OpenAI is integrating ChatGPT for Healthcare with Epic’s electronic health record system, giving clinicians read-only access to authorized patient information such as clinical notes, medications, encounters, conditions, and laboratory results. The integration can also place ChatGPT directly inside supported EHR workflows so clinicians can review patient information without leaving the chart.
OpenAI announced the integration on September 1. Epic says its software contains records for more than 325 million patients, giving the integration access to a large base of healthcare organizations where administrators enable it.
Clinicians Can Review Patient History Inside ChatGPT
The Epic connection allows clinicians to ask questions about patient history, recent changes, medications, laboratory results, specialist recommendations, referrals, and unresolved issues. ChatGPT can summarize information from the authorized record and point users back to supporting chart data.
In supported deployments, clinicians can also access ChatGPT directly inside the EHR interface for tasks such as pre-visit reviews and clinical timelines. OpenAI said the connection is read-only, meaning ChatGPT cannot write information back into the patient chart.
Access requires an administrator-configured Epic EHR application, an individual Epic login, and the clinician’s existing chart permissions. OpenAI said organizations should also have the appropriate Business Associate Agreement and workspace configuration before using the integration with protected health information.
New Plug-In Connects Nine Public Healthcare Sources
OpenAI is also introducing Healthcare Public Data, a read-only plug-in that searches nine official healthcare sources. These include ClinicalTrials.gov, PubMed, DailyMed, CMS Coverage, and RxNorm, along with other datasets covering medications, providers, healthcare facilities, and Medicare information.
Healthcare teams can use the plug-in to review clinical trial eligibility, medication identifiers, coverage policies, research, and provider records. Eligible ChatGPT for Clinicians users in the U.S. can access the public-data plug-in, while eligible ChatGPT for Healthcare and HIPAA-enabled Enterprise workspaces can use both Epic and Healthcare Public Data in ChatGPT and Codex.
OpenAI said physicians evaluated 4,363 responses across 27 clinical use cases, including medication review, handoff summaries, pre-visit review, and clinical timelines. Physicians rated 99.1% of those responses as safe, although OpenAI continues to state that ChatGPT is not intended to replace clinical judgment for diagnosis or treatment.
The company also said physicians across 60 countries have reviewed more than 700,000 model responses covering 26 medical specialties. Last month, OpenAI expanded Health in ChatGPT to eligible U.S. consumers aged 18 and older.
Featured image credits: Wikimedia Commons
For more stories like it, click the +Follow button at the top of this page to follow us.