
Tales from the Wards 2.0
Six real cases. Could AI Have helped?
Reorienting the conversation in medicine and health care
Live Last read · last published · next check

Six real cases. Could AI Have helped?
If you grew up in the 1990s, is your brain already singing the next line?A&Ox2 is a reader-supported publication.

From Waymo's safety data to the launch of our AVERT Study

Profound Sadness for the State of Science

Prediabetes is useful as a risk marker. It is a signal to take metabolic health seriously. But metformin for prediabetes is a different claim. It says that changing the timing of a diagnosis is itself enough to justify medication. That is where the logic breaks.

A potpourri of clinical pearls from my practice

An Example of Thickened Fluids in Dementia
Rethinking language and responsibility in medical training

Hyponatremia reliably predicts bad outcomes. But does correcting it fix them? A new randomized trial makes that uncomfortable truth harder to ignore.

A satirical commentary on valuing what is easiest to measure, not what matters most

Why doctors should abandon the new shiny drug for MASH in favor of a far superior drug class.

If the goal in academia is exchange, networking, and collaborations, then we should stop pretending the hotel ballroom is the only or even the best way to achieve that.
Defining the expertise at the heart of hospital medicine

The Message of Eric Topol’s Peter Attia Takedown
We recognize great doctors when we see them, but our systems measure the wrong things

The humility of uncertainty & WikiGuidelines by Dr. Spellberg

Watch now | Why Following Guidelines is Not Synonymous with EBM

Watch Dr. Spellberg now | UCSF Division of Hospital Medicine Grand Rounds at San Francisco General Hospital

The Implausibly Large Effect in the Recent Time-of-Day Immunotherapy Trial

What Moderna's data tell us about the efficacy of their mRNA vaccine