Thread by @ProfDFrancis

This is the funniest thing on earth.

Now we know why the finance industry are in such a mess. The people who charge you money to lose your money, are busily doing it.

Moronic commentary too. Let me explain.

https://twitter.com/CNBCnow/status/1250890382516379652

A starter question.

How is this relevant to the utility of rEmDe (lets call it by half a name until the company provides the missing half of the trial)

Now I know why Syed Gardezi was laughing at me in the hospital this morning.

"Hey Prof, the Remdesivir people got your message about not having an unblinded control arm!"

"Oh?" I said, flattered that I was so powerful and impactful. "I'm sure it wasn't just me. That other guy, John Mandrola, probably contributed 1 or 2%. But of course mainly it was me."

Syed walked away laughing. "Yes, definitely no unblinded control arm any more".

Incidentally when not making me make a fool of myself, Syed's other talent is advanced heart failure.

He has done a BJCA Cardiology Fellow Online Tutorial on heart failure, available shortly. (If it becomes available during this thread, can someone Whatsapp me the link?)

Only when I came home did I realise that the excitement was about a patient group that didn't have an unblinded control group, but that was because they didn't have any control group at all.

From my insider sources (google.com) I have been able to piece together the following.

A research doctor was talking to fellow researchers about experiences of case series of patients who had received rEmDe.

It is useful to know their heads don't fall off, nor do they grow an extra leg.

But only slightly useful.

In that conversation she said this:

Someone recorded the conversation, which was intended for intelligent people.

And then put it in the hands of stupid people.

And these stupid people utterly botched the interpretation of what was said and passed it on to even more stupid people.

Like an Amway corporation of idiocy.

Any Cardiology Registrar would not be so foolish as to fall for this nonsense.

Because they make themselves intelligent through education. For example Syed Gardezi and the BJCA, whose HF tutorial is now on line at

bjca.tv/aiovg_videos/a…

Let's test Dr Mullane for intepretative skill.

"Most our patients have already been discharged"

Is this "good news"?
Is this "great"?

You miserable, miserable bastards.

Of course is good news!

Who the hell can be unhappy that patients get better and go home?!?!

What are you, coffin manufacturers?

Mr Jones? He's gone home

Mrs Shah? She's gone home

Mr Lin? He's gone home too.

That's the best thing EVER.

For those patients.

A thought experiment.

If only 2% of patients die after taking a drug, what does that tell you (leaving aside confidence intervals, stats, etc)

This was a DOCTOR treating patients, reporting that almost everyone had got better and gone home, and only two had died.

This is really good news for those patients and she was happy to tell intelligent, educated healthcare workers this great news.

She did NOT say the drug works.

It was not good news for the drug.

It was good, great, wonderful news - for the patients who got better.

Just like my followers in Francis-ism, who raided pharmacies worldwide a couple of days ago, and each started themselves on whatever drug took their fancy.

Only 2% of *them* died.

The rest are doing just fine.

Doesn't make it a good idea.

I would expect better of Matthew Herper

Here is the story he and Adam Feuerstein wrote:

statnews.com/2020/04/16/ear…

In a way it is clever.

Here is what it says. I have deleted an obviously untrue statement, if it related to StatNews!

Oh dear.

Agree?

This is my one criticism of our colleague.

It is not hard.

One arm trial = not HARD, but IMPOSSIBLE to deduce effectiveness.

Here is what I mean

if x is 9406702394687230497534

what is x squared?

And now this

if x is 23

what is y squared?

However

(1) She was speaking to intelligent people who knew that hard was an abbreviation for "impossible"

Just like I call @LCC_DrMalik "You twit" as a shorthand for "Your imperial majesty and my illustrious colleague"

(2) Everything else she said was spot-on-perfect, and there is NO excuse for anyone to misunderstand it.

This proves that rule whose name I have forgotten.

You know, where the more stupid you are, the easier you think things are?

Now, clash of the titans.

Which is more intelligent?

In the blue-checked corner,
Contestant number 1.

Let's hear it for Eric Topol!
Rah rah rah!

And in the beige corner,
Contestant number 2.

This brick.

Which is brainier?

Well, the brick spoke on behalf of University of Chicago School of Medicine.

It was clear, courteous and accurate, as you might have expected of a brick from such an elite institution.

Big Eric, mmmmmmm, not so hot, in the analysis department.

Maybe I am being harsh?

OMG how embarrassing.

I have trolled myself.
Sorry, I mistook @brianSkorney for a thicko.

I am not used to being out-wit-ted.

Poor old Matt Herper and Adam Feuerstein.

Right now @ethanjweiss let me in to their press conference about their story.

They are absolutely failing to understand that they can NOT deduce whether the drug has ANY benefits.

They are saying "Gilead should not be surprised by our article. We checked our sources and figures etc etc."

Actually Gilead would be very surprised.

They probably thought, like me, that these guys were sharp.

Poor guys. I better delete this thread before I they get off their call and get really angry with me.

My respect for them has gone up massively having joined their live video chat

Despite me asking them nasty questions and trolling them while they tried to answer, they managed to keep a straight face throughout.

And look,

They are now saying "It is really hard to use synthetic (historical) controls. Randomized controlled trials are the greatest discovery of medical science."

Spot on!

So they do know that the one-armed trial is completely meaningless.

OK, so the onward silliness must have come from other people.

One more point to avoid confusion.

The moronic commentary was the TV commentary, not the journalists'.

The journalists were simply wrong in their deductions.

The TV people managed to add another layer of confusion to make it moronic. But they aren't scientists.

This is how bad stuff happens.

Thank you for Dr Bansilal for prodding me about something.

This "SEVERE COVID" trial being discussed.

What percentage of people were ventilated at time of randomization?

Meanwhile I personally don't blame the company for increasing the sample size

Effect sizes on EVENT endpoints take thousands of patients, if you want to be able to detect even (say) a reduction in mortality from 25 to 20%.

So em-biggening the trial gets a thumbs up from me.

(Should still add a bag of normal saline in the control arm, though, guys! There's still time. Until the last patient is randomized, there's still time.)

In the above quiz,

The first 10 votes, 6 out of 10 got it right.

So, not that severe, after all.

So maybe not that astonishing that almost everyone got home safe and sound.

*That* is why we have a control arm (if outcome is not death or a blood test, a *placebo* control arm).

To stop ourselves getting nasty surprises that the wonderful outcome on the drug, would be just as wonderful without it!

Good night, and please buy shares in Gilead*.

(* Sellers need someone to sell to)

I am glad I don't have @SameerBansilal as my enemy.

He has dug up the definition of "Severe" in the severe trial.

Oh dear...

I understand this to be the correct answer, by the way.

Mo Elshazly @mbelshazly, I am so very glad you asked this!

I was wondering that but *I* did not want to admit I found it confusing.

Let me tell you a story.

... when I was an intern, there were 3 types of cardiac arrest

1. VF/pulseless VT
2. EMD
3. Asystole

Nowadays, when I mention this, people look at me like I am insane.

"You can't just make up a third type of cardiac arrest! There are only two types, shockable and non-shockable."

"Wasn't there a third one in there somewhere?"

"No, don't be ridiculous."

So when people report that a phase 3 trial has only one arm, I want to leap out and say "No, it's gotta have two arms. One arm you give it, and the other you don't".

But I hesitate. It might be like cardiac arrests.
Stuff might have happened that I hadn't kept tabs on.

But I am glad someone else agrees with me that there seem to be some "phase inflation" going on.

If a one armed-trial can be a phase 3 trial, anything can be a phase 3 trial, surely?

As far as I recall,

phase ONE = injecting it into a few medical students, to see if they die or anything

phase TWO = giving it to patients, and seeing if some TESTS get better (not death or hospitalisation, but EF or BNP or something)

phase THREE = actually finding out if it has a net benefit on patients, in some endpoint that can get the drug a licence. (e.g. death/hosp, or chest pain, or treadmill time etc)

Which obviously means 2 arms!

And obviously placebo if possible.

Only excuse for not having placebo is outcome cannot be influenced by any other intervention that clinicians carry out (e.g. death, when there is no intervention thought to affect death).

Wikipedia says this. Admittedly it is not God, but it is fairly clear and if there was no need for there to be two arms, someone like me would have tweaked it.

The only explanation I can think of is that their term III is a mathematical formula.

Where I = the cube root of 2

So III = I * I * I = 2 ^(3/3) = 2

Therefore phase III trial would be a synonym for phase 2.

That must be it.