Andreas Scherer

Big healthcare AI news yesterday as Tempus AI received the first-ever CPT code for an algorithm-only analysis of previously-sequenced, laboratory developed test transcriptomic data. TLDR: AI molecular phenotyping has a new business model. The PurISTSM test uses RNA profiling to identify molecular subtypes in patients with advanced pancreatic ductal adenocarcinoma, which may inform first-line therapy selection and management. While I'm skeptical that payer balance sheets are going to solve the "who pays for healthcare AI" problem, the reimbursement pipes for clinical AI are maturing. I am surprised to see more progress for usage-based vs. performance-based models given the many challenges with FFS economics. It will be interesting to see how payers assess the "medical necessity" and therefore reimbursability of clinical AI as these models come to market. https://t.co/eLe18SmihV

420

9 Comments

M13 is thrilled to lead the Series A in RadiantGraph which is focused on personalization in healthcare. As we continue to invest in early stage companies it is abundantly clear that the biggest indicator of success is the founder and his or her ability to execute on their vision of the future. In this case, I don't think I have seen much better founder-market fit than we have with founder and CEO Anmol Madan. We believe in his vision: in order to have the best patient outcomes in healthcare we need increased personalization and AI is the breakthrough to get us there. Anmol started one of the pioneering digital health companies Ginger and then was Chief Data Scientist at Livongo and Teledoc. He has done it all: early stage, late stage, public company. And he deeply understands the nuances of the healthcare market from all angles which is critical for success. While there are many players building healthcare-specific AI capabilities, RadiantGraph is focused on outcomes: The platform is already processing personalization models for more than 3.5 million people, despite launching just last year. Excited to partner with my friend Adam D'Augelli and the great True Ventures again and Morgan Blumberg who has been deep in the healthcare market with me for many years. More on our thesis in the blog below. https://lnkd.in/evDDmC3W

72

4 Comments

There are two key takeaways we learned this year attending Meeting on the Mesa Cell & Gene Conference 2024 #CGMesa2024. 1.     The first that stood out from the opening keynote to the general session panel discussions to individual company break-out presentations, is hearing about all the therapeutic advancements in drug development and FDA approvals through successful commercialization for so many patients in need suffering from difficult to diagnose and difficult to treat (often life threatening) diseases.    Sitting in the audience quietly saying “Yeah, that was UBC, we did that work”. UBC has helped over 60% of the cell or gene therapies brought to approval and available in market.  UBC drives the top therapies in on-going support after approval. 2.     The second is that visibility is as important as performance. We all think that the quality of our work should speak for itself. And it does, but it is self-limiting. There are so many more patients and colleagues that can benefit from your work. There are numerous important reasons more people should be aware of you and what they should additionally consider when choosing to work with a partner. Being the quiet, unsung hero has its great moments. There is gratitude and encouragement as we work together to minimize suffering in the world. More people and the patients they serve can benefit if they know who you are. More to come regarding what else is essential to consider when choosing a strategic partner in healthcare. Pictured with Brett Huselton, SVP, Strategy, Partnerships & Innovation at UBC. Reach out to learn more about UBC's Evidence, Safety and Patient Access Services. #UBC #realwordevidence #RWE #RWD #safetyandriskmanagement #patientservices

48

4 Comments

Regulating #AI is a sticky wicket. As is often the case, California is ahead of the curve. SB 1047 is the first legislation to impose safety mandates for large language models. I'm as eager as anyone to see some guardrails put in place, but the risks of getting it wrong are not trivial. Solid reporting here via @jacoberobbins of PitchBook #artificialintelligence #regulation #machinelearning

10

1 Comment

You should read Steve Usdin’s insightful essay on #FDA's approval of Elevidys’ DMD label expansion: https://lnkd.in/g2jK2A5i Beginning with Janet Woodcock’s reversal on Sarepta’s Exondys 51, Steve believes a series of “center director’s decision[s] to override agency staff” means “FDA has adopted a new Peter principle: when reviewing a proposed therapy for a terrible disease, a center director can cite any glimmer of positive data to justify approval.” But instead of caprice, this can be viewed in a benign light as embodying a patient-centric regulatory philosophy: As Congress intends and patients desire, the Agency is leaning in –hard in this case– for patients with life-threatening diseases and few therapeutic options to enable access, clinical practice, and evidence generation (more below vv), absent overriding safety concerns, judging that this more than balances the risks that efficacy may ultimately be equivocal in the broader patient population. Like Duchenne patients, FDA is taking risks. This is precisely why Pat Furlong calls this “a monumental leap forward in our collective efforts to end Duchenne muscular dystrophy” and Muscular Dystrophy Association deems it “a major source of hope for countless families affected by this relentless disease.” Dr Peter Marks and his staff reached profoundly different conclusions– what Steve calls an “idiosyncratic” decision that “disregard[s] the competence of long-serving staff.” Even though Marks has nothing but respect for his staff, Steve asks how reversals like this will impact their morale. Steve rightly notes “the problem is that Marks didn’t describe his disagreement with CBER staff as a matter of policy” and here in fact is Dr Mark’s policy-driven conclusion: “[t]he benefit to risk considerations are favorable, taking into account the existing uncertainties, such as the ultimate duration of response, which is not yet known, as well as the significant unmet need.” Finally, though, I don’t see any connection between the decision and a new ‘ultra-rare’ disease review pathway. That is apples to oranges. With a US prevalence in the tens of thousands, DMD is rare but not ‘ultra-rare’. And, if anything, Dr Mark’s decision proves a new pathway is not only inapplicable, but unnecessary. Setting aside the risks of legislating, FDA is fulfilling –here and in cancer– decades of congressional intent to bring novel treatments to market faster. What is crucial –which Steve points out– is timely completion of confirmatory trials; FDA must use their new accelerated approval authorities to their fullest to ensure this. We should read Dr Mark’s decision memo as an invitation for public discourse like Steve’s on FDA use of its regulatory authority– not for flawed legislation, whether ultra-rare, right-to-try, or promising pathways. FDA is not perfect. But clearly, they are trying hard to drive #innovation into the lives of patients with #rarediseases. #kendallsquarepolicy

28

1 Comment

How do we minimize the number of protocol amendments and increase the operational feasibility of our clinical trials? This article below points out the benefits of using data, conversations and real world data to enhance the feasibility of the protocol design and inclusion/exclusion criteria: 1) Are we being too strict on our criteria in order to "control" for data variability? 2) Do are criteria allow for the intended patient population in the real world (if not, they need to) 3) How realistic is your design in how the study sites actually work? How they see and treat patients? How is your company ensuring the design meets these requirements? https://lnkd.in/gzJzFd-3

4

1 Comment

In case you missed it, last week the Coalition for Health AI (CHAI) released a draft consensus framework for responsible health AI. Developed with input from over 100 contributors representing a diverse network of healthcare stakeholders, this guide proposes actionable evaluation criteria throughout the AI lifecycle—from identifying use cases to deployment and monitoring. Key examples outlined in the guide include: • Predictive EHR Risk Use Case (Pediatric Asthma Exacerbation) • Imaging Diagnostic Use Case (Mammography) • Generative AI Use Case (EHR Query and Extraction) • Claims-Based Outpatient Use Case (Care Management) • Clinical Ops & Administration Use Case (Prior Authorization with Medical Coding) • Genomics Use Case (Precision Oncology with Genomic Markers) The draft framework is open for public review and comment for the next sixty days. Please use this form to submit feedback: https://lnkd.in/eYughFYC Kudos to the CHAI team led by CEO & President, Dr. Brian Anderson, MD, on this milestone. Eric Horvitz Jennifer Goldsack John Halamka, M.D., M.S. Michael Pencina Micky Tripathi Nigam Shah Suchi Saria Troy Tazbaz #healthcare #ai #artificialintelligence #generativeai https://lnkd.in/e7VRv6fD

125

1 Comment

Did you miss this very interesting session on next generation #ELN and #FAIRdata in #biopharma and other related industries? Really worth a listen, and you can still catch it on the link below!

4

Do not miss this webinar by ZONTAL and 20/15 Visioneers on how you can take #digitalization and #AIreadiness further in your R&D (and ultimately manufacturing) ecosystem

10

This is perhaps where the "U" turn for DTx starts? I'm not suggesting we will enter the highly bullish times of 2018-2021 but I think like a lot of technology cycles, this may serve as an inflection point as we look back on this 2-4 years from now. In addition to some of the additional coverage guidance around breakthrough treatments, we may see folks filing back into this space over the next 12-18 months. #dtx #digitaltherapeutics #FDA #reimbursement

4

Creating a therapeutic platform with multiple applications? EverGlade Consulting can help you obtain BARDA funding to bring your versatile solutions to market. #innovation #health #biotechnology #scienceandtechnology #research #science #publichealth #covid19

1

The third and final key pillar to the most disruptive and impactful effort to streamline pre-clinical stages of drug discovery and development is articulated below in response to another great question posed about the DSDC.

7

1 Comment

#EHR vendor Epic Systems is partnering with various genomic companies to integrate genomic data within EHRs to accelerate faster diagnosis. Epic is starting with the interpretation workflow, importing VCF files into the EHR, and then outputting results reports directly into each patient's record. Following Rady children`s institutie(#RCIGM), #invitae, #myriad_genetics, and #Foundation_medicine, this new collaboration with #GeneDx will make #rapid_WGS a reality for #NICU patients. #RapidWGS #Raredisease #Genetictesting #EHR #diagnosis #3billion https://lnkd.in/gHfS3bHv

6

1 Comment

I spent this morning in Nashville with Christine Mohrhaus Hale, MD, MBA and Fred Barton as a guest of HealthCareTN speaking to employers about #AI in #healthcare, and what #AI means in terms of solving the problems they care about. It got me thinking about #languages, and transferring #meaning from one language to another. As a mathematician who worked in sales, as a native Russian speaker who also learned English (and later French) and studied German and Chinese, I can tell you: translating meaning from one language to another is really difficult. Translating AI applications in biotech to enterprise level cost problems for employers was a new one for me. But where good translation ends, collaboration begins. Employers looking at healthcare today have hard dollar spending challenges and workforce productivity challenges that chatGPT cannot solve. But employers can be asking hard questions of their healthcare spending that biotech tools (like diagnostics, hugely enabled by AI and machine learning) CAN solve. 'What percent of my drug spend is actually effective and enabling disease improvement, and what percent is not effective? How do I make the effective number go up?' 'Why am I paying for so many late-stage cancers, does that have to be the status quo, or can we find these cancers earlier?" Those are questions that not only have answers, they have diagnostic solutions in the form of actual tests employers can cover, adopt, and use. AI is an enabling tool across all of human activity, and thus all of healthcare. It is our human challenge to seek, define and refine the problem statements we are seeking solutions for. It is then the job of technology to make those solutions as seamless and high-performance as possible, so we can move on to our next mountain. https://lnkd.in/epyc8gie

19

2 Comments

💥 New Episode: Second Opinion with Caroline Pearson, Executive Director of Peterson Health Technology Institute (PHTI)! Remember that bombshell PHTI report showing many digital diabetes companies aren't producing meaningful improvements in patient outcomes—and are even driving up healthcare costs?? We had a candid chat with Caroline about how the industry can do better. Some takeaways: 📊 Prove Your Worth: "Digital" isn't a value proposition—it's table stakes. Whether you're driving clinical outcomes or providing patient support, back your claims with evidence and be crystal clear about what you're actually delivering. 🏥 Healthcare is Unaffordable: Caroline talks about the tough trade-offs we make every day due to finite resources. If we’re going to spend healthcare dollars, let’s make sure it’s on something that delivers real value. 🤝 Collaboration is Key: Caroline emphasizes the need for companies to work together—don’t compete when it comes to building evidence. Let’s all push towards better solutions for patients! Check out the full episode here and let us know what you think: ᯤ https://lnkd.in/gUmfWSPq  https://lnkd.in/g7FdqZhs ⏵ https://lnkd.in/gnGH7dpt Turpentine Christina Farr Luba Greenwood, J.D.

53

2 Comments

One key reason why healthcare technologies like AI, telemedicine, and software haven't structurally lowered costs or broadened access for the entire system after decades of experimentation (excluding pure financial engineering of payer contracts) is the limited reimagining of care delivery itself, with asynchronous telemedicine as a notable exception. In the same way that digitizing physical paper failed to exploit the capabilities of a new computing medium for word processing, hosting physician-patient conversations over video does not transform the actual delivery of care. In telemedicine, many of the inputs (e.g., clinician time) and therefore cost structures of in-person care persist, which is why we often see fee-for-service telemedicine companies struggle to achieve economies of scale (i.e., growing revenues non-linearly while growing costs linearly). Until we deconstruct the components of a medical appointment – from intake and triage, to chart review, history gathering, physical examination, diagnosis, treatment, and follow-up – we will continue to be underwhelmed by the disconnect between the transformative potential of these technologies and real-world patient/system impact. I’m optimistic that the emergence of reliable and performant medical AI agents will urge us to perform this deconstruction and redesign the fundamental building blocks of a “medical appointment.”

215

38 Comments