ZDNET’s key takeaways
- Athenahealth patients can consult Anthropic’s Claude.
- The partnership provides the AI with access to sensitive health information.
- Athenahealth claims the data cannot be used to train Anthropic models.
A major US healthcare technology provider just teamed up with Anthropic to connect patient health records with Claude.
Athenahealth, a healthcare technology provider serving over 20% of the US population, launched a Model Context Protocol (MCP) compatible server that connects to Anthropic’s Claude. The server, named AthenaPatient MCP, allows the AI model to gather information from health records and answer related questions. The process is secure and requires patients to sign in with their account credentials
AthenaPatient MCP is available only to Claude Pro and Claude Max users, which cost $20 and $100 per month, respectively. Not all Athenahealth patients will have their data shared with Claude, and not all Claude Pro and Claude Max users must opt in. Users must give Claude their authenticated Athenahealth credentials, can choose which health information Claude can access, and can revoke access at any time, according to the press release.
Also: The wearable health boom is creating a data overload for doctors – what happens next
Once authorized, Claude takes patients’ information like lab results, vitals, medications, health conditions, immunizations, visit history, diagnostic reports, allergies, procedures, and care plans into account when answering questions.
The AI partnership does not provide medical advice vice president of product management, data & ecosystem at Athenahealth, who responded to ZDNET in an email. It also should not replace a clinician’s professional judgment or medical advice. “Patients should contact a qualified healthcare professional with questions about diagnosis, treatment, or changes to their care,” Lambson said
It appears that the AthenaPatient MCP tool is designed to help patients better understand the information provided by health professionals, rather than to give them medical advice. So, one may ask AthenaPatient MCP about their latest blood test results, such as “What was my most recent red blood cell count?” but users shouldn’t expect Claude to explain how an abnormally high or low count may affect them.
Not the first of its kind
Partnerships between big tech and big healthcare aren’t new. In fact, the collaboration between Athenahealth and Anthropic follows a larger trend across the health tech industry. Earlier this year, Microsoft teamed up with Mayo Clinic to develop an AI model for healthcare; Boston Children’s Hospital forged a partnership with OpenAI to develop an enterprise AI system for use across research, clinical, and administrative teams, according to a press release.
Also: Before you buy a smartwatch or smart ring, consider what you’re giving up
A growing number of Americans are consulting AI for medical advice. A recent survey from Pew Research Center found that 34% of US adults have talked with a chatbot for medical advice. This partnership makes the process slightly more secure by authorizing patient account credentials before granting the model access to health records.
Health data is extremely valuable, and tech companies are increasingly interested in partnering with healthcare providers to access it. Much of the “AI could cure cancer” messaging we hear relies on these humongous companies obtaining enough health data to train their AI models and link disease data to patient data. This is not to say, however, that solving this complex issue is the motive behind the partnership between Athenahealth and Anthropic.
Also: Asking AI for medical advice? There’s a right and wrong way, one doctor explains
A more reasonable explanation for such a partnership is that an AI that can answer patient queries could eliminate some of the admin work doctors do between patient visits, like answering questions about test results, giving Athenahealth doctors some time back.
Will partnerships like Anthropic and Athenahealth’s improve patient outcomes? That remains to be seen, but a recent survey by Pew Research Center found that Americans aren’t that confident on the matter. The largest shares of respondents still preferred that a human oversee AI when it’s used in a clinical context.
