Saturday, March 28, 2020

Fact Checking Dr. Birx

The US Coronavirus Response Coordinator Dr. Birx made a number of questionable statements at the  daily Coronavirus Task Force Briefing on March 26. This post analyzes some of these statements.

Background on Dr. Birx

Dr. Deborah Birx is a well-respected medical researcher who currently serves as the "Coronavirus Response Coordinator" in the office of the US Vice President. She worked in the US Army from 1985 to 2005, and has held various leading roles in the CDC and the US government since 2005. This included different positions related to HIV/AIDS. Trained as an immunologists, Dr. Birx was well qualified for HIV-related roles, since HIV specifically targets the immune system.

In the following analysis of Dr. Birx statements, I will focus on specific statements that are incorrect. I encourage you to check the transcripts or the video of the conference to understand the context of the statements.

False Statement #1

On March 26, Dr. Birx gave the impression that computer models were not able not match the numbers observed in reality.
"... the predictions of the models don’t match the reality on the ground in either China, South Korea or Italy."

Rating


False

Analysis

There is a large number of studies that specifically tried to model the epidemics in different countries, including Italy, South Korea, and China. These studies had to take into account the effect of government measures like "test and track", quarantine, movement restrictions, and so on. Some studies specifically tried to estimate how effective different interventions are, and have started to provide some insights that will be useful to tailor future interventions. However, virtually all studies agree that strong interventions are needed to contain the epidemic, and keep deaths and human suffering to a minimum.

False Statement #2

"There are a large group of people who are asymptomatic, who have never presented for any test."

Rating


False

Analysis

Apparently, Dr. Birx is referring to one particular study by a group of zoologists in Oxford that had received a lot of media attention in the days before. In this particular study, the researchers found that the observed epidemics can also be modeled if the model assumes that there is a very large number of  asymptomatic individuals. If this would indeed be the case, then the majority of people in strongly affected countries would already be infected, and the epidemics in these countries would soon be over, even without any government intervention.

This study in very problematic because it ignores a large number of facts that we know about the epidemic. A well-known epidemiologist summarized "why it can't be trusted". In addition to the reasons he gives, the assumption that many asymptomatic cases exist completely ignores the data we have from "test and track" approaches in different countries, where great efforts were (and still are) undertaken to identify all possible contacts of every infected person, and test and/or quarantine everyone that had contact. One example is the intense tracking of cases in Singapore, which includes the use of antibody tests to identify people who have already recovered from the virus.

False Statement #3

On March 26, Dr. Birx stated that the researchers in the UK had "adjusted" the model which had prompted the "15 Days to Slow the Spread" guidelines, which reduced the number of predicted deaths from 500,000 to 20,000. Here is what she said (from video of the briefings, starting at 2:30:45):
"Many of you saw the recent report out of the UK about them adjusting completely their needs. This is really quite important. Remember that was the report that said there would be 500,000 death in the UK and 2.2 million deaths in the United States.
They've adjusted that number in the UK to 20,000, so half a million to 20,000.
We are looking into this in great detail to understand that adjustment."

Rating


False and misleading

Analysis

The statement that the ICRF has "adjusted" numbers is false. The numbers Dr. Birx refers to can both be found in the study that was published on March 16. Here is this section of table 4:

The study explains clearly that the expected number of death is 510,000 if no interventions would happen (first column, "Do nothing"). It then shows that this number can be reduced to 19,000 deaths by a combination of interventions (third column, PC_CI_SD): closures of schools and universities (PC), isolation of all infected patients (CI), and social distancing (SD). Have a quick look at the second column, which shows what is predicted to happen when schools and other public places remain open: the number of predicted deaths jumps more than 4-fold. For the larger United States, the corresponding number would be about half a million deaths even with social distancing.

This was shown in the initial study. The data convinced the British government to dramatically change its plan of action: rather than using only social distancing measures limited to the most vulnerable groups, the UK implemented a large range of mandatory and strict measures to stop the spread of the virus, and avoid running out of ICU beds.

The results of the study were shared with the US government even before the publication; the study was publicly available since March 16.

The response by the US government has been to issue "guidelines" that are intended to stop the spread, called "15 Days to slow the spread". It's a 2-page document; here is how page 1 starts:

Everything on the first page is directed to sick people, their relatives, or members of risk groups. On page 2, there is advice for everyone that include "Avoid social gatherings in groups of more than 10 people". All those guidelines are voluntary. Experience from many countries all over the world, including the US, shows that voluntary guidelines for social distancing are often ignored. This severely limits the effectiveness of social distancing. The US guidelines correspond to the second column (CI_HQ_SD) in the table above. According to the ICRF model, these restrictions would lead to more than 400,000 deaths in the US,  assuming they were generally followed.

For comparison, here are some of the differences in the restrictions that the British government has imposed to keep the corona virus epidemic at a level that will not overwhelm hospitals:

  • Everyone is required to stay home, except for "very limited purposes" like shopping for essentials
  • Most businesses are closed, including restaurant, bars, and hotels
  • Public venues like playgrounds and churches are closed
  • All public gatherings of more than 2 people are prohibited, with minimal exceptions (e.g. for families that live together)
  • Workers must work from home whenever possible, and employers "should take every possible step to facilitate their employees working from home"
  • All restrictions are mandatory and will be enforced by police and health departments

Overall, the British restrictions are much more severe than the restrictions in US, and doubtlessly will be more effective.

I leave it to you to speculate why Dr. Birx made multiple false statements within one briefing. But if she had wanted to support the current strategy of minimal interventions by undermining the credibility of computer modeling, she could not have done a better job.