ChatGPT for Healthcare unveils Epic integration as part of data consolidation strategy
OpenAI’s ChatGPT for Healthcare division announced this week that it will be deploying a new integration that gives its large language model access to medical records stored by Epic Systems, the largest electronic health record vendor in the country.
Epic’s EHR holds data on nearly all Americans—325 million patients. The connection will allow providers to utilize the AI to examine those records as they care for individual patients, OpenAI said in a statement. But beyond that, it will allow ChatGPT and the EHR to work together as part of a single workflow, effectively adding the AI chatbot to the record system during clinical encounters.
Much like other AI that is pulling from medical records, the integration will allow a provider or clinician to access details that are important to a patient encounter: current prescriptions, known medical history, any record of allergies, lab results, and other documentation for past patient care encounters.
According to OpenAI, ChatGPT can automatically summarize much of that information, allowing providers to more easily glean insights. Many of the features of the integration will allow a provider to do this, while adding to the existing medical record and developing a care plan, without closing a patient’s chart.
The company said this will allow providers to focus more on patients, and aid in their decision making by ensuring they have all available information on the person they’re caring for. It is part of a larger effort by ChatGPT for Healthcare to bring complete medically relevant data on everyone to a single place.
“Healthcare organizations need AI that works across the systems and information central to care and operations. Patient context, medical evidence, public healthcare data and organizational knowledge often live in different places,” the company wrote. “Connecting these sources in a governed workspace helps teams find the right information, understand it in context and put it to work across the business.”
To that end, OpenAI confirmed a new Healthcare Public Data plug-in will be able to access information in databases used by the government and community organizations, including ClinicalTrials.gov, PubMed, certain Centers for Medicare & Medicaid Services records and more.
In some cases, this data can be used to help providers to understand what options are available for a patient, as well as provide them with knowledge about emerging treatment protocols, in addition to up-to-date details on what devices and pharmaceuticals are in the regulatory approval pipeline.
“ChatGPT for Healthcare already helps teams answer clinical questions and synthesize medical research with trusted clinical search across thousands of medical sources,” OpenAI wrote.
“Teams can work with specific records, fields, identifiers, and versions while focusing the task on these authoritative sources,” it added. “This makes it easier to compare and verify precise information such as trial eligibility criteria, medication identifiers, coverage policy versions, and provider records without searching each source separately.”
Pilot launches find success
The full suite, which is now open to all customers, comes after it piloted the tech with a limited number of health systems—one of which was University of California, San Francisco (UCSF) Health.
In the statement from OpenAI, the academic health network provided some details on how it has been using ChatGPT for Healthcare:
“As a pilot partner, we’re exploring how the new EHR integration with ChatGPT for Healthcare can help clinical teams understand what has changed and what matters most across a complex patient record,” Suresh Gunasekaran, President and CEO, UCSF Health, said. “By bringing relevant information together more quickly and comprehensively, the technology has the potential to reduce time spent synthesizing data and give clinicians more time with patients.”
“We’re also engaging frontline teams to validate these capabilities in practice and help shape where they can add the most value.”
OpenAI initially announced its healthcare-specific ChatGPT workspace in January.
Since then, the company said pilots have shown the AI is relatively safe and accurate. The company said over 4,300 responses from physicians revealed that 99.1% of the time, the AI gave a safe response.
The lack of perfection means a human will still need to be involved. The company has been adamant that its AI is not designed to replace review by an actual provider.
Chad is an award-winning writer and editor with over 15 years of experience working in media. He has a decade-long professional background in healthcare, working as a writer and in public relations.