Paulius Mui, MD

@curious_vii @firefliesai @AnthropicAI while I am excited, I am also reminded that best tech doesn't win in healthcare. distribution rules: who > how. also you should charge your phone 🪫

This may literally be the biggest opportunity in health tech right now. Buy a telehealth practice, give all the providers an AI notetaker & Claude Pro after signing BAA's with @firefliesai & @AnthropicAI, and I'd bet you could see transformative results in less than a quarter. 👀 @pauliusmui @ryanstellar

@RebeccaCoelius @DrDeepMD @DrDeepMD all of this is a symptom of a larger issue in healthcare that we give ourselves permission to ignore. i'm still an optimist that the right people in leadership can drive much needed change

I am biased from the start because I worked for Farzad years ago and he’s one of the most mission driven and competent (esp in the rules of CMS) leaders I have ever met, as are his major investors. That aside: - coding for true complexity is core to the entire model working. Getting upset that we haven’t historically been fully documenting health issues or paying primary care properly, and now that we are everyone looks a lot sicker, is not the same thing as fraud. - you had better be doing something different and better based on the new and appropriately coded patient populations. Aledade is one of the few organizations that has shown the ability to help practices transform to do this at scale, not just play more financial subdelegation games and ACO purchases like others…

𝑾𝑯𝑰𝑺𝑻𝑳𝑬𝑩𝑳𝑶𝑾𝑬𝑹 𝑬𝑿𝑷𝑶𝑺𝑬𝑺 𝑹𝑨𝑴𝑷𝑨𝑵𝑻 𝑴𝑬𝑫𝑰𝑪𝑨𝑹𝑬 𝑭𝑹𝑨𝑼𝑫 Im really interested to see how this plays out, could have major ramifications to the entire VBC space. Any of us in the know, have seen the “gray zone” practices - almost all the “value-based care” biz models run on boosting denominator revenue as much as possible. I generally have a lot of respect for Aledade and Farzad, but I do worry this is the canary in the coal mine and more docs/insiders will come forward with examples of less than savory tactics in risk adjustment. It’s the only way to actually win with risk-based contracting. Appropriate coding and correcting under coding is one thing. Actively chasing and boosting suspect codes without really providing more care is another. I think all the major players are massively exposed here. The hard work of actually reducing cost of care by taking better care of patients is a lot more difficult than just pushing docs and using AI to code the hell out of everyone. @CMSGov is starting to take note and will crack down. We’ve already seen a big push by moving from v24 ➡️ v28. If one of these rulings goes in favor of doc vs MSO, it will kick off a cascade of others. I think the reckoning is coming,..

welcome to the #pinksocks tribe @pauliusmui! the world is full of good! when u believe it, u see it. keep doing that!🌍💖😊 #HLTH2023 ✨ #TX

People like to talk about the importance of "vision," but ultimately running a startup is like doing science: you have to follow the truth wherever it leads. Vision is only valuable insofar as it predicts where you'll find truth.

Check out @DevotedHealth wonderful approach to primary care. It is all based on solving for the "the one thing." It is simple and works better. A Clinical Pathway to Well-Being: Putting Patient Priorities at the Center of Care ajmc.com/view/a-clinica… via @AJMC

Dr. @oziadias argues that the best work in cross-disciplinary fields such as medical machine learning often arises when a single person acquires a deep understanding of both disciplines as opposed to two experts in separate fields being brought together. nejm.ai/ep8

Full body MRI screening is currently a "thing" thanks to Kim Kardashian. From both a societal and individual level, FB-MRI is problematic. To understand why, consider the "epidemic" of thyroid cancer in Korea, beginning in 1999. nejm.org/doi/full/10.10… (1/10)

Shopify's product team has a neat framework for clarifying responsibilities on projects. It's called the AAA framework. Each team member is assigned one of three roles: 1. Aiming: responsible for strategy and direction of what we are building 2. Assembling: responsible for bringing the right people together and keeping them on track 3. Achieving: responsible for the day-to-day work of getting shit done (GSD) like design, code, etc. As @glencoates shared with me, "I think most companies think of hierarchy and jobs basically as this sort of single line of leadership downward, based on how senior you are. But a few years ago we introduced the AAA framework. You don’t want to take on a leadership job and suddenly be responsible for aiming when you’re really passionate about assembly, or the other way around. It’s helped us put people in the right roles and not just have one dimension of leadership that everyone has to conform to. On projects, it’s really good to be able to know who’s the aimer, responsible for the strategy. It may not be the most senior person in the room; it may be the staff designer or the staff engineer. It’s sometimes good to know that the team’s manager actually isn’t the aimer, which is by design. Then there’s usually dedicated product ops and program management people who are responsible for the assembly. And of course, the achievers, who actually do the work." Read more about @Shopify's product development process in the thread below...

@dvansickle @PeterAttiaMD help us get there!

If my primary care doctor is any indication, we’re not prepared for the era of @PeterAttiaMD and the emerging strategies to quantify and lower long-term individual risk

📄 20 studies that Internal Medicine Residents should read in the first year 👇 1/ 🏃‍♀️ RACE II trial: Lenient resting HR < 110 was non-inferior to a strict HR < 80 in patients with atrial fibrillation. 🧵

product-led growth is hard to come by in healthcare, with the divine hammer of privacy regulation waiting to smite you at almost every step. i thought through the question: where is it possible, if at all, to do bottoms-up distribution?

I administered the Montreal Cognitive Assessment (MoCA) to the image-generating AI program DALL-E this morning. Thing really went off the rails. Let's see how it did. 🧵

We @OscarHealth are publishing our use cases, prototypes, ideas & research notes on using LLMs in healthcare. AI research moves too fast for companies to do this work behind closed doors, so we’re sharing ours for others to read & comment: oscarai.notion.site. 1/

Opportunities for AI & Healthcare classes and resources 👩🏻‍⚕️👩🏻‍💻🤖 cc: @pauliusmui linkedin.com/posts/pauliusm…

@rikrenard is the Hemingway of health tech, sassy and insightful. read his stuff 👏

An overlooked aspect of AI that I try to point out to corporate leaders is that a vast pool of talent just came online as both competitors & collaborators Folks in 169 countries have GPT-4 for free via Bing Creative Mode. They can now write flawlessly, get code help, learn, etc.

@_btlee_ @B_Madden4 let the leapfrogging begin. honestly exciting times!! and hopefully care teams and patients will win in the end

more thoughts on this in this collab with @B_Madden4 @pauliusmui here quiteafewclaims.substack.com/p/ai-has-enter…

1/ With AI deployment being easy, full-stack providers like @CarbonHealth can perform: - deeper integration of AI-enabled features compared to vendors (ex. in app vs plugin; fully utilize preexisting in house EHR data) - dogfooding + potential commercialization of products

We are happy to announce our participation in the upcoming Hackathon event where we will proudly serve as the API backbone from June 9th to June 11th. Shout out to @pauliusmui for organizing! 📅 Applications are due June 1st for teams hackathon.xprimarycare.com

Graphic designers are history. AI creates beautiful logos in seconds. Here's how to create any logo using AI (for free):

👀 Ready to hack Nabla? 💡 Participants in the Healthcare + AI Hackathon with XPrimary Care will be able to use Nabla Copilot to bring to life tools to revolutionize clinical workflows. 🚀 We can’t wait to see what they come up with.