Ok, folks, why the difference?

John Mandrola, MD It's great to be an optimist when you're running a clinical trial. But during the design phase you need to be a pessimist, or have a flexible budget and use a sequential design that doesn't pretend to know N.

An important observation👇🏻 The majority of *major* cardiology trials are under powered to sort signal from noise. Just as Frank Harrell points out nearly every week. This is bad for our field.

I can’t explain this observation. Seriously. I can’t.

I’m not an oncologist and a lot these studies are intimidating due to the hard names of drugs. But. If I’m diagnosed with cancer, and a doc recommends a Rx, I’m googling it along with “Vinay Prasad”. Incredibly important work for patients and society. Kudos John Arnold

Darren Dahly - on Bluesky and Substack I shudder whenever I read “clinical trial emulation”. How do you emulate blinding or guarantee that all confounders are measured? How do you start the clock at the same time for every patient?

I really enjoyed this piece. Perhaps Walter Kirn approaches a modern-day Mark Twain. And gosh do we ever need that now.

This is such an excellent summary image. Perfect for BP and glucose.

Also. Inherent in this image is exactly why quality measures more often worsen healthcare quality

hs-troponin has become a useless clinical test. Ordered indiscriminately, nearly always positive regardless of situation. Can you tell I’m on general call?

I'm up to diagnostic tests! A two-parter -- there is just too much to get into one post. If you ever have confused yourself on sensitivity and specificity, this is the post for you!
sensible-med.com/p/improving-yo…

Agree completely with John Mandrola, MD. Additional issues:
1) DEDICATE-DZHK6 1 year events are too high !!! TAVR= 5.4% (1% in PARTNER 3) and SAVR= 10% (2.9% in P3). Data from Germany seems to do worse than other RCT (remember GARY registry?)

An enjoyable listen. Impressive thinking and speaking. Mr Hughes… I’m a fan but worry not I’m not that conservative. :)

TAVI is a great innovation, but the title of this editorial in NEJM on the one-year results of DEDICATE-DZHK6 trial is not good

1 yr is too soon to judge TAVI vs surg AVR (experts know that)

Another prob is making readers download a separate document to see COI

COI should be…

Me too. Another “old fashioned” thing is to go to radiology and discuss scans w radiologists in person. They are great teachers.

Very nice explanation of analytical choices.

Analytical choices are probably the biggest problem for observational studies - and one of the most widely ignored.

Can lead to massive harm.

You lost me at Cochran review. COVID confirms all guideline writing groups to have significant biases. Some are good, some are bad. Very hard for patients to do this themselves (see Alex Berenson and spine surgery.)

Find a doctor you relate to and trust, get second opinions for…

Drs Alexander and Hanson have used many thousands of words debating the utility of healthcare. Both make great points.

Obviously healthcare helps people. Obviously much of healthcare is low value.

Love this advice from RH on finding low-value care

You have to liste to the last episode of TWIC from John Mandrola, MD talking about Complete trial
nejm.org/doi/full/10.10…

Thanks for the mention. Indeed there are reasons to pursue rhythm control. Nearly always for symptoms. But. But.

Rhythm control is a) hard, b)confers risk, c) requires expertise

EAST is one of the most over-interpreted trials in #Afib space (less than 20% rhythm control arm…

Perhaps another important heterogeneous treatment effect 👇🏻👇🏻 (Andrew Foy )