Rohan Ramakrishna, MD on LinkedIn: That Message From Your Doctor? It May Have Been Drafted by A.I. | 15 comments (2024)

Rohan Ramakrishna, MD

Neurosurgeon and Co-Founder at Roon

  • Report this post

Do clinicians & patients want to see AI-generated content in their Epic MyChart inbox? The Times recently reported on the use of AI in managing Epic MyChart Inbox Messages. So I became curious and decided to analyze the comments from readers – both patients and providers - to see how they were reacting. The result? AI in healthcare messaging is not welcome, at least not yet! Over 70% of commenters expressed deep concerns, ranging from privacy to the lack of humans in the loop. Here’s what I learned:1. Transparency is Essential: Several health systems do not seem to be disclosing the use of AI in these responses, per reporting on the topic. Most commenters indicated they wouldn’t trust AI-generated replies. "I don’t message a doctor to have some AI reply to my messages,” one person wrote. “How do we know that doctors are even reading them?"2. AI does not save doctor's much time in this setting: Given the volume of messages and the need to review the AI-generated content, doctors are not seeing much time savings. Most doctors in the comments felt MyChart’s messaging overload added hours to their already packed days. "After a 10-hour clinic day, we stay another 2 hours to work through MyChart messages,"My POV - if the AI doesn’t give you at least a 50% productivity gain, you will not care or take the time to implement it. 3. AI Mistakes Can Be dangerous: The implementation of AI in MyChart presupposes a human in the loop. This process, however, is highly susceptible to automation bias, where people will just trust the AI rather than provide detailed oversight. A study cited in the article found that 7% of AI-generated responses posed severe risks if not corrected. Even if it was 0.01%, this is a huge problem given the scale of messaging. "AI assured me I had my hepatitis shots,” another commenter wrote. “This was completely false." This brings up another question: What level of error are we willing to accept in AI, when it comes to health? People are forgiving of doctors who make honest errors; With AI, I doubt it.4. The Human Touch Matters: According to the article, AI is edging into doctor-patient relationships. Most commenters expressed concerns that AI erodes the personal connection between doctors and patients. "Even if AI were flawless, do you want to automate one of the few ways we’re still interacting with each other?"5. Structural Reform: A majority of commenters pointed out that the real issue lies in the broader healthcare system, including consolidation, underfunding, and unsustainable workloads. MyChart messaging amplifies and highlights the existing access problem in healthcare. We have infinite demand for health information and services, but a limited supply of experts. Seems like the only solutions are increasing the supply of doctors(Hard) and/or reducing/diverting the demand for health related queries(easier?).What’s your take? Would love to know. Drop me a comment of your own. https://lnkd.in/eu7y6u66

That Message From Your Doctor? It May Have Been Drafted by A.I. https://www.nytimes.com

50

15 Comments

Like Comment

Sarah M Worthy

CEO and Founder at Door Space Inc.

9h

  • Report this comment

I don't want Epic to have any of my patient data. They have proven over and over they will sell our data, not secure it

Like Reply

1Reaction

Spencer Dorn

Vice Chair & Professor of Medicine, UNC | Balanced healthcare perspectives

1d

  • Report this comment

Nice Rohan. Though the NYT article is underwhelming, I agree with much of your analysis. Applying AI to write drafts turns us into editors rather than writers. Yet most of us write faster than we can edit. The UCSD study showed that family docs who used AI spent 20% LONGER responding to messages.Also, although there are common patterns and themes, patients' messages are usually pretty idiosyncratic. So, many AI drafts are not helpful and go unused.Finally, as you wrote, most people want to hear from their own physician (or care team) -- not a bot (or a doctor who sounds like a bot).There's value to using AI here, but we've been applying it to the wrong part of the process. It's better to use AI to (a) ensure the message is routed to the right person/teams (often not the physician), (b) to label the message (to help teams prioritize and distribute the work), and (c) to provide context about the patient and their needs (rather than having to review the EHR for it). Then let a human write the actual response.

Like Reply

12Reactions 13Reactions

Quinn Shean

Tusk Ventures

19h

  • Report this comment

Good points. Re: need for transparency.... we need to be judicious about where we are mandating such disclosures so as to truly alert patients when they should be aware (avoid consent fatigue) and avoid unnecessary mistrust, particularly as the technologies progress where they yield time-savings and require less editing.If the response is automatically generated and sent, or there is real-time chatbot communication, perhaps the disclosure is warranted. But should transparency be necessary where AI help generate an answer if the provider reviews it and/orrevises it? Do we require a disclaimer on patientcommunications when ascribe was usedor if another professional generated a first draft? Adisclaimer suggests somehow the provider did not support the content or that a patient should be more skeptical than had the provider drafted without any AI tools.

Like Reply

1Reaction

Adam Carewe MD

Virtualist | CMO Heidi | Fractional C-Suite | VC Advisory | Angel Investor | Emerging Technologies | Digital | AI | Board Certified in Clinical Informatics & Family Medicine

1d

  • Report this comment

What if patients could get their answers from AI without engaging their personal doctor? What if that AI was an extension of their doctor? Spoke like their doctor? Acted just like the doctor the patient knows so well. We need AI to help solve the existential problem coming: Way more demand with an aging population coupled with a physician pipeline that is not growing (and a staggering number of medical students who don’t even want to touch a patient in direct patient care). We need AI solutions to combat this problem. We spend too much time talking about the current state. Why is doesn’t work now. Let’s work towards the future state. And I stand by my conviction that AI will nearly replace everything what a doctor does today in a matter of years. That includes all doctors eventually in time. We will get there and legacy fee for volume healthcare needs to go away with it to reach this reality. Value based AI enabled care is the way.

Like Reply

1Reaction 2Reactions

Alice Chen

National physician leader in health advocacy and public health

1d

  • Report this comment

My experience is that patients send a message to their doctor because they want the reassurance and counsel of a person who knows them. When I write to a patient about their cholesterol results, I'm taking into account the conversation we had about their new resolve eat healthier or how they are too tired at the end of a long workday as a cleaner to exercise or how worried they are because their mother recently had a stroke. When a patient asks several questions about a referral, it signals to me that they are really concerned about it, so I know to bring it up proactively when I next see them in person. There may be some simple messages that AI can respond to, but so much of our communication as doctors and patients is innately human and relational and trust-building that I wouldn't feel comfortable outsourcing much if any to a machine.

Like Reply

4Reactions 5Reactions

Wesley S

CEO @ Proteotype | Dad | On a Mission

1d

  • Report this comment

I guess the only acceptable limit would be AI helping doctors write at the speed of thought.

Like Reply

1Reaction

Dennis Deruelle, MD, FHM

Chief Medical Officer at Mendota Health

1d

  • Report this comment

Humans are not doing a good job of helping patients and could use the right kind of help Just call a doctors office or use your portal to navigate an active issue and your will see the contact hours it takes to accurately get something doneWe need help from AI because we are not doing a good job currently I am a stanch supporter in transparency about when and where and how AI is being used however Patients will welcome AI in concert with their doctors if it leads to an improvement

Like Reply

2Reactions 3Reactions

Madhubala Radhakrishnan

Founder mCURA

1d

  • Report this comment

Fully Agree .. it erodes the personal connection.

Like Reply

1Reaction

See more comments

To view or add a comment, sign in

More Relevant Posts

  • Rohan Ramakrishna, MD

    Neurosurgeon and Co-Founder at Roon

    • Report this post

    🚨 New Edition of Good Medicine!Our latest edition of Good Medicine dives into an essential and timely topic—truth in medicine—with none other than Dr. Vin Gupta, MD MPA , Chief Medical Officer at Amazon Pharmacy and a leading medical contributor at NBC.In a world flooded with misinformation, especially in healthcare, Dr. Gupta shares his unique journey from the frontlines of military medicine to the forefront of medical journalism. From treating soldiers in challenging environments to combating false health narratives on national television, he offers lessons we can all apply—whether we're doctors, patients, or simply seekers of truth.👉How can healthcare professionals develop a trustworthy voice that patients want to hear? Check out the latest Good Medicine to hear Dr. Gupta’s take. 🩺💡https://lnkd.in/g4dSHuvPThanks as always to Christina Farr, Joshua Lang, and Naintara (Tara) Rajan!

    A conversation with Dr. Vin Gupta rohanramakrishna.substack.com

    29

    1 Comment

    Like Comment

    To view or add a comment, sign in

  • Rohan Ramakrishna, MD

    Neurosurgeon and Co-Founder at Roon

    • Report this post

    As a surgeon, writing has become such an unexpectedly fun part of my career. Which is why I’ve been so excited to be the author of Good Medicine. If you’re not already subscribed, sign up today for interviews with some of the best clinical minds. Our latest interview is with Dr. Alice Chen , who spoke with me about the importance of the physician advocate. We have more of a voice than we think; we just need to find the right ways to harness it. Alice has bridged the medicine and policy divide, and has advice for those who are looking to do the same. You can check out the article in the link below. Special Thanks to Christina Farr, Joshua Lang, and Naintara (Tara) Rajan for their invaluable assistance.

    A conversation with Dr. Alice Chen rohanramakrishna.substack.com

    37

    2 Comments

    Like Comment

    To view or add a comment, sign in

  • Rohan Ramakrishna, MD

    Neurosurgeon and Co-Founder at Roon

    • Report this post

    Good Medicine has been a wonderful excuse to talk to my peers and colleagues about not just the work they're doing now, but also what got them to this point. Through my latest interview with Maven CMONeel Shah, I learned that we had a lot in common in terms of what brought us into medicine in the first place. From there, Neel developed a strong passion for women's health, which he views as fundamentally underserved, and in really thinking deeply about how clinicians should be supporting their patients at such vulnerable times of their lives. As he told me, it's really about three things: Competence, empathy, and trustworthiness. We have to do what we say we'll do - and do it well - and we need to take the time to listen. At Roon, we just launched into a new area of fertility, so this discussion felt particularly top of mind for us. Read our full conversation here:https://lnkd.in/eUrNzQ7s

    A conversation with Dr. Neel Shah rohanramakrishna.substack.com

    38

    2 Comments

    Like Comment

    To view or add a comment, sign in

  • Rohan Ramakrishna, MD

    Neurosurgeon and Co-Founder at Roon

    • Report this post

    Thank you Tara Lipinski for being on Team Roon and sharing your Story with PEOPLE Magazine | PEOPLE.com! https://lnkd.in/gxSvvTq6

    16

    Like Comment

    To view or add a comment, sign in

  • Rohan Ramakrishna, MD

    Neurosurgeon and Co-Founder at Roon

    • Report this post

    There's been a lot of buzz about how AI is more empathetic than doctors. But this notion oversimplifies a deeper issue. It’s not that our doctors lack the desire to connect; it’s that we’ve burdened them with relentless documentation and administrative tasks. At Roon, we've seen first hand how compelling and empathic MDs are when given the time and space to be themselves. Jennifer Lycette, MD makes this powerful and important argument in Stat Newshttps://lnkd.in/e-fisRDn

    Why a chatbot might seem more empathetic than a human physician https://www.statnews.com

    43

    2 Comments

    Like Comment

    To view or add a comment, sign in

  • Rohan Ramakrishna, MD

    Neurosurgeon and Co-Founder at Roon

    • Report this post

    Big News: I’m so proud to announce the launch of Roon’s newest condition: Roon Fertility!This is especially personal to me as IVF was key to building my own family. Even with a medical background, the experience was at times confusing and isolating. I remember being in those waiting rooms with so many other aspiring parents. The uncertainty, the fears…I think anyone who goes through this process will tell you how tough it can be.It turns out this journey is really common, with fertility issues affecting one in six adults worldwide. And so at Roon, we wanted to create a curated, free resource where you can get expert answers to all your questions about the fertility journey, in the form of short-form easy to consume video.This was built by over 100 experts from specialties like Reproductive Endocrinology, OBGYN, Mental Health, Urology, Integrative Medicine and more. You can explore a wide range of topics, from medical procedures like Egg Freezing, IUI, and IVF, to Mental Health, Financial & Insurance considerations, Diet & Nutrition, Surrogacy and more. And, you’ll hear from people who have actually gone through these journeys like Olympic Gold Medalist Tara Lipinski and Girls Who Code CEO Reshma Saujani. Combined With experts from esteemed institutions like Johns Hopkins, Harvard, Northwestern, Weill Cornell, Yale, UCSF and Columbia, as well as leading fertility clinics, this is the best, most comprehensive resource ever made for navigating the fertility journey. And we’re just getting started. If you're on your fertility journey, visit us at roon.com or find us on the iOS App Store(we also support GBM, ALS and Dementia). We've been there, and we're excited to make navigating fertility easier for you.

    • Rohan Ramakrishna, MD on LinkedIn: That Message From Your Doctor? It May Have Been Drafted by A.I. | 15 comments (38)

    93

    7 Comments

    Like Comment

    To view or add a comment, sign in

  • Rohan Ramakrishna, MD

    Neurosurgeon and Co-Founder at Roon

    • Report this post

    Amazing new role at Roon. Check it out !

    17

    1 Comment

    Like Comment

    To view or add a comment, sign in

  • Rohan Ramakrishna, MD

    Neurosurgeon and Co-Founder at Roon

    • Report this post

    I'm so thrilled to announce the launch of my newsletter, Good Medicine! 🩺 As doctors, we live in a world with a cascading set of challenges. And so with this monthly publication, I hope to shine a bright light on how the attributes that make medicine great impact healthcare today, as told through the perspective of some of its leading minds. For our first post, I had the opportunity to interview Dr. Robert Wachter, Chairman of Medicine at University of California, San Francisco (UCSF). We discussed his role as a leading voice in the public sphere, medical misinformation, and the perils and promise of AI. You can find this free post here. https://lnkd.in/e4ctRsiBA huge shoutout to Christina Farr for her invaluable support in bringing this to life, and to my Co-Founder and co-interviewer at Roon, Vikram Bhaskaran.Hope you enjoy it! Who should be featured next?

    • Rohan Ramakrishna, MD on LinkedIn: That Message From Your Doctor? It May Have Been Drafted by A.I. | 15 comments (47)

    48

    5 Comments

    Like Comment

    To view or add a comment, sign in

Rohan Ramakrishna, MD on LinkedIn: That Message From Your Doctor? It May Have Been Drafted by A.I. | 15 comments (51)

Rohan Ramakrishna, MD on LinkedIn: That Message From Your Doctor? It May Have Been Drafted by A.I. | 15 comments (52)

3,105 followers

  • 39 Posts
  • 1 Article

View Profile

Follow

More from this author

  • #HLTH Conference 2023 Roundup! Rohan Ramakrishna, MD 11mo

Explore topics

  • Sales
  • Marketing
  • IT Services
  • Business Administration
  • HR Management
  • Engineering
  • Soft Skills
  • See All
Rohan Ramakrishna, MD on LinkedIn: That Message From Your Doctor? It May Have Been Drafted by A.I. | 15 comments (2024)
Top Articles
Latest Posts
Recommended Articles
Article information

Author: Zonia Mosciski DO

Last Updated:

Views: 6207

Rating: 4 / 5 (51 voted)

Reviews: 90% of readers found this page helpful

Author information

Name: Zonia Mosciski DO

Birthday: 1996-05-16

Address: Suite 228 919 Deana Ford, Lake Meridithberg, NE 60017-4257

Phone: +2613987384138

Job: Chief Retail Officer

Hobby: Tai chi, Dowsing, Poi, Letterboxing, Watching movies, Video gaming, Singing

Introduction: My name is Zonia Mosciski DO, I am a enchanting, joyous, lovely, successful, hilarious, tender, outstanding person who loves writing and wants to share my knowledge and understanding with you.