Off The Record


Quick Advice for the CIO: When is Speech Recognition a Viable Choice?

From the Desk of Kendall Tant, CEO, iData

 

With the continued migration toward the almighty EHR, it is incumbent on the CIO to make sound, cost-effective choices in Clinical Documentation choices.  Oh, and 98% isn’t good enough.  Accuracy is still King.    You’ve heard the chattering about Speech Recognition (SR) – and you’re wondering if it merits a closer look.

As always, iData’s here to sort through the clamor and offer some direction.

Of course you’ve already encountered SR technology.  Your own voice was captured and directed to the appropriate desk when you called the bank today and spoke your selections.  It works – right?  Fair enough.  But the stakes are high here.  There’s a big difference between, “no history of heart attack” and “blisters on his back.”  You’ve got to work with a partner you can trust.

Voice Recognition:  How it Works

Traditional Clinical Documentation (Medical Transcription) works like this: the clinician’s data is recorded via handwritten, hand-held device, laptop, phone or computer workspace.  The information is translated, typed, coded and captured into a report the clinician approves.  Essentially, Speech Recognition (SR) skips the step of typing.  But it doesn’t mean a real live human isn’t necessary.  Here’s how SR works:

1.     The clinician dictates and records patient information into a handheld device.

2.     A voice file is sent directly to a voice recognition server.

3.     After the file is transferred, editors review the notes for accuracy.

4.     Finally, clinicians then review these corrections and sign the stamp of approval.

With SR, The Medical Transcriptionist’s (MT) task becomes that of editor, instead of transcribing verbatim. M*Modal’s Speech Understanding™, for example, fosters medical documentation by transforming narrative into electronic documents that are structured, clinically encoded, searchable, and shareable.

Check out an example of how M*Modal’s software converts data into a structured report:

http://www.mmodal.com/technology.jsp

Many have identified the traditional MT’s job as dying a slow death in light of the evolution of SR.  We say – not so fast.

While some hospitals and practices champion the efficiency and cost savings associated with SR technology, it is just one of a range of solutions we offer at iData.  SR technology is evolving.  There are still issues to tackle.  A clinician’s mumbling, ancillary noise, pronounced accents, and poor speech habits can compromise the accuracy of the data.  (Again – a human touch is still required!)

For the time being, SR offers a cost-effective choice to practices where the data is often repeated and easily categorized and coded, in specialties such as emergency medicine and radiology, for example.  As with any other technology, SR can be assimilated into a range of solutions, so the accuracy so vital to “meaningful use” is safeguarded.

In a large hospital, for example, a “hybrid” approach works well, in which some departments choose SR and assign editors to scour the reports for accuracy.  Other departments may choose to stick with the traditional route, merging the options for a solution that improves efficiency and reduces costs, while preserving the integrity of the data.

As always, we welcome your feedback.  Leave a comment, and keep the conversation going.  Have you used SR technology?  What works in your practice?  (And don’t forget to sign up for our RSS feed, so you can receive blog posts in your mailbox!)

Best,

Kendall

 

 

Leave a Reply

iData image