“selected by Trump purely because his pandemic position, essentially one of zero intervention, conforms to the alternate reality of Trump’s administration”
Scott Atlas, an advisor to the White House coronavirus task force, was selected by Trump purely because his pandemic position, essentially one of zero intervention, conforms to the alternate reality of Trump’s administration. A neuroradiologist and NOT an expert in infectious diseases, he was recently rebuked by the usually restrained Fauci as lacking “insight or knowledge or experience” for the advisory role. Atlas appears to be pursuing a libertarian ideology in ensuring all his analysis conforms to a worldview that is wholly anti intervention. He recently endorsed a call by a libertarian think tank — the Great Barrington Declaration — to let the virus rip, which would result in millions of more deaths.
Proof of his inadequacies were well in view as far back as April in his woeful misreading of antibody research, demonstrating that conspiracy theorists are not the only ones who suffer from numerical illiteracy and critical blindness. The results of two research studies into the prevalence of antibodies were then widely disseminated. A medical research team from Stanford University tested 3,330 residents of Santa Clara County, California and concluded that by early April 2.5 to 4.2% of the county were or had already been infected. A second study by University of Southern California (USC), that included some of the authors of the previous study, estimated between 2.8 and 5.6% of the Los Angeles adults had been infected. This study consisted of 864 tests. In an article with over a million shares Atlas joined in with a handful of others to argue the research implications proved Covid-19 was far more widespread than the nationwide swab-testing then indicated. So apparently widespread for Atlas to assume fatality rates were 20 to 30 times lower than concurrent WHO estimates.
Blinkered interpretations took the Santa Clara study as proof of a fatality rate for the county of just 0.12 to 0.2% people. Those with a libertarian agenda like Atlas, were quick to apply these stats to national and global fatality rates and to par Covid-19 with the less ominous flu with its death rate of around 0.1%. Atlas, with others, all too quickly packaged this framing into an existing narrative that measures to contain the virus were unnecessary and a case of destructive state overreach.
The studies, however, and their wider implications, were riddled with errors, false assumptions and biases, tantamount to idiocy in how far short they fell of appropriate expertise. Firstly, the Santa Clara study, recruiting its participants off Facebook, left itself exposed to a selection or ‘consent bias’ by being prone to drawing out healthier members of the population (and people who were not cocooning) and encourage participants with erstwhile Covid symptoms, who being previously unable to get tested, to take advantage of the opportunity afforded by the study. Facebook and online recruitment are also liable to be more exclusionary against older members. The Santa Clara study did not even adjust properly for age — a core factor in fatality rates. (The USC study made claims of greater representativeness despite its smaller sample but failed at the time to publish its methodology).
As well as that, the test kits used bt the Santa Clara study are prone to significant errors in antibody measurement. The Santa Clara study, used an antibody test (not approved by the FDA) with a sensitivity rate of 80.3% (i.e. a measure of test’s ability to identify the presence of antibodies) and a specificity rate of 99.5% (i.e. a measure of the tests ability to identify the absence of antibodies). The margin for errors meant that the study could produce enough false results to widely skew the study’s small number for antibody carriers. For example, with a 99% specificity, the 100 antibody positives a study identifies from testing 1000 participants could in fact be made up of 50 false positives. The test kit for the Santa Clara study, it is reported, could show up 2 false positives for every 371 true negative samples. With just 50 positive results from 3,330 participants this is liable to make any implications drawn from the positive findings meaningless. However, this basic mathematics completely escaped Atlas’ reasoning.
Atlas’s discussed the studies as though proof of the lower-than-flu fatality rate for Covid-19 on a universal or global scale. Leaving aside propensity for higher rates of infection than the flu and the deeply erroneous findings on which these claims were made, the anchoring of the studies in two locations does not render them suitable for nationwide or global representations. Santa Clara, for example, was a coronavirus hotspot unlike other USA counties back then.
These interpretations were bundled in with claims (based on obscene statistics that related [undercounted] age-related Covid deaths to total population) attaching almost the entirety of the threat to the over 70s, as well as people with underlying conditions. Through a privileged lens, opinion pieces, by conservatives like Dr. Scott Atlas, manage to view older people, and people with underlying conditions, as distinct from the general ‘healthy’ population. This ignores the widespread existence of people with (often hidden) underlying conditions. The prevalence of such conditions is known to be increased — with severe Covid-19 consequences — by air pollution, which, varying nationally, discredits national implications drawn from localized studies. Underlying conditions can also be worsened by racial and healthcare inequalities — i.e. one of the reasons suggested for disproportionate black, Asian and minority ethnic (BAME) fatality rates, along with their disproportionate role as urban essential workers.
The additional simplicity of the idea that we can just cocoon older people also ignores socioeconomic and cultural contexts that affect the degree to which multi-generational shared living occurs (i.e. households with grandparents), as has been suggested as one of the reasons for higher fatality rates in Italy and Spain (although questionable in regard to widespread covid-19 deaths in care homes). Reopening society and the economy and forcing people back to work, arguably, has exposed more older people and caused more fatalities.
A critical reading of findings, from studies like that of USC and Santa Clara, needed to also explain inconsistencies with real-world fatality rates. In real terms, even when compared with their total populations (and not just the total infected population) Covid-19 hotspots like Lombardy, Italy and New York had already passed the fatality rates incorrectly drawn from the antibody studies.
It is clear from Atlas’ outright wrong standpoint on this and everything the man has printed and said since, he is not fit to advise anyone. But that, it seems, matters not to a Superspreader President bent on willing into existence an alternative Covid-19 dreamlike reality in this nightmare he has created for the American people.