Gabriel Brat

REMINDER: The way to attend the leading event on AI in medicine @SAILhealth #SAIL24 May 7–10 in Río Grande, Puerto Rico, is to submit an abstract by this Friday 1/19 to present at podium or poster! Top selections receive travel awards from @NEJM_AI. Visit sail.health/event/sail-202…

Never regret a day in your life. Good days give happiness, bad days give experiences, worst days give lessons, and best days give memories. —Professor Richard Feynman

Join us next week Tuesday, October 11th @ 4:00 p.m. EDT for our Harvard Surgical HSR Speaker @bratogram: "The Surgeon-AI Partnership: Leveraging Clinician Intuition to Improve Algorithms" csph.brighamandwomens.org/gabriel-brat/ Register here: bit.ly/3yb66V8

@docmartin22 @hayfarani Going from 30 to 10 pills for a lap chole is not driving transition to illicit opioids for vast majority of patients. For most, it’s just more concordant with needs. That fear is addressed by personalized multimodal prescribing over simple reductions for high risk pts/procedures.

@hayfarani Except nobody does. They just get up on the podium and congratulate each other for prescribing 10 pills instead of 30.

Dr. Singer, you are wrong. Data: 1. IPOP study: 91% of US vs 5% of non-US patients get opioids after same surgeries & in much larger amounts 2. Patients use <37% of their prescribed opioids after surgery 3. Diversion of excess prescribed opioids to community- well documented

@elliotthaut @TSACO_AAST @ashlie_nadler @injurydr @Sunnybrook @MartyMakary @HeidiOverton @hayfarani @TopKniFe_B @JosephSakran @DrJtrauma @kadtraumamd @MarkBicket Most likely and depends on the reference used, if any. Our Annals paper this month shows opioid consumption for multiple procedures, including laparoscopic surgeries, is lower than even popular consensus recommendations. Personalization is important. journals.lww.com/annalsofsurger…

@TSACO_AAST @ashlie_nadler @injurydr @Sunnybrook Really nice work about #opiods for emergency surgery in @TSACO_AAST. Are we prescribing too much after laparoscopic surgery? #opioidcrisis @MartyMakary @HeidiOverton @hayfarani @TopKniFe_B @JosephSakran @DrJtrauma @kadtraumamd @MarkBicket

#ThrowbackThursday How do you determine the amount of opioids to prescribe upon discharge? @ashlie_nadler @injurydr @Sunnybrook highlighted the need to standardize discharge opioid rx practices. Re-read #OpenAccess ➡️ bit.ly/3sqDXWw #SoMe4Surgery #opioidcrisis

Once upon a time, we didn’t have MRIs, we didn’t have electronic medical records, we didn’t have dermatoscopes. Medicine is always evolving with new tools. AI algorithms are tools that can help us - we have to figure out which ones are actually helpful and adapt to using them.

I won’t poo poo any optimistic sign when the hospital is so full of sick people.

Happening now -- it's not to late to join the webinar! Hear from members of 4CE ("foresee"), the international Consortium for Clinical Characterization of COVID-19 by EHR. Speaking now: @bratogram on clinical perspective. Next up: @avillach & @ftbourgeois on pediatric studies.

1. Read and write more. 2. Don't hesitate to admit when you're wrong. 3. Be comfortable changing your opinion. 4. Find a mentor. 5. Stay teachable. 6. Make mistakes and learn. 7. Don't get offended easily. 8. Ask questions. 9. Spend time with nature. 10. Stay humble.

📰Just out…our Fall/Winter issue of "Inside Surgery!" Read about alum Dr. Martin Dib, Dr. Gabriel Brat’s research, faculty & trainee achievements, new clinical services, our newest research fellows, & much more! bit.ly/3CXxbM0 @BIDMChealth

Looking forward to this conversation with friends and colleagues. I’m always ready to evangelize about the bright future of data science in surgery.

In their new blog, @Jayson_Marwaha, @jhalamka, @bratogram, + William J. Gordon from @BIDMChealth, @MayoClinic, + @harvardmed discuss why administrative + billing codes are foundations upon which data is generated + used to inform public health decisions. bit.ly/3kP8UkG

An insightful piece. We should always be asking the “Hamming question—‘what are the most important problems in your field, and why aren’t you working on them?’” Researchers and Founders blog.samaltman.com/researchers-an…

@c3K @zakkohane @dunemovie Didn’t people say that about the 1984 version?

@zakkohane @dunemovie Looks like an amazing cast, excellent director, strong art design, and a budget to do the books justice. Hopeful🤞

Sand and worms. The still could end up disappointing but the trailer is promising @dunemovie bit.ly/3kHgNsI #ButlerianJihad

🏥🚨New surgical opioids preprint out - targeted learning quantile regression with EHR predictors to estimate nonresponse-adjusted procedure-specific #opioid prescribing guidelines w/ @bratogram @Jayson_Marwaha @KortneyR10 et al. @BIDMCSurgery @HarvardDBMI - via @ildiazm’s method

The @surginformatics lab at @BIDMCSurgery has been busy. Excited to be presenting projects at @AmCollSurgeons #ACSCC21 and more at #QualitySafety21 on safe post-surgical opioid prescribing, surgical consent, and other topics! facs.org/clincon2021

@AndrewLBeam @kat_heller @tw_killian Variable meaning/use of diagnosis codes is why I focus on procedure codes for surgery (as well as being a surgeon). Surgery codes have their own problems, but they are consistently documented and are great interventions or outcomes. Less fumes when cold steel is required.

@kat_heller @tw_killian +1 to this Our mantra has recently been try to infer physiology as best you can instead of predicting exhaust fumes from the the healthcare system

This is one reason why I suggest moving away from the use of labels, particularly in medical settings.

We need to remember: We’re at war with a virus — not with each other.

@MaartenvSmeden We had the same thought so we’ve been building the technology. Covidclinical.net Automatically collected patient data + sharing among international group of experts = 4CE consortium. 7 countries, 300+ hospitals, hundreds of data scientists. @zakkohane @i2b2tranSMART

It makes me sad thinking about how many lives we might have saved if we had the technology to automatically record hospitalized COVID-19 patient data and the will to share that data with organized international groups of experts

To beat COVID-19, we need to collaborate across borders, languages, and different coding systems. Our 4CE consortium of (now) 7 countries just published 1st efforts. Inspiring that >100 scientists/clinicians worked so hard with good will and trust. nature.com/articles/s4174…