friseb

Los médicos ofrecemos un servicio, el de aplicar una ciencia a una persona en particular. No somos artistas.

Among patients with atherosclerotic cardiovascular disease, targeting an LDL cholesterol level below 55 mg per deciliter led to a lower 3-year risk of cardiovascular events than targeting a level below 70 mg per deciliter. Full Ez-PAVE trial results and Research Summary: nejm.org/doi/full/10.10…

110.000 papers que contienen referencias inventadas solo en 2025

This is a whopping effect — the intensive-lowering group had ~15% lower LDL cholesterol levels than the conventional group, for a 33% reduced risk of a cardiovascular event They did this through more high-intensity statin use, more ezetimibe use, and slightly more PCSK9 inhibitor use Safety was equivalent, except that the intensive-lowering group had FEWER events of creatinine elevation (i.e. better kidney function) The lower the LDL cholesterol, the better

Retraction—Today, we retract an unsigned 1977 commentary suggesting talc powder containing asbestos was not harmful. The Lancet was informed that the author had undisclosed competing interests and breached publication ethics. /4

Software horror: litellm PyPI supply chain attack. Simple `pip install litellm` was enough to exfiltrate SSH keys, AWS/GCP/Azure creds, Kubernetes configs, git credentials, env vars (all your API keys), shell history, crypto wallets, SSL private keys, CI/CD secrets, database passwords. LiteLLM itself has 97 million downloads per month which is already terrible, but much worse, the contagion spreads to any project that depends on litellm. For example, if you did `pip install dspy` (which depended on litellm>=1.64.0), you'd also be pwnd. Same for any other large project that depended on litellm. Afaict the poisoned version was up for only less than ~1 hour. The attack had a bug which led to its discovery - Callum McMahon was using an MCP plugin inside Cursor that pulled in litellm as a transitive dependency. When litellm 1.82.8 installed, their machine ran out of RAM and crashed. So if the attacker didn't vibe code this attack it could have been undetected for many days or weeks. Supply chain attacks like this are basically the scariest thing imaginable in modern software. Every time you install any depedency you could be pulling in a poisoned package anywhere deep inside its entire depedency tree. This is especially risky with large projects that might have lots and lots of dependencies. The credentials that do get stolen in each attack can then be used to take over more accounts and compromise more packages. Classical software engineering would have you believe that dependencies are good (we're building pyramids from bricks), but imo this has to be re-evaluated, and it's why I've been so growingly averse to them, preferring to use LLMs to "yoink" functionality when it's simple enough and possible.

A modern call by the founders of mendelian randomization (MR) to turn back to basics and run MR only if genetic variants used as instruments are real proxies of the environmental/drug/lifestyle perturbation we want to test. Some thoughts on the flood of MR papers👇

I was so pleased to be part of a great team writing a commentary on the draft FDA guidelines on Bayesian clinical trial designs. Should have been done 30 years ago, but better late than never

Today we are introducing 2 key features to JAIGP: AI Review & Open Prompting. AI review is part of 5-step process where papers get feedback from @reviewer3com & are evaluated based on their ability to address that feedback. This means they are not stuck in an endless AI review loop. Open prompting is a newer idea. We are making all prompts we used to create JAIGP open, and also, opening up the journal's rules to the community for suggestions. You can suggest the prompt we should run next! So, if you have opinions about AI generated papers, you can share them directly with us at jaigp.org/prompts

Are arterial webs limited to the carotid? In our case series + systematic review, we describe extra-carotid webs and propose the term “cervical arterial web”—a unified entity that may represent an underrecognized cause of cryptogenic stroke. @ENossek @BellevueHosp @nyugrossman @nyulangone @svinsociety @SVINJournal @AHAScience doi.org/10.1161/SVIN.1…

Ok this figure is pretty intimidating...

At biostroke.org, we sat together and thought about what needs to be done to advance #multidisciplinary #stroke #biomarker research in a pragmatic and meaningful way. This @TheLancetNeuro paper summarizes what we came up with. sciencedirect.com/science/articl…

Get ready for this. These Chinese investigators report a cohort of 121 malignant MCA infarction patients age >60, 47 of whom underwent drill-assisted sterotactic aspiration of the infarcted tissue from 24-72 hrs after onset. Check out the CT images and the 70 mL of nectrotic brain tissue in the image below to give you an idera of what they were doing. Retrospective, not randomized -- but the results were great. Herniation: 47% vs 9% Death: 65% vs 9% Favorable mRS (0-3): 3% vs 26% Not only #curingcoma –– cheating death. pmc.ncbi.nlm.nih.gov/articles/PMC12…

Adding more GPUs will never make a machine conscious. Nobel Prize-winning physicist Roger Penrose just dismantled the entire AI race’s core assumption. Right now, the industry operates on one belief. Build massive data centers. Scale the models. AGI will just “wake up.” Penrose destroys this completely. Penrose: “There is this sort of view that once you make a computer complicated enough or something, it suddenly becomes aware. I just don’t believe that. There’s no reason to believe that.” A machine can compute better than any human alive. But computation is not awareness. Penrose: “There is something quite different involved in understanding things, in being aware of things, of feeling things, which is not part of computations.” We’re confusing rule-following with actual intelligence. Penrose: “The keyword is the word ‘understanding.’ You can follow rules alright, but we don’t understand what we’re doing. The understanding is the key point.” Models today are exceptional at processing data. At mimicking logic. But true understanding requires consciousness. Penrose: “It doesn’t make sense to say of a device that it understands something if it’s not even aware of it. There is something much more profound in being conscious of something.” And here’s what should terrify every AI lab on earth. Penrose: “I believe that the brain is following the laws of physics, sure. We don’t have a good picture of the laws of physics.” Penrose: “Quantum mechanics is not an answer to the way the universe operates. It’s a partial answer. It’s incomplete.” We’re trying to engineer synthetic consciousness using classical computation. While biological consciousness likely operates on physics we haven’t even discovered yet. The race to AGI isn’t just an engineering problem. It’s a frontier science problem. The labs are hiring engineers. The problem might require physicists who don’t exist yet.

#TransientGlobalAmnesia was associated with high-grade styloidogenic jugular vein compression (~70% stenosis) in patients with small internal jugular veins, suggesting a novel imaging risk factor. ja.ma/4sbKrFV

Original Article: Phase 1 Study of Rezatapopt, a p53 Reactivator, in TP53 Y220C–Mutated Tumors (PYNNACLE study) https://nej.md/3OIQC5P Science behind the Study: Restoring Function to a Variant of p53 in Solid Tumors https://nej.md/3N0pQW8 #Oncology #Genetics

Excited to share one of the first examples of agents transforming clinical trial & EHR data analysis at scale. Biomni leverages 12M+ patient EHRs at Mount Sinai to scale complex trial emulation end-to-end:

@bobvarkey @IlliasulK I will question 20-65yo from 1983 population to adjust the ESR on a patient of 70, today.

@IlliasulK Thank you

70-year-old woman. ESR = 40. CRP = 1.2 mg/dL. Is it normal ? Inflammatory markers scale with age. ESR Men → Age / 2 Women → (Age / 2) + 10 So at 70: Normal ESR ≈ 45. CRP (mg/dL) Men → Age / 50 Women → (Age / 50) + 0.6 At 70 (female): 70 / 50 + 0.6 ≈ 2.0 CRP 1.2? Not impressive. Most “elevated” labs aren’t pathology. They’re physiology •aging •context. Medicine needs better algorithms. Treat the system. Not the scare number. @DrAkhilX @IhabFathiSulima

Are connectome-based network mapping methods and the >200 papers that have used it invalid? New paper out in @NatureNeuro says YES. nature.com/articles/s4159… I have concerns about this new paper's methods and conclusions, but am biased. What do others think?

Andexanet alfa pulled from the US market. This drug, that cost at ~$20k a dose, never showed real benefit, and caused stroke. The story is a fascinating example of misguided ethical thinking, and the human weakness for the power of narrative. This is a wild ride, get ready. 🧵

In the latest issue! Phase 1/2a clinical trial of hESC-derived dopamine progenitors in Parkinson’s disease dlvr.it/TPmg8g