Dave Wright’s Post

"AI will transform medicine and the health care system— for better or for worse, depending on how it is built and applied." Our Co-Founder and Chief Scientific Officer Ziad Obermeyer testified before the U.S. Senate Committee on Finance to discuss how the government, private sector, and researchers could advance commonsense approaches that both protect patients and promote AI innovation in healthcare. While there are several specific things that programs under this Committee’s jurisdiction could do to ensure that AI produces the social value we all want, Ziad highlighted three approaches in particular: 1️⃣ "AI developers must be transparent about the output of their algorithms. If an algorithm predicts health care costs, the developer should not be able to claim that it predicts “health risks” or “health needs”—unfortunately, many cost-predictors currently do exactly this." 2️⃣ "Algorithms’ outputs should be evaluated for accuracy in a completely independent dataset, both overall and in protected groups, in keeping with good machine learning practice [so algorithm purchasers, users, and regulators] can focus on one simple question: is the algorithm predicting what it's supposed to predict, accurately and equitably?" 3️⃣ "AI products should be valued and reimbursed according to established principles from health economics and outcomes research. If an AI results in an earlier diagnosis of heart attack or breast cancer, for example, that generates value to patients in the form of life-years, and to the health care system in the form of downstream costs avoided. The sooner public programs lay out what they are looking for, the sooner the market can deliver safe and effective AI products to solve the urgent problems they face." The full Committee hearing is available here: https://lnkd.in/eXBGjN4q

In Union Healthcare Insight's recent Board Briefing on where the smart money is going in 2024, I made a bit of a buzzkill comment that for my money, much of the near term benefits offered by generative artificial intelligence (gen-AI) would be in the administrative space, in particular provider revenue cycle and health plan claims management. If you read the trade press or literally anything that came out of Vive this year, that might be a surprise, given the attention that clinical applications are getting (or more to the point, patient care innovations). And it's not like we at Union are bearish on the potential for AI to diagnose, interact, and even make recommendations or decisions about patient care. Far from it. But let's follow the money. Literally. Our chief research officer, Yulan Egan, recently published a report on the Future of AI in healthcare (members can access in the link) that includes a feasibility calculator as a decision guide for AI investment. And honestly, revenue cycle and its counterparts tick most of the boxes, as seen in this week's Slide[s] of the Week. And yes, I know that administrative functions don't seem to offer much excitement, but AI has the potential to make new leaps forward in documenting, preparing, submitting, evaluating, denying, appealing, and resolving claims. In fact, it's all reminiscent of the gains in automation made in the mid-2000s, which drove a revenue cycle arms race, and resulted in drastically lower AR days for providers, higher denial rates for payers, and a whole lot of new costs for everyone. But when one considers the vital role days cash on hand plays in the ability to secure debt and invest in the new capabilities, it's hard to overstate the impact that cash acceleration/deceleration can have in our industry. So I'll say that the excitement factor really is there, if you know how significant it can be. One caveat that could alter my view here: AI's use in MA denials has received some recent political scrutiny, and invited a few lawsuits, so we could eventually see some regulatory hurdles there. Which is a long way of saying I'm doubling down on what I said in the webinar, with our reasoning laid out more fully in this week's slides. And by the way, I am aware of how narcissistic it looks to publish a slide with my own quote on it (it's the second one in the link). But for what it's worth, I didn't make it. Our chief education officer, Amanda Shoemaker Berra, thought it a good encapsulation for her upcoming AI Bootcamp. Sign up if you're a member (link below)! (And if you want to make fun of me anyway, go ahead, because you can't insult a narcissist.) #healthcareai

Medical errors are of critical concern as they can result in fatalities, staggering financial settlements and reduced patient confidence. What are the root causes behind medical errors and how can new tech such as AI help reduce risk for one, and all? Get the answers in our latest white paper: https://lnkd.in/gg_4J2dP #HARMANDigitalTransformationSolutions #HealthcareSafety #MedicalErrors 

Artificial intelligence creates tangible information

Yesterday, the House Energy and Commerce Committee held a hearing titled, “Leveraging Agency Expertise to Foster American AI Leadership and Innovation.” Here are a few key takeaways.   * The discussion highlighted the importance of a comprehensive data privacy standard that addresses information outside of traditional health records protected by HIPAA. The emphasis was on developing a nationwide standard to safeguard health information that falls outside of HIPAA regulations. * The integration of AI in healthcare aims to enhance patient care, streamline administrative tasks, and support physicians. However, there's a delicate balance to strike to avoid diminishing the role of healthcare professionals and maintain the doctor-patient relationship. * The FDA has been actively involved in approving AI applications in healthcare, with a notable increase from 400 to almost 700 approvals in a short span. The agency has received commendation for its thoughtful approach and commitment to patient safety. The Office of the National Coordinator for Health Information Technology (ONC) and the FDA work closely to coordinate requirements for AI-enabled health technologies. They ensure transparency and avoid conflicting rules to prevent stakeholder confusion. You can watch a recording of the full committee hearing here: https://lnkd.in/geFxUv4E #telehealth #ai

Did you know that currently, some 4.5 billion people lack access to essential healthcare services? It's a staggering number, but there's hope. Detailed data can help identify and address gaps in healthcare provision, while machine learning and artificial intelligence can help with disease detection and precision therapies. Let's work together to ensure everyone has access to the healthcare they need. #Healthcare #Data #AI #ML

Looking at the way this change will affect Life Science and Healthcare as a whole, it is a huge opportunity for all stakeholders both sides of the pond, to embrace the journey starting now. Although we all understand that, especially after Covid, the meaning of data privacy has changed its value, we are now more conscious what could be the trade-off of having those Electronic Healthcare Records produced and available for a greater good: for in the end it is Societies’benefits we are talking about, and based on those updated EHRs to have AI/ML support the quest for better and faster care.

https://lnkd.in/gSYsqwee This is the best talk I saw two weeks ago at the "AI in Health" Conference @ Ken Kennedy, Rice. It starts a little slow but he's really going through the entire process, from the economic case to the nurse workflow to the AI and how it helps. One of the things that has really stuck with me is this idea of taking using AI to analyze video and audio from potential stroke victims and then alert nurses. Man, that sounds cool and really valuable but...if, as a health system we want to be able to intelligently purchase these products, we don't just need to be able to verify that they work, we also want to be able to compare the performance between different AI products. What we'd really like is to have a year or two worth of this video and audio, along with some EHR metadata, hide the outcomes from a potential AI algorithm we'd like to buy/license, and then test that potential AI against that historical data. But that would involved efficiently storing thousands of hours of video and that's...not an insignificant technical challenge.

For those looking at the US as market for their AI/ML device, this webinar gives an overview of the current US regulatory landscape to be navigated #AI #machinelearning #SaMD #FDA https://lnkd.in/eMpzuxZa

If you're watching #CaLeg for #HealthIT bills, it's all #AI all the time: - SB 1120 (Josh Becker, sponsored by the California Medical Association): Physicians Make Decisions Act - AB 2013 (Irwin): Artificial Intelligence: Data Transparency - AB 2058 (Weber) : Automated Decision Systems - AB 2885 (Bauer-Kahan) : Artificial Intelligence - AB 2930 (Bauer-Kahan): Automated Decision Tools #artificialintelligence #healthinformationtechnology