athenahealth launches tool to connect patient records to Anthropic Claude

Bridging records and AI

athenahealth has introduced a new enterprise server integration called athenaPatient MCP. This tool is built on the Model Context Protocol, which is an open standard for connecting data sources to AI models. It allows patients to link their official electronic health records directly to Anthropic’s Claude.

The initiative addresses a growing trend in how people manage their personal health information. A Pew Research Center survey from August 2026 found that 34% of U.S. adults already use artificial intelligence assistants for health-related tasks. These activities range from exploring symptoms to seeking interpretations of medical diagnoses.

By providing a direct connection, athenahealth aims to help patients get more value from their own healthcare data. The goal is to reduce the complexity of medical records and help individuals take a more active role in their own care. The integration is currently available for Claude Pro and Claude Max subscribers in the United States.

How the MCP middleware works

The athenaPatient MCP functions as a governed middleware layer that sits between the medical record and the AI. Instead of using brittle screen-scraping or dumping an entire medical history into a prompt window, it serves as an on-demand data broker. This means the system only retrieves clinical records dynamically when a patient asks a specific question that requires historical context.

The architecture follows a principle of zero static pre-loading. This prevents the patient's entire medical history from being loaded into Claude’s training set or prompt window upfront. Instead, the server structures the data only at the moment it is needed to answer a user's query. This method ensures the AI is grounded in verified clinical history rather than general patterns.

The protocol uses a standardized client-server convention transported over JSON-RPC. This technical framework allows for a more stable and secure way to move information between the health record and the AI model. It provides a structured way for the AI to reference discrete data fields without exposing the entire database at once.

Scope of clinical data

When a patient uses the connector, Claude can reference a wide variety of discrete data fields from their chart. This includes laboratory results, vital signs, and active medication regimens. It can also access problem lists, immunization histories, and outpatient procedures.

The integration goes beyond simple lists to include more complex narratives. The AI can reference encounter summaries, diagnostic imaging narratives, and longitudinal care plans. This breadth allows for a more comprehensive conversational interpretation of a patient's health journey over time.

By pulling from these specific fields, the tool provides a more accurate context for the AI's responses. Patients can ask questions about their specific history, and the AI can pull the relevant data points to form a coherent answer. This is designed to be more reliable than a patient trying to interpret these documents on their own.

Security and data privacy

Security and consent are managed entirely by the consumer rather than the service provider. To use the tool, patients must authenticate using their official athenahealth patient portal credentials. They must also select specific clinical data scopes and have the ability to revoke the model's authorization at any time.

The interaction layer is designed with strict governance to protect personal health information. The system operates under a zero-data-retention commitment, meaning the information retrieved via the endpoint remains restricted to the active session. This prevents sensitive data from entering the foundation model's pre-training corpora.

This setup ensures that the AI does not learn from the patient's private medical data. By keeping the data segregated, athenahealth and Anthropic aim to provide a responsible way to use AI. It allows for the benefits of conversational AI without compromising the privacy of the underlying medical records.

The future of patient care

The launch of athenaPatient MCP represents a significant step in integrating AI into the patient experience. Paul Brient, chief product and operations officer at athenahealth, noted that the goal is to help patients get value from their own healthcare data. He highlighted the importance of working with a partner that focuses on both power and responsibility in AI.

This tool marks a shift away from static data exports, which can be difficult for non-experts to navigate. By creating a conversational bridge, athenahealth is moving toward a model where patients can interact with their records in plain English. This could potentially reduce the burden on healthcare providers who often have to explain the same records repeatedly.

As more patients begin to use AI for health-related purposes, these types of integrations will likely become standard. The success of athenaPatient MCP will depend on how well it balances ease of use with the rigorous demands of medical data privacy. It sets a precedent for how other healthcare providers might connect their proprietary data to large language models.

Try it today

If you are a Claude Pro or Claude Max subscriber in the United States and have an athenahealth account, you can check your Claude account settings to see if the athenaPatient MCP connector is available for activation.

Watch

How other EHR providers will respond to the Model Context Protocol standard.

Whether more healthcare providers will adopt zero-data-retention commitments for AI tools.

How patients will react to the accuracy of AI-generated interpretations of their clinical history.

athenahealth Launches athenaPatient MCP to Bridge Official Patient Records into Claude