CDC's risk benefit slide is inaccurate
Malleable science will lead to our demise
Some types of science are more malleable than others. Randomized trials, if honestly reported, can be impacted by the inputs: control arms, post protocol care, concealment or blinding issues, and drop outs, but otherwise are solid.
Cost effectiveness studies are a different animal entirely. There are dozens of assumptions that go into them, and they can bend in any direction. That’s why we found, “Having a study funded by a pharmaceutical company was associated with higher odds of a study concluding that a drug was cost-effective than studies without funding (odds ratio, 41.36; 95% CI, 11.86-262.23).”
The CDC’s recent modeling exercise arguing that young people are better off with bivalent boosters is shown here. It is the most malleable design of all. And just like Pharma, the CDC is an untrustworthy analyst. They already made the policy. Now they need the analysis to validate them. Yet, sadly, their mistakes, lies and exaggerations having robbed them of credibility. Take a look at the Figure.
The first thing to say is that it is 0 myocarditis… so far. The booster will definitely have myocarditis. I bet it will close to 1 in 10k, the estimate Sharff had for the last booster. The CDC’s harms balance is dishonest.
The second thing to say is that the rate of hospitalization is incorrect. These rates assume the rates of Dec 2022 last forever— they won’t b/c natural immunity— and do not exclude incidental covid19 hospitalizations. Specifically:
That’s bad. Obviously this inflates the upside of vaccines. Elsewhere the authors provide a version of the slide that tries to correct for this, and halves the numbers, but the deeper problem is that the taxonomy of for vs with covid hospitalizations is underdeveloped, and even this correction is likely a distortion
Next, the authors assume the Vaccine Effectiveness from VISION study, which is a test neg case control. But using this design with low uptake— BTW no one wants this bivalent booster anyway— creates volatile estimates. The VE estimate is basically made up as well. It assumes equal presentation patterns for illness, which is likely untrue. A randomized trial of 100k kids would give accurate estimates.
(Side note: Ironically, the ridiculously large benefit the authors postulate means that the trial is very doable)
So what is wrong with the CDC’s analysis?
The harms side of ledger is inaccurate
The benefits side is overstated
The vaccine effectiveness is a volatile estimate
Basically everything is wrong, and the conclusion is the absolutely most optimistic scenario, which SURPRISE, SURPRISE confirms the political decision their boss already made.
Granting EUA to bivalent boosters for most adults was a huge mistake. It was an abdication of FDA’s duty and the CDC has failed Americans to push it. Without precise estimates of VE (which come from RCTs), honest projections of future hospitalizations, and honest tallies of myocarditis— the CDC continues to push dishonest propaganda.
Trust in public health has been shredded. Joe Rogan didn’t do it.