A Head Full of Coronavirus Research, Part 67

I know, I know, it is Thanksgiving, but as usual I am up very early and this post was mostly done last nite.

I first just want to re-emphasize a finding in the last episode of this series, in regard to seroprevalence.  In Minnesota, that data, particularly the comparison of antibody prevalence to detected cases, would appear to suggest that right now we may have had up to 1.5 million cases.  Those cases, as you will recall from the age/case distribution chart, are largely among the large segments of the population between ages 15 and 40.  These are the people with the most contacts and therefore highest likelihood of both exposure and infecting others.  That percent of Minnesotans infected would drastically slow transmission opportunities, and as I have said, with substantial hedging that I repeat here, I believe we likely have plateaued and will be headed down, subject to whatever havoc Thanksgiving plays with the numbers.

This occurred without any intervention from government and we should absolutely not allow politicians to take any credit for this.  The virus and the population are following their own rules and that shapes the epidemic.  The Governor should have been patient and once again made a panicked, herd mentality move.  In doing so he caused further economic damage and terrorization among the population.   And here is another dubious milestone, the state has now tested half its citizens at least once for CV-19.  To what end is very unclear, since all that testing made no difference in stopping or shaping the recent case wave.

Finally, the Powerline Blog people have been incredibly generous to me is helping spread my information and data.  Scott Johnson there has had his travails with getting DOH to allow him to participate in briefings and ask questions.  He finally, after having to sue, is getting three questions a week answered.  You can find those questions and answers here for this week.  Kind of enlightening what they say.  (PL Post)

The Lancet study focused on viral loads, viral shedding and duration of shedding, and compared CV-19 with the original SARS and MERS.  (Lancet Study)   79 studies on CV-19 were included, but 58 of these came from China and the vast majority involved only hospitalized patients.  No study found the presence of viable virus longer than 9 days after symptom onset, although viral shedding could occur for a longer period of time.  The highest viral load in the upper respiratory tract occurs during the first week of illness.  The peak appears to coincide with symptom onset.  Several studies suggested viral loads were similar at the start of infection among symptomatic and asymptomatic persons but faster viral clearance in the asymptomatic group, suggesting a shorter period of potential infectiousness.  Many of the key questions only had a handful of studies directly on point, which may limit the utility of the meta-analysis.  The results support the notion of a shorter isolation period of ten days and would suggest that asymptomatic persons are actually not likely to be the source of significant spread of CV-19.

This Science article also focused on items related to transmission.   (Science Article)   It comes from China and involved detailed contact tracing.  The authors discovered that 80% of transmission came from 15% of primary infections, indicating substantial variability in susceptibility and/or infectiousness.  Transmission risk increased with duration of exposure.  Lockdowns increase transmission risks in families and households but isolation and quarantine reduce risks across all types of contacts.  Household contacts have the highest risk of transmission followed by extended family, social and community contacts, with health care contacts having the lowest risk.  Transmission risk was highest around the time of symptom onset.  Children were at significantly less risk of transmission than were adults, with the very elderly at highest risk.  The interval between both infection and symptom onset between an infector and infectee was estimated at around 5 days.  The authors estimated that around 63% of all transmission occurred before symptom onset, but believed this might have been affected by lockdown measures and early isolation of cases.  In an environment where such measures are not widespread, it is likely a much lower percent.

The New York Times weighs in with a story regarding scientists questioning whether small gatherings, like those people would have at Thanksgiving are actually the source of much spread.  (NYT Story)  I mentioned that Minnesota got an unfavorable notice in this story in a post last night.  It is interesting how post-election we are seeing sources like the NYT suddenly questioning some of the more hysterical reactions to the epidemic.  And Minnesota, along with the Dakotas, gets singled out again for failure to address more serious areas of concern, like nursing homes.

And this paper questions whether widespread testing of asymptomatic people may actually hinder efforts to limit transmission.  (Medrxiv Paper)  The authors address the notion that creating a program that repeatedly tests asymptomatic people will slow spread.  They  find that these programs can result in an overall increase in infections due to delays in giving results and people being unsure what behavior to engage in while awaiting results.  It is just a modeling study, but it points to real world issues, including that in most places PCR testing results aren’t available for at least two days.  Of course, one might question the whole raison d’etre of massive screening for a pathogen which has so little population level morbidity.

Some places are reporting very high levels of seroprevalence.  Qatar for example, according to a recent paper, among its general work force, which is quite young, has a prevalence of 60% almost all asymptomatic.  (Medrxiv Paper)  That is an astounding level.