such a drop in admissions for their “big 3” emergencies (heart attacks, strokes, and appendicitis) that it is making doctors suspicious. 

“It cannot be that we suddenly have 30% fewer strokes than usual because of corona” says one — so they suspect patients are staying away when they shouldn’t, out of fear of contracting COVID19. “In 2018 there were 210,000 heart attacks and about 300,000 strokes in Germany. That these numbers have suddenly contracted because of the Corona-epidemic, nobody in the medical community believes.”

(3) (Hat tip: Alex W.) Quartz India wonders why the remarkably low toll in India. A young population pyramid is a plus, but against that stand two minuses: multigenerational families and high incidence of chronic diseases even among fairly young people. Then again, the weather being very hot and humid (bad for the virus), universal BCG vaccination, and broad (hydroxy)chloroquine use in areas where malaria is endemic could all be factors. (Incidentally, monsoon season in India is flu season there, so we could see a surge then.)

Related, however, a recent preprint claims that the BCG differential is “an illusion created by testing”:

(4) Finally, could the serine protease nafamostat (an anticoagulant that also has some antiviral properties) be a drug candidate for COVID19? A Japanese group shows in vitro evidence in this preprint:

UPDATE: Matt Ridley, popular science writer and member of the House of Lords, gives a layman-friendly overview of COVID19 drug candidates in the special 10,000th issue of The Spectator.