Next-Generation Voice Recognition for EHRs

Samuel L. Jackson makes it look so easy. iPhone™ in hand, the actor manages all his activities just by talking to “Siri” in a popular TV commercial.

Fortunately, with recent advances in voice recognition technology, physicians and other medical providers can look like superstars, too.

Physicians are natural storytellers. Sometimes it’s easier to update a patient’s documentation by simply speaking to describe what’s happened. Voice recognition technology can capture spoken words and turn them into accurate text that can be easily reviewed and approved.

Using voice recognition technology saves time. Providers can quickly document patient care and outcomes while easily navigating between screens in the EHR. Plus, it saves money because there are no transcription fees.

Perhaps most importantly, physicians are able to build better patient relationships because, for example, they can record the story of how Joe Smith hurt his back, not just the specifics of the injury.

That’s why we’ve added voice recognition technology into Allscripts solutions.

Improving on the standard

Still, for the technology to be truly effective, we know there are some voice recognition issues specific to our industry:

•             The solution needs to recognize thousands of medical terms.

•             The solution needs to work in loud environments.

We’re working closely with our clients and partners to solve these remaining issues.

Meanwhile, to give our clients a choice, we work with two voice recognition technology partners. The first option is Dragon Medical 360 | Network Edition from Nuance Healthcare with highly accurate speech-to-text and command-and-control capabilities designed for clinical settings.  In addition, Nuance’s Clinical Language Understanding leverages advancements in Natural Language Processing (NLP), medical Artificial Intelligence (AI) and speech recognition technology to create tailored solutions for healthcare informatics.  Clinical Language Understanding technology can be used to derive patient information from free text and to produce actionable data used to improve patient care and streamline workflow.

We also launched a new voice recognition solution in partnership with M*Modal. As part of the development process with M*Modal, one of the most exciting features is called Speech Understanding™. As the title suggests, the solution provides a deep understanding of the context to help identify meaning and intent. It’s a real-time transcription of the physician’s narrative at the point of care.

M*Modal Fluency Direct is also tuned for greater accuracy from the voice profiles of 200,000 physicians. It adjusts on the fly to differences in accents, dialects, cadence and sub-specialty terminology.

So, while Sam Jackson is planning a date by talking to his phone, physicians can simplify input of data into the EHR and improve patient outcomes. That’s a role every practitioner loves to play.

Future Possibilities

One intriguing application of this type of technology is not related to voice recognition at all:  applying this type of technology to existing text documents.

Suppose a new patient came in with a thick folder from the previous doctor. Of course we can scan those documents, but what if we also “read” the text into the EHR (commonly referred to as OCR)?  And if you apply this type of technology to the now recognized words – can you automatically pick out the patient’s Problems, Medications and Allergies? Of course they would still need to be validated – but isn’t that a better place to start the new patient’s records?  How much time could you save by saying, “are these correct?” instead of “type everything in”?

Are you using voice recognition technology in your EHR today? What do you wish it did better or differently? Help us continue to provide the best solutions. Leave your thoughts below.


About the author

Stanley Crane is Chief Innovation Officer for Allscripts. In his more than 30 years of healthcare and consumer-related software experience, he has led the development of award-winning software programs including electronic health record, electronic prescribing, web-based medication sales, online physician education, resource scheduling, financial systems, materials management, medical translation software and voice recognition dictation systems. Previous to his healthcare experience, Stanley was involved in Silicon Valley, where he held positions with many well-known software companies. As the General Manager of Lotus cc:Mail, he created the first remote mail products. He was also the Vice President of Engineering at WordStar International, and Director of Applications at Ashton-Tate, managing their Macintosh products as well as dBase IV. Before that, Stanley was a founder of two Internet startups – MaxMiles, an automated frequent flier mileage aggregator, for whom he built the first versions of the product; and Shopping@Home, a company that was acquired by Allscripts in 1999 to support medication sales.

3 COMMENTS on Next-Generation Voice Recognition for EHRs


Mike says:

06/25/2012 at 1:59 pm

Looks Good. Great work Samuel L. Jackson


dan says:

06/27/2012 at 6:52 pm

Very interesting, good to see people discussing this finally.


Primary Care says:

07/19/2013 at 1:20 pm

This is brilliant. There’s no reason we shouldn’t be converting over the tech breakthroughs used in consumer devices, like voice recognition, to our healthcare needs. Sometimes the solutions seem so obvious yet it still takes the innovators to connect the dots. Keep it coming


Your email address will not be published. Required fields are marked *