Another day, another CDC myocarditis study

Ah, the Center for Disease Control is at it again with another study meant to justify vaccinations in children. They continue to use the same old tricks, too. I will jump right to the conclusion as it is important to note that this study is completely irrelevant as it was done almost exclusively in a period where now extinct variants were dominant, so when the authors conclude “These findings support continued use of mRNA COVID-19 vaccines among all eligible persons aged ≥5 years”, they are merely selling vaccines, not following any semblance of science. But even if the variants that dominated 2021 were not extinct, their conclusion is still silly upon closer examination.

Study setup:

The study looks at data from 40 US Health Care facilities from January 1, 2021 to January 31, 2022. Health outcomes of 15,215,178 persons ≥5 years old were observed¹. The authors used five cohorts within the study population: “infected” (814,524 infections), “single dose mRNA” (2,548,334 doses), “two doses mRNA” (2,483,597 doses), “unspecified dose mRNA” (1,681,169 doses), and “any dose mRNA” (6,713,100 doses). They examined three outcomes myocarditis and pericarditis (with 7 and 21 day windows) and MIS (7, 21, and 42 day windows). Individuals that had an infection/vaccination 30 days prior or following a vaccination/infection were excluded.

Unvaccinated inclusions

On first glance, one might conclude that there are no unvaccinated individuals included in the study. That does not appear to be true as there are no exclusions from the infection cohort based on vaccination status except for the aforementioned exclusions. In other words, some of the individuals may have been infected but not vaccinated.

While excluding the unvaccinated as a separate cohort is strange as we lose any sense of a base rate in the population, the real eyebrow raiser is combining both the vaccinated and unvaccinated in the infection cohort. The choice to have separate cohorts for both groups is so obvious and natural that one has to wonder what is being hidden in the data by not separating them. Much could be learned by seeing if the rates differ among the vaccinated and unvaccinated upon infection, and the data was clearly available. It would take all of two minutes and a few limits of code in order to be included in the study.

Vaccine enthusiasts would, of course, also like to have seen these cohorts broken down by vaccination status. After all, we have heard that the vaccine is supposed to help prevent “severe COVID”. In my opinion, the proof for that has been rather weak. Even in the randomized control trials, the authors were careful in their wording on that point as there simply was not a large enough sample size of severe outcomes to draw any such conclusion. There needs to be a very high degree of certainty for making such a claim that, due to differences in vaccinated and unvaccinated populations, has not been satisfied in any real way. Even in the midst of, say, the delta waves the vaccinated and unvaccinated who were hospitalized were dying in similar proportions, and, in fact, the prevention of severe outcomes could only be said to come from prevention of cases if anything.

If, however, this study found no decrease or even an increase in myocarditis among the vaccinated infected, it would be a bad look for the vaccines. The narrative would not collapse, but it would be weakened. People might actually have to look up the definition of severe COVID, which revolves around oxygen saturation in the blood and has never claimed to involve outcomes like myocarditis. I am sure the authors, and the CDC, would like to avoid people catching on.

Where have all the boosted gone?

Unspecified dose, believe it or not, does not include the boosted. It includes only unspecified first or second doses. Excluding booster doses is an interesting omission considering this study likely serves the purpose of convincing people that the risks of the virus outweigh the costs of the vaccine, and thus, they should continue to get dosed. But there are reasons to believe that more doses lead to worse outcomes, and I suspect the booster may be worse than we realize.

I am willing, however, to give the authors the benefit of the doubt. Perhaps they were more interested in younger populations and the number of people boosted in those populations was too low for meaningful results. The authors never sufficiently articulated that point. But, hey, why not assume the best?

Windows

The authors included, really, four separate windows in their analysis. They dropped those infected and vaccinated, or vaccinated and infected, within 30 days from the study. In a way, that seems like a reasonable exclusion, but the choice of 30 days seems arbitrary. Stranger still, it does not even align with the windows they chose for myocarditis and pericarditis or even the MIS window. No justification was given for this choice at all. I would have at least liked to see what happened to people in those days for illustrative purposes.

The choices of windows for myocarditis and pericarditis are equally strange. No justification was given for, say, 21 days as a window, and justification is needed. Adverse events are not limited to 21 days, and yes, one might say that there needs to be a cutoff somewhere. But the logical way to justify a cutoff date would be to look at whether the rate of these complications is increasing, decreasing, or leveling off. Ideally, we would want to see the cutoff point being after cases of myocarditis or pericarditis have leveled off in this population to reach some sort of base rate. If rates are increasing on day 22, that may signal that either adverse events take time to kick in with the vaccine or take time to be diagnosed.

The authors seem to understand that it can take time to diagnose certain issues because they arbitrarily choose a 42 window for MIS, allowing time for a late diagnosis — as if myocarditis and pericarditis are always caught quickly, but MIS is not.

In fact, the inclusion of MIS in general is a red herring as they do not calculate the rates of MIS after vaccination, only infection. They do this while noting that “MIS often occurs in the absence of prior positive SARS-CoV-2 test results; these cases were not captured in the infection cohorts”. Yikes. What is the reason for that? Why did they note explore whether MIS is caused by the vaccine? How many cases is “often”?

42 day window

Further to the earlier point about adverse events should be leveling off when choosing a window, the authors do calculate the 42 day risk-window for myocarditis or pericarditis. Hidden in a footnote is this:

The incidence of myocarditis or pericarditis in [the 42-day] risk window was 4.0, 37.1, 19.7, and 12.8 cases per 100,000 for males aged 5–11, 12–17, 18–29, and ≥30 years after a first dose of an mRNA COVID-19 vaccine; 4.7, 39.4, 16.8, and 12.7 cases per 100,000 after a second dose; 12.9, 33.4, 31.3, and 25.3 cases per 100,000 after an unspecified dose; and 6.5, 37.1, 22.0, and 15.8 cases per 100,000 after any dose.

We will briefly discuss additive probability later, but I want to point out, say, the 12-17 year old cohort. From a first dose, the risk per 100,000 is 37.1, and for the second dose that is 39.4. So the probability of a male 12-17 getting myocarditis from the vaccine would be 76.5 per 100,000 or about 1 in 1307.

(Though, this may be increased by individuals receiving their second dose in this time period)

Infections

The critical failure of the study is in underestimating the infected cohort. Testing in homes, schools, community sites, or pharmacies were not included. Furthermore, any infections without tests were not represented.

This is not a limitation of the study — it is an invalidation of the entire premise of comparing the rates between vaccine injuries and infection outcomes. Such failures are commonplace in these types of studies. The authors have no idea what the denominator on the infection cohort is, yet represent it as fact. No conclusion on cost benefits can be drawn from such a methodology. PERIOD.

I’ll add more on how the cost benefit analysis should look like later, but first…

Additive probability

In all of the charts posted by the authors, one can note that the rate of myocarditis and pericarditis is lower in the “any dose” category. The authors do admit that people are counted twice here, and thus, the rate in the any dose category is nonsensical.

People being vaccinated twice do not have one chance at an adverse event, but two, and from all accounts the probability at an adverse event after the second dose is higher. We need to be adding probabilities here, not averaging them. If, for example, a male in the 12-17 age group has a 3.3 in 100,000 rate at getting myocarditis or pericarditis after the first dose and a 35.9 in 100,000 rate after the second dose, then the rate would be 39.2 in 100,000, not 19.6 in 100,000.

This is, of course, not exact as the chances of having an adverse event are not likely to be mutually exclusive between doses (or AEs would not increase after the second), and people who had an adverse event on the first dose would probably not get the second dose..

How do we get to a cost benefit of vaccinations?

First of all, we need to remember, far more people get vaccinated than get infected. That is why comparing infected people to vaccinated people based on rates is insane. Either the chance of being infected in a given time period, say a year, is similar to what this study suggests, which is about 815,000 per 15.2 million people minus the excluded infections… which suggests the chances of actually being infected are low to begin with, or infections are much higher and the rates of myocarditis or pericarditis from infections are far lower (and thus, the risk from vaccination is far higher than the risk from infection).

The things we need to consider in a cost benefit analysis would be, roughly:

1. The chance of being infected over a given time period, which will vary for everyone based on risk taking behavior, location, et cetera.

2. Some function of vaccine efficacy. We would need to factor in decreasing efficacy over time, decreasing efficacy with variants, negative efficacy immediately after taking the vaccine, and the possibility of negative efficacy from OAS.

3. Vaccines efficacy against serious outcomes. That is an unknown and vaccines may make serious outcomes more likely with ADE scenarios.

4. Risks from vaccines including all subsequent doses required. For all intents and purposes, this risk seems to increase with each dose.

5. And risk of a severe outcome from the virus.

That is the very least of what is required. If this study captured all infections that occurred that year, the chance of being infected would be around 5.4% or slightly higher due to infections that did not make it into the analysis.

Let’s pretend two doses of the vaccine was sterilizing for the length of the study, so we do not need to worry about a complicated function as in (2), (3), or (4) (if subsequent doses are included). Even a staunch vaccine enthusiast should not complain about those assumptions because they clearly weigh heavily in the vaccines favor.

So … the vaccine had 100% efficacy and all the cases occurred in the unvaccinated. So an unvaccinated 12-17 year old male would have a 5.4% chance of being infected and, if infected, would have a 64.9 in 100,000 chance of myocarditis or pericarditis. Which, in the given year of the study, means their risk would be around 3.5 per 100,000. Two doses of the vaccine, on the other hand, has a risk of 39.2 per 100,000 in the first 21 days.

Do you believe the image below and get the vaccine? Or do you think for yourself and reject it?

1. Their data set included more than 15 million people, but their analysis did not include anywhere near this number. The authors claim 28% of individuals were included.