EHR vendor Cerner is expanding its relationship with AWS for added AI and machine learning capabilities as well as tools like a digital scribe that can translate speech to text.
AWS re:Invent 2019 showcased news about the tech giant's push into healthcare, including a new layer to the Cerner-AWS partnership.
The EHR vendor is furthering its partnership with AWS, already the vendor's preferred cloud provider, specifically around machine learning and AI, Cerner CEO Brent Shafer said while speaking at the AWS re:Invent 2019 conference in Las Vegas.
By applying machine learning to de-identified EHR data in the AWS Cloud, Cerner will build models to predict causes of hospital re-admissions. Cerner is also using a new voice transcription tool from AWS to build a digital virtual scribe that captures information during a patient visit.
Tanuj Gupta
Tanuj Gupta, M.D., vice president of Cerner Intelligence, said Cerner's initial partnership with AWS focused on migrating its customers to the public cloud. Since then, the EHR vendor has been exploring how to deepen the relationship, from modernizing Cerner's infrastructure to using AI at scale.
At AWS re:Invent, Cerner provided an initial look at its virtual scribe, which Gupta said is currently in alpha testing. For patients who choose to have their appointments recorded, software will capture their conversation with practitioners and use it to fill out the patient chart, adding information about allergies and medications as well as health problems.
Cerner is using Amazon Transcribe Medical, a machine learning service from AWS, to enable the virtual scribe, according to Gupta. The goal of the product is to eventually create a medical note, he said.
"It keeps the doctor from having to do the data entry while speaking with a patient," said Gupta, whose team focuses on AI and machine learning. "When the doctor goes to the chart, he or she can accept or reject whatever was filled out."
Gupta said the long-term benefits of the Cerner-AWS partnership will be the ability to scale AI and machine learning algorithms, as well as apply big data analytics to find new patterns in patient care.
Big tech is teaching us how to leverage big data; we are teaching big tech how to take a 'do no harm' mentality.
Tanuj Gupta, M.D. Vice president, Cerner Intelligence
"What we haven't done as a healthcare industry, to any large degree, is leverage big data for insights," Gupta said. "Big tech companies outside of healthcare, they have done that to a huge degree, and we have seen a huge benefit from doing it. When you look at Amazon, they say, 'People like you ordered this.' Well, we don't do, 'Patients like you who have similar conditions improved by doing these things.' We don't do that in any big way yet."
Gupta believes a partnership like Cerner-AWS advances not only the field of healthcare, but the field of big tech as well.
"Big tech is teaching us how to leverage big data; we are teaching big tech how to take a 'do no harm' mentality," Gupta said. "I think a partnership between a company like Cerner and AWS could be very good because you can leverage the benefits of us having a healthcare mentality and the benefit of big tech understanding how to leverage big data."
Jeff Becker
Jeff Becker, senior analyst at Forrester Research in Cambridge, said the next layer to the Cerner-AWS partnership is indicative of a trend among EHR vendors wanting to do more with voice and AI capabilities, which can be enabled by tech giants like Amazon.
"Everybody wants to do more with voice and machine learning, it's not just Cerner," Becker said. "Meditech just partnered with Google Cloud Platform, Cerner is partnered with AWS, Epic's not saying it, but they're partnering with Azure."
Amazon Transcribe Medical
Becker pointed to Amazon Transcribe Medical, which AWS introduced at re:Invent, as an example of how tech giants are meeting healthcare's needs head-on. AWS described the medical transcription service as a "new speech recognition capability of Amazon Transcribe."
Transcribe Medical is designed to convert speech from clinicians and patients during an office visit into text, reducing the amount of time clinicians spend doing clinical documentation. According to the announcement, Transcribe Medical has been trained to understand medical terminology, enabling clinicians to "speak naturally," and it notes that the service is HIPAA eligible.
Becker said Microsoft and voice technology company Nuance are also building out a clinical voice-to-text product called Project EmpowerMD, a product that's like Transcribe Medical. He said it's refreshing to see tech giants addressing long-standing issues in the healthcare community.
"These are big technology players and they are interested in things like physician burnout," he said. "It's great that some of our industry problems are finally getting that level of awareness from the major tech players because we need it. We need tech to take healthcare seriously and start helping us solve some of these problems."