Second Opinion Summit takes flight

This week was our inaugural Second Opinion Summit, and it was unbelievably energizing. We brought together 175 folks from across the health care industry to get real about some of the biggest issues facing providers, payers, technologists, investors, and of course, patients.

Within that summit, I hosted a panel on state AI sandboxes that provided insight into what these programs offer and what they don't. We primarily talked about Utah's AI sandbox, which is the only one currently in action. Launched through the state's Department of Commerce, the Utah AI Sandbox has been figuring out how to safely pilot AI-driven health care products as it goes. It is true evidence that states are still labs of democracy.

In January, Utah gave the green light to Doctronic to allow AI to renew a list of commonly prescribed drugs. Now, the program has two active pilots, both of which involve AI prescribing, and a handful approved for a later start date. The state is also now asking pilot participants to work with a third party to generate evidence to support the performance of their AI pilots. But questions remain about whether the approach is working and if it's actually enticing to health tech companies.

Panelists included:

— Virta Health's COO Amit Shah
— Sword Health's clinical and regulatory lead Dr. Fernando Correia
— Vega Health's director of products and services Joanne Kim
— Coalition for Health AI's CEO Brian Anderson
— Innovaccer policy lead Lisa Bari

Key takeaways:

• Health tech companies see the Utah pilot as something to watch, but not necessarily something to participate in. Many health tech companies don't use AI to practice medicine. Rather, AI supports doctors or helps direct patients; it's not actually delivering care and therefore wouldn't really need a suspension of rules to operate.
• The main reason to join a regulatory sandbox is for marketing. Companies say the pilot represents an opportunity to show off how well their products work.
• Broadly, the Utah sandbox is regarded as a positive step, allowing companies to test out specific AI-driven products and regulators to understand AI better. Companies like the sandbox approach over regulation, because they say it's too early to regulate AI and there are already rules that regulate health care.
• However, companies are worried that every state will develop its own sandbox. Many on the panel agreed that states should consider developing cross-state compacts that let companies expand their AI pilot programs to other states.
• While Utah has brought in third-party evaluators to monitor the performance of AI products, it doesn't have a mechanism for ensuring AI isn't over-prescribing. That would require an outside physician to oversee or review prescriptions.

If you were unable to join us for our first Second Opinion Summit, have no fear; there will be more. We'll also be sharing more about the other sessions that took place at our summit very soon. Stay tuned!

Now onto the news…

—Ruth