Two Good Pieces on "Long Covid"

 Here is one in Stat News titled Setting the Record Straight: there Is No Covid Heart. The title is a little misleading. They make very clear that there are some rare, severe complications that lead to serious heart problems. But it is far, far less common than some early reports and scare-mongering indicated. The piece pionts out something that I noticed months ago when people were waving around the study of Ohio State University athletes who'd recovered from covid (15% of whom supposedly had heart abnormalities). There was no control group for this study! If you scanned a bunch of random people, you'd find a lot of "abnormalities". So you can't just scan a bunch of non-random people (say, people who'd recovered from covid), detect some abnormalities, and declare that covid is the cause. The Stat News piece points to a study showing a similar prevalence of heart abnormalities in athletes without covid. 

Here is one by Stuart Ritchie titled Does Long Covid Really Exist?. Ritchie is known (well known?) for being a debunker of scientific and statistical fallacies. (His recent book is titled Science Fictions: How Fraud, Bias, Negligence, and Hype Undermine the Search for Truth. Just to give you a flavor of what he's known for.) I don't exactly care for Ritchie's tone in this piece. This part, for example:

In rhetorical terms, Long Covid seemed the perfect stick with which to beat the Covid Sceptics — it added extra weight to our case by bolstering the already scary death statistics, and was the perfect comeback to a breezy “let it spread” attitude. So perfect that I hesitated while typing it out. Could it be too good to be true?

It's not totally clear he's doing a mea culpa, that he's acknowledging he shouldn't have been grasping for a club with which to bludgeon the skeptics. And he leaves it as an open question. He does link to the Adam Gaffney piece that I posted about recently:

Still, as the medical scientist Adam Gaffney has argued, it’s likely that some substantial proportion of people reporting Long Covid are actually people who’ve never had the virus. Which might help us understand why the numbers on Long Covid are so weird. Some sources argue that “10-30%” of people who have had a Covid infection go on to experience it — which is itself already quite a range. But look at a UK study released this week (which hasn’t yet been peer-reviewed and is in preprint form). The researchers — some of whom are colleagues of mine — were able to dig into electronic health records from the NHS, and produced a startling figure. Of the 1,199,812 people they found who’d had a positive test for Covid, been hospitalised for Covid or been otherwise diagnosed with Covid, just 3,327 had also reported Long Covid — that’s 0.27%, a different universe from the other numbers.

I can pick nits, but it's a good sign that Ritchie is publicly doubting the truth of this rhetorical tool in the covid policy fight. I'll count it as a win, at least in the sense of "influential thought leaders updating their beliefs." 

________________________________________

This FAQ website by Ritchie and some of his colleagues was set up to "debunk" claims from "covid skeptics." He links to it at the top of his piece. I actually think it reeks of bad faith. See entries like "Claim: Children don't spread the virus." And "Claim: Covid is only a problem for the elderly and vulnerable." See what this website is doing? It's converting a claim about relative risks and propensities into categorical true/false claims. "Children don't spread the virus" isn't absolutely 100% true, in the sense that it's impossible or it never happens. But it's still a reasonable short-hand for guiding policy. Transmission in US schools has been negligible. If we're discussing the policy question of whether to re-open schools, the "dumb skeptic" who categorically denies that children transmit the virus is closer to the truth than the person who insists that this is a huge problem. Also, see the table of infection fatality rate risk by age on this page.

Immediately below this table it says:

These numbers undermine the idea that only the "oldest-old" are at high risk of death from Covid, although of course they are at a substantially higher risk than younger groups (this has the effect of raising the average age of death, obscuring the fact that many younger people still die of the disease).

Do they seriously think people are claiming that the 0.004%  and 0.068% figures are literally zero? Someone isn't listening very well. You can discuss relative risk and say things like "demographic X doesn't have to worry about Y" without literally claiming that Y poses zero risk to X. Rounding figures these small down to zero is a lot closer to the truth than insisting that they're so large you should turn your life upside down to avoid them. Risk calibration in the first case is much closer to how we treat normal hazards in the background of our daily lives. And why do these people always insist on 65 as some magical cut-off? 

According to the Office for National Statistics, there had been 10,603 deaths involving Covid-19 among the under 65s in the UK by the week ending 15 January 2021.

Most of these so-called covid skeptics will acknowledge that there's some kind of smooth (but dramatically sloped!) curve of mortality risk, and that there are a lot of infirm people in the 55-65 category. Indeed, "elderly and vulnerable" implies some acknowledgment of risks to younger people with pre-existing conditions. Props to Ritchie for updating his priors on "long covid", but the FAQ created by him and his colleagues is just terrible.