Discussion about this post

User's avatar
Paul Fischer's avatar

Yes yes so what. I read this study years ago. There are many such studies. The authors, once again, statistical illerates. Here's what I wrote about it when it came out:

This JAMA paper may look impressive because of its size, but the flaws are obvious once you scratch the surface. The population is almost entirely older men with heavy comorbidities, so the results don’t generalize. The authors treat vaccine choice as if it were randomized, but it was actually determined by clinic availability and preference. That means selection bias is baked in, and not fixable with the statistical matching they tried.

The outcomes are equally shaky. They rely entirely on billing codes in administrative records to define “adverse events,” which is notoriously unreliable for clinical detail. Their censoring scheme, cutting off follow-up once infection occurs, assumes infections were equally detected across groups, which is unlikely. And of course, statistical illiteracy rears its head with the sheer number of outcomes tested without correcting for multiple comparisons, and some of their “findings” are almost guaranteed to be statistical noise. What they actually show is that older sick men have lots of health events, not that one vaccine is clearly safer than the other.

Expand full comment
Antoinette Fay Ryser's avatar

I’m confused. Why would vaccinated patients be generally healthier? Is the implication that people who care about their health are more likely to get vaccinated? That certainly doesn’t apply to me. I refuse vaccination because i care about my health.

Expand full comment
110 more comments...

No posts

Ready for more?