Mintu Turakhia, MD MS

#EPeeps I isolate PE in PersAf when:

Excited to do this with some great leaders at #ACC2026. Thank you, @Kpakanati3 and @ACCinTouch.

@EJSMD @ftrae @SergioPinski @drjohnm In the US, we have conflated “nice to have” with “must have.” Essentialism requires only catheters.

@SergioPinski @drjohnm ICE makes more sense to me than 3D map. Suspect one main issue is lab dependence on mappers to do any ablation. Before I gave it up, I watched ablation costs skyrocket due to unnecessary mapping and regulatory barriers to catheter reuse. All this from industry influence.

How far tech has come. I remember ablating CTI in 1995 w a 4 mm non irrigated catheter and…gulp… no 3d mapping

Delays in Orlando due to storms. Interestingly, Dubai still listed as the only flight on time.

You can’t generalize whether you should categorically not switch patients from warfarin to DOACs from these studies. Here are the the most important factors: 1. Time in INR therapeutic range (TTR) and time above INR range. This remains the gold standard process measure linked to outcomes and why specialized warfarin clinics save lives. I don’t see TTR used here. 2. DOAC dose-patient mismatch. GFR is incredibly important, and some DOACs with high renal elimination also have the longest half life and are twice daily (dabigatran). CKD is associated with worse time in INR range. Going from well-managed warfarin to DOAC overdosing is a problem. This can happen when GFR is close but has not hit a threshold for dose adjustment. Going from poorly-managed warfarin to appropriate DOAC dosing is rational based on the totality of safety evidence. 3. Concomitant aspirin. Should be stopped unless there is a very strong case not to. TL;DR - when dealing with edge cases, use pharmacology, math, and judgement to tailor to your patient rather than a measure of central tendency of observational studies of highly-selected patients that may not generalize to yours.

@MKIttlesonMD IMO, it’s less about being a direct endorsement and more about validating the narrative that the candidate offers in their own candidate statement. Promotion is harder when the candidate’s self-view differs from those in field.

So, if basically no one ever writes a negative letter for academic promotions, doesn't that make the recommendation letter component of the process performative and a waste of the letter writers' time?

@hvanspall 🙏

@leftbundle Thank you for your kindness, Mintu. And warm congratulations on your upcoming keynote at ACC.

He was a mentor But departing from evidence, he had my dad w new angina wait 4wks for CABG: Cath/IVUS had shown 70% LM, tight ostial LAD+LCx lesions, RCA occlusion My beloved dad died before CABG The ensuing silence left me w #lessons I carry everyday1/ acpjournals.org/doi/abs/10.732…

@hvanspall Thanks for sharing and resurfacing. I’m sorry for your loss and for having to live with the “what ifs”. It’s brave of you to share and important that people see it. ❤️‍🩹

The preventable #death of my dear dad & the disturbing #UnspokenCurriculum shaped me to - better listen to patients - pay attention to details - choose evidence over bravado - err on the side of caution - be truthful about mistakes - forgive #Lessons acpjournals.org/doi/abs/10.732…

Privileged to deliver the Dr. Louis F. Bishop Keynote at #ACC26 and share the stage with some absolute giants in the field. Thank you, @AmericanColl.

Congratulations @rohan_khera @CathieBiga @leftbundle & everyone👏👏👏👏👏 Looking fwd to your @ACCinTouch #ACC26 Keynotes & vision!

@nmarrouche I was registered and with tickets and hotel booked — but had to change plans for Family First. Big weekend for both kids. I’m so sorry to miss. Pics look amazing, and team iRhythm will still support and represent!

Why not at #wafib2026 @leftbundle !?

@DrJasonAndrade @netta_doc @nmarrouche @westernAFib @jhurwitz55 @deering_f @Nishaki1 @KTamirisaMD @EJSMD It was never meaningful. The method of Kaplan and Meier was developed for incomplete survival data in life-threatening disease — not 30 seconds of afib. It’s now used for any time-to-event variable because it shrinks sample size and trial costs.

@netta_doc @nmarrouche @westernAFib @jhurwitz55 @deering_f @Nishaki1 @KTamirisaMD @EJSMD Fair. But then you can’t rely on time to recurrence as an outcome as it becomes even more meaningless.

💥Starting by off his talk with a bang! “WE SHOULD JETTISON THE BLANKING PERIOD” No other trials have a blanking period! We wear 2 hats: clinician & scientist- We can use judgement, *and* we should count all recurrences. #WAFib2025

Who needs ECG on a watch when you can instead go to a bar and be master of the universe?

@ross_prager @venkmurthy In theory sure. But there is no reason to choose a cutoff that isn’t the 100/100 point, which is why it seems counterintuitive.

Here's a more full explanation. AUC looks at all cutoffs, not just one. Example test scores: Diseased patients: 71, 72, 73, 74, 75 Non-diseased patients: 10, 20, 30, 40, 50 Every diseased patient has a higher score than every non-diseased patient = perfect separation, so AUC = 1.0. Now pick a single cutoff of 74: Sensitivity = 3/5 = 60% (you miss the two lowest diseased scores) Specificity = 5/5 = 100% The ROC curve does reach the top-left corner (100% sensitivity and specificity) at a different cutoff (e.g., 71), which is why the AUC is 1.0. Sensitivity = 0.70 just reflects where you chose to operate on that curve, not the shape of the curve itself.

I've thought about these measures for decades and I don't understand how this can be true AUC of exactly 1.0 means sensitivity and specificity both must also be 1.0 exactly

After leading AI & Health at @UHN — Canada’s largest hospital network — for the past few years, this is the first time it genuinely feels different. A cardiologist just placed 3rd out of 13,000 at Anthropic’s hackathon. He built it between hospital shifts. In a week. An AI agent that follows patients home — reverse scribe, full history, devices, evidence — all in one place. A few years ago this would’ve taken a full team and months. Now? One doctor. That’s the shift.

Turn on audio. Another stellar post from @CHP_Truckee. But their real gift is for all they do to keep us safe. Thank you.

25-40% of clinical trial costs come from an excessive quality-check process that the FDA itself has recommended against for over a decade. Great piece from a former FDA official on why the whole system is stuck in a "too big to fail" loop. learninghealthadam.substack.com/p/why-clinical…

@ctreada @cljack Or you could build enterprise tools for the schools to create more effective, concise communication.

@cljack I've been building this watching parents struggle w/ the most stressful time of the day -- 6am! I would love your feedback! parentpoint.ai

I'm trying to train Claude to read the weekly emails from my kids school and reliably summarize them and print a list of action items. It is losing its damn mind and rapidly spiraling into madness. I feel vindicated

@cljack The solution is to for the schools (not the parents) to use AI to make their emails clear and concise. Or move to slack.