Misinformation Goes Viral

Misinterpretation of a flawed study on the influenza vaccine provides raw meat to anti-vaxxers.

A recent preprint out of the Cleveland clinic seems to show that the influenza vaccine increases your risk of influenza. A closer look reveals a fatally flawed study, which, if anything, suggests exactly the opposite.

To summarize: 

1. They compare two groups of employees of the Cleveland Clinic, those who received the influenza vaccine (82%) and those who sought an exemption (18%). 

2. As hospital employees, they are aware of the extent to which their work puts them at risk of exposure. It is highly likely that those who sought an exemption from the vaccine were at lower risk of exposure. The investigators fail to even acknowledge this possibility and make no effort to determine differences between the vaccinated and unvaccinated beyond very crude categorizations such as the facility in which they worked . 

3. After 100 days, the authors found higher influenza rates in the vaccinated. By far the best explanation for these results is that study is biased in exactly the manner I just described. Read on.

4. They provide no plausible explanation as to how the inactivated vaccine puts one at increased risk of influenza 100 days after vaccination. In fact, they try to sidestep the issue entirely in their conclusion by suggesting their only finding is that vaccine was not effective.

5. This is a trivalent, inactivated virus. There are only two ways it could increase your risk of influenza. It could impair your immune system, but there is no evidence of that ever. Anywhere. Zero. The only other way it would make you more likely to get the flu is if the virus were improperly inactivated, which would cause flu within days of vaccination, not months later.

6. That means the ONLY plausible explanation for a significantly higher risk of influenza in the vaccinated is a significantly higher exposure risk in the vaccinated. Ergo, the sample is biased. 

7. It is notable that the infection rate for influenza among the vaccinated was only 2.5% in a high risk setting for infection. There have been many studies done that showed average reductions in illness of about 50%.

8. The difference is most likely to be a reflection of much lower risk of exposure to the virus among the unvaccinated. In other words, those who knew they were at high risk were much less likely to seek an exemption.

9. In sum, the best explanation for their results is that their sample was biased. If the vaccine was 50% effective, their findings would be consistent with those seeking exemptions from the vaccine having half the risk of exposure of those opting to receive it.

Unfortunately, this possibility was lost on the usual suspects spewing the usual nonsense.

Berenson had no less than six posts about this, one of which got over 600,000 views. But X algorithms may never show an explanation of the study to people who want to believe the Alex Berensons of the world. Even if they did, understanding the fatal flaw requires far more explanation and reader attentiveness than the headline grabbing conclusions of the authors, particularly in a character limited world. Meanwhile, this deadly misinformation virus is out in the wild and will cause disease (and even death) as surely as an actual virus.

Shame on these authors for their cavalier sloppiness (or intentional misinterpretation of the data) and shame on their amplifiers. What this study does best of all is demonstrate why we do randomized, controlled clinical trials, something the anti-vaxers like Berenson love to crow about but fail to understand.