Amazon Web Services Offering Direct-To-EHR Transcription

Amazon Web Services has introduced a new line of business focused on tackling in on one of the most vexing day-to-day problems physicians face. The behemoth tech firm is rolling out a version of its new speech recognition technology designed to help developers integrate medical transcription functions into relevant applications like EHRs.

The new offering, Amazon Transcribe Medical, leverages AI to capture not only dictated documentation but also conversations with patients and translate what it hears into text. It comes with an API that can be used with voice-enabled applications and any device with a microphone on board.

To dictate, users call the API to open a secure connection over WebSocket protocol and start passing an audio stream to the service, which in turn sends back a stream of text in real-time, according to a press release announcing the launch.

Cerner has already begun working with Amazon to integrate Transcribe Medical technology into its own digital voice scribe. Using the technology, the digital voice scribe will “listen” to clinician-patient interactions and capture the dialogue in text form.

In addition to leveraging Transcribe Medical, Cerner is working with AWS to find ways to reduce patient readmissions and reduce the time physicians spend documenting patient visits. The partners are using de-identified patient data to pinpoint factors causing repeat hospitalizations.

Now, with the new announcement, it seems likely Amazon will be calling on its favorite health IT vendors to see whether they, too, are ready to take its transcription service for a spin. It’s hard to imagine many vendors turning it away, though it’s possible some will balk if it relies too heavily on its history outside of healthcare. A proposed deal with a hard-punching Amazon is as close to an offer you can’t refuse as it gets.

What takes all of this beyond yet another self-serving corporate announcement is the larger pattern emerging here. This announcement is just one of a handful of healthcare-related news coming out of Amazon over the past year or so. For example, in late 2018, Amazon Web Services rolled out Comprehend Medical, a tool using natural language processing and machine learning to pull together relevant information from unstructured text, and seems poised to keep hacking away at big problems in health data analytics.

The e-retailing colossus has also been working aggressively to build on $753 million acquisition of online pharmacy PillPack, and has shown a willingness to get down in the mud and fight traditional pharmacies for control of the business.

More recently, it announced the launch of Amazon Care, a primary care program bringing together telehealth and access to in-person doctor and nurse visits. While Amazon Care is still at the pilot stage, it’s being watched closely by potential competitors given its resources and consumer visibility.

Now, we’ll get a look at how Amazon does going head to head with voice transcription giants like Nuance, which is working with Microsoft to deploy “ambient clinical intelligence,” a bundle of services which include capturing clinical conversations doctors have with patients and integrating transcriptions into EHRs.

It’s worth remembering that Amazon, like its competitors, is just getting rolling on a product niche which is likely to go through many changes as AI matures. Still, as I’ve said before, the sheer mass, influence and resources it brings to the table make anything it does worth a look. If any Big Tech company can bust through the walls that hold back outsiders from health IT success, Amazon can.

About the author

Anne Zieger

Anne Zieger

Anne Zieger is a healthcare journalist who has written about the industry for 30 years. Her work has appeared in all of the leading healthcare industry publications, and she's served as editor in chief of several healthcare B2B sites.

1 Comment

  • Great to see Amazon join the other tech titans Microsoft and Google to provide medical speech recognition tools as a backend platform for health tech companies. Great evidence that #Voice as a workflow is an ideal modality for clinical documentation. As speech recognition has become commoditized, the real challenge now is the ability to convert speech to intent and derive meaning from the conversation into a highly accurate clinical note that tightly integrates with the EMR.

Click here to post a comment