A Head Full of Coronavirus Research, Part 8

First, another paper suggesting that since mortality from coronavirus disease is extremely bifurcated, maybe the models should take this into account in looking at spread.   (Medrxiv Paper)   The researchers, who are physicists, assumed that one part of the population, 10% to 20%, has high susceptibility and the remainder has lower susceptibility.  Their model suggests, as others have, that when you model this scenario, population immunity is reached more quickly.

Next up, another paper along the same lines, suggesting that maybe we ought to adjust models for varying susceptibility in the population, and if we do, the level for population immunity drops dramatically.   (Medrxiv Paper)   The authors begin by appropriately questioning the standard model assumption that the population is homogenous in susceptibility.  They assumed host susceptibility was directly proportional to the the transmissibility rate.  They also assumed that those who were more susceptible to infection were more likely to transmit it to others.  If the beginning transmissibility factor is around 2.5, which is what it seems to be for this coronavirus, then a homogenous population would reach population immunity at around 60% infection, but assuming some heterogeneity lowered it to under 50%.  They then tested the theory on Stockholm’s results to date and suggested that a bimodal distribution of susceptibility/infectiousness existed, in about an 80% to 20% ratio.  In other words, most people aren’t very susceptible or infectious and a few are much more so.  In this model population immunity could be reached at as low a level as 18%.

The next paper identifies changes in rates of cases in children and adults.  (Medrxiv Paper)   The study covered about 6500 children and 46,000 adults.  The rates of positive tests among children was consistently lower than that for adults, but at the start in was 7 times lower and at the peak it was 3 times lower.  One conclusion the authors drew was that after the lockdown, adults began transmitting to children.  Really smart to close schools.

This paper deals with the contribution of superspreaders; people who may infect many other people.   (Medrxiv Paper)   These authors modified the typical model to account for different types of contacts–home, work and other.  Work includes school for people at those ages.  They assumed that 10% of the people are 50 times more infectious than other people.  Their conclusion was that if we can control the superspreaders, everyone else can live a pretty normal life.  In particular, the home and work contacts could continue as normal if the other bucket, which tends to include large gatherings, is controlled.

Another paper on whether and under what circumstances a population immunity strategy is acceptable.  (Medrxiv Paper)   The paper is from Japan and the authors said that determining a population-based immunity strategy should depend on the number of deaths incurred and the possibility of exceeding available health resources, which might cause more deaths.  They find that when the population is divided into a high risk and a lower risk population, it benefits everyone if the disease is allowed to spread in the low risk segment at a rate that does not tax health resources while the high risk segment is protected.  Obviously I like this paper because it is consistent with the strategy I advocate, but I advocate that strategy, as do these authors, because it is consistent with the nature of this epidemic and is the lowest risk method for controlling the epidemic.

How infectious is coronavirus in a symptomatic patient?  This statement from the Singapore health authorities address that issue.  (SP Paper)   They declared that it appeared that a patient was infectious for about 5 to 7 days after the patient became symptomatic.  They also found that all patients have stopped viral shedding by around day 15 after symptom development.

What happens if you age-adjust models for susceptibility.  This paper looks at that question.  (Medrxiv Paper)   The Phillipines was the setting and that country has a relatively young population.  The authors first looked at the effect of quarantining infected persons quickly.  They then looked at a model adjusted for age-based susceptibility.  As in Minnesota, people were put in ten year cohorts, including 80 and over.  Weirdly, they defined all groups with respect to the one with the highest number of infections, instead of the highest number of deaths or hospitalizations.  But even with this flaw, the age adjusted model indicated lower levels of infection.

This paper covers antibody development in health care workers in Italy.  (Medrxiv Paper)    Almost 4000 employees of seven hospitals were tested in the Lombardy region.  Antibody testing was used, which I am starting to wonder about, given that T-cells may play a larger role in defense against this specific pathogen.  Altogether, 87% of the employees were negative, 2% were equivocal and 11% were positive.  From 3% to 43% showed positive antibody reactions, depending on the hospital location.   7.6% had positive infection tests.  About 20% of the positive tests were in workers with no symptoms.  Age and sex appeared to play a role in the development and strength of antibodies.

Next up, antibody responses in patients in New York City.  (Medrxiv Paper)   These authors examined the severity of disease in 240 patients and the subsequent strength of antibody development.  Antibody development appeared to be stronger in patients with more severe disease.  They also tested 11,092 patient samples for which disease severity, or even the presence of disease was unknown.  47% tested positive.  Very young children had low rates of positivity, but teens and young adults had relatively high ones.

Last up, another population immunity paper, this one from Italy.  (Medrxiv Paper)   The authors sought to estimate when population immunity would be reached, based on data from Italy, with a starting transmissibility of 2.5 and varying susceptibility by age.  The variation was in relation to death risk.  They found that around 43% of the population would be infected when population immunity would begin.  Interestingly, however, they had the highest rates of infection in the older groups to get there, which would produce a lot deaths.