The Peterson Health Technology Institute launched an artificial intelligence task force to puzzle out the value of in-demand AI technologies for healthcare delivery organizations.
The task force has been operational for six months, Caroline Pearson, executive director of the PHTI, said in an interview. It will be led by Prabhjot Singh, M.D., Ph.D., a physician and co-founder of CHW Cares, which sold to Oak Street Health in 2022, and Margaret McKenna, former chief technology officer at Devoted Health.
Both Singh and McKenna are advisers to the PHTI.
There are about 60 people on the task force from a dozen healthcare systems, including UC San Diego Health, Intermountain Health, Mass General Brigham, Providence, Ochsner Health and MultiCare. Pearson also said there are many C-suite executives on the task force including CEOs, chief financial officer and chief information officers.
“They're not AI cheerleaders,” Pearson said. “They're just trying to run effective, efficient healthcare systems.”
The PHTI is well known for its splashy reports on digital health technologies, where it has compared the clinical evidence of popular commercial solutions head-to-head. It tackled diabetes, hypertension and musculoskeletal solutions in 2024.
The AI task force will release a report on AI-powered ambient scribes and revenue cycle management tools later this quarter. Pearson said it won’t be a typical PHTI report because AI is still in a nascent adoption phase in health systems compared to the maturity of digital health tools.
The task force also aims to put out more reports on AI cost efficiency in the future.
“One of the biggest, fastest growing areas of AI, of course, is with ambient documentation tools,” Pearson said. “And so we really wanted to talk to health systems who are rapidly adopting these tools, to say, 'How are you thinking about these AI tools? How are you measuring how this is working and and whether it's worth it to you financially?'"
The AI task force is chiefly looking at how healthcare delivery organizations are assessing the effectiveness of ambient scribes and revenue cycle management tools. Many of the companies claim to reduce administrative burdens on providers and diminish "pajama time" for doctors.
“We want to kind of ride along with that adoption and understand where are systems seeing benefits, and where are they experiencing challenges, and how are they measuring it?” Pearson said.
A goal of the upcoming report is to provide guidance to healthcare organizations on how to track the efficacy of the tools. This could be particularly helpful for less resourced organizations that often do not have large IT teams that can track the impact of a solution.
“We're going to put forward guidance about how systems can and should begin to track the impact of these solutions on their companies and on their hospitals,” Pearson said.
The PHTI is keeping its focus narrow, especially for the first report. It also is not going to engage in any AI validation or testing, Pearson noted.
“There's lots of folks thinking about many other facets of AI and how it's designed, how it's built, whether it's equitable, all of those things which are very important,” Pearson said. “This is really like, where AI hits the health system on sort of our core values around affordability … We want to adopt and embrace technology, and we really want that technology to produce better outcomes at lower costs.”
In talking about what metrics health systems are using to track the impact of AI tools, vendors have offered to provide more reports to systems if they are seeking specific data they don’t have, Pearson said.
Costs start to increase dramatically for healthcare organizations once they have already implemented an AI vendor on documentation, for example. The companies quickly put down roots and can offer other services, which exponentially increases the cost to healthcare providers.
“AI scribes are the tip of the spear by which these companies are entering the systems,” Pearson said. “But once they are there, you start to see these vendors quickly spreading out into other parts of the business, whether it be staffing, whether it be rev cycle management, any number of places,” Pearson said.
She continued: “If we can get clarity on sort of those ancillary areas where AI is having an impact, you could potentially save smaller providers a lot of time and money.”