Skip to main content

Why Is Oregon Still on Lockdown?

A blog post gaining traction in Oregon was brought to my attention. The post makes some dangerous claims, and argues that Oregon should be reopened. I understand the desire to reopen as soon as possible, but the claims made in the post are wrong and dangerous - read below to understand why.

Where are we in the cycle of the disease?

"Professor Hinkley", the author of the post (I'm not providing a link because I don't want to spread false information) claims that the number of hospitalizations from COVID-19 began decreasing before the lockdown began. To support his claim, he annotates a graph published by the Oregon Health Authority on April 5th:
OHA COVID-19 Graph annotated by Hinkley

The biggest problem with Hinkley's analysis is the annotation he wrote over in his graph (and I have circled below): as of April 4th, illnesses between 3/25 and 4/4 may not be reported yet.


Hinkley says the number of cases started to decline even before Oregon implemented "Shelter In Place" on March 23rd.  At first blush, it does look that way. But he ignores the fact that many people who developed symptoms between 3/25 and 4/4 have not yet been picked up by the system. The number of cases on these dates will rise over the course of the next 10 days.  Let's look at a graph from March 28th as an example:

OHA Data Report Released on 3/28

Look at the number of cases whose symptoms onset was March 22nd: 22 patients were reported with 6 hospitalized, 7 not hospitalized, 9 other patients whose status was unclear.  Now look back at the chart from April 5th:  the number of patients who developed symptoms on March 22nd increased from 22 to 41!  

This is an important caveat to remember when looking at COVID-19 data: the number of cases detected always lags the number of cases that have actually occurred. This is especially true when the disease has an incubation period as COVID-19 does. Similarly, the number of deaths caused by the infection lags behind the number of cases by 2-3 weeks. 

The unfortunate reality
"Professor Hinkley" (who, as far as I can tell, has no background in medicine, epidemiology, or public health), claims that "Locking down an entire state and causing 20-40% unemployment for a virus that was already on the tail end of its arc" is "madness". If it were true that the virus was on the "tail end of its arc", then I would agree. Unfortunately, we are not anywhere near the end of our time dealing with this virus. As soon as we re-open society, the virus will come roaring back. The economic devastation of shelter-in-place is enormous. But the alternative is worse: overwhelmed healthcare systems leading to millions dead across america. 


Comments

  1. I'm just curious. You kept referring to the author as "Professor Hinkley." Why the quotes around his name and title throughout?

    You also claimed that not continuing "shelter-in-place" will lead to "millions dead across america." What evidence do you have to support this claim? Having studied the data very closely and extensively, the data-based estimates I've seen (and run) seem to be more in the 100-200k range, and that is without "shelter-in-place."

    ReplyDelete
    Replies
    1. Thanks for the comment - I put the author's names in quotes because it is a pseudonym which they admit to on their webpage.

      Regarding the "millions dead" line, in hindsight I wish I had left that line out. It was inflammatory. However, it's not necessarily wrong. The Imperial College of London model predicted that if no mitigation measures were taken, 2.2 million Americans would die.

      On a fundamental level, if there is no vaccine and no non-pharmacologic interventions (like social distancing), then the virus will infect 70-80% of the population before herd immunity kicks in and protects the remainder of people. Assuming a case fatality rate of 1%, that would result in 2.6 million deaths.

      Delete
    2. What evidence do you have to support your claims about (1) a 70-80% morbidity rate, and (2) your prediction of 2.6 million dead?

      I've looked very closely at the stats coming out of the cases and studies around the world. The general infection rate is in the 10-20% with close, confined exposure lasting many days (varying by age distribution). Further "case fatality" rates are generally not based on infection rates -- since these are rarely known -- but upon actual medically- (and often laboratory-) confirmed cases. What exactly do you mean by "case fatality"?

      Delete
    3. 70-80% is not the morbidity rate, but rather the "herd immunity threshold", which is the % of people who would need to be infected before the virus would no longer be able to spread.

      Because the virus is so widespread, it will continue to infect more people around the globe until it reaches the herd immunity threshold. Fortunately, we can take measures to reduce the rate of spread: social distancing, increased handwashing, wearing masks, case tracking, etc. This will slow the rate at which people are getting infected. However the spread will not completely stop until we have a vaccine.

      I'm not certain what you mean by "general infection rate". Do you mean the number of people around the world who have been infected by the virus? In that case I can tell you with certainty the number is much less than 10% -- it is closer to 1% right now, even when accounting for asymptomatic people and those who haven't been able to get tested. The rate varies a lot from country to country, and city to city.

      The typical measure of how infectious a virus is called "Ro" (R-zero). If Ro is 5, that means one infected person will go on to infect five other people. Estimates of Ro for COVID-19 have ranged from 1.4 to 5.7. The most recent estimates (when more data has been available) have been on the high end of that range. It's important to note that Ro is different from R. Ro measures how infectious a virus is if nothing is done to stop it. R, on the other hand, measures how infectious a virus is at a given moment in time, and takes into consideration countermeasures to fight the virus (like social distancing).

      Case Fatality Rate is defined as the odds of a person dying once they become infected. You are right that it is difficult to determine the case fatality rate of COVID-19 right now because we aren't detecting everyone that become infected (because some people are asymptomatic, and others who have mild symptoms aren't able to get tested). Furthermore, there is disagreement about whether patients who are completely asymptomatic should be included in calculation of the case fatality rate. Still, there have been several small populations that were extensively tested (people on the Diamond Princess Cruise ship and the citizens of Vo, Italy). We can extrapolate the case fatality for the general population based on these smaller populations; most experts looking at this data have estimated the case fatality rate for symptomatic patients to be between 1 and 2 percent.

      I can't put links into comments but happy to provide references if you'd like.

      Delete
    4. You've written a couple times now that 70-80% of people will (eventually) be infected by this (to reach "herd immunity"). It does seem to be the case that 70-80% of people need to be immune to something for "herd immunity" -- at least that is the commonly-accepted number-range -- but this doesn't mean that all of those people must have been infected, does it?

      Your appeal to the Diamond Princess Cruise ship is interesting. Assuming a 100% exposure among those on the ship, what was the overall infection rate, and what was the actual (crude) mortality rate as a percentage of all people exposed to the virus?

      Delete
    5. That's correct: 70-80% of people need to be immune. There are only two ways to be immune to the virus: become infected and eventually recover, or be vaccinated. No one has any sort of natural immunity against this virus. Social distancing buys time so that we can get a find a vaccine and immunize everyone before they become infected.

      It's not reasonable to assume 100% exposure on the cruise ship -- once the virus started to spread, passengers were ordered to stay in their rooms, limiting exposure to the virus. If the cruise had gone on without a quarantine, eventually 70-80% of people would have become infected.

      Everyone on the ship was tested, regardless of whether or not they showed symptoms. 712 people (19% of the ship) tested positive, 9 people died (1.3% of people who tested positive). There are additional caveats to the Diamond Princess data which I'm happy to discuss.

      Delete
    6. How do you know that no one has any natural immunity to coronavirus? What evidence do you have to support this claim?

      Delete
    7. That is a great question, and one that requires a dive into immunology, biochemistry, and some of the studies on covid-19 that have already been done. I'm trying to put together another post on this topic, hopefully I'll have it up in the next day or two

      Delete
  2. The "Professor" has made another post and an Oregonian friend of mine has shared sharing it.

    ReplyDelete
  3. A friend posted the follow up article and I was curious because she is an educated person with a background in healthcare, and that article is a prime example of armchair epidemiology. Then I wanted to do more research. I clicked on the FB, Instagram, and Twitter links, and they all led to squarecash. Then I clicked on on the hamburger menu to find about more about this "professor." This is total BS. I asked my friend how she knew this person, whether or not their first name was professor, etc. She told me it was a pen name. Dude, if you purport some expertise, and you are blantantly critiquing public health policy, show me your real name and credentials. Also, if we take care of the people and the public health crisis, the economy will improve. If we just focus on the economy, everything will suffer.

    ReplyDelete
  4. It turns out to be some antivaxer Republican guy in rural southern Oregon who has some community college credits.
    I guess he managed to misrepresent himself and some graphs enough to go viral, woo.

    ReplyDelete
  5. Hi Paul, I'm a reporter with the Portland Mercury. I've been seeing Professor Hinkley's articles being shared by influential Oregonians and was hoping I could talk to you about the flawed logic and your debunking. Please email me at blair [at] portlandmercury.com if you're up for a chat. Thanks!

    ReplyDelete

Post a Comment

Popular posts from this blog

Why is Coronavirus Antibody Testing Important? Is Antibody Testing Available Yet?

There are a few different types of tests for COVID-19. Currently the most common is a PCR test, which detects genetic material from the virus. PCR tests are great because they are highly specific - they have a very low rate of false positives. If the test is positive, you can be certain you have COVID-19. But the test only works on patients who are sick and still have the virus in their body. What if you want to know if someone had COVID-19 in the past? That's where antibody tests (also known as serology) comes in.

For more details, keep reading.

How Many COVID-19 Tests Can The USA Run Each Day?

The short answer is:  at the time of this writing (March 18th, 2020), the US is able to perform 36,810 COVID tests daily. This number is increasing daily, and updates can be found on twitter at @COVID2019tests. More details below.

After an awful start, US testing capability has dramatically increased over the past two weeks, and should continue to increase as lab testing companies like LabCorp and Quest Diagnostics ramp things up. When we began testing for COVID in the united states, the CDC required that each patient have two swabs performed - one swab of the nostrils and one of the mouth. Last week the CDC got rid of that requirement which effectively doubled our testing capability. Unfortunately, we are still dramatically underdetecting COVID in the United States; several epidemiologists have calculated that we need to be performing about 100,000 tests per day in order to catch up with the disease. The @COVID2019tests twitter provides semi-regular updates on our testing capability.