Breaking News

OpenAI’s Color cancer copilot, AI regulation battle brews, & Done’s top leadership arrested

June 18, 2024
Reporter, STAT Health Tech Writer

Good morning! Lots of health tech news today — including, unfortunately, more than one instance of business misconduct. Also, if you can find precious time to relax this summer, we'd love to hear what health, medicine and life sciences books and podcasts you're discovering. 

As always, news tips and thoughts go to mohana.ravindranath@statnews.com

Artificial intelligence

OpenAI collabs with Color on cancer treatment

In today's buzzword bingo, ChatGPT developer OpenAI is working with population health and testing startup Color Health to use OpenAI's latest large language model, GPT-4o, to flag missing diagnostics and suggest pathways for cancer screening and treatment. The so-called "co-pilot" tool, the companies say, can crunch patient data from medical notes, including family history and individual risk factors. It can also churn out answers to questions like "What screenings should the patient be doing?"

The companies say there's always a clinician "in-the-loop," meaning it's always reviewed before any of it reaches the patient. Still, as STAT has reported, experts still haven't gotten a handle on the scope of automation bias: the extent to which automated suggestions could influence clinicians' judgement, especially if it's typically, but not always, offering seemingly accurate assessments. Read more from OpenAI here. 


Basecamp hunts rare proteins at the world's end

This week Brittany Trang takes us inside London company Basecamp whose mission is helping communities understand and preserve their local biodiversity, all while feeding genome sequences to AI models in the hopes of generating new enzymes: proteins enabling reactions in a biological system, like digestion and metabolism. In recent preprint, Basecamp and Institute of Molecular Biology of Barcelona researcher Noelia Ferruz describe a new AI model that they say can generate new enzymes. Read more on the company, and one of its field researchers' unexpected link to Jason Momoa, here


Protecting startups' interest in AI regulations

(Julie Yoo/Andreessen Horowitz)

Earlier this month I reported on startup ecosystem leaders like Y Combinator and Andreessen Horowitz defending so-called "Little Tech" in Washingtonand it looks like the battle for influence over Washington's potential AI regulations is only picking up. In an opinion for STAT this week, a16z's health general partner Julie Yoo takes particular aim at the Coalition for Health AI, the industry body led by the likes of Google, Microsoft, and Mayo Clinic and which has proposed a nationwide network of health AI assurance labs to ensure quality and safety. 

"This review process would put big tech companies, which are themselves developing AI models for health care, in control of market entry," Yoo writes. "By seeking to regulate the health AI industry on behalf of the federal government, established big tech companies could create an uneven playing field for newer or smaller companies." What do you think? Read more here; thoughts welcome. 



lizzy's device digest

Stimwave CEO gets 6 years in prison

MedicalImplantFraud_Illustration_MollyFerguson_052223-2048x1152

The former head of a medical device company that sold pain relief devices outfitted with useless plastic parts has been sentenced to six years in prison by a New York judge, our devices reporter Lizzy Lawrence writes. If you need a refresher, check out Lizzy's investigation into Stimwave from last year. 

Led by Laura Perryman for 9 years, the company was aiming to score higher reimbursement from payers — the dummy parts let the doctors collect $18,000 more than they should have, and nearly 8,000 devices had the fake part.

"Laura Perryman callously created a dummy medical device component and told doctors to implant it into patients," U.S. attorney Damian Williams said.  "Perryman breached the trust of the doctors who bought her medical device, and more importantly, the patients who were implanted with that piece of plastic." Read more on her sentencing here. 


Telehealth

Done's top leadership arrested on fraud charges

In other misconduct news, top leaders from a closely watched telehealth startup known for prescribing ADHD-medication Adderall online were arrested and charged with fraud, the Justice Department announced late last week.

Done founder Ruthia He, and clinical president David Brody, were charged with allegedly taking part in a scheme to prescribe and distribute controlled substances online — along with submitting fraudulent reimbursement claims for stimulant prescriptions, I reported along with Katie Palmer.

The company is perhaps best known for its social media ads hawking ADHD treatment; it's also one of a small group of startups boldly seeking to profit off relaxed regulations for drug prescriptions during the pandemic. Read more here. 


Report: One Medical yields patient safety issues

The Washington Post has an investigation into Amazon primary care tech property One Medical revealing several concerning patient safety issues, including routinely routing elderly patients to a call center that seemed ill-equipped to handle medical issues that required more attention. Among them: One patient called to report a blood clot and swelling, and the call center scheduled an appointment for the future instead of flagging the issue for more immediate examination. Read the investigation here. 


industry forecasts

PTSD treatment company bets on slow and steady

My colleague Mario Aguilar has reported on digital health companies' herculean  struggles to reach profitability; this week he spoke with one company betting that a slow and steady approach might be a better path to success than raising a lot of money and promising investors unrealistic revenue growth. Joe Perekupka, who heads Freespira, a company developing software- and device-based treatment for panic and post-traumatic stress disorders, told Mario it's signing on a few thousand patients at a time — but it's already heading toward a break-even point by the end of next year, and positive cash flow sometime in 2026.  Read more here


Cybersecurity 

Drug changes for cyber hiring on the horizon?

Late last week I dropped by the CyberMed summit in Washington, where a small group of technical luminaries gathered to discuss the underplayed cybersecurity risks to health systems. Among speakers was the White House's Office of the National Cyber Director official Seeyew Mo, who teased a federal cyber workforce hiring change that could potentially open the floodgates to solve historical shortages: A pending update would modify questions related to candidates' drug use — long considered a barrier to entry for otherwise qualified candidates — to ask about their use of drugs that are widely illegal, instead of drugs that might only be illegal in some states. I'll be tracking the update and its implications for health. 


More around STAT
Check out more exclusive coverage with a STAT+ subscription
Read premium in-depth biotech, pharma, policy, and life science coverage and analysis with all of our STAT+ articles.

What we're reading

  • Even doctors are falling into the medical bill trap, NYT 
  • Oncologists look to blood tests for circulating tumor DNA, but questions remain, STAT 
  • Medicare will recalculate quality ratings for MA plans, The Wall Street Journal
  • Amazon offers drug prescription program to Medicare members, Fierce Healthcare 

Thanks for reading! More on Thursday - Mohana

Mohana Ravindranath is a Bay Area correspondent covering health tech at STAT and has made it her mission to separate out hype from reality in health care.


Enjoying STAT Health Tech? Tell us about your experience
Continue reading the latest health & science news with the STAT app
Download on the App Store or get it on Google Play
STAT
STAT, 1 Exchange Place, Boston, MA
©2024, All Rights Reserved.

No comments