A new suite of AI agentic tools from Amazon Web Services (AWS) is available to assist healthcare organizations with freeing up staff from administrative tasks like patient scheduling and medical data summaries. 

Amazon Connect Health includes AI agents that are patient-focused and agents that have provider functionalities. The patient-focused tasks include verification of patient identities and scheduling patient appointments. The provider functionalities include summarizing patient medical histories, creating clinical notes from patient-provider conversations, and pulling out medical codes from clinical documentation.

For patient scheduling, the AI agents can be customized regarding requirements for patient verification and appointment scheduling based on physician/practice/organization preferences. For example, before scheduling an appointment, the agent will check to see if prior authorization is required and ensure it has been approved by the patient’s insurance.

The scheduling AI agent is also trained to detect frustration on the patient’s part if there is difficulty in scheduling an appointment and will route the patient to a human agent if needed. Caller requests for a human will be routed to a human agent.

On the provider side, the AI agent can create patient summaries from EHR and information exchange sources to save providers from having to dig through charts at or prior to a patient visit. The suite includes an ambient documentation assistant to create notes of clinical visits. This task is an offshoot of Amazon’s HealthScribe tool, with the added ability to create after-visit summaries and recommend appropriate codes for billing and diagnosis.

It’s important to note that all clinical documentation created by Connect Health AI agents requires human review for approval. In addition, there are clinicians on staff at AWS who analyze AI outputs.