Voice Recognition Software: Making Technology Work for You

Doctors have long used pen and paper to record their patients’ histories, physical exams, and treatments prescribed. However, in the past 10 years, the electronic medical record has been gaining popularity as a method of recording patient data. This highly efficient and effective addition to the doctor’s armamentarium requires physicians to enter information into the record by either typing it in or using a mouse or stylus on the touchscreen of a tablet PC. One solution for avoiding these laborintensive

data entry methods is the use of voice recognition software, a relatively new technology, in concert with an electronic medical record. Although voice recognition software has been available since the ‘80s, it didn’t achieve a critical mass of users early

on because the programs available at the time were woefully inaccurate, which required users to heavily edit transcribed notes and records. Users also had to devote considerable time to “training” the programs to reliably recognize medical terminology. Currently, however, top-end voice recognition software programs boast greater than 95% accuracy and require little or no editing after an initial training and acclimation period. Furthermore, the best programs also come equipped with medical dictionaries with thousands of words commonly used in daily practice by physicians and specialists, eliminating the need to build the program’s vocabulary from scratch. Th anks to these and other advances in voice recognition software (VRS) technology, implementing a program into your EMR will improve the effi ciency of your practice and reduce your overhead expenses. Voice recognition software, like most innovative technologies, has had a predictable pattern of adoption.

A quick look at the phases of technological adaptation will tell you where you and your practice fall on the curve. If you have already implemented VRS in your practice, congratulations, you are leading the charge. According to some reports, only about 5% of US physicians use speech recognition to generate text in their offices. By our estimations, our practice is currently in the phase of the “early adopters,” joining other physicians on the cutting edge of technology who are careful but eager to accept change, given the eff ectiveness of VRS. We use Dragon Naturally Speaking v9.0 Medical Editionby Nuance in our office. Dragon requires a Pentium-class PC and 512MB of RAM. The software is easy to implement and simple to install; after a brief, 15-minute session to acclimate the program to your speech, you are ready to dictate. This software also comes with a helpful tutorial, which teaches the user to master the commands in a matter of minutes.

The program does require some patience in the beginning while you train it to learn the nuances of your speech patterns. However, it will quickly learn to accurately transcribe words common to your vocabulary. The software makes fewer mistakes the more you use the program. Overhead costs are a major concern for all practices, large or small. In years past, our office was paying as much as $10 per dictation, even though some of the letters were as short as a few sentences. At more than 100 dictations each month per physician, we were paying more than $1,000 a doctor for a transcriptionist to re-write things that had already been created in the electronic medical record. We were saving time, but at a cost of more than $12,000 each year per physician.

Dragon Naturally Speaking Medical Edition retails locally for about $1,500. It doesn’t take a math genius to calculate that the program will more than pay for itself in just a few months. Everyone is familiar with the adage that “time is money.” In medicine, this is especially true. VRS programs would be next to worthless if they constantly made mistakes with unfamiliar and arcane medical terminology, requiring physicians and staff to make correction after time-wasting correction. To avoid this, the most effective VRS programs for healthcare come with a pre-installed medical dictionary, eliminating the need to spend valuable physician time to train the program to recognize words like “pessary” and “cystoscopy” (words common to our urology practice). Our program even recognizes new drug names, both trade and generic, as program updates are periodically made available via the Internet.

What about the intangibles, or those soft benefits that can’t be measured in dollars or minutes saved? Many of our patients, particularly those who grew up in less technology-saturated times, are impressed at their doctor’s ability to speak to a computer and have their records or prescriptions printed before their eyes. With the use of template letters and procedure notes, records can be completed and sent to a referring physician before the patient leaves the exam room. Our patients frequently comment that we have the “latest and greatest technology.” It reinforces the idea that our practice is on the cutting edge and delivers the highest possible quality of care.

Patient satisfaction leads to more referrals and also to the best and most cost-effective form of marketing: word of mouth. Early versions of VRS required special microphones and computers with a lot of processing power. However, given the recent advances in audio and processor technologies, physician users are now free to use laptop computers, tablet PCs, desktop computers, and wireless microphones, streamlining the process without sacrificing the quality of dictation.

With the advent of Bluetooth™ technology, any of the current wireless phone headsets can be used with a Bluetooth™ USB antenna adapter. Our practice currently uses the following tools: Dragon Naturally Speaking v9.0 Medical Edition, laptop and tablet PCs with Windows XP operating system, a Plantronics Audio 910 Bluetooth™ USB antenna, and the Jawbone Bluetooth™ wireless headset by Aliph.

The bottom line

The time has come for physicians to merge the same advances in technology used in the clinical care of our patients into the management of our practices and offi ces. The use of electronic medical records in tandem with voice recognition software offers physicians an opportunity to document care at the point of service to capture data that will help in the management of patients, to avoid duplication of tests and procedures done by other healthcare providers, to improve the efficciency of the practice, and finally, to reduce overhead expenses while increasing profit margins. These technologies are readily available, affordable, and have been tested by early adopters, with many of the bugs and glitches worked out so that the implementation

process is nearly seamless.

Neil Baum, MD, is a physician in New Orleans, LA and the author of Marketing Your Clinical Practice —Ethically, Eff ectively, and Economically.

Erik Anderson, MD, is a graduate of Tulane University School of Medicine and is now PG-1 at Brigham and Women’s Hospital in anesthesiology.