You’re reading the web version of STAT’s Health Tech newsletter, your guide to how technology is transforming the life sciences. Sign up to get it delivered to your inbox every Tuesday and Thursday.
Carbon Health founder steps down as CEO
Carbon Health CEO and co-founder Ellen Bali announced in a tweet over the weekend that she was taking a step back from the primary care technology company to return to Udemy, an education company she co-founded, as chief technology officer. “Udemy is my first child. It’s special to me and I’ve always served on the board so I was never separated from the company,” Bali said. Carbon’s chief operating officer, Kerem Ozkai, will take over as CEO, and Bali will remain chairman of Carbon’s board, the post said.
It’s just the latest senior management change at a company focused on a hybrid business model that combines the best of virtual and in-person care. Carbon has since been increasingly investing in incorporating AI into the homegrown health records it sells to other provider groups. Myoung Cha, Carbon’s former president and chief strategy officer, joined Verily earlier this year as chief product officer.
Abridge to deploy AI technology at 40 Kaiser Hospitals
Competition in medical scribing technology is heating up, with companies like Nuance, Abridge, Augmedix, Suki and Nabla vying for hospitals’ business. In a big win for Abridge, the startup’s ambient documentation tool is now available to doctors at 40 Kaiser Permanente hospitals and more than 600 clinics in multiple states across the country. “Reducing administrative tasks helps doctors focus on patients and creates an environment for effective communication and transparency,” Ramin Davidov, head of Southern California Permanente’s medical group, said in a release.
In a recent update to STAT’s Generative AI Tracker, my colleague Katie Palmer wrote about how health systems are testing these AI scribes.
Separately, Kaiser has published an interesting set of principles for responsible use of AI, which I encourage you to read and let me know what you think.
Could the FDA use AI to monitor drug safety? This research team thinks so
There are many proposals to use large-scale language models to facilitate the delivery of real-world healthcare — scouring medical literature and health records, providing patients and clinicians with easy-to-understand answers to complex questions, and maybe even providing diagnostic results in the future — but what about helping federal regulators?
This week, a team of researchers, including two from the Food and Drug Administration, put forward a provocative proposal in JAMA Network Open: “What if we used LLM to sift through unstructured text in electronic medical records, social media posts, insurance claims, and other sources for data about the safety of specific drugs once they’ve hit the market?”
I spoke with lead author Michael Matheny, a professor of biomedical informatics at Vanderbilt University and principal investigator at one of the few FDA innovation centers for Sentinel, the FDA’s electronic system for product surveillance and evaluation. Among the tidbits of trivia he shared in our interview was that while there are concerns about hallucinations, the risk of error and how detrimental it is to patient safety varies by situation.
“You can ask a large language model, ‘Did you have diabetes?’ but diabetes may not be relevant to the risk of adverse events. For someone who is taking a certain medication, it doesn’t matter that they have diabetes. If there’s a 10% error rate in detecting diabetes, it’s probably not going to have a big impact.” [on safety analyses.]”On the other hand, if the adverse event is bleeding after taking a medication, and the only way to detect that bleeding is through natural language processing, and only 10-20% of nosebleeds are severe enough to warrant a trip to the emergency room, and the rest are just reported to a doctor, then that detection is very important because your entire analysis depends on detecting that adverse event.” Read more.
New Lawsuit Filed Against FDA Over Lab-Developed Testing Rule
And speaking of the FDA, a trade group representing molecular pathologists has filed a lawsuit against the agency for overstepping its authority in trying to regulate lab-developed tests, writes Lizzy Lawrence. This is the second lawsuit against the effort; the American Clinical Laboratory Association filed a similar suit in May.
“Because the FDA does not have the authority to regulate LDTs, we filed this lawsuit to ask the Court to invalidate the FDA’s rule and avoid serious and harmful disruptions to clinical laboratory medicine,” Maria Arcilla, president of the Society for Molecular Pathology, said in a statement.
As Lizzy writes, the FDA hasn’t historically regulated lab-developed tests, but as products like misleading prenatal genetic tests and Theranos’ fraudulent blood panels become more complex, the potential for patient harm increases. Read Lizzy’s full article to find out more.
AI pharmaceutical company Recursion’s path to industry dominance
Biotech colleague Allison DeAngelis takes a deep dive into Recursion Pharmaceuticals, a company that’s rising to the forefront of the increasingly competitive field of AI-guided drug development. The company is a partner with Nvidia and recently planned to acquire Exscientia. Founded by a med school dropout, an e-commerce entrepreneur, and a University of Utah scientist, Recursion struggled to attract biotech venture capital interest at first and hasn’t always been successful. “We’ve been laughed out of quite a few rooms,” co-founder and CEO Chris Gibson told Allison. Allison spoke with more than 30 of Recursion’s leaders, investors, former employees, and analysts.
“Recursion is a startup that plans to both conquer the world and survive,” says Viswa Colluru, a former employee who now leads the learning-based biotech company Enveda Biosciences. Read more.
Cancer testing company Grail announces job cuts
Another industry darling also continues to face challenges. Grail, best known for its blood-testing company Galeri (it is a partner in celebrity-backed consumer-testing company Function Health, whose fundraising we covered in our newsletter earlier this summer), has announced major layoffs that will affect about a third of its current and planned employment, reports my colleague Jonathan Worthen. That represents about 350 existing roles and 150 vacant positions. The company announced the cuts on the same day as it released its second-quarter earnings, and they are its first since being sold by DNA-sequencing company Illumina in June.
As Jonathan writes, Grail’s test sells for around $1,000 out-of-pocket, but it has no regulatory approval and is rarely covered by insurance. Clinical trials may change that, but the UK’s National Health Service has already announced that preliminary data from a recent study is not enough to launch a large-scale pilot to test Grail’s potential use in routine health care. Read more from Jonathan.
What we’re reading
Five health issues to watch at the Democratic National Convention, STAT Health technology policy roundup for Q2 2024, FierceHealthcare Clinician perspectives on potential health care cost savings and feasibility of practice shifts, JAMA
Source link
