Dive Brief:
The Joint Commission is partnering with the Coalition for Health AI to develop guidance for deploying artificial intelligence tools in healthcare, the organizations said this week.
The safety and quality standards group will work with CHAI, a nonprofit that aims to set guidelines on AI adoption, on playbooks for using the technology in the sector. The first guidance should be released in the fall, and the Joint Commission will later release a certification program for evaluating responsible AI deployment.
The guidance aims to help health systems safely adopt, manage and monitor AI tools in healthcare settings — which isn’t an easy task, Dr. Brian Anderson, CEO of CHAI, told Healthcare Dive. “It takes real resources from a people and from a technology standpoint, and being able to do that in a financially sustainable way has been a real challenge for even some of the biggest health systems in the U.S.,” he said.
Dive Insight:
Founded in 2021, CHAI has 3,000 member organizations, including academic medical centers, rural health systems, technology firms and startups.
The group has developed resources for AI in the sector, like a guide to development and deployment as well as model cards describing AI tools and a registry where healthcare firms can access them.
CHAI’s work will inform the playbooks developed with The Joint Commission — an important partner given its broad reach with hospitals, Anderson said. The group accredits more than 23,000 healthcare organizations and programs in the U.S., according to a press release.
The partnership comes as AI has become an increasingly exciting technology for the healthcare industry. AI tools could help hospitals cut down providers’ heavy burden of administrative work, like documenting patient visits, assisting with prior authorization requests or sifting through clinical data, proponents say.
But there are risks, including AI hallucinations, inaccurate information or biases embedded in models that could worsen health disparities. Additionally, health systems will have to continue to monitor their AI tools over time, as the assumptions underlying the models could change and worsen their performance.
That’s a significant undertaking for providers, Anderson said.
“These were systems that were rolling out pilot programs and AI governance processes for one or two or three models, and they were quickly finding that these kinds of approaches cost between $1 [million] and $2 million,” Anderson said.
The guidance developed by CHAI and The Joint Commission will likely help tackle some of this work, like how to set up governance processes and committees, the technical infrastructure needed to monitor performance and the value of a strong partnership with model vendors, he said.
The playbooks should also apply to different types of providers. For example, a critical access hospital will have fewer resources at its disposal, so guidance could recommend it have a smaller governance team while working with a referral hospital partner, Anderson said.
The first guidance will be available in the fall, with the certification program coming later. The program will be the “next chapter” of The Joint Commission’s Responsible Use of Health Data Certification rolled out in 2023, Dr. Jonathan Perlin, president and CEO of The Joint Commission, said via email.
