Founder, MIT Sloan, coach, fmr Chief Med Officer & Medicaid med dir, Top Voice 4x, data nerd, digital health, MedEd, humanist, Tweets personal, RT🚫endorsement

Joined May 2009
13,430 Photos and videos
Pinned Tweet
Replying to @usnehal
My experiences: -as child in hospital -as #womenshealth patient -caregiver/advocate for my WOC mom w/ #COVID19 in #ICU -#MedEd -#FQHC (#NYC #Boston) -#globalhealth Inform how: I am a doctor I ask❓on #data & standards #MedTwitter #research #innovation #WomenInSTEM #WomenInTech
As doctors we must understand: asymmetry of power/knowledge in the doctor-patient relationship vulnerability of patients the sacred trust they put in us by sharing their stories, bodies, hopes, fears, pain An x-ray is not just an x-ray. Is someone’s body part #medtwitter
4
42
108
The lived reality vs the way the system collects data.
A medical school professor walks into clinic with her 12 year old son. He has 18 specialists. She does not fill out the intake form anymore. She hands the front desk a one-page medical sheet and says, "This is the current state of affairs," because the complexity will not fit in anything a clinic prints. That is Kathleen Muldoon, a med school professor and the parent of a medically complex child. She and co-guest Jonathan Cottor came on the podcast to talk about what pediatric medicine misses when it treats the diagnosis as the whole job. Kathleen is the one teaching your future colleagues what humanistic, whole-person care looks like. She is also the one who gets up in the middle of the night to reposition her son, because he cannot move on his own. Here is what your 15 minute visit does not see. The alarms that go off all night. The cold mornings in Arizona when his whole body tenses and nothing moves the way it should, which changes his mood before you ever lay eyes on him. The communication device she has to fight for so people can finally hear that her son can answer his own questions when he is given the chance. The two siblings absorbing the labor and the mood of a household run on triage. The third grader asked at school whether his brother was going to die. The six months of logistical work it takes to arrange a single weekend away. She has a spider web diagram she shows med students that maps all 18 specialties and how they interact, because none of them talk to each other. The diagram is not a curriculum exercise. It is the operating manual of her actual life. We have built a pediatric system that treats medical complexity as a list of diagnoses to manage. The families managing those diagnoses are running a 24/7 operation with no shift change, no relief team, and a clinic visit that asks about A1c but not about whether anyone in the house has slept. The intake form has a field for allergies. It does not have a field for "the parent across from you has not slept through the night in years." The reframe worth bookmarking: care is not just clinical, it is relational. When you care for the caregiver, you are caring for the whole family. In a 15 minute visit, that is three questions. Do you have respite care. How do you care for yourself. What would make it easier for you to keep doing this safely. Those three questions are the cheapest, highest-leverage intervention in pediatric medicine that almost no chart prompts you to make. They take under a minute. They cost nothing. They tell the family across from you that the person doing the most work in their child's care, the one with no medical license and no shift end, has finally been seen by the system that prescribed it. Listen to the full conversation on The Podcast by KevinMD. Link in the replies. Which of those three questions have you been avoiding asking, because you are afraid of the answer? #ThePodcastbyKevinMD
4
215
Perhaps… (this is going to be a radical idea), we should put the money towards actual patient care activities, not Power Point decks by MBA in suits.
US nonprofit hospitals spent $7.8 billion on management consultants from 2009 to 2023, but contracts were not associated with meaningful changes in finance, operations, or quality of care. 🧵 ja.ma/4d46zfq
3
148
Spring in Boston
20
77
Now this is hilarious and brilliantly creative. Love it.
These kids came up with a unique performance for their school Talent Show doing Synchronized “Air Swimming!”
1
109
Umbereen S Nehal, MD, MPH, MBA retweeted
Twenty years ago, longevity medicine did not have a name. No standards, no curricula, no consensus on what counted as evidence. Today, Advanced Research Projects Agency for Health (@ARPA_H) has committed $144 million through its PROSPR program to develop therapeutics that target the biology of aging directly. The @xprize Healthspan competition is mobilizing global talent. Governments from Saudi Arabia to the U.S. are funding aging research at unprecedented scale. The field has reached an inflection point. The field has reached an inflection point. The people building it are gathering in Cambridge on May 26 and 27 for the Aging Code Summit, part of the inaugural Boston Tech Week. I'm speaking on Day 2. A few of the conversations I'm most looking forward to: @manoliskellis (MIT) on the computational biology of aging Li-Huei Tsai (The @MIT_Picower Institute for Learning and Memory) on the latest in neurodegeneration Kristen Fortney @kpfortney (@bioagelabs) on translating aging biology into therapeutics Sharon Rosenzweig-Lipson (@lifebiosciences) on the science behind a leading longevity company @mahdi_moqri (@agingbiomarkers Consortium) on where aging clocks are headed Amy Proal, PhD @microbeminded2 (@polybioRF Research Foundation) on chronic infection in aging Saranya Wyles, M.D., Ph.D. @drwyles_derm (Mayo Clinic) on regenerative dermatology and skin longevity Jens Eckstein @AkikoaCom (Hevolution Foundation @hevolution_f) on capital meeting the science. My Day 2 talk focuses on evidence-based best practices in longevity medicine, the framework I've been developing across clinical practice, Harvard Medical School teaching, and advisory work with longevity organizations worldwide. The question I keep returning to: which interventions actually deliver outcomes, and how do we tell. Day 2 also features the pitch competition, where top seed and Series A companies compete for a one-year residency at The Engine, MIT's tough-tech platform and one of the most consequential launchpads in longevity right now. If you work where this field is being built, you should be in this room. Register: lnkd.in/eYZX5NH2 Joining an extraordinary faculty including Christin Glorioso MD PhD @DrGlorioso (NeuroAge Therapeutics @NeuroAgeTX), @JamieHeywood (@AldenScientific), Fiona Miller (@quadrascope), @Dr_RayMak, José Navarro Betancourt, MD, PhD, Justin Taylor, Noriko Yokoi, Ph.D., Daniel Dacey, SpringBehrouz, Raghav Sehgal, PhD @rv_sehgal, Jay Luthar, MD, DipABLM, @HillaryLinMD, Umbereen S. Nehal, MD, MPH, MBA @usnehal, @tomzuber, Robin Mansukhani, Fernanda Cerqueira, David Hall, Yeh-Chuin Poh, Salah Mahmoudi, and @RutaLaukien Updates at: linkedin.com/in/agingdoc More at linkedin.com/in/agingdoc #BostonTechWeek #LongevityMedicine
1
3
11
1,083
Woah. How is this legal?
“This new Google experiment is not a matter of cosmetic adjustments — it radically transforms journalists’ words.” 🔗 rsf.org/en/usa-google-claimi…
1
101
Lifestyle & personal choices? 👍 Actual operations & staffing matter as well. Both/and
In healthcare, no amount of personal wellness practices can substitute for appropriate staffing, humane scheduling, efficient clinical workflows, adequate administrative support, or a culture of psychological safety.
1
83
If you want to fix healthcare, don’t start with the ED. Start with everything that forces people to come here.
2
98
Thing is, the cognitive load of working with a tool that gets half the answers wrong (same as a coin toss), is high. #AI and #tech that is this faulty adds to burnout. Eventually, the human will get exhausted and fail to catch an AI error.
Two studies tested ChatGPT, Gemini, and other AI tools on health-related questions. In one, they got almost half the answers wrong. Companies are improving AI health tools, but experts say they should supplement, not replace, medical professionals. wapo.st/4mKEJt4
4
321
Umbereen S Nehal, MD, MPH, MBA retweeted
Every system that was regulated, either explicitly or implicitly, by the fact that they were effortful for humans (letters of recommendation, lawsuits, government filings, essays) will break.
🚨New preprint! We find evidence of LLMs enabling people to file lawsuits without lawyers (filing "pro se") at historically unprecedented rates in federal courts.👇 1/n
56
189
1,282
279,127
Hear, hear
Four million physician hours are spent on mandatory modules each year. They represent not just clinical time lost but “evenings reclaimed and cognitive load reduced.” Redundant training and low-value modules should be a thing of the past. jamanetwork.com/journals/jam…
162
Am here for this content, not what is trending right now.
In Nara Park, Japan, the deer are so calm you can walk right up and hand feed them under the blooming cherry blossoms 🌸🦌
2
2
10
189
Umbereen S Nehal, MD, MPH, MBA retweeted
Are you deploying AI in a clinical setting? We’ve expanded our #AIMI26 Call for Abstracts! Inviting submissions for posters & short oral presentations (“What’s Working Now: Real-World Deployments in Health AI”) Apply by Apr 10: stanford.io/47wkuJu #StanfordAIMI #HealthAI
4
15
46
4,688
This needs to be painted.
春の浜名湖は美しい
1
2
49
When I use #GenAI as an assistant, I’m never sure if it’ll end up causing me to spend three times as long on it or be more efficient. (It does tell me that I have “high standards” so it may be that I expect too much accuracy from it while most people accept the errors.)
AI was supposed to save time. Instead, many workers are having to spend time and effort prompting, checking, and fixing flawed output. f-st.co/7Fc8fu4
2
69
Umbereen S Nehal, MD, MPH, MBA retweeted
This is huge. - Publishers running your submitted books through AI systems adds your books to the AI training data without your consent. - Your book of 300-500 pages is being reduced to an AI summary and not "read" by editors. - An AI summary will be what your book is accepted or rejected on. - If the publisher tests your book to see if it was written with AI, it will get a false positive as the AI will recognise the book from it's training data, due to this process. This has to stop and publishers must be forced by agents and writers unions to sign agreements for every book, prior to 'reading' stating that they will not put submitted books through AI systems without written consent.
Curtis Brown (@CWAgencyUK) has voiced concern about editors using ChatGPT to assess confidential manuscripts, as more literary agencies introduce clauses around Artificial Intelligence into contracts 👇
65
852
2,244
90,804
Umbereen S Nehal, MD, MPH, MBA retweeted
Research shows that the hardest work in deploying agentic AI in a clinical setting is the “sociotechnical aspects” — rather than tasks like prompt engineering. These findings were distilled into five “heavy lifts” that are necessary for success in deploying AI agents in any setting. Learn more: bit.ly/40uxOda
4
33
90
5,285
When you build, train, or tune a model, what are you optimizing for? In #health and #mentalhealth, not missing a crisis can be more important than accuracy.
AI can miss mental health crises, and that’s dangerous. This study builds a guardrail that detects psychiatric emergencies in text with ~99% sensitivity, aiming to catch what other systems miss. So, why is this important? Most AI systems optimize for accuracy. This one optimizes for not missing a crisis. nature.com/articles/s41746-0…
1
2
123
Last week I was part of a #biotech delegation to the Netherlands for #womenshealth. We were welcomed by the U.S. Ambassador on the first day.
1
9
491
Sightseeing in Amsterdam
5
50