For this example I'm going to use systolic blood pressure but it can be anything that was highlighted above by @ProfDFrancis. In the table above, the parentheses mean they found an interquartile range for systolic BP of 125-133 - or 50/100 of their subjects fell in this range. 
This is a problem. In a much larger study across the German population the IQR is much larger. For example, if we take the 41-50 age range from this larger study (similar range to the JAMA study) they found a mean of 130.5 and a standard deviation of 16.8.
Assuming a normal distribution curve (which I think is a safe assumption) you have about a 19% chance of finding someone in Germany with a systolic blood pressure within this range.
Again, going back to the JAMA study they said they found roughly 50 people out of 100 that were within the 125 - 133 mmHg range for systolic blood pressure. But based on the larger German study, you would expect to only find 19 people out of 100. So what are the chances of 50?
Well…assuming a Poisson Distribution (again I think a fair assumption) this is what the curve looks like - this means while finding around 19 people is the most likely, there are other feasible outcomes as well. But where is 50? It doesn't even warrant a place on the x axis.
Again, the JAMA COVID cardiology study is full of statistics like this - they are either wrong, a result of selective bias, or worse, completely made up. I will gladly admit fault if any of my numbers are wrong here - but they follow pretty basic statistical theory.
Shamelessly tagging @ClayTravis , @kylamb8, @justin_hart, @AlexBerenson, @kerpen to hopefully read/RT this thread to bring it to light - and have smarter people than me check my work. But if it's right - they need to retract this study immediately.
