Amazon Web Services (AWS) has introduced AWS HealthScribe, a generative AI-based service aimed at automating clinical documentation. This service allows health IT providers to use a single API to create transcripts, extract important details, and generate summaries from clinician-patient interactions, all of which can be entered into the Electronic Health Record (EHR).

AWS HealthScribe enables health IT developers to integrate generative AI capabilities into their applications without the need to manage the underlying machine learning (ML) infrastructure or train large language models (LLMs). Initially, the service will support generative AI capabilities for general medicine and orthopedics specialties.

The main goal of AWS HealthScribe is to facilitate the responsible deployment of AI systems by providing the source of each line of generated text from the original conversation transcript. This ensures that providers can review the notes before entering them into the EHR, ensuring accuracy and reliability.

According to Bratin Saha, the vice president of machine learning and artificial intelligence services at AWS, the aim is to empower healthcare professionals to focus more on creating innovative clinical care and research solutions for patients, while reducing the time spent on building and maintaining health data capabilities. Documentation is often a time-consuming process for healthcare professionals, and AWS HealthScribe aims to use generative AI to alleviate this burden.

Several health IT vendors have already committed to using HealthScribe, including 3M Health Information Systems (HIS), Babylon, and ScribeEMR. These partnerships aim to transform clinician workflows and streamline clinical documentation and billing processes, ultimately benefiting healthcare organizations and improving patient care.

Article written by Hannah Nelson



EHR Intelligence