Okay, first, my cohort trends update, I inadvertently copied the Minnesota deaths column instead of the CDC one after saying I was going to use the CDC actual week of death numbers. I have it fixed and when my web person gets it inserted into the post, I will republish it, partly because with the right number of deaths in a week it is even clearer how the hospitalization and death rates continue to trend down. But wait for the testing normalized version, which should be ready this weekend. That is very revealing.
Also on the deaths front, I took my first look at the CDC deaths for Minnesota since the updating pull I did last Friday. Already since then I have identified one death posted to a week over two months ago, so the reporting of old deaths continues. I will give a regular update every day or two. We can begin to see how many deaths are added each day by the CDC and compare that to Minnesota date of report numbers.
And given my recent morbid fascination with death, here is a study from Italy looking at CV-19 death certificates. (SSRN Paper) Over 5300 death certificates from the peak of the epidemic in the region were examined. The authors found that CV-19 was the underlying cause of death in 88% of cases, usually accompanied by pneumonia and respiratory distress. Comorbidities were present in 72% of the deaths. The authors suggested that this meant a large number of healthy people died, but I think it is far more likely that the certifier didn’t bother to list comorbidities that weren’t perceived as directly contributing to death.
In case you think that it isn’t possible that hospitalization numbers fail to accurately reflect CV-19 disease trends, this study should disabuse you of that view. (Hosp. Article) The authors reviewed over 300 hospitalizations in England out of concern that there may be miscoding of whether they were related to acute CV-19 disease. They found that as the epidemic progressed an increasing number of admissions treated as CV-19 ones were actually readmissions of people who had CV-19 some time ago and were being readmitted for an unrelated purpose and admissions which were not for CV-19, but CV-19 was diagnosed during the stay. By the end of the study period, these accounted for over half the admissions attributed to CV-19, but wrongly. Need the same kind of study here.
More indications of reductions in needed medical care comes from England. (EHJ Article) The researchers looked at major cardiac procedures in 2020 compared to the previous two years. An alarming 45,000 fewer procedures were performed this year. These are not minor procedures, they are typically necessary to avoid serious deterioration in health. The authors express grave concern about the future implications for mortality.
Another antibody study from France, this one finding a prevalence of about 5% by the end of the survey period in May. Prevalence was significantly lower in children. Multiple assays used, which is good, but sensitivity may have been set too high to pick up all positives. (Medrxiv Paper) Prevalence was again many times reported cases.
Another antibody prevalence study, this one from a city in Mexico. (Medrxiv Paper) The prevalence was a very high 30%, with about a 21% prevalence among people who never had symptoms. I think it is a pretty safe bet that this city Veracruz, will never have another serious swell of cases. Enough people are infected to prevent widespread transmission.
And one more antibody study, from New York City blood donors, so not random. (Medrxiv Paper) In July prevalence was 11.6%, almost all from asymptomatic people.
For those of you interested in learning more about T cells role in adaptive immunity, this is an excellent and understandable paper from the great folks at the Center for Evidence-Based Medicine. (CEBM Article)