I know of a 40,000 patient clinical trial where the SAS programmer determined the key secondary outcome as positive if Y=time to stroke < maximum follow-up time. Y=missing -> positive outcome because with SAS missing < anything. The error still stands.

Our latest in ERJ: Beyond other conceptual flaws, race-specific spirometry references face practical issues because technicians often misclassify race and ethnicity of visible minority patients which introduces further bias in clinical algorithms: bit.ly/4qozHUq @europeanrespsoc.bsky.social

Where should we host next year's symposium? Please let us know by filling in our feedback form or commenting below: #ML4H2025

Your universe is a photocopy of a photocopy of a… on.ft.com/4ahkNKa

We have a new paper in Science Advances proposing a simple test for bias: Is the same person treated differently when their race is perceived differently? Specifically, we study: is the same driver likelier to be searched by police when they are perceived as Hispanic rather than white? 1/

Allowing doctors to practise in both public and private systems solves what exactly? What Alberta is proposing may sound good superficially but makes no sense in the Canadian context, by @picardonhealth.bsky.social www.theglobeandmail.com/opinion/arti... via @theglobeandmail.com

It’s not too late to register!

Not too big. Not too small. #ML4H2025 is just right for deep conversations and building meaningful connections. Come get your healthy dose of ML research and vitamin D in sunny San Diego! Register today: ahli.cc/ml4h-register/ It might be your favorite conference of the year!

Online registration for #ML4H2025 closes in 6-7 days! Don't wait. Secure your spot before November 21st and join us in San Diego: ahli.cc/ml4h-register/ In-person registration will be available during the event at Westin San Diego Gaslamp.

Social Determinants of Health and Racial Disparities in Lung Function Environmental exposures explain a larger proportion of racial disparities in lung function than previously believed aminad.bsky.social @news.ubc.ca @donsin40.bsky.social @msafavi.bsky.social 🔗 tinyurl.com/43vb6ubz

"Trapped in a ChatGPT Spiral." 🌀 Important work by @kashhill.bsky.social & @dylanfreedman.nytimes.com on how chatbots have a tendency to endorse conspiratorial and mystical belief systems. This shows again how conformist LLMs can be. Worth a listen 👇 www.nytimes.com/2025/09/16/p...

Some health-data companies are selling access to patient records to pharmaceutical companies without patients’ knowledge, according to a new study, by @hannay.bsky.social www.theglobeandmail.com/business/art... via @theglobeandmail.com

Is it (race as a social determinant of health)... or is it (racism as a fundamental cause of health inequities")?

🗞️Hot off the press in @atsblueeditor.bsky.social! Our analysis shows that unfavourable social and environmental exposures explain a much larger share of racial disparities in lung function than previously thought. Details: doi.org/10.1164/rccm... @msafavi.bsky.social @awgaffney.bsky.social

🎉 #ML4H2025 decisions are out! ✅ 74 papers accepted in the Proceedings track ✅ 71 papers accepted in the Findings track Huge congrats to all authors! Register today and join us in San Diego for an amazing conference: ahli.cc/ml4h-register/

⏰ Don’t miss the Early Bird rates (ends Oct 15)! AI & the Future of Medicine 📅 Nov 15–16 | 📍 UBC Vancouver Day 1: Free Day 2: $15 (academics) / $150 (industry) 🎟️ Register now tinyurl.com/2ja3n5j8 ℹ️ More info tinyurl.com/44vayx4m #DataHealthUBC #AIandMedicine #MITCriticalData

Q: The US spends almost 2x on healthcare - why? Is it the 💰prices, or ⬆️use? A: It's (still) the prices. We pay 2x prices (and often use less: fewer in-person physician visits, shorter hospital stays) www.healthsystemtracker.org/chart-collec... #Quality #Affordability #Healthcare

A Spanish foundation using an artificial-intelligence tool to screen grant applications says it makes the peer-review process more efficient, but other researchers say it’s a breakdown in trust. www.nature.com/articles/d41...

📢 Submissions are OPEN for #ML4H2025! Main Symposium and Demos Track now accepting papers. Deadline for both is Sept 8 (AOE). 🔗 Symposium: openreview.net/group?id=ML4... 🔗 Demos: openreview.net/group?id=ML4... Join us in sunny San Diego on Dec 1–2, right before @neuripsconf.bsky.social

📢 The #ML4H2025 Author Mentorship Program is now open for sign-ups! 🐣 Mentees: get expert feedback & level up your submission 🦉 Mentors: share your wisdom & support the next gen It’s like co-writing, minus the authorship drama! Sign up 👉 ahli.cc/ml4h/mentors...

🚨 Machine Learning for Health (ML4H) is back and better than ever! 🌴 Join us in San Diego on December 1–2, 2025, right before @NeurIPSConf! 📢 Call for Papers is LIVE → ahli.cc/ml4h/call-fo... ⏳ Deadline: Sept 8 — don’t miss it! 🔁 RT and follow @ahli-cc.bsky.social for updates!

Lung function reference equations are designed to estimate healthy lung function using data from healthy volunteers. However, in patients with respiratory diseases, "healthy lung function" is an unobservable counterfactual. www.atsjournals.org/doi/abs/10.1...

Are you a statistician or health-economist who wants to develop new methods and who is interested in evaluating the role of diagnostic & prognostic prediction models and AI? We have an opening for a PhD position at Maastricht University! www.academictransfer.com/nl/jobs/3503...

🫁⚖️ Can lung tests ever be truly race-neutral? A new study shows no spirometry measure fully avoids racial bias. Fairness depends on context – different definitions of 'race neutrality' clash in practice. 🔗 www.atsjournals.org/doi/10.1164/... #SciComm #HealthEquity 🧪

Breath testing has more than a century of racist assumptions baked into it because it was originally used in the U.S. to argue Black people have inferior lungs, says UBC public scholar Amin Adibi. 🫁

The Blue Journal thanks Amin Adibi, MSc, for his contribution to the March issue Is Achieving a Fully Race-Neutral Approach to Lung Function Classification Even Possible? @suryapbhatt.bsky.social www.atsjournals.org/doi/full/10....

📣UBC Medicine Datathon brings together students and professionals from diverse health disciplines and data-related fields to tackle healthcare challenges through data-driven approaches. We invite you to participate! To register and for more info: datascienceandhealth.ubc.ca/events/ubc-m...

New paper alert!! What makes a ‘good’ decision with AI? We offer an ethical framework that does not require explainable AI, factors local model performance, and centres patients’ best interests ebm.bmj.com/content/earl...

The gap between First Nations and non-First Nations life expectancy in Alberta is 19 years, and wider now than any other point on record. This is a provincial and national emergency. #ableg www.stalbertgazette.com/local-news/f...

🚨BREAKING. From a program officer at the National Science Foundation, a list of keywords that can cause a grant to be pulled. I will be sharing screenshots of these keywords along with a decision tree. Please share widely. This is a crisis for academic freedom & science.

Still thinking about this really thoughtful piece. doi.org/10.1164/rccm...

This is at the heart of almost every AI-enabled disaster: Scaling is not the only way to improve, it's just the easiest; training on copyrighted work, ignoring data quality, overlooking evaluation validity, etc are not actually"necessary evils", but just represent the path of least resistance.

Impossibility of Race-Neutral Spirometry? Our latest in @atsblueeditor.bsky.social brings an #algorithmicfairness perspective to spirometry. We show that no single spirometry output can simultaneously satisfy all intuitive definitions of race-neutrality. @msafavi.bsky.social doi.org/10.1164/rccm...