Glonce

OpenAI adds Epic health records and public data access to ChatGPT

productsThe Glonce editorsOpenAI🔥68Glonce rating68Impact65Novelty60Relevance85Surprise40Players85Credibility90
OpenAI adds Epic health records and public data access to ChatGPT

OpenAI has introduced two new capabilities for ChatGPT for Healthcare: an electronic health record (EHR) integration with Epic, and a Healthcare Public Data plugin. The Epic integration brings authorized patient context, appointment notes, lab results, medications, and specialist documentation, into ChatGPT so clinicians can ask questions such as what has changed since a patient's last visit, which recent lab results to review before an appointment, or what follow-ups and referrals remain unresolved. ChatGPT pulls together the relevant information from the authorized record, summarizes recent developments, and links back to the supporting chart entries. The integration works two ways: clinicians can bring EHR context into ChatGPT directly, or, in supported deployments, ChatGPT can be embedded into the EHR layout itself so the assistant is available without leaving the patient chart.

The Healthcare Public Data plugin adds dedicated connectors to nine official public healthcare sources; the announcement names five of them: ClinicalTrials.gov, CMS Coverage, RxNorm, DailyMed, and PubMed. It lets teams work with specific records, fields, identifiers, and versions from these sources rather than searching each one separately, for example checking trial eligibility criteria, medication identifiers, or coverage policy versions. OpenAI gives examples of use: a research team identifying actively recruiting trials on ClinicalTrials.gov, a pharmacy team confirming a drug's latest label and warnings on DailyMed, or a population health team combining research, trial data, and Medicare coverage information for a diabetes-prevention program.

OpenAI says it partners with hundreds of physicians across 60 countries, 49 languages, and 26 medical specialties to help define, measure, and improve ChatGPT's health responses, and that these physicians have reviewed more than 700,000 model responses to date. To validate the new EHR integration specifically, physicians evaluated ChatGPT's responses across 27 clinical use cases, including pre-visit review, clinical timelines, medication review, and handoff summaries, and rated 99.1% of 4,363 total responses as safe across all use cases. In a separate two-round evaluation, physicians reviewed responses to clinical questions drawn from large U.S. healthcare datasets across five connected data sources; for each of the five, more than 93% of responses were rated as having "good" or better accuracy.

Access differs by account type. ChatGPT for Healthcare customers must have a workspace administrator turn on the EHR integration and the Healthcare Public Data plugin. ChatGPT Enterprise customers need to contact their OpenAI account team to confirm eligibility and configuration for a Regulated Workspace. Eligible U.S. ChatGPT for Clinicians users can install the Healthcare Public Data plugin individually, but the EHR integration itself is not available on individual accounts. OpenAI positions the underlying workspace as governed for HIPAA-compliant use, combining the healthcare tools with enterprise controls such as role-based access, single sign-on, and audit logs, and, with an applicable Business Associate Agreement, allowing the same workspace to run ChatGPT Work and Codex alongside the new plugins.

Key facts

  • OpenAI launched an Epic EHR integration for ChatGPT for Healthcare, bringing authorized patient context (visit history, labs, medications, specialist notes) into ChatGPT, either as a source ChatGPT queries or embedded directly in the EHR layout.
  • A new Healthcare Public Data plugin connects nine official public healthcare sources; the announcement names five: ClinicalTrials.gov, CMS Coverage, RxNorm, DailyMed, and PubMed.
  • Physicians evaluating the EHR integration across 27 clinical use cases rated 99.1% of 4,363 total responses safe.
  • In a separate two-round accuracy evaluation covering five connected data sources, more than 93% of responses for each source were rated "good" or better accuracy.
  • OpenAI says it partners with hundreds of physicians across 60 countries, 49 languages, and 26 medical specialties, who have reviewed more than 700,000 model responses to date; the EHR integration is not available on individual accounts, only through ChatGPT for Healthcare or Enterprise workspaces.

Why it matters

This wires ChatGPT directly into the clinical workflow through Epic, the EHR system many U.S. hospitals and health systems already run on, rather than leaving the assistant to work from whatever a clinician types in. Paired with structured, direct access to public reference sources like PubMed and CMS Coverage, it pushes ChatGPT for Healthcare further from a general chatbot toward an enterprise clinical tool meant to sit inside existing chart-review and research workflows.

Who it affects

ChatGPT for Healthcare customers get both new capabilities once a workspace administrator turns them on. ChatGPT Enterprise customers have to go through their OpenAI account team to confirm eligibility and set up a Regulated Workspace. Individual U.S. ChatGPT for Clinicians users can install the Healthcare Public Data plugin on their own, but the Epic EHR integration is restricted to organizational accounts and is not offered to individuals at all.

How to use it

Enabling the features requires a workspace administrator (for ChatGPT for Healthcare) or direct contact with OpenAI's account team (for Enterprise, to configure a Regulated Workspace). With an applicable Business Associate Agreement in place, the same governed workspace can also run ChatGPT Work and Codex alongside the new healthcare plugins to support HIPAA-compliant work, backed by role-based access, single sign-on, and audit logs. No pricing or cost figures for either capability appear in the announcement.

How solid is it

OpenAI backs the release with two internal physician evaluations rather than an independent study: 4,363 ratings across 27 clinical use cases put the EHR integration's safety rate at 99.1%, and a separate two-round evaluation across five connected data sources put per-source accuracy above 93% for "good or better" responses in each case. The physicians involved are OpenAI's own network of partner clinicians, and the announcement does not disclose which five data sources were tested for accuracy or give a breakdown by source.

Risks and caveats

A 99.1% safety rating on 4,363 responses still leaves roughly 40 responses physicians did not judge safe, which is not negligible in a clinical setting. The announcement names only five of the nine public data sources actually connected by the plugin, leaves the identity of the other four unstated, and does not specify which five sources were used in the accuracy evaluation. No launch date beyond "today," no rollout timeline for Enterprise or individual access, and no pricing are given.