ChatGPT Health Adds Epic Integration for Clinicians to Access Patient Data

OpenAI is expanding the capabilities of ChatGPT Health by integrating it with Epic’s electronic health record (EHR) system, enabling clinicians to access and review patient information through AI-powered tools.

The integration enables healthcare professionals to import patient data and ask questions about medical records using ChatGPT. OpenAI said the connection can help clinicians review appointment notes, laboratory results, medications and specialist documentation, while also summarising information and preparing for future visits.

In some healthcare systems, ChatGPT will be integrated directly into existing EHR workflows, allowing clinicians to use AI features for pre-visit reviews and clinical timelines without leaving a patient chart. OpenAI said the integration provides read-only access to health records, meaning the AI system does not write information back into medical records.

ChatGPT Health expands healthcare data tools.

OpenAI is also introducing a Healthcare Public Data plug-in that can retrieve information from sources including ClinicalTrials.gov, CMS Coverage, RxNorm, DailyMed and PubMed.

The company said the tool can help healthcare organisations organise information related to clinical trial eligibility requirements, medication identifiers, coverage policies and provider records.

OpenAI said organisations with a Business Associate Agreement can also use ChatGPT Work, Codex, apps and connectors within their workspace for healthcare-related workflows that require compliance controls.

OpenAI adds healthcare AI safeguards.

The Epic integration follows OpenAI’s broader expansion of ChatGPT Health, which became available to U.S. consumers earlier this year. The company said users submit hundreds of millions of health-related questions to ChatGPT every week.

Despite expanding into healthcare systems, OpenAI has said AI tools are not intended to diagnose medical conditions or replace clinical decisions. The company said it evaluated responses from physicians across multiple clinical use cases, including medication reviews, visit preparation and patient handoff summaries.

OpenAI reported that its evaluation collected more than 4,300 physician responses across 27 clinical scenarios and found that 99.1% of responses were considered safe. The company has also faced scrutiny over cases where users alleged harmful health-related guidance from ChatGPT.

In one recent case, a Florida pastor filed a lawsuit alleging that ChatGPT provided dangerous advice, according to reporting on legal claims over the chatbot’s health recommendations. OpenAI has continued to emphasise that AI systems should be used as support tools rather than as substitutes for professional medical judgment.

The Epic integration represents another step in OpenAI’s effort to bring AI tools into healthcare workflows while maintaining limits around patient safety, data access and clinical responsibility.