Ben Millette

Banner image with screenshot of scientific article from nature Medicine, as well as two panels from the study method and results

⚠️ Despite all the hype, chatbots still make terrible doctors. Out today is the largest user study of language models for medical self-diagnosis. We found that chatbots provide inaccurate and inconsistent answers, and that people are better off using online searches or their own judgment.

There are few absolute rules in medicine but «never treat sinus tachycardia, treat the cause» has extremely few exceptions.
#emimcc #cardiosky #medsky

A big hand with the word 'No' underneath

Some interesting ideas in here some of which is consistent with my practice, some of which (epoprostenol) is entirely new to me!

Refractory septic shock: our pragmatic approach

AKA the “Tommy’s Tonic”

Now updated for 2026

link.springer.com/content/pdf/...

Thoughts welcome @damienbarraudmd.bsky.social @redneeraj.bsky.social

Great discussion here that reminds me of the old medtwitter

My own practice is that it is default the home team's responsibility to have the treatment escalation conversation. I will undertake it myself on certain occasions when e.g.
-my opinion is at variance with home team decision
-time is a factor
-additional complexity to the case

Intensive care medicine ST3 scoring matrix that references POCUS multiple times

This year's scoring matrix for ICM ST3 applications contains 3 references to POCUS. Looks like we think this is important for the practice of ICM and maybe should be on the curriculum? 🤔

bsky.app/profile/benm...

The impact of skin tone on performance of pulse oximeters used by NHS England COVID Oximetry {at}home scheme: measurement and diagnostic accuracy study Objectives To assess the impact of skin tone on the measurement and diagnostic accuracy of five fingertip pulse oximeters used by patients in the NHS (National Health Service) England COVID Oximetry {...

Fantastic to see this published at last

An objective assessment of the efficacy of home pulse oximeters on different skin tones

Ping @iwashyna.bsky.social

Open to comments from all #foamed #medsky

www.bmj.com/content/392/...

A tweet saying: Asked the burd in Krispy kremes for 5 Nutella donuts and she says "have you got any nut allergies" aye pal I'm planning suicide by donut

See also:

I mean we are absolutely in a place now where the only solution to this information disorder is for everyone to constantly evaluate the source of information. Never trust a chatbot, but also don't believe a video unless you know and trust where it comes from.

Unfortunately... that's a lot of work.

James Cameron: Great at making Movies™, rubbish at marketing

I dunno, I kind of think Lung is the slightly boring sister that no-one wants to dance with that actually has more to her than meets the eye

Very true, but even with that- Heart is a more difficult accreditation to achieve! To have more than twice as many accrediting compared to Lung- there must be some motivating factors there too...

I used to think lung was quite widespread

Then I looked at NEAT-ECHO again and did the math

Remember that with FUSIC heart any cardiac physiologist can be a supervisor- this hugely increases the pool…

Flow is SO much better during IA surely there are some lessons to be earned here for optimising consultant time? (Heavily specialty dependent of course)
bsky.app/profile/benm...

New accreditation registrations: Top three modules: 756 Heart 380 Lung 59 Cactus Lung

Am I alone in being surprised by how dominant Heart is over all the other FUSIC modules? I feel like Lung is the low hanging fruit, accessible to a wider range of professionals esp AHPs.
Is there better provision of Heart mentors or is it just sexier? #pocus #emimcc

a man is standing in front of a whiteboard with the words nice flow written on it ALT: a man is standing in front of a whiteboard with the words nice flow written on it

It really is remarkable how much better flow through the hospital is on strike days. Maybe having a few more senior decision makers around on normal days will pass a cost benefit analysis?

Screenshot from the series pluribus showing Rhea Seehorn being told "that's the opinion of every medical doctor on earth"

Trying to think of the differential for this presenting to cardiology... 🤔

Yeah they say it's to get the IR team more procedures to get over the learning curve but it's definitely easier to recruit if you tell patients they have a 2/3 chance of being enrolled to the sexy new treatment

Had this exact thought re: 2:1!
If you don't think the patients will have equipoise, then you have a PR/ information provision problem, not a randomisation problem

True, but it's more precise than our previous ham-fisted attempts at immunomodulation in sepsis! Maybe this is something we can iteratively achieve as we better define sepsis phenotypes?

CCR Down Under 2025 Original Article | Dec. 9, 2025 | NEJM.org Ketamine or Etomidate for Tracheal Intubation of Critically Ill Adults Figure 1. In-Hospital Death, According to Trial Group. The NEJM identity sits at the bottom.

In a randomized trial involving critically ill patients undergoing intubation in EDs or ICUs, the use of ketamine for the induction of anesthesia did not lead to significantly lower in-hospital mortality than etomidate. Full trial results: nej.md/44EOlxB

@criticalcarereviews.com

Bear hugging a friend who I hadn't seen in ages, both of us in full PPE in the middle of a covid ICU

Trimethoprim-sulfamethoxazole carries a very low (but real) risk of causing ARDS.

Duration of therapy is a risk factor (>7 days).

Could likely be minimized by avoiding unnecessarily long courses.

TMP-SMX is usually well tolerated, but can wreak havok (SJS, ARDS, drug-induced meningitis). #EMIMCC

www.fraserhealth.ca/news/2025/No...
V interesting- spinal anaesthesia still seems to work in microgravity! I look forward to the full publication of these results