National Conference of State Legislatures’ Post

Simplify medical information for your patients. It's not just a courtesy, in today's digital age, it's a necessity. And in today’s digital age, you need Digital Health AI solutions from DeliverHealth. Learn more about partnering with us: https://bit.ly/DHAIdig #patientexperience #patientcare #AI #digitalhealth

Healthcare professionals have access to AI tools and these systems are playing a more prominent role in everyday patient interactions and decision making. The article argues that governments are lagging when it comes to regulating AI systems and that some tools (chatbots and summary tools for patient health journaling) fall outside of regulatory framework entirely. This could impact individual patient care or bias patient care for entire groups of people. It's going to be more critical for local healthcare organizations to maintain population-specific validation programs and to continuously monitor AI systems of all kinds. This is primarily for patient safety but to also to meet accreditation and the regulatory standards of the future.  https://lnkd.in/dEp4mge8

Interested in learning how AI can enhance our health and well-being? Join IQVIA today at the OntarioMD Digital Health Conference where we explore the numerous ways AI can revolutionize various aspects of patient care. View the conference schedule here 👉https://bit.ly/3ZnOWlE #OMDDHC2024!

Just Launched! Check out our new blog: Tenasol Data Bytes to hear from the leading voices of #AI and #data in #healthcare. #MedicalRecords account for only a small portion of a #patients #wholehealth picture. Read our first blog, "A Comprehensive Guide to Healthcare Data Sources", to learn how additional #datasources outside a record, and implied information extracted with #machinelearning from within a medical record, create a more wholistic view of a patient.

With #RADV audits starting back up, health plans can mitigate financial penalties from extrapolation by leveraging clinically tuned AI to identifying net new HCCs. Learn more: https://ow.ly/UUTT50Rij2L #medicareadvantage #riskadjustment

Had the opportunity to chat with Dr. Keith J. Dreyer following his session on the future of AI in Healthcare. Great things going on at Mass General Brigham surrounding foundational models, reshaping patient care, diagnostic accuracy and more. Another key topic was about the importantance of addressing the ethical implications and regulatory challenges that come along with innovation. So much to discuss but the industry is trending in the right direction. Thanks for your great insight, Keith! #himss24 #analytics #healthcareinnovation

Here's the end of this article - but I hope you will be moved to read the full text. To say there's lot of uncertainty in this area is ... an understatement? "... the FDA should be handling AI validation within its own walls and that makers of AI systems should at a minimum have to show that they improve outcomes for patients, regardless of who does the oversight. He noted the failure of an AI system from electronic health records firm Epic to detect sepsis, a sometimes fatal reaction to infection, that had slipped by regulators. The company has since overhauled its algorithm and a spokesperson for the FDA said that it doesn’t disclose communications with specific firms. But the incident has left many in health care and technology feeling like the agency isn’t using its current authorities effectively. “They should be out there policing this stuff,” said Ayers." It's a bit facile to say what the US FDA should be doing when the what is still undefined ) changing. I am reminded of a very old ( as in Black&White) Superman episode where intrepid reporter Clark Kent is hounding Inspector Henderson to do something about the episodes villain. Henderson replies " The Law, Kent - the law! What has he done to break the law?" Even harder to answer when there is no Law yet . #fda #AI #artificialintelligence #ehr https://lnkd.in/emGr5Ytr https://lnkd.in/emGr5Ytr

Spot on, John English, HCCP. It's easy to say that the FDA should do more but when the "what" is still so ill-defined, doing more may turn into a misapplication of time/effort. The FDA has established a way forward, though, to focus on how to arrive at a definition of the what. In their 2023 discussion paper “Using Artificial Intelligence and Machine Learning in the Development of Drug and Biological Products,' (well worth the read btw), the FDA clearly define well-established problems w/AI along with a focus towards defining standards. From the discussion paper: "AI/ML algorithms have the potential to amplify errors and preexisting biases present in underlying data sources and, when the findings are extrapolated outside of the testing environment, raise concerns related to generalizability and ethical considerations. Additionally, an AI/ML system may exhibit limited explainability due to its underlying complexity or may not be fully transparent for proprietary reasons. These concerns have resulted in a focus on developing standards for trustworthy AI that address specific characteristics in areas such as explainability, reliability, privacy, safety, security, and bias mitigation." For those of us in pharma who are availing ourselves of AI/ML, we can begin to address these very real problems by encouraging (requiring) the use of software development best practices. For starters, we can reduce the opacity of the LLMs used, and define what reproducibility looks like.

AI will undergo three phases of increasingly complex integration

It’s strange to think that this paper will probably be cited from time to time for the rest of my life and beyond. It describes, with the best data set that exists, the changes (improvements) in patient safety in the USA in the 10 years before Covid-19 arrived. The annoying thing is that the JAMA is editors didn’t allow us to say this, and they required that we add a peculiar semi-disclaimer as the last sentence of the abstract and the paper to say that more research was needed. But we and the editors knew very well that there would be no additional research forthcoming because no one had better data than we did about these years. Although maybe someday AI will be able to go through hundreds of millions of patient records from that period and develop some additional findings. https://lnkd.in/euBSXQJW