As with all of these types of surveys, the data out of New York is preliminary, not peer-reviewed and we have to view it with a grain of salt, if not a jaundiced eye.
Having said that, here are the results of the state-wide sample testing conducted recently in New York, as presented by Governor Cuomo today:
An estimated 13.9% of the New Yorkers have likely had Covid-19, according to preliminary results of coronavirus antibody testing released by Gov. Andrew Cuomo on Thursday.
The state randomly tested 3,000 people at grocery stores and shopping locations across 19 counties in 40 localities to see if they had the antibodies to fight the coronavirus, indicating they have had the virus and recovered from it, Cuomo said.
With more than 19.4 million people residents, according to U.S. Census data, the preliminary results indicate that at least 2.7 million New Yorkers have been infected with Covid-19.
As of a few days ago, when this survey was being conducted, New York state had about 250,000 confirmed cases, or about 1.28% of the population, so that works out to about a 10X multiplier of total cases vs. confirmed cases.
This survey contradicts the preliminary results of 2 surveys conducted in California, which claimed somewhere between 25 and 80 times as many actually infected vs. confirmed tested people. Both of those studies have been widely criticized and even debunked for basic errors.
There are others out there — Chelsea, MA found 30% of test volunteers were positive for antibodies at a time when 2% of the city had tested positive, yielding a raw multiplier of 15X. German researchers found a 7X multiplier. So throwing out the CA surveys, we have a spread of 7X, 10X and 15X, which looks to be starting to settle somewhere around 10%.
So is this survey better or more accurate than the others? We just don’t know.
There are problems with getting a ‘representative sample’ for testing, and because no group of people is truly representative of a larger population, researchers are required to fiddle and diddle (Johnny Most or Chick Hearn?) the numbers to make them ‘more correct’.
Since getting tested is voluntary, people’s decision to get stuck with a needle is influenced by whether they want to know the result, whether they think they might have been infected, and even whether they trust the gubmint.
There are lots of other potential problems, including specificity issues (does the test only show positive for Covid-19, or for other coronaviruses like the common cold?). So none of these surveys are likely to be particularly accurate, but as always, we can only work with the data we are given.
What we can take away from all of these tests, I believe, is to be pretty certain that the number of actually infected people is much higher than the number of confirmed tests.
We’ve known that, intuitively, ever since we found out that many people never show symptoms.
Combined with the knowledge that those asymptomatic people are contagious for around a week or more, even while they have no idea they are infected and spreading a deadly disease, we can begin to get a better picture of just how dangerous this virus is…. and just how dangerous it is to throw our country open before our active case count is low enough to contain through measures like fever screens, testing and contact tracing.
I would like to hope that the growing body of evidence of this will help our decision makers keep us safe.
Unfortunately, that doesn’t seem to be happening in some states as they rush to re-open themselves to the spread of this deadly virus.
Oh, well, at least you Georgians will be able to go bowling tomorrow!
Cheers.