TEFCA Reaches 1.7B Transactions Amid AI-Ready Data Push

More than 1.7 billion transactions have now taken place through the Trusted Exchange Framework and Common Agreement (TEFCA) as federal health technology officials focus on making that data usable across the healthcare system. 

“TEFCA connects more than 100,000 organizations with more than 1.7 billion transactions under common expectations for exchange. And as that exchange grows, the question changes. It is no longer simply, ‘Can we move data?’ Increasingly, the question is, ‘What can we do with it now?’ Because data liquidity isn’t the end goal, better healthcare is,” said National Coordinator for Health IT Thomas Keane. 

The milestone comes as the Centers for Medicare & Medicaid Services marks just over a year since launching its Health Technology Ecosystem, which officials said has also expanded rapidly as the agency works with industry, patients and caregivers on new ways to make health information more accessible and useful.  

What’s Next for TEFCA

Keane said the health IT industry is reaching an “inflection point” after years of concentrating on digitizing healthcare and connecting systems that were never designed to communicate with one another. The next phase, he said, is about making information “computable.”

He said new certification capabilities can bring patient-specific prescription benefit information into the prescribing workflow, while electronic prior authorization capabilities can allow clinical information to move through modern APIs. 

“The goal is to make unnecessary administrative work begin to disappear into the infrastructure,” he said. “Real-time prescription benefit and prior authorization are different use cases; one is about affordability, one is about administrative burden. But underneath, they depend on the same thing: information that is standardized, that is computable and that is available when it is needed. That is the shift from moving information to putting information to work.”    

One Year of Progress for the Health Technology Ecosystem  

CMS’s Health Technology Ecosystem has expanded considerably since its launch more than a year ago, according to Amy Gleason, U.S. DOGE Service administrator and CMS deputy administrator. It has grown from about 60 companies that initially made commitments to more than 900 companies and 300 patients and caregivers participating in work groups. The groups focus on price transparency, real-time benefits, modern scheduling, clinical trial matching, bulk FHIR, pharmacy systems, provider assistance and diagnostic imaging. 

One of the Health Technology Ecosystem’s initiatives, the Medicare App Library, now includes more than 30 apps certified by CMS, enabling patients to get their records, share them with their doctors and use AI to get personalized advice.  

The next main focus, Gleason said, is “one of the biggest pain points in healthcare” — prior authorization. In May, CMS launched a working group to streamline the electronic prior authorization process for medical items and services. It faces a Jan. 1 deadline when certain health plans regulated by CMS must implement and maintain a prior authorization API.

AI’s Role in Health IT  

Gleason said AI in healthcare is only as good as the data it can reach. She cautioned against using AI to just digitize the healthcare system’s failures.

“We can use AI to speed up prior authorization, or coding, or documentation. Those things do matter, and they do make a difference to the people working in the system,” she said. “But if we put AI on top of fragmented data and disconnected providers, we get faster fragmentation and more efficient bureaucracy. If we add intelligence to a system that still relies on the patient to be the glue, then we just digitize its failures.” 

Once data is reliable and interoperable, Gleason described a future in which patients could have a trusted health companion available 24/7 that works from their medical records and other information they authorize it to use. Such a system could help patients understand their health, determine when they need care and recognize when a human clinician should become involved. 

She identified four requirements for making that possible: earning public trust, building the underlying data foundation, establishing clear rules and paying for technologies that demonstrate better outcomes and lower costs. 

“Say you get a new lab result, your companion works from your actual record with your permission, and it knows all about you because you’ve told it things, you’ve given it your wearable information. It has your medical records, and it can figure out whether you need to wait or whether you need care tonight. And it knows when to hand you to a human,” Gleason said.