Satya Dash

13/ We also emphasize this in the paper: many of our benchmarks depend on carefully curated cases (by physicians!). Real clinical data is messy, incomplete, and contradictory in ways that structured cases aren't. Performance in actual clinical workflows may be lower.

@ethanjweiss The appC3 LOF carriers have lifelong reduction in apoB. It would interesting to see longer-term data.

@SatyaDash009 it is a reasonable question though of course lots of caveats...

I am surprised that nobody is surprised by these results. It feels quite earth-shattering to me but what do I know? At 6 months, olezarsen reduced triglycerides by 63.9%, remnant cholesterol by 71.9%, and apolipoprotein B by 16.0% over placebo, with no difference in LDL-C. And yet no difference in plaque at all at one year? None? So many questions and so few answers but I think this is one of the most interesting null results I have seen in a long time.

Let’s cool our jets on assuming that the multitude of benefits are due to weight loss or wt loss alone. It’s very possible, but I won’t forget results of Harmony. Different population (diabetes) but remarkable CV risk reduction with zero weight loss thelancet.com/journals/lance…

Another win for the lower is better crowd… Intensive LDL Cholesterol Targeting in Atherosclerotic Cardiovascular Disease | New England Journal of Medicine nejm.org/doi/full/10.10…

When a patient is in the hospital, doctors used to use paper notes to communicate with one another. In the EHR era, the note stopped being a clinical communication tool and became a billing and compliance artifact. The results have been a disaster. One JAMA study found notes got 60.1% longer from 2009 to 2018, while redundancy rose 22%. ONC has explicitly acknowledged that clinicians use templates to stuff notes with unnecessary information into the chart to meet billing requirements, creating note bloat. The clinical note was no longer a method of communication. It was a billing document. So hospitals layered secure chat on top just to communicate the actual plan of care. And even that workaround is not working. A 2024 JAMA study found more secure messaging was associated with more time on the phone, not less. Doctors needed to call to clarify the now constant message stream. Another study found higher messaging volume was associated with higher odds of errors. More messages means a higher cognitive load with most of the information being low-importance. This increased cognitive load leads to more errors. We took what should have been efficiency improving technology, a computerized chart, and so over-regulated it and misaligned incentives that it has led to harmful downstream effects. Now, please don't do this with AI...

REVIEW ARTICLE Bariatric surgery in the era of modern obesity pharmacotherapy by Ally Zhao @SatyaDash009 Obesity and Endocrinology, Volume 2, Issue 1, January 2026 Open access 👇 doi.org/10.1093/obendo…

Financial conflicts of interest are a *much* bigger problem in alternative medicine than they are in mainstream medicine.

I can't respond directly here because I guess I am blocked by Marion. So I will take this opportunity to sing it from the mountaintops: PLEASE DO NOT TAKE MEDICAL ADVICE FROM THE INTERNET!!

Hey look! @nejm moved my writing to a new spot. Here's the first post, a rant about a particularly annoying requirement for those of us in work in hospitals (link below) 1/3

@doctorveera Congratulations! Where will you be based?

After a decade abroad in research, I am happy to be home 🇮🇳. I am excited to share that I am joining the AI healthcare startup Wellytics as Chief Scientific Officer. I trained as a physician in India. I moved abroad to pursue human genetics and later focused on genetics-driven target discovery and drug development at Regeneron. Returning now to build in India feels deeply meaningful. At Wellytics, I will lead R&D and build our genomics division. - We will establish large-scale Indian genomic datasets to power drug discovery. - We will build a world-class human genetics and target discovery team. - We will collaborate closely with academic geneticists across India. - We will provide tools, training, and support to strengthen human genomics research in India. Wellytics is digitizing Indian healthcare and making it AI ready. We will combine AI, clinical data, and genomics to generate real-world evidence and build international-grade genetic association resources for India. Excited to help build a genetics-driven drug discovery engine in India!

I am genuinely gratified that the NYT has at long last published an opinion piece questioning whether major American medical associations abandoned scientific rigor in pediatric gender medicine. For years, we were told “the science is settled.” We were told that every major medical association supported youth medical transition as safe, necessary, and lifesaving. But the real story isn’t in Opinion. It belongs in the paper's news pages — where reporters should be investigating how U.S. medical associations developed policies that projected certainty, suppressed scrutiny, and sidelined dissent. This is not a cultural controversy. It is a medical scandal. When professional authority is used to overstate evidence and shield weak science from challenge, that is institutional malpractice. Those who provided the imprimatur for pediatric “gender-affirming care” will have to answer for it. Paywall-free link to @jessesingal’s NYT op-ed: nytimes.com/2026/02/24/opi…

Excited to share one of my favorite genetic discoveries made at the Regeneron Genetics Center. We went looking for genetic clues about why some people smoke more than others and found something in an unexpected place: the genomes of Indigenous Mexicans. 1/

There should probably be a Community Note on this, stating that the article the estimate comes from used a totally invalid modeling choice and got a nonsense result. Statins actually extend lifespans by a year or more if you start early.

It's worth reading the replies to the article: bmjopen.bmj.com/content/5/9/e0… One of them predicted, and the other led to an actual, much better thought out re-estimate: tandfonline.com/doi/10.2217/fc… Projecting from the relevant trials, extensions over a lifetime were denominated in years:

This IS pretty much misinformation. The numbers are correct. But they don’t mean what they are taken to mean. Stats is confusing, yes even for smart doctors. These small numbers are NOT the actual life extension effect of a statin on an individual. They are population averaged.

Stats lesson #2 for statin denialists: NNT and ARR are not intrinsic properties of a drug. They are functions of RRR, baseline risk and time horizon. Anyone using them to say statins provide no clinically meaningful benefit is dead wrong. If the underlying event rate is low, NNT will be high even when the drug produces a real and consistent proportional risk reduction. This is why risk stratification matters. The same treatment will have very different NNTs in low-risk versus high-risk groups, purely because baseline risk differs. So quoting a “high NNT” to claim a treatment doesn’t work is a category error. For an individual, expected benefit depends on their own baseline risk. Remove risk stratification and time frame, and NNT becomes a context-free number that tells you far less than people think.