Power of Voice Recognition Software

Good heavens! What’s this? Print!

Charles Dickens, Somebody’s Luggage

Television is now using speech recognition to instantly record in print the comments of what people are saying as they speak.  The print appears below peoples’ words as they speak.   This translation of talk into print is occurring on cooking shows, on panel discussions, late night and comedy shows, and TV series required translations into other languages.

 This is a giant step forward.  It supplements speakers’ comments and allows their spoken words to be retrieved after editing (editing is required because voice recognitions translation are imperfect).

Print + Algorithms - Basis for Power of Net

Print and the power of software to preserve print instantly and permanently, over the Net is basis is the basis for the transition to the Information Age.  Clinical algorithms make this possible.

Take Google.  Its software allows the Net to preserve print and to store endless amounts of  print on the cloud. And to make it available to all who wish to read it.    

Creature of Print 

I am a creature of print.   I have written twelve books, 4300 blogs, nearly 1300 tweets, thousands of editorials and articles, and committed perhaps 2 million words to print.  I am guilty of seborrhea, or, more precisely, printorrhea.  Print is the need of my soul.

Data Can’t Replace Voice Narrative

As a doctor, I am sensitive to the lack of printed narratives in electronic health records.    Doctors like to dictate summaries. They do not like clicking data entries or reading reams of data.   With your voice, you can succinctly sum up a patient’s history and the summary your findings.   You can’t do that with data.

That’s why I wrote the following blog back in 2010, in which I wrote about the work of Nuance, a speech recognition company.

Previous Blog

For Fans of Electronic Health Records

Nuance, a very slight difference in meaning, feeling, or tone.

Dictionary definition of Nuance

I have never been a big fan of electronic health records. EHRs lack nuance. With EHRs physicians can’t express themselves in plain English, just in data bytes. EHRs too often generate unreadable numeric gibberish. They fail to pass the test of useful narrative information.

So much for my electronic angst.

As always, I may be wrong. Now, there may be a technological breakthrough. In the March 14, 2010 NEJM, Drs. Gorden Schiff and David Bates of Harvard write “Can Electronic Clinical Documents Help Prevent Diagnostic Errors?”

Their answer is "Yes". Improved speech recognition technology now makes it possible for physicians to clearly describe and communicate the patient’s story without typing or handwriting, while at the same time, as a bonus, decreasing diagnostic errors.

The two authors then list other benefits of EHRs speech recognition technology.

• Gain access to information in narrative context

• Record and share clinical assessments in plain language

• Maintain a dynamic and current patient history

• Maintain problem lists

• Track medications

• Track tests

• Coordinate and control care

• Enable follow-up

• Provide follow-up to clinicians upstream

• Offer second opinions

• Increase efficiencies

The biggest benefit of computerized speech, from my perspective, is that speech recognition allows physicians to tell the patient’s story, past and current, without scanning reams of data. Perhaps I am impressed with Nuance Healthcare software because I believe in old-fashioned story telling and the power of narrative.

In any event, here’s how Nuance Healthcare, speech recognition software developers, and a disruptive player in the HIT space, interpret the work of Drs. Schiff and Bates.

“Dr. Schiff and Dr. Bates struck a particularly relevant chord with their paper on the impact that electronic clinical documentation can have on preventing diagnostic errors,” said John Shagoury, executive vice president and general manager, Nuance Healthcare. “More than 150,000 physicians use our speech recognition technology to document patient encounters without having to type or handwrite. The majority of these doctors will tell you that speaking their medical notes is not only faster, but it allows doctors to include more information on their patients. It’s wonderful to see the free-text narrative, alongside EHR point-and-click templates, being recognized as highly important and valuable to improve patient care, as well as to improve physician and patient interactions. One customer of ours, The Fallon Clinic, saw the quality of medical notes improve by 26 percent when they were created with speech recognition.”

I close with this verse, which tells the story more succinctly than prose

With EHRs, there really nothing like human speech,
to capture nuances beyond ordinary data’s reach.
There’s more to telling a patient’s story beyond data,
Which unwittingly may produce unexpected errata.
So now of EHRs I can speak and preach.