Revisiting "Covid is a leading cause of death in children" ... AGAIN

An updated version of the paper I fact checked in June 2022 has been published in JAMA and is making waves, again.

This has been the paper that just won’t die, and now it’s published in JAMA Network Open. I fact-checked an early flawed pre-print in June 2022, after the CDC used it in the VRBPAC and ACIP meetings to get Covid vaccines in children under 5. At the time, the CDC used it to claim Covid was a top 5 cause of death in children of every age group (<1, 1-4, 5-9, 10-14, and 15-19). That was obviously not true, as anyone who knows anything about pediatric Covid mortality should have immediately realized. I’m still so disappointed that anyone believed that, much less people at the CDC and their supposed expert advisors. How am I more familiar with the data than they are?!?

The pre-print was then revised in late June 2022 as a result of my critique, and I posted a Twitter thread responding to their revisions, which I have since archived on Substack. In November 2022, Eric Feigl-Ding and other Zero Covid activists re-hyped the pre-print again, presumably to fear-monger about kids as we headed into winter. I posted another Twitter thread about the pre-print in November to provide some more context. Now that it’s been revised yet again and published in JAMA, pushed by Eric Topol on Twitter as well as many other “experts”, and covered on CNN and elsewhere in the media, I wanted to update my response to it.

Hopefully people are starting to understand that these kind of papers and stories are designed to push an agenda…

So who’s behind this study?

• It’s written by a team of Zero Covid proponents in the UK (many from Imperial College of London), including Seth Flaxman, Deepti Gurdasani, and Juliette Unwin. At the time of their initial pre-print, they weren’t even aware how to find underlying cause of death numbers for Covid in CDC WONDER until I sent them a link.

• The team is made up of mathematicians and computer scientists, not medical doctors. They specialize in machine learning, statistics, epidemiological modeling, etc., not pediatrics or infectious disease.

• Many of these same UK authors partnered with the CDC on a 2021 modeling study to estimate how many children have been “affected by orphanhood” due to Covid (which they define very broadly) and again in a 2022 follow up paper. Both were promoted on social media by the CDC and got a lot of media attention.

• They likely focused on pediatric deaths in the US instead of their own country because the death rates are higher in the US than the UK, likely due to our generous counting of Covid deaths. In addition, the UK health agency has been involved in studies of their pediatric deaths in detail, providing reassuring data about the low risk to children (here, here, and here). Meanwhile, the CDC has taken the opposite approach, refusing to study the circumstances surrounding pediatric deaths, and instead repeatedly pushing fear about pediatric hospitalizations and deaths in order to encourage childhood vaccination.

• In emails between the CDC and corresponding author Seth Flaxman that were obtained via FOIA by Epoch Times, two different people at the CDC actually apologized to Dr. Flaxman for the negative attention their pre-print received after it was used in the ACIP meeting, and one thanked the team for their “important work.”

Is Covid actually a leading cause of death?

To be clear, nobody is claiming Covid is THE leading cause of death in children and young people. That’s very different than saying A leading cause of death, which just means it’s one of many top causes.

As for whether Covid is one of the leading causes of death is a matter of semantics, not math or data. The CDC has a default database query that ranks the 15 leading causes of death based on the underlying cause of death on death certificates. They categorize deaths into various pre-defined groups. One of those groups is “accidents” which includes car accidents, drug overdoses, drownings, and other accidental deaths. (If that group were split up, all three of those individually would be more than Covid deaths in this age group, but that’s how the CDC ranks them.) Using CDC WONDER, you can query any age group, location, and time period you want. For the ages and time period the authors selected, Covid is the 8th leading cause of death in children, making up ~2% of deaths in that age group. Based on that, they can call it a “leading cause of death”.

But what that realistically means is another story. Is the 8th best scorer on your favorite sports team considered a “leading” scorer? That probably depends on who you ask, and how the points are distributed among players. What if that player only scored ~2% of the team’s points that season? It’s really just a matter of perspective. Ranking just doesn’t tell the story very accurately. But it is an effective way to stir up fear about children and Covid, and bring out all the “experts” who fought to keep schools closed, sports cancelled, and children masked in perpetuity.

What does the study actually tell us?

Let’s break down the data the authors presented. The paper includes this table of the top nine causes of death from 2019, plus Covid from August 2021 - July 2022 (more on those dates later). You can see that accidental deaths are NINE TIMES Covid deaths, and murders and suicides are each more than triple Covid deaths. And that’s compared to 2019 deaths from those other causes. (This is important.) So while the “Top 10”-style ranking is a way to exaggerate the effects of Covid on mortality in this age group, when you look at the numbers this table really doesn’t tell a compelling story of Covid doom. (And that’s even without the caveats about the dates that I mentioned.)

Now let’s look at the ages affected… Notice that the Covid death rate in infants is much higher than other age groups. This is because infant deaths are always on a very different scale than other ages, so mixing them in is a bit misleading. Likewise, Covid deaths are higher in older teens than in younger teens, so including 18 and 19 year olds is also a questionable choice. (Pediatric is normally defined as ages 0-17.) Mary Pat Campbell, a life actuary, addresses these age range decisions in more detail on her Substack about this paper.

The authors also present a graph showing what month the deaths occurred, which is also very telling. They chose dates that included the worst months of the pandemic for ages 0-19 — the Delta and Omicron waves (August 2021-February 2022) — which I marked with a red line below. You can see that deaths before and after that time are much lower. The authors changed the time period at least twice since their initial pre-print (which included deaths from the whole pandemic, both cumulative and annualized) to select the 12 month period with the highest Covid deaths at the time. Note that they could have used Sept 21 - Aug 22, but didn’t. Comparing August ‘21 to August ‘22 makes it pretty clear why they made that choice. As each of the peak months fall out of the 12-month window, the total will drop further and further.

“You’re just a Covid minimizer who doesn’t care if children die!”

I get this a lot from Zero Covid activists. It’s disgusting. I’m a mother. I love children. Of course I don’t want to see children die. In fact, I care a lot about pediatric deaths, including ones that don’t list Covid on the death certificate. And one of the things that I think gets overlooked is how deaths from other causes have increased in the past few years, likely in part due to the massive disruption in our society.

The study authors compare deaths to 2019, but the total deaths for several of leading causes of death have changed. I made the graphs below a few months ago to show how some of these other causes of deaths have increased since 2019 (I used the earlier time period of April ‘21 - March ‘22 from the pre-print). Accidents increased by 1798 (24%) and homicides increased by 1105 (40%)!

Accidental poisonings (a subset of Accidents) had more than doubled. And how many people realize drowning deaths outnumber Covid deaths in this age group? And drownings increased as well during the pandemic. It’s hard to know for sure why, but we know that many kids missed swim lessons during the pandemic and other kids were left at home unsupervised because they couldn’t go to school.

In Closing…

My concern is that the singular focus on Covid as a significant harm to children and young people is misguided and dangerous. We’re ignoring other things that are bigger risks to children while too many in public health and the media obsess about one virus, and distract parents from other risks that they should consider. When you look at things more holistically, it’s clear that Covid deaths in children are rare compared to other risks that people think very little about. When my daughter was young, I was far more cautious about car seat safety than I have ever been about Covid safety in the past few years. But the CDC can’t sell vaccines to reduce car crash deaths. (Based on my Twitter searches, the last time the CDC tweeted about car seats was 2019, for what it’s worth.)

And none of the issues I’ve discussed above even get into how generously we count Covid deaths in the US, especially in children, which I consider a whole separate issue that I will write more on another time. (The study authors actually suggest we’re undercounting pediatric deaths, but there is no indication that’s true.)