Drowning in Coronavirus Research, Part 95

I am hoping the tide is turning on giving children a normal school experience.  When a prestigious, very liberal magazine like Science says children need to be in school, that helps.  (Science Article)   The article repeats the pretty well-established facts that children both are infected less often and less seriously and that they do not play a significant role in transmission.  But more importantly, the authors note what everyone should be focused on–the immensely greater harms to children, especially minorities and low-income children, from not being in school.

And in a related story, Europeans keep looking north enviously and wondering why they didn’t follow Sweden’s example instead of China’s.  France, despite an increase in cases, is opening up schools more, again in recognition of the importance of an in-person education.    (French Article)

Another study popped up related to viral load and testing.  (Medrxiv Article)   This one related cycle number to likelihood of patient death in a hospital.  Over 1000 patients were included.  Average higher cycle number was found in patients who survived than those who died, indicating the importance of knowing cycle number and likely viral load when treating a patient.

A couple of articles relating to the question of pre-existing cross-reactive or other immune responses to CV-19.  The first comes from the British Medical Journal.   (BMJ Article)   One of the most interesting parts of the article, which summarized the body of research on CV-19 cross-reactivity, was a parenthetical box relating to the swine flu epidemic of 2009, in which researchers realized that the reason that epidemic wasn’t as bad as originally forecast was that cross-reactive T or B cells were found, especially in older persons.  So here we have a recent example of mis-forecasting an epidemic due to failure to consider this factor, yet all our epidemiology and immunology experts didn’t take this into account early on in the CV-19 epidemic.  The authors also note the bad assumption about random mixing and even susceptibility of the population and how new models are challenging that.  They also discuss the role of T cells, which are rarely the subject of prevalence surveys.

This paper was looking at the prevalence of pre-existing immune responses.  (Medrxiv Paper)  The study comes from Boston University and included a cohort of people whose samples were drawn from before the epidemic.  The researchers used a new, more sensitive antibody assay.  No pre-existing antibodies were found in persons over 70, but they were found in lower age groups.  They also detected antibodies in samples after the start of the epidemic but with no symptoms or actual diagnosis of CV-19 infection.  Many of the pre-existing antibodies were to the nucleo-capsid protein.  Antibodies were weaker to the receptor binding domain.  These studies are suggestive of the potential impact on spread and seriousness of illness from prior immune responses.

While some studies suggest heterogeneity in susceptibility to infection and infectiousness, and in contacts, may lower population immunity thresholds, some prevalence surveys find very high levels of persons with antibodies, which would suggest transmission continuing at a rapid rate for some time.  This study came from Manaus, in Brazil, and found a prevalence of 52% within a month after epidemic peak, and a current likely rate of 66%.  (Medrxiv Paper)   These rates tend to occur in densely populated and poorer areas of countries.  I tend to think less about a threshold for population immunity and more about what combination of factors slows transmission significantly.  In some circumstances a low prevalence could accomplish that; in others it may take high prevalence.

And here is another study trying to guess how many people have actually been infected in the US.  (SSRN Paper)   Lot of complicated math, lot of variation across the country, but also apparently a lot of undetected cases.