Big AI
OpenAI’s health policy wishlist
OpenAI recently published a health policy framework that offers “principles for responsible AI adoption that can operate within the institutions we already have, support better care, and drive improved experiences.” The blueprint follows the company’s push into health with new products for consumers and health care providers.
As Brittany Trang reports, key proposals include using enforcement and rulemaking action to make health data more portable, including data from wearable devices; and further clarifying what constitutes low-risk AI use cases that don't require Food and Drug Administration review. The former is largely in-line with current government initiatives like CMS' Health Technology Ecosystem initiative and recent information blocking threats from HHS. FDA in January offered refreshed guidance on some AI, but the future for AI regulation is murkier. All of the proposals, former regulators told Brittany, would be very good for OpenAI.
Read more here
medical devices
Dexcom teases refreshed Stelo app and 'new CGM system'
During Dexcom's earnings call last week, CEO Jake Leach teased an update to the app for Stelo, its continuous glucose monitor that's available without a prescription. The refresh will introduce “a more consumer-friendly feel, more AI-driven personalized insights, and additional food logging capabilities, including detailed macronutrient information.”
In October last year, Dexcom disclosed over $100 million in revenue from Stelo in its first year on the market. Leach said Dexcom is planning an international launch of Stelo this year as well as “a new CGM system that is designed to further extend our market reach.” Expect more details at the company’s investor day next week.
Overall, Dexcom reported $1.2 billion in first quarter revenue, 12% over the year before. The major market mover for the company will be a CMS decision on coverage for its G7 CGM for people with type 2 diabetes who are not on insulin.
spicy take
Should preventive body scans come with stronger disclaimers?
There has been pushback to startups and established radiology providers aggressively marketing preventive MRI and CT scans that might identify hidden health conditions. But I was amused by a perspective in JAMA this week that advocates for more substantial informed consent before elective MRI screenings. The pithy proposal by Matthew Davenport, from the University of Michigan, and Scott Reeder, from the University of Wisconsin, would tell patients that that medical guidelines and studies don’t support this testing; that there’s a high-likelihood that the test will find something that will cause distress; that it’s unlikely finding cancer this way will help you.
The physician-academics go on:
“There are no studies showing that undergoing this test will improve the quality or length of your life. Based on what we know, if you undergo this test, you are more likely to be harmed than helped due to complications from efforts to diagnose and treat what we find and the low likelihood we find something that can improve the quality or length of your life. On average, any apparent benefit is likely illusory, even for many cancers.”
Read the whole viewpoint for detailed citations backing up the language.
No comments