Harlan Krumholz

I've seen trying to educate people through the Internet to see if that helps.

And there was a publication about five years ago in The Lancet in trying to use strategies to educate the public to do these features.

But generally, we only think about them now.

And in about a few months' time, if this has gone away, it will go in the basket again and disappear, won't it?

And it will be put away.

And so there tends to be poor quality evidence informed by case control and cohort studies and a complete lack of randomised trials.

Look, that was really interesting.

The first thing is to say Peter Opshaw said experience of use of steroids in SARS.

But it's interesting, the Lancet produced what I consider is a really important piece of evidence, a sort of rapid review, you like, of the evidence that does not support corticosteroid treatment.

And in there, there was one.

One study that was about MERS.

There were four studies I identified on SARS.

They also cited the influenza and increased mortality.

That's in a systematic review and a trial on RSV in treatment.

And basically across the board, steroids is a bad idea.

And I think this is a really useful piece of evidence.

That allows you to look at it and think, not just experience, this is an evidence base right now.

The current best available evidence says don't use steroids in coronavirus unless it's for another indication, a clear indication, such as asthma or one of the other reasons to use it.

Well, look, this is a really good example of evidence-based medicine.

In the steroids, there's clinical experience that might say to use steroids.