Prof Darrel Francis ☺ Mk CardioFellows Great Again Profile picture
Cardiologist, Scientist. I separate taking my work seriously (I do) vs myself seriously (I don't) Ideas my own (best ones stolen from my amazing PhD students)
Anil Gopinath Profile picture UnChatUnChat Profile picture Jean-Luc Szpakowski Profile picture Steve Granier Profile picture Rui 1less🚘 10more🫀 Profile picture 17 added to My Authors
Jun 9 11 tweets 3 min read
How exciting !
I might have an Omicron !
Mar 28 8 tweets 3 min read
The pandemic has made it really clear to me that trend to have patients make their own health care decisions with advice from us, is well intentioned but harmful.

A great many people *don't realise* they have no idea how to decide wisely. This unfortunate citizen thinks that this graph is what they should use to decide whether to have a vaccine.

the doctor has a choice of explaining about RCTs versus irrelevant 3d colour graphs, telescoping into a few minutes what it took years to grasp, or just sigh and move on.
Mar 16 10 tweets 2 min read
Which is more? Which is a bigger discount?
Jan 11 24 tweets 4 min read
Businesses are financially incentivising vaccination by reducing sick pay for people insisting on not participating in community vaccination. Some people advocate personal freedom with no responsibility to the community which makes it possible to live our lives so freely.

I personally think I drive best when drunk as a skunk.

Why should anyone tell me what I do with my body?
Jan 9 54 tweets 12 min read
Lots of people seem to be tweeting at me super-confidently that "Vaccines don't work".

Funnily enough, none of the people who go through any of my courses in clinical research stats are so overconfident in their interpretation. Image Mop up any seminars on probability and risk that you haven't already done,

and if you've done them all, repeat the low-scoring ones aiming for the "gold" 100% status!

Jan 9 4 tweets 1 min read
Great question This is not a question of taking a statin or not, where the benefit is entirely for themselves.

People should have a free choice, just as they have a free choice to inject morphine, smoke crack cocaine, drive while drunk as long as they don't collide with anyone.
Dec 27, 2021 7 tweets 4 min read
Good morning UK!

I see Elsevier and Springer have finally done something to show the magnitude, and direction, of interest in science.

They thank you for all the funding you have provided them. and are now blocked for UK users. The difference between the music pirate sites, and the science distribution site sci-hub, is that the AUTHORS of the music do NOT want their music to be accessible for free, whereas the authors of the scientific papers DO.

How have we allowed ourselves into this mess?
Dec 25, 2021 8 tweets 3 min read
Merry Christmas!

Who can give the shortest, clearest explanation for why angle A + angle B must equal angle C?

(From the Master of puzzles, Martin Gardner) Image The winner is @leoshmu Leo Shmuylovich, who is entitled to 50% off the $24,995 annual subscription to! (invite code: stats )

Step 1.

Imagine reflecting the "B" diagonal upwards. What would happen if you connected the free end of it to the bottom left? Image
Dec 20, 2021 14 tweets 4 min read
A beautiful example of the advantages of conventional and log scales.

Here is the conventional scale data for London.

Clearly the cases in late 2021 produce
- much less hospitalization
- very much less deaths
then the cases in 2020. (Fingers crossed that this is due to Omicron being more nose-centric than lung-centric; or vaxxed people becoming cases but not super-ill)

So on the right, is the blue "death" curve really a scaled-down version of the red curve? Or just a random wiggly line that is just smaller?
Dec 19, 2021 31 tweets 7 min read
Thanks to @adnanalkhouli for alerting us to this elegant study from the ingenious team at Aarhus!

I don't think the central illustration is because it causes people to focus on the wrong thing.

But I love the study so want to do a micro-thread! Image What patients with HCM complain to their doctors about is:
Dec 5, 2021 17 tweets 4 min read
In a patient presenting with myocardial infarction,

PCI is In patients presenting with unstable angina due to coronary disease,

Revascularisation is
Nov 7, 2021 5 tweets 2 min read
This is awesome on so many levels.

Can you count the levels? Hat tip to @DrAhranArnold for grabbing this before the telegraph deleted it in shame.

1. What is real world data? What are the alternative worlds where data come from? Telegraphia?
Nov 7, 2021 7 tweets 2 min read
Yesterday, I tweeted that specialty societies rarely say "don't do our technique, in situation X".

Today @CardiacJoshi found a counterexample: In between my tweet and Joshi's, I was speaking to David Oxborough, R&D lead of Brit Soc Echo, about a UK-wide Echo AI research programme we are planning.

(I obviously didn't mention my little faux-pas of being rude about societies. I hope he is not on Twitter.)
Nov 3, 2021 5 tweets 2 min read
This is why I could never stand for election for the society of X or y technique. And even if I was crazy enough to do so, I would certainly never be elected !

You must publically advocate that the technique is wonderful and amazing.

If I was president of the E wave society... I would have to consistently argue,

"Measuring the E wave is the fundamental axiom of cardiology."

"E waves contain more information than the entire rest of the echo!"
Nov 2, 2021 8 tweets 2 min read
I have combined Guy gadboits and Voids answers to come up with a pictorial solution. You need the two breaks to be on opposite sides of the stick (since if they are on the same side, one piece will be greater than 1/2, so the other pieces can't match its length).
Probability 1/2
Oct 23, 2021 7 tweets 2 min read
This Sam Loyd question is generating lots of very nice equation-based answers, but let me try a method, derived from Zugzwang et al, to see if I can do it without so much algebra, just by thinking. Suppose we CHANGE the question to this:

"The hour hand and minute hand are SUPERIMPOSED on each other. How long till this superimposition next happens?"

Here is my approach. Suppose it is now just-after-two o'clock. When will this next happen?
Sep 29, 2021 18 tweets 4 min read
Finding the stats tweetorials hard?

Some easier questions.
Examine the quote tweets on this tweet, and answer the questions that follow.

Here is the claim:
Sep 14, 2021 56 tweets 11 min read
When your patient adverse symptoms on statins, what should you do? Here are 4 choices
Sep 11, 2021 11 tweets 2 min read
Everyone who dies of Covid today, got it from Suppose on average it takes 10 days from a person being infected, to that person passing it on to the next person in the chain.

Let's say we are about 500 days into the pandemic.

If I get it today, how many people long is that chain that led to me?
Sep 10, 2021 8 tweets 3 min read
This is very funny.

Does anyone know if this person is actually a doctor?

Or has somehow just managed to get through medical school without noticing how science works?

If someone makes an allegation, one can

(a) ignore it,
(b) conduct experiments to test for it. If you ever feel a bit dim or slow witted, just click the link below which lists hundreds of commentaries on the above tweet.

Click on it, scroll to a random point, and realise that actually, you are very fortunate to have received an education.

Jul 6, 2021 4 tweets 1 min read

I was just about to say, "Looks fine to me” ... ... when I suddenly noticed this.