Panda: Pandering to the public's ignorance
By Scott Sumner
Andrew Gelman has a post discussing a website called “Panda”, which provides a wealth of misinformation about Covid-19. What makes the site of interest is that its board contains some pretty big names, including former Trump advisor Scott Altas, as well as some Stanford University professors:
The board also includes, among others, Stanford medical school professor Jay Bhattacharya, Stanford biology professor Michael Levitt, and Michael Yeadon, a retired pharmacologist and drug company executive who, according to the website, “believes the pandemic was over in the summer”?
Gelman points out that until a few days ago the site was discouraging people from using Covid vaccines:
There was also this, from the organization’s webpage entitled, “You asked, we answered,” under the heading, “Would you have the vaccine yourself?”:
As for any other medication, a vaccine must be shown to be safe and effective before it is introduced to the general public. Vaccines take 10 to 15 years on average to be developed. . . .
Currently, there is no one for whom the benefit would outweigh the risk of these vaccines—even the most vulnerable, elderly nursing home patients.
. . . I guess this statement was a bit of an embarrassment after one of the members of the Panda scientific advisory board publicly stated that he and his mother had received the vaccine. The above link is from 22 Jan 2021, courtesy of the Internet Archive. Go to that page now and that whole section has been removed.
OK, fine. But . . . also no acknowledgment of their earlier ridiculous statement.
And this is just the tip of the iceberg. Even the revised statement is loaded with mistakes:
the mortality overall is relatively mild compared to past severe pandemics such as the 1918-19 Spanish flu and several more recent influenza pandemics such as the Hong Kong flu of 1968 and the Beijing Flu of 1993. The UK government even declared that “[a]s of 19 March 2020, COVID-19 is no longer considered to be a high consequence infectious disease (HCID) in the UK”.
Given that the overall statement was revised within the past week, I’m not sure why they still rely on estimates from March 2020. In any case, Covid-19 is an order of magnitude worse than the Hong Kong Flu of 1968. There was very little social distancing in 1968, and without social distancing the death toll from Covid in the US would already exceed a million. (About 34,000 Americans died of the Hong Kong flu, although the number would be several times larger today, as there are now far more older Americans.)
The low mortality across the South East Asia and Oceania super region is likely driven by other factors, possibly prior immunity.
I don’t think there’s any evidence that prior immunity explains the success of Australia or New Zealand. A recent outbreak in Melbourne spread rapidly before being brought under control, and of course Wuhan was devastated back in January. Does anyone seriously believe that all of China except Wuhan had natural immunity? (Almost all Chinese Covid deaths were in the Wuhan area.) Yes, some countries may have some natural immunity, but it’s disingenuous to minimize the role of behavioral changes, which obviously played a huge role in China, Australia, and elsewhere.
We are unaware of any studies using sound methodology that show a benefit for masks in the general population. The only COVID-19-specific mask study using sound methodology found no significant impact of mask wearing on the spread of the disease.
If you follow the link you find a Danish study that did not even test whether masks help to slow the spread of the disease. To do so, you’d have to test whether mask wearers are less likely to spread the disease. Did they even read the abstract?
The fatality rate in most people infected with SARS-CoV-2 is very similar to that of the flu. COVID-19 is less severe than the flu for children and young people and more severe than the flu for the elderly with severe underlying illness.
I’d call this misleading, albeit not false. It’s true if by “elderly” you mean a 55-year old man. However for older middle-aged people, especially men, Covid is far more dangerous than the flu. Indeed it’s not even close.
And this is just ridiculous:
On the other hand, it has been observed that winter respiratory mortality patterns are usually associated with a single dominant pathogen at any time, so it could be that (this year at least) COVID-19 has simply supplanted influenza and is, in the main, taking the lives that would have previously been lost to influenza.
New York and New Jersey already have more than 65,000 Covid deaths, despite widespread social distancing, and yet they contain less than 10% of the US population. The entire US usually has far less than 65,000 flu deaths each year.
This is also extremely misleading, if not outright false:
There is no clear evidence in the literature showing that asymptomatic transmission is a major driver of the pandemic. The poorly supported theory that suggested this, was the main logic behind lockdown policies, which in any event have been shown to have no beneficial effect on death curves.
The primary worry was that presymptomatic people would spread the disease, but according the Panda those people are not “asymptomatic”:
An asymptomatic person is one who never develops clinical symptoms at all (no sneezing, coughing, fever, loss of taste or smell). This is distinct from a presymptomatic person, who begins to show symptoms after the incubation period of a few days.
A meaningless distinction. Almost every average person would assume the term ‘asymptomatic’ applies to the presymptomatic. People without symptoms often spread Covid.
However, many countries are recording COVID-19 official deaths if there is past evidence of a positive PCR test, or the patient is considered “probable” or “presumed’ to have COVID-19, even where the cause of death is clearly unrelated and symptoms are not present. This generous diagnosing can inflate the number of deaths in the data. Countries categorize deaths as “COVID deaths” using different criteria, so comparisons of such statistics are of questionable validity.
In fact, excess death data suggests that most countries have severely undercounted Covid deaths, and also that the excess deaths cannot be explained by other factors like suicide or people not getting cancer screenings.
Gelman suggests that this website has links to the conservative movement. One thing I’ve noticed over the past year is that conservatives seem obsessed with minimizing the severity of Covid-19, and also seem interested in showing that measures to prevent Covid-19 (such as masks) are not likely to be effective. This “head in the sand” approach has done a great deal to discredit the entire conservative movement with the well-informed part of the population. That’s a shame, as there are areas (such as economic policy) where conservatives have lots of good ideas. But they are rapidly losing votes among the college educated part of the population, and this sort of misinformation doesn’t help.