Californian study finds very high 35 to 1 ratio of actual infections to officially confirmed infections in Silicon Valley

Californian study finds very high 35 to 1 ratio of actual infections to officially confirmed infections in Silicon Valley, by Steve Sailer.

Here’s today’s much-awaited PDF preprint (not peer reviewed) by a Stanford team that performed blood antibody tests on a fairly representative sample of 3,330 Santa Clara County [Silicon Valley] residents on April 3-4. Stanford professor authors Bendavid, Bhattacharya, and Ioannidis have been prominent skeptics of the recent doom-oriented conventional wisdom. …

Results: The unadjusted prevalence of antibodies to SARS-CoV-2 in Santa Clara County was 1.5% (exact binomial 95CI 1.11-1.97%), and the population-weighted prevalence was 2.81% (95CI 2.24-3.37%). Under the three scenarios for test performance characteristics, the population prevalence of COVID-19 in Santa Clara ranged from 2.49% (95CI 1.80-3.17%) to 4.16% (2.58-5.70%). …

So, using antibody tests, the rate of COVID-19 infections was found to be about 1.5% in the sample. Still nowhere near the 60%+ required for herd immunity.

Scaling the sample up to the whole county suggests that there are between 48,000 and 81,00 people who have been infected. But Santa Clara County has had only 1,870 official confirmed cases. So the ratio of actual infections to officially confirmed infections is about 35 to 1. This is much higher than the 2 to 1 or 4 to 1 found in previous  studies.

But the study probably overestimates the number of infections, possibly substantially:

Even a study like this one that made an effort to reduce self-selection will still have problems with self-selection, although it’s hard to guess how they will bias results. They probably got a lot of people who signed up for the drive-thru because they are worried they might have the virus, perhaps because they are out and about. …

Also the study left out age, which seriously compromises the result:

[Commenter Philip Neal] The authors of the study adjust their raw data for zip code (essentially class), sex and ethnicity but not for age.

The Silicon Valley population is quite unusual:

It has a very high percentage of world travelers, but it also went heavily for work-from-home fairly early in March and has a population that is way above the American average in the cognitive power to grasp and follow new rules. …

[Commenter Elli] Or maybe they are all skinny and hardy with high vitamin D levels, and substance abuse and mental illness or its treatment are protective.

I used to live and work in Silicon Valley. It is notable for the relative lack of fat people, smokers, and people with underlying health problems. They are generally not the sort of people coming down with COVID symptoms.

And the antibody tests have a LOT of false positives. If you test positive there’s a 1 in 2 chance you were never infected.

Commenter Just Sayin’:

This leads me to the conclusion that an individual’s reaction to a SARS-CoV-2 infection is strongly mediated by both genetics and pre-existing conditions. Most people are lucky and have either no response to a SARS-CoV-2 infection or a very mild one. An unfortunate minority are genetically predisposed to a severe reaction. Pre-existing conditions may amplify an individual’s response to a SARS-CoV-2 infection. …

If these rather reasonable conclusions are correct then the SARS-CoV-2 virus will continue burning its way through human populations until a significant proportion those predisposed to a severe response to infection are removed from the gene pool and herd immunity emerges. This is basic Darwinian theory. It’s been operating for thousands of years.

And if that is true then all the economic, social, and political chaos created by current, stringent public health measures may be for nought. They may just delay the inevitable culling of the herd that occurs when a new pathogen appears on the scene.

The statistics of false positives mean there is a reasonable chance that none (i.e. zero) of the people in the study were infected.