|
IT'S THE TALK OF
THE MEDICAL WORLD
By Karl Barksdale © 2000
(Courtesy
SpeakingSolutions.com)
According to Larry Gibson, this is the year that medical
and dental professionals will investigate an expanding product line of
voice recognition solutions. He is totally comfortable with the idea that
voice recognition will reshape data input and record keeping in the medical
and dental professions. Speech recognition and other specialized
productivity software will make doctors and dentists more efficient while
saving billions in transcription and record-keeping costs.
Larry is the founder and former CEO of Dentrix Dental Systems
(www.dentrix.com) and the current President of Practice Management
Technologies for Henry Schein (www.henryschein.com), the multibillion
dollar medical and dental solutions conglomerate. Henry Schein and Dentrix
are on the cutting-edge of automated and ecommerce medical solutions.
The Dreaded Periodontal
Exam
In the complex world of voice-related software, a single example can illuminate
the implications of speech recognition in surprisingly simple terms. For
instance, take the dreaded periodontal exam.
In this soft tissue examination, the dentist or dental hygienist probes
with a sharp instrument deep into a patient's gums -- several times in
fact. Probes are made into various portions of the gums looking for signs
of gum disease. It's a slow process because the results of each probe must
be meticulously recorded by the hygienist. This requires stopping and
starting the exam over and over again. Ouch!
This is where speech recognition comes in.
Before speech recognition, a periodontal examination and its
accompanying paperwork might take between 30-45 minutes to complete. With a
speech-driven system, the examination can conclude in about ten minutes.
Needless to say, with today's rising dental costs, this represents quite a
savings.
With carefully designed speech-enabled software, the hygienist can
conduct the examination while simply speaking the results as they go along.
With voice software, there is no stopping to record results. As the
hygienist speaks, the patient's chart data is automatically updated. A
common phrase might be, "2, 3, 2, bleeding suppuration." While
this may seem like babble to you and I, it's important data for patients,
insurance companies, dentists, and even lawyers.
There is little need for a computer monitor during the exam, because the
dictated information can be synthesized and read back to the hygienist
through their headset. This audio-feedback process improves accuracy.
Dentists are sticklers for accuracy. Mr. Gibson quickly points out that,
"Anything less than 97 percent accuracy will destroy the whole
thing!"
To insure greater accuracy, the lexicon for this exam is reduced to
include only the vocabulary required to complete the examination
successfully. While the average person may have over 60,000 words in their
basic vocabulary, for periodontal exams, a mere few thousand words are
sufficient. This smaller lexicon dramatically reduces the possibility of
error. (These smaller lexicons, or dictionaries, are often called models.
Uniquely designed models are used in other medical specialties like
radiology, cardiology, and pediatrics.)
A dentist or hygienist can
enroll and start using the system in about ten minutes. With one hour of
practice, most are proficient. According to Gibson, acceptance by
professionals for this specialized speech solution has been nearly 100
percent.
Dramatic Changes in Medical Transcription
Dental care isn't the only area where tremendous progress is being made.
In our interview for the Balance Sheet, Larry opened our eyes to a wide
variety of speech recognition solutions.
For instance, medical transcription is an expensive and time-consuming
process which requires the doctor to dictate, a transcriptionist to enter
text, and a doctor to edit the final transcripts looking for errors.
According to Mr. Gibson, transcription costs may easily be the second or
third highest expenditure a doctor must pay out. It is a costly and time-consuming
process, and despite everyone's best efforts, errors can be made. In the
field of transcription, speech recognition technology is:
- Lowering costs by
eliminating transcription charges
- Increasing accuracy
- Saving time
- Making it easier to
comply with insurance requirements and government regulations
- Reducing Carpal
Tunnel Syndrome (CTS) and Repetitive Strain Injuries (RSI) in the
medical profession
These compelling arguments are moving medical professionals away from
traditional keyboard input devices towards speaking solutions. We were able
to corroborate Larry Gibson's insight with a startling study suggesting
that the days of the medical transcriptionist is rapidly coming to close.
In a recent study of speech recognition in 26 medical units at the University
of Texas Medical Branch reports over a 90 percent voice technology
acceptance rate among the residents, fellows, and doctors. This
overwhelming acceptance has dramatically reduced the number of medical
transcriptionists required to process the hospital's nearly 14,000 monthly
reports. In fact, the number of transcriptionists has dropped from 100 to
less than 10 in under two years! While it is sad to see 90 percent of any
job category eliminated, the trend is clear.
Speech-recognition is removing transcriptionists from the loop while
saving time for the doctor, who can now dictate and edit medical records
and charts in real-time. Combine speech-recognition technologies with
computer screens that respond to the touch of a stylus, and you have an entirely
new way of keeping these vital records. And in today's medical environment,
were doctors and dentists are being asked to cut costs at every turn, the
cost of medical transcription is simply too high.
The Pivotal Year, 1997
Gibson relates that several significant things happened in 1997 to
change the balance of computing power in the favor of speech recognition
technology forever:
- The widespread
availability of 200 MHz or higher capacity Pentium II processors
- The falling cost of
RAM
- Improvements in low-cost
speech recognition headsets
- Improvements in
continuous speech recognition software
Until the availability of Pentium II 200 MHz and faster processors,
speech recognition on the Windows platform was impractical. Also, voice
requires 64 MB or more of RAM to dictate digitally. Speech processing power
has improved even more with the powerful Pentium III level computers which
often come with 128 MB of RAM. When it comes to speech recognition, speed
and memory are vital.
Quality microphones and sound cards are also essential. Additionally, in
many medical and dental applications, practitioners are moving to wireless
models. The need for wireless headsets is quite understandable. Any
instrument that comes in contact with either a patient or practitioner must
be sterilized or removed from the medical facility in hazardous waste
containers. Long cords are a liability, often getting in the way of the
other tools dentists, doctors, and nurses depend upon. So, highly efficient
wireless headsets will be the norm.
Summary
At a time of rapidly escalating medical costs, it is nice to see at
least one technology that can actually lower our dental and medical bills.
But change can bring both heartache and opportunity. To the professional
medical transcriptionists, who will soon find their jobs described in the
same breath as punch card operators, Larry offers some consolation. He
doesn't see speech technology reducing the number of people needed in
dental and medical offices. Rather, he suggests that jobs change and evolve.
Those who learn the new technologies will be more productive and ultimately
more employable.
Mr. Gibson encourages Business Educators to teach speech recognition so
that medically oriented students will be prepared to work in modern medical
and dental office facilities. Students must be comfortable speaking clearly
to computers. They must learn enunciation skills and how to avoid and
correct mistakes.
Larry encourages instructors to "figure out" the curriculum
and to start teaching speech recognition as soon as they can. Quite simply,
"This is the year."
|
