Thread by @MaxKennerly

Gah. Well, congrats to the headline-seeking researchers behind that VAERS / myocarditis study, they found the audience they were looking for!

But let's talk for a second about the study.

/1

VEARS is cool, just like FAERS is cool. It's a simple system for collecting adverse event reports from anywhere. Typically, these are used to flag events with disproportionate reporting frequency, which can then be analyzed with other methods that have validated medical data.

/2

I will grant the authors of the paper ( medrxiv.org/content/10.110… ) one thing, they set up a link to make it easy to comb through the VAERS reports included in their analysis: bit.ly/CAEmRNA

So let's take a look at some of them.
/3

The core problem the authors claim to have identified involves boys 12-17 who were hospitalized with myocarditis, pericarditis, or myopericarditis. Every single VAERS report I discuss below is a boy 12-17 who was hospitalized and was included in the article's analysis.
/4

VAERS ID 1328253-1. Notice what little information it provides. I've highlighted how, despite two days in the hospital, they never got any information about the vaccination, all of which would've been on the vaccine record card. (Check your own, you'll see.) That's odd.
/5

VAERS ID 1336480-1. It does not mention myocarditis, pericarditis, or myopericarditis. It does not include lab values or MRI results. It just says "elevated troponins." Read @han_francis, a pediatric cardiologist, for why that's not enough:
/6

VAERS ID 1347549-1. Missing vaccination info. Conflicting info on whether hospitalized, or just seen in emergency room. No diagnosis of myocarditis, pericarditis, or myopericarditis. No MRI.
/7

VAERS ID 1351002-1. Missing vaccination info. Conflicting info on hospitalization. No diagnosis of myocarditis, pericarditis, or myopericarditis. No MRI.
/8

The study had 202 hospitalized boys age 12-17 total. Those four I discussed above are 2% of that entire dataset.

They're hardly alone. VAERS 1360831-1 says "myocarditis," but not much else. Doesn't even say which dose of the vaccine it was. No MRI.
/9

Even the comparatively good reports include gaps and tantalizing missing information. VAERS ID 1377873-1 is missing what state it was in. But the bigger problem is the pending MRI. We don't know what they were eventually diagnosed with, nor their outcome.
/10

I didn't review all of the reports, these are from a brief skim of the first three dozen they listed. Point is, VAERS is useful, but it can't do what the authors claim, there's just not enough reliable data to derive incidence rates, much less do a harm-benefit analysis.

/end