The Big 10 report on COVID relies heavily on this paper which found rampant abnormalities among normal controls and had many statistic that make no sense.

Time to retract or correct this paper. https://twitter.com/ProfDFrancis/status/1294370975702822914

TBH I'm not a football guy so whether they put off a year doesn't really affect me, but the Big 10 is a majorly respected organization and many Americans get their news from @espn.

Unfortunately much of the messaging centers around a flawed paper.

More information about statistical aberrations here: https://twitter.com/ProfDFrancis/status/1288246926392070145?s=20

TBH the statistical issues are concerning but equally troubling in my mind is the high prevalence of abnormalities in the "control" arm.

Controls are supposed to be people who don't have known cardiac problems.

Nonetheless, 15% of the risk factor matched controls have pericardial effusions (fluid around the heart) in an amount that would be considered moderate or potentially severely abnormal.

Everyone has a little bit of fluid around the heart, we believe it helps lubricate the heart motion among other things.

Sometimes we can just barely see that fluid on scans like MRIs. We call that a trace effusion.

When it is a bit more than trace, we call it a mild pericardial effusion.

Now occasionally a normal person can have a mild pericardial effusion, but it is not textbook normal.

>10 mm defines a moderate pericardial effusion. So in this study 15% of the so called normal patients had effusions that would be moderate or even severely abnormal.

Likewise, 15% of the normal control subjects have enhancement of the pericardium.

We believe this indicates inflammation of the pericardium.

Also, about 1/6 of the normal controls have LGE or scar in their heart muscle.

I suppose this is possible, but seems like quite a lot for normal people to me.

In the end, I'm not sure whether college football should be played or not.

To me there are more substantial and important things in the world.

That said, papers like this should not be used for policy decision.

At least until they are fixed/corrected.

The idea of using highly sensitive tests like MRI in elite athletes can be problematic.

This is a nice review article of the types of abnormalities previously described in athletes. https://onlinelibrary.wiley.com/doi/full/10.1002/clc.23360

Also, extreme exercise like marathon running can elevate some of the very same blood tests which are used to diagnose myocarditis. These abnormalities can persist for a few days.

https://journals.lww.com/acsm-msse/Fulltext/2011/10000/72_h_Kinetics_of_High_Sensitive_Troponin_T_and.2.aspx

Of note, both these same MRI abnormalities and blood abnormalities can look very similar to those of myocarditis!

In this thread @ProfDFrancis makes it clear - the JAMA Cardiology COVID MRI paper is not worthy of basing anything much on. https://twitter.com/ProfDFrancis/status/1294981630852050944?s=19