Or people who chose to get vaccinated are generally more cautious and think more about their health and may be even more careful drivers etc., therefore they would die less often from traffic accidents compared to irresponsible drunk drivers who didn't bother to get vaccine.
It is so hard to distinguish cause from effect in these studies.
They tried to control for that by choosing a control group that got a different vaccination: “The team studied 6.4 million people who had been vaccinated against Covid-19 and compared them to 4.6 million people who had received flu shots in recent years but who had not been vaccinated against coronavirus. They filtered out anyone who had died from Covid-19 or after a recent positive coronavirus test.”
Well, of course! Flu shots were never politicized to the stupid degree covid vaccines are politicized, so the covid-vaxxer(?) group will show a clear political leaning that's different from flu-vaxxers. One can expect all sort of fun correlations.
(I assume nobody is crazy enough to think that a covid vaccine will statistically protect them from, say, a drunk driver.)
This elucidates one potentially strong source of bias in vaccination studies. The first to line up for voluntary vaccination are likely healthier and higher in socioeconomic status.
You perform an observational study, not adjusted for "baseline health status, underlying conditions, health care utilization, and socioeconomic status". You get a 3.2x difference in mortality. Still, you decide to publish it. Great job!
10 comments
[ 2.6 ms ] story [ 30.2 ms ] threadIt is so hard to distinguish cause from effect in these studies.
(I assume nobody is crazy enough to think that a covid vaccine will statistically protect them from, say, a drunk driver.)