Fact-checking a COVID-19 rumor
In this newsletter, I will be fact-checking a COVID-19 rumor. As rumors go, it's not an especially believable one. My broader purpose is to discuss why rumors like this take hold, and how they might be dislodged. As you'll see, I believe that statistics play a role in the problem as well as the solution.
The rumor
Since COVID-19 vaccines first became available, one of the rumors that keeps popping up is that the vaccines are killing pro athletes. One week ago today, for example, Senator Ron Johnson (R-Wis.) complained on a podcast that "we’ve heard story after story. I mean, all these athletes dropping dead on the field, but we’re supposed to ignore that." Four days earlier, John Stockton, the NBA Hall of Famer, said that "There’s 150 I believe now – it’s over 100 professional athletes dead, professional athletes, the prime of their life, dropping dead that are vaccinated, right on the pitch, right on the field, right on the court."
Evidence for the rumor
On the internet you can find a list of 603 vaccinated athletes who have experienced cardiac arrests and/or died. That's it. That's the only evidence. (More on this list shortly.)
Assuming you could trust the list, why would COVID-19 vaccines kill athletes? Well, the side effects of vaccination include myocarditis (inflammation of the heart muscle) and pericarditis (inflammation of the outer lining of the heart). In the heat of competition, an athlete with one of those problems might have an increased risk of a serious cardiac event.
In short, even if the rumor strikes you as ridiculous, there is evidence, albeit weak, that can be cited on its behalf. This makes it a little more dangerous than, say, the rumor that Bill Gates is using vaccines to implant microchips in our bodies. I don't think any Senators have pushed that one.
Lets look at the stats.
Debunking the rumor, part 1
As I said, the only "evidence" that vaccines are killing athletes is a list, one that’s published on website called Goodsciencing.com. (In Appendix A, I discuss whether this site is a credible source of information. My conclusion: It's not. Not even close.)
Good Sciencing's list was drawn from a mix of news reports and personal blogs. Most of the reports simply indicate that an athlete – presently active or retired – suffered cardiac arrest and/or died, sometimes while playing their sport, sometimes not. No evidence is provided that vaccines were to blame; rather, whoever compiled the list simply assumed these athletes had been vaccinated in recent months. Many of the stories don't even mention vaccinations. The most famous name I spotted on the list is MLB Hall-of-Famer Hank Aaron, who died in his sleep on January 22, 17 days after being vaccinated. The medical examiner who performed an autopsy found no signs of cardiac arrest or adverse reactions to vaccination. Aaron, age 86, was judged to have died of natural causes.
Assuming without evidence that vaccinations kill athletes illustrates the "post hoc ergo propter hoc" fallacy – i.e., the assumption that because X occurred before Y, X must have caused Y. Hank Aaron died after getting vaccinated, but that doesn't prove the vaccine contributed to his demise. He also died after January 21. It's safe to say that January 21 didn't cause his death.
In sum, no credible evidence supports this particular rumor. In the next section, I provide some data that speak against it.
Debunking the rumor, part 2
Vaccines rarely cause cardiac problems or fatalities in any group, including athletes.
For instance, there are currently 2,132 reports of myocarditis or pericarditis among vaccinated Americans 30 or younger (see Appendix B for details). Over 54 million people in this age group have received at least one dose of the vaccine. In other words, less than 1 in 25,000 young Americans who've received a vaccine subsequently developed a heart problem. (To put that number in perspective, the total number of professional athletes in America's five most popular spectator sports – football, baseball, basketball, ice hockey, soccer – has been about 5,000 per year in recent years.)
Myocarditis and pericarditis following vaccination tend to be mild conditions and resolve on their own. Although estimates vary, it's safe to assume from current data that fewer than 1 in 300,000 people will get vaccinated, develop one of these inflammations, and then experience cardiac arrest.
The Good Sciencing list consists of athletes who had cardiac arrests and/or died, and who were presumably vaccinated. The only national database in which physical problems are specifically linked to vaccines is the Vaccine Adverse Event Reporting System (discussed further in Appendix B). VAERS includes, among other things, reports of 14,506 fatalities attributed to vaccinations. The actual number is probably lower. Health care providers are legally obligated to report adverse events following vaccination, even when there's no evidence that vaccines caused them. Moreover, anyone can file a report, and some reports have been removed owing to falsification.
Still, even if we assume that vaccines caused every single fatality in the VAERS system, the rate of fatality following vaccination would be less than 1 in 100,000.
I could go on, but you get the idea. Adverse effects of COVID-19 vaccines are exceedingly rare. You might wonder though: Is there something about athletes that makes them more prone to myocarditis or pericarditis following vaccination, and then more prone to dying from one of these problems? There's zero evidence for this. No scientific data, no reports from league officials on rising fatalities in their sport (for any reason), and no known medical reason why being an athlete would increase the chances of inflamed heart tissue following vaccination.
Why do rumors persist?
Why do people cling to rumors when supporting evidence seems weak (and disconfirming evidence is plentiful)? And, what can we do to encourage a reasonable consideration of the data?
Here are four very different reasons why rumors are sticky:
1. Limited input.
Some folks don't rely on many sources for their news. If the one the news outlet you listen to claims X, you belong to a subreddit that promotes X, and your spouse and best friend say X, then of course you'll believe X. There's nothing unreasonable about that. Even if "X" is the rumor that vaccines are used to implant microchips into peoples' bodies, it's reasonable to believe that if it's all you hear. After all, my belief that Saturn is a planet is reasonable, because I've never heard otherwise. I’m not able to directly verify the fact myself; I've simply decided to trust people I consider experts.
As far as I can tell, people who only rely on a few sources for their news often don't fully appreciate how those sources can be biased. They either fail to understand the technology (e.g., they don't realize their news feed is driven by an algorithm) and/or they lack an appreciation for how "facts" may be driven by ideology.
2. Confirmation bias.
Confirmation bias refers to the way we seek information consistent with our beliefs, while overlooking contradictory information. I say "we", because studies have shown that everyone, including scientists, are prone to this bias.
Here's a personal example: My father believed that women are terrible drivers. (Throughout my childhood he was on active duty for the Air Force, so we spent a lot of time in the car.) Every time he saw – or thought he saw – a woman drive poorly, he would loudly reiterate his belief in female vehicular incompetence. This illustrates confirmation bias, because my father never pointed out men who drove poorly or women who drove well. (Studies suggest that he might not have even noticed these counterexamples.) Confirmation bias also ensured that my father's belief would get stronger with each new confirming instance. And, if he had heard a rumor that "Jane Doe" caused an an accident, he would’ve immediately believed it.
Analogously, if someone assumes that vaccines are dangerous, they might start a list of vaccinated athletes who experience health problems. Over time, their list will grow, and so will the strength of their belief. In time, even the sketchiest rumor will be viewed as credible. This might explain why the Good Sciencing list includes blog entries in which the writer blames vaccines without providing evidence.
Confirmation bias has created a lot of mischief during the pandemic, because it directly undermines the informal statistical thinking that helps us sort through COVID-19 data. For example, the CDC’s data on COVID-19 hospitalizations show that during any given week since December 2021, COVID-19-related hospitalizations occur among more than 80 out of 100,000 unvaccinated people, but among fewer than 6 out of 100,000 vaccinated people. Even though these stats clearly show that vaccination reduces hospitalization rates, confirmation bias leads some anti-vaxxers to say that vaccines don't work, because they keep hearing about vaccinated people who get hospitalized.
3. Pride.
People don't like to be told that they're wrong. They don't want you to correct their rumors, point out flaws in their logic, or ask them to replace confirmation bias with statistical reasoning. And, what people believe about certain topics may be central to their identity, so when you attack their beliefs, they feel you're attacking them.
4. The challenges of probabilistic thinking.
Cognitively speaking, it’s easier to say that vaccines "work" (or that they "don't work") than it is to treat effectiveness probabilistically – i.e., in terms of how much vaccination affects the risk of infection. The difficulties of doing so are exacerbated when numbers are filled in. (For example, if a vaccine is 90% effective, we can’t predict that 10 vaccinated people out of 100 will get infected. Rather, we can only predict that infection rates among vaccinated people will be 90% lower than among unvaccinated people. As I pointed out in an earlier newsletter, even reputable journalists struggle with that kind of distinction.)
There's also a fear factor. Probability, by definition, implies uncertainty. For some people, it's more comforting to think that vaccines work than to think that they work XX% of the time, even if XX is a large number.
Solutions
How do we dispel unsubstantiated rumors?
To start with, it might help to remember that rumors stick for very different reasons. Each source of stickiness I described above calls for a different strategy.
In an ideal world, you’d confront each source directly. You would tell the person who believes that vaccines kill athletes: You're listening to the wrong news. Or: You're a victim of confirmation bias. Et cetera. Then you would cite the relevant statistics.
Clearly this would never work in the real world. Here's what psychologists and communications experts recommend instead: Start by acknowledging the other person's point of view. Paraphrase what they say, without judgment, and comment on the merits of their perspective – or at least acknowledge any points of agreement (e.g., shock over the death of athletes). Try to identify common goals. Then invite them to consider your views. Present the stats carefully (e.g., acknowledge that vaccines do have adverse effects before noting that these effects are rare). Be tactful.
As I wrote the last paragraph, I couldn’t help thinking that it still sounds awfully idealistic. In my experience, conversations like this rarely end well. What we need is more than casual conversation. We need to talk to people before they get entrenched in particular ways of thinking and attached to particular news outlets. We need to talk to them more than casually, and more than once. In short, I think that quelling unfounded rumors (including fake news, conspiracy theories, etc.) is best accomplished educationally. Consider:
—High school students are capable of reasoning abstractly, but they're still young, and their views tend to be relatively malleable.
—By high school, students will already have been exposed to instruction that fosters critical thinking, as well as other kinds of skills that promote resistance to unfounded rumors. (For example, the curricular standards in Texas, as in many other states, require that students be introduced to the rudiments of media literacy beginning in elementary school.) In short, there’s a pedagogical foundation to work with – one that evidently has room for growth.
A concrete example
Here I want to suggest that a high school U.S. Government teacher could easily develop a public health unit, with a focus on COVID-19, that meets many of the curricular requirements for that class and would serve to ward off a variety of unsubstantiated rumors. (This is just an example, by the way. I'm not claiming it could be used to improve on any lesson plan teachers currently use.)
Following are excerpts that I've cherry-picked from some of the Texas curricular standards, but any state standards would do. To make things simple, I've added my own numbering system so I can refer back to them later.
In a high school U.S. Government class, students are required to...
"...identify the purpose of selected independent executive agencies, including...the FDA" (Standard 1)
"...evaluate government data using charts, tables, graphs, and maps" (Standard 2)
"...describe the potential impact of recent scientific discoveries and technological innovations on government policy" (Standard 3)
"...evaluate the validity of information, arguments, and counterarguments...for bias" (Standard 4)
"...describe the factors that influence an individual's political attitudes and actions" (Standard 5)
Here's a simple way these five requirements could be combined in a unit on public health for a U.S. Government class, using an integrated curriculum approach (i.e., one in which material from other classes is incorporated):
Part 1:
Following a suitable introductory exercise, the teacher explains the purpose of the CDC (Standard 1). The teacher then logs into a CDC website, where COVID-19 data are presented in accessible form (Standard 2). Here students can explore and, with the teacher's guidance, interpret stats on infection rates as well as side effects. Next, the teacher discusses vaccine research which led the CDC and other government agencies to recommend vaccination (Standard 3). The teacher now describes the FDA's emergency authorization process (Standards 1 and 3). Finally, students revisit the CDC data and reflect on differences in infection rates according to vaccination status, as well as the incidence of adverse side effects (Standard 2).
Part 2:
With assistance from the students, the teacher assembles and shares with the class a variety of quotations on vaccination effectiveness and side effects excerpted from news stories, social media posts, and statements from elected officials. The teacher ensures a mix of accurate reportage and unfounded rumors. Students evaluate the validity of the quoted statements by systematically comparing them to the CDC data they examined in Part 1 (Standard 4). Students then reflect on who would be exposed to the quoted material, and how it might shape their attitudes and actions (Standard 5).
In my opinion, exercises like this, integrated into the high school curriculum, would promote knowledge of government, while increasing resistance to unsubstantiated rumors of any sort, by teaching students, among other things, about the importance of data sourcing as well as the proper interpretation of simple statistical data.
Appendix A: Is Good Sciencing any good?
The Good Sciencing website describes itself as run by "a small team of investigators, news editors, journalists, and truth seekers, now backed up by others, who are discovering pieces of information that we can investigate. It doesn’t really matter who we are....We’re doing this anonymously because we’ve seen people viciously attacked and threatened for doing things like this."
Even before looking at the website's content, that passage raises a bunch of red flags. (Anyone can call themselves "investigators." We're all "truth seekers". And who are those "others"? It really does matter who you are. Etc.)
The website itself cites a lot of scientific studies, but after 20 minutes on the site, I was unable to identify a single study that wasn't misrepresented in some way. In essence, the entire website is an assemblage of half-truths and lies (some of which have been repeated by U.S. Senators).
For instance, one section is entitled "47 studies confirm ineffectiveness of masks for COVID...", but that's not what any of the studies show. Here's a typical example: All that's stated on the website about one of the studies is a quote from the study itself: "We included three trials, involving a total of 2106 participants. There was no statistically significant difference in infection rates between the masked and unmasked group in any of the trials”. What the study actually showed is no differences in surgical infection rates (not COVID-19 infection rates) for patients being operated on by teams who were masked vs. unmasked. These were clean surgeries, and infection rates were low (about 4% overall). In short, the findings are clearly irrelevant to how well masks prevent the spread of COVID-19.
Appendix B: VAERS
The Vaccine Adverse Event Reporting System (VAERS) is a database maintained by the U.S. Department of Health and Human Services. Anyone – medical professionals, patients, families, etc. – can submit a report to this database. As HHS cautions repeatedly on the VAERS website: "A report to VAERS generally does not prove that the identified vaccine(s) caused the adverse event described. It only confirms that the reported event occurred sometime after vaccine was given. No proof that the event was caused by the vaccine is required in order for VAERS to accept the report. VAERS accepts all reports without judging whether the event was caused by the vaccine."