According to Dr. Daniel Erickson, co-owner of Accelerated Urgent Care in Bakersfield, Calf., his urgent care clinic has completed 5213 COVID-19 tests, of which 340 (6.5 percent) were positive. He and his business partner, Dr. Artin Massihi, during a press conference, further say that 12 percent of all California COVID-19 tests have been positive. And here is where they make their FATAL FLAW in their methodology.
They extrapolate that finding (12 percent of those tested were positive) and conclude that 12 percent of the California population is positive. That assumption is flawed and doesn’t even make common sense.
The problem here is known as SAMPLING BIAS in Epidemiology. The patients tested are NOT a represented sample of the general population. They had symptoms that prompted testing. Their findings cannot be generalized to the general population. It’s the equivalent of trying to estimate the average weight of Americans by measuring the weight of individuals visiting a weight loss clinic.
Based on this flawed assumption, they conclude that 4.7 million Californians have had COVID-19 and use that figure in the denominator of their fatality rate. They incorrectly figure the case fatality rate as 1400/4.7M or 0.03 percent. This false claim of a minuscule fatality rate is why they are getting attention.
Erickson clearly does not understand the problem with sampling bias as he goes on to try to estimate the population prevalence from the proportion of positive tests in New York State, the USA, Spain, Sweden, and Norway.
He correctly points out that physicians take classes in microbiology, molecular biology, and immunology. Medical training also includes basic epidemiologic principles, among which this is about as basic as it comes.
To give you an idea of how implausible a fatality rate of 0.03 percent is, let’s look at New York City. NYC has a population of 8.4M and has experienced 12,500 deaths. If you make (the incorrect assumption) that everyone in the city has had COVID-19, you get a fatality rate of 0.14%…nearly five times the fatality rate of Dr. Erickson’s California calculation.
Now, since we assumed that EVERYONE in the city was positive, this gives us a floor rate. The rate cannot be any less than 0.14 percent. Of note, the case fatality rate of influenza is about 0.1 percent. So, the absolute floor case fatality rate of coronavirus is greater than influenza.
As I have stated before, I do not know if it is safe to open things up or safer to keep things locked down. While there is evidence to support each argument, I lean toward the latter.
However, Dr. Erickson’s evidence is complete FICTION. He is entitled to his own opinion, but not to his own facts. You can’t just make stuff up. YouTube keeps removing his video for violating its terms of service. Now, I have mixed feelings about this. I feel that YouTube, as a non-government business, has the right to make decisions on what they host; removal of content does not constitute a violation of constitutional rights, as I understand it. That said, I don’t favor removing opinions from the public discourse.
It’s less clear to me what to do with demonstratable false claims, which may have a significant public health impact. Although I may not agree entirely with its removal, YouTube is trying to remove FALSE information, not UNPOPULAR opinions.
Side note, they repeatedly point out that they are in a better position to know what is going on because they are seeing patients, unlike those epidemiologists in an “Ivory Tower” who have not seen patients in decades. I am a Pulmonary & Critical Care Physician and Epidemiologist. I am writing this at the end of my shift, seeing patients in a COVID ICU.
