Jonah Feldman MD, FACP’s Post

AI will have the greatest positive impact in #medicine and #healthcare only if it is integrated alongside strong organizational behavior-based initiatives and through team-based quality improvement pathways. I was dubious that the AI goal of “better clinical note writing“ could be attained until reading this transformative study from NEJM Catalyst. #aiinmedicine

It’s not too late to book! We can’t wait to host our event ‘AI and how it fits into Healthcare’ next Wednesday at Banyan, Corn Exchange Manchester. You don’t just have to be in the medical sector to have an interest in this. It is very much going to be based on a personal level. Charlotte Lewis Mills & Reeve, Andre Fialho MySense.ai, Niels Peek The University of Manchester and Noura Al Moubayed Durham University Evergreen Life are having a very open and honest discussion about the benefits and indeed the pitfalls of AI. - How do they see the future of the Healthcare system with regards to AI? - The limitations - the evidence and the quality of data - What exactly is AI in Healthcare, it doesn’t necessarily mean artificial - Save money, save beds, save lives. Can it do this? Pop onto the link below to book. See you there! https://lnkd.in/dYKAhk4Q

I recently came across this insightful article and found the findings fascinating. It outlines the results of a comprehensive survey conducted by the American Medical Association, showcasing how 65% of physicians acknowledge the significant advantages of AI tools in non-clinical settings. These findings align with our ongoing efforts at Bupa to integrate AI and digital solutions across various aspects of healthcare, beyond just clinical applications. The potential for AI to optimise scheduling, enhance patient experiences, streamline operations, and even support education and research is incredibly promising. As we continue to explore and adopt AI within Bupa, particularly through our digital health offering, Blua, it's encouraging to see the broader medical community's enthusiasm for such technologies. These advancements not only promise to improve operational efficiency but also to make healthcare more personalised and accessible to everyone. #DigitalHealth #Innovation #AritificialIntelligence

And as f/u to the Health Misinformation Post: A new survey shows physicians are optimistic about using generative AI in healthcare clinical decision-making and patient care, BUT HAVING A TRUSTED CONTENT SOURCE IS KEY TO ADOPTION. Key findings include: 91% of physicians would have to know the GenAI SOURCE MATERIALS were created by doctors and medical experts. Source TRANSPARENCY also matters. 89% of physicians said they would be more likely to use GenAI in clinical decision- making if the technology vendor was transparent about where the information came from, who created it, and how it was sourced. 54% of physicians believe GenAI will save them 20% or more time looking for data when making clinical decisions. 81% of physicians say GenAI can improve care team interactions with patients. To be successful, the AI industry and healthcare leadership would benefit from taking physician input into consideration when assessing GenAI solutions for their care teams. The foundation of any new technology and its successful implementation is being able to trust the information sources and the vendor. 

Potential of AI in healthcare

𝐀𝐈 𝐢𝐧 𝐇𝐞𝐚𝐥𝐭𝐡𝐜𝐚𝐫𝐞: 𝐏𝐫𝐨𝐦𝐢𝐬𝐞 𝐚𝐧𝐝 𝐂𝐡𝐚𝐥𝐥𝐞𝐧𝐠𝐞𝐬 [𝐔𝐒 𝐆𝐀𝐎 𝐑𝐞𝐩𝐨𝐫𝐭] The future of medicine is increasingly intertwined with artificial intelligence (AI). A 2022 report by the 𝐔𝐒 𝐆𝐨𝐯𝐞𝐫𝐧𝐦𝐞𝐧𝐭 𝐀𝐜𝐜𝐨𝐮𝐧𝐭𝐚𝐛𝐢𝐥𝐢𝐭𝐲 𝐎𝐟𝐟𝐢𝐜𝐞 (𝐆𝐀𝐎) explores the exciting potential of machine learning for medical diagnostics, but also highlights some key challenges to consider. 𝐇𝐞𝐫𝐞 𝐚𝐫𝐞 5 𝐤𝐞𝐲 𝐭𝐚𝐤𝐞𝐚𝐰𝐚𝐲𝐬 𝐟𝐫𝐨𝐦 𝐭𝐡𝐞 𝐆𝐀𝐎 𝐫𝐞𝐩𝐨𝐫𝐭: 𝐀𝐈 𝐟𝐨𝐫 𝐈𝐦𝐩𝐫𝐨𝐯𝐞𝐝 𝐂𝐚𝐫𝐞: Machine learning tools are being developed to enhance patient care, potentially leading to better diagnoses and treatment plans. 𝐁𝐞𝐲𝐨𝐧𝐝 𝐃𝐢𝐚𝐠𝐧𝐨𝐬𝐭𝐢𝐜𝐬: While the report focuses on diagnostics, AI has the potential to streamline administrative tasks and optimize workflows within healthcare institutions. 𝐂𝐡𝐚𝐥𝐥𝐞𝐧𝐠𝐞𝐬 𝐭𝐨 𝐎𝐯𝐞𝐫𝐜𝐨𝐦𝐞: Despite the promise, hurdles exist regarding product development, data quality, and integrating AI into clinical practice. 𝐓𝐡𝐞 𝐍𝐞𝐞𝐝 𝐟𝐨𝐫 𝐏𝐨𝐥𝐢𝐜𝐲: Regulatory frameworks by the FDA and other agencies might be needed to ensure the safe and effective use of AI in medical technologies. 𝐎𝐧𝐠𝐨𝐢𝐧𝐠 𝐆𝐨𝐯𝐞𝐫𝐧𝐦𝐞𝐧𝐭 𝐈𝐧𝐭𝐞𝐫𝐞𝐬𝐭: This GAO report is part of a series on AI in healthcare, demonstrating the government's commitment to understanding and shaping this evolving field. This report offers valuable insights for healthcare professionals, policymakers, and anyone interested in the future of medicine. Authors and Collaborators: Julia Adler-Milstein Nakul Aggarwal Mahnoor Ahmed Jessica Castner Andrew Gonzalez Steven Lin Kenneth Mandl Michael Matheny Hugo Aerts Pat Baird Barbara Evans Richard Frank, MD, PhD Maryellen Giger Kenneth Goodman John Halamka, M.D., M.S. Eric Horvitz Mia Levy Anil Parwani Lily Peng Bruce Pyenson Berkman Sahiner Robert Sparrow Eric Topol, MD Hayden Y.-T. Huang Matthew Hunter Jehan Chase Jonathan Felbinger Nathan Hanks Eric Larson Eric Lee Anika McMillon Yesook Merrill Ben Shouse Michael Steinberg George Bogart Robert Copeland Leia Dickerson Kaitlin Farquharson Charlotte Hinkle Monica Perez-Nelson #AIinHealthcare #MachineLearning #MedicalDiagnostics #USGAO

While the potential impact of #AI in medicine has been long discussed, real-life, clinician-facing applications of AI have only recently become a reality. AI-assisted chronic disease management, diagnostic support, and administrative work (such as documentation, billing, and patient messaging) have significant potential to improve medicine and to take some burden off physicians allowing them to focus on physician-level #patientcare The use of #clinicalai is part of a broader shift in medicine toward “#digitalhealth” where many aspects of medical care are conducted remotely, mediated by a technological intermediary, leading to potential improvements in efficiency and access. These developments stand to make a substantial impact in primary care, a field that is currently grappling with high rates of physician burnout, inadequate compensation, and a growing shortage of physicians. However, there is concern that if AI is poorly integrated it could exacerbate the “disconnect between professional values and the realities of primary care practice”. Despite so much on the line, there is limited literature on PCP views toward AI. Much of the research that does exist has taken place in a purely theoretical context exploring AI in general terms with physicians that did not have experience using AI-powered systems. This paper on "Navigating the doctor-patient-AI relationship - a mixed-methods study of physician attitudes toward artificial intelligence in primary care" is valuable for its nuanced insights into PCP attitudes towards AI in primary care, highlighting the importance of addressing both technological and people-and-process concerns for effective AI integration in healthcare. By highlighting specific #aiusecases, the paper highlights new concerns and attitudes that would remain hidden when discussing AI in general. Failure to engage end users in the design of #aipowered #digitalhealthtools leads to inefficient or unsuccessful integration of these tools into clinical workflow, leading to added clinician burnout and even patient harm. Link to the paper: https://buff.ly/3TCRfxS

Exploring the Intersection of AI and Resident Physicians: Shaping the Future of Medicine Artificial Intelligence (AI) is fundamentally transforming the landscape of healthcare, marking significant strides in how we approach medicine today. But what does this mean for resident physicians, the future vanguards of medicine? As a soon to graduate resident physician myself, I see AI not as a replacement for my knowledge and acumen but as a powerful ally. It offers tools that enhance diagnostic accuracy, personalize treatment plans, and even predict patient outcomes. However, the true potential of AI in healthcare hinges on how well it's integrated into our training and practice. Resident physicians must be at the forefront of this integration. We need to understand AI's capabilities, limitations, and ethical implications. By doing so, we can ensure AI is used to augment our clinical skills, not undermine them. There’s more to the conversation than just developing a working understanding of AI; We must actively participate in its development - providing critical feedback, contributing to data interpretation, and ensuring its alignment with the patient-centered care we plan to practice as the future of medicine. So as we step into this brave new world of AI powered healthcare let’s seize the opportunity to ensure AI is a partner that helps us deliver exceptional, personalized, and evidence based patient-centered care. The future of medicine is not just about technology; it's about how we, as physicians, harness it. #ArtificialIntelligence #ResidentPhysicians #FutureOfMedicine #HealthcareInnovation

🌟 Med-Gemini: Pioneering AI in Medicine Exciting news from Google Research and DeepMind! Their recent publication on Med-Gemini introduces a highly specialized suite of AI models designed to revolutionize medical diagnostics. This paper, backed by the formidable prowess of Google’s technological ecosystem, marks a significant leap forward in the application of AI within healthcare. ✨ Highlights: • Top-Tier Performance: Achieving state-of-the-art results in 10 of 14 medical benchmarks, Med-Gemini showcases exceptional capability, particularly in understanding and integrating multimodal data for comprehensive analysis. • Practical Utility: Beyond theoretical metrics, Med-Gemini extends its utility to real-world applications such as generating medical summaries and assisting in complex diagnostic scenarios, potentially enhancing both patient care and administrative efficiency. Yet, the journey from a controlled research environment to clinical settings is fraught with hurdles. The paper openly discusses the need for extensive real-world testing and validation to ensure safety and reliability—critical in a field where decisions impact lives directly. ⚠️ Concerns to Address: •Real-World Application: How will Med-Gemini perform outside lab settings? • Ethical Implications: The integration of AI in healthcare raises significant ethical and privacy concerns that must be carefully navigated. While Med-Gemini represents a promising advance in medical technology, its integration into healthcare systems must be handled with care to truly benefit patient outcomes. With Google's track record, there's much to be hopeful about, yet equally, a cause for rigorous scrutiny. 🔗 Read the full paper: https://lnkd.in/e3DQuAVa #MedGemini #HealthcareAI #DigitalHealth #EthicalAI

This research evaluated Generative AI to enhance patient comprehension by transforming medical discharge summaries into more approachable language. 🔍 Using OpenAI, the research focused on the understandability and readability of inpatient discharge summaries from NYU Langone Health, leading to patient-friendly versions of these critical documents (PEMAT). 🧬 This bridges the gap between medical jargon and patient understanding, crucial for active involvement in one's health care decisions and improved outcomes. ♻️ Yet, the study advocates for cautious implementation due to safety concerns (Readability scores were higher across the board but 18/50 summaries were factually incorrect or incomplete). Thus, physician oversight is recommended as a step toward integrating generative AI with patient safety at the forefront. "𝐓𝐡𝐞 𝐚𝐩𝐩𝐫𝐨𝐚𝐜𝐡 𝐝𝐞𝐦𝐨𝐧𝐬𝐭𝐫𝐚𝐭𝐞𝐬 𝐡𝐢𝐠𝐡𝐞𝐫 𝐮𝐧𝐝𝐞𝐫𝐬𝐭𝐚𝐧𝐝𝐚𝐛𝐢𝐥𝐢𝐭𝐲 𝐬𝐜𝐨𝐫𝐞𝐬 𝐚𝐧𝐝 𝐥𝐨𝐰𝐞𝐫 𝐠𝐫𝐚𝐝𝐞 𝐥𝐞𝐯𝐞𝐥𝐬 𝐢𝐧 𝐀𝐈-𝐭𝐫𝐚𝐧𝐬𝐟𝐨𝐫𝐦𝐞𝐝 𝐬𝐮𝐦𝐦𝐚𝐫𝐢𝐞𝐬, 𝐦𝐚𝐫𝐤𝐢𝐧𝐠 𝐚𝐧 𝐚𝐝𝐯𝐚𝐧𝐜𝐞 𝐢𝐧 𝐩𝐚𝐭𝐢𝐞𝐧𝐭 𝐜𝐨𝐦𝐦𝐮𝐧𝐢𝐜𝐚𝐭𝐢𝐨𝐧 𝐰𝐢𝐭𝐡𝐢𝐧 𝐡𝐞𝐚𝐥𝐭𝐡𝐜𝐚𝐫𝐞 𝐬𝐞𝐭𝐭𝐢𝐧𝐠𝐬." - 𝐙𝐚𝐫𝐞𝐭𝐬𝐤𝐲 𝐞𝐭 𝐚𝐥. #healthinformatics #artificialintelligence #patientcare #medicalai #healthtech