Who says small community hospitals don’t use cutting edge IT, or that age is a negative factor in healthcare professionals willingness o adopt it? Who says increased efficiency through IT leads to lay-offs, or that speech recognition technology is best suited to American-born doctors?
Brooks Memorial Hospital in Dunkirk, N.Y., is smashing misconceptions left and right with its creative and successful use of EXSpeech from Stratford, Conn.-based Dictaphone Corp. The hospital’s journey down the speech recognition road began several years ago, as it was replacing its text product and was offered a promotional opportunity by Dictaphone that would, in part, reduce its initial investment in speech technology.
Brooks Memorial Director of Medical Records Cheryll Rogers leaped at that one. “Brooks Memorial’s administration is very progressive. Our president has talked for years about speech recognition technology, so when a promotional opportunity made it possible to purchase 10 speech licenses, we thought that would be a great introduction to the technology. At the time, we also thought 10 licenses would be plenty.”
When the time came to select a transcription text product, Brooks Memorial Hospital involved their transcriptionists in a month-long process of examining several vendors’ text products before officially signing on with their selection of Dictaphone’s EX-text. It was a short but intense period of hands-on due diligence. In the end, Transcription Supervisor Sandy Congdon says the hospital chose the Dictaphone product primarily for ease of use, but also for the reassurance of their support system. “Doctors dictate 24/7, so transcriptionists need support 24/7, too,” she says. Those promotional-opportunity benefits to add on the speech recognition product also were an incentive.
Experience and Skepticism
Founded in 1898, Brooks Memorial is a community hospital about 30 miles outside Buffalo, situated on the shores of Lake Erie, with 99 beds and a staff of more than 400 employees. The hospital provides a full array of services such as inpatient and outpatient surgery, cardiopulmonary services, cancer and radiation therapy, obstetrics, physical therapy, radiology and imaging, intensive care and sleep-disorder services.
The transcriptionists at Brooks Memorial aren’t novices. As unit supervisor, Congdon has 27 years of experience; one of the transcriptionists in her unit has 24 years, and the other three have about five years each in the field. For many years, Brooks Memorial transcriptionists fit the stereotypical image–the headset-and-foot-pedal equipped transcriptionist keyboarding away and turning physicians’ dictations into typed texts.
When the transition to EXSpeech occurred, “I was the number one skeptic,” says Congdon. “I’d been doing transcription for 27 years, and all of a sudden, I had to learn something new. I was well aware of speech recognition products on the market from seeing them at conferences, but I thought I would be retired before having to learn one.”
She and the hospital’s other transcriptionists retained their skepticism well into their intensive, Dictaphone-provided training in the fall of 2003, when they began to slowly embrace the technology. From the transcriptionist’s perspective, Congdon says, there were two initial hurdles to overcome with the software: elimination of the ubiquitous foot pedal and having to learn certain keystrokes to trigger specific commands. Beyond that, she says, the technology isn’t difficult, and at Brooks Memorial, it has helped to morph typing transcriptionists into medical text editors–and even an entrepreneurial unit of sorts–in a short time.
Selecting Dictators
Brooks Memorial management chose to utilize a backend speech server so that deploying the system would pose nothing new for physicians. Physicians dictate just as they always have, and the product of the voice files appears as text on the transcriptionist’s monitor. With her headset, she listens to the dictation as she visually reviews the written text for accuracy, using up to a half-dozen keystrokes to interrupt the voice file and edit directly into the text for 100 percent medical accuracy. The software injects into the text cues and symbols such as cubes or musical notes to indicate a segment of text it’s not certain about and that warrants a transcriptionist’s attention. Transcriptionists edit the digitized text for medical accuracy as well as grammar and punctuation.
Congdon stresses that speech recognition technology won’t provide a knowledge base of medical terminology where there is none, but for the trained transcriptionist with a medical terminology background, the software is an efficiency boon.
In the beginning, Rogers, Congdon and the transcription team selected the physicians they thought would make the best speech dictators for the trial run, and found they had even more physicians than they needed who were interested in using the technology. They also found that the physician’s country of origin wasn’t a significant factor in whether or not he would be a successful dictator, and that both American-born and foreign-born physicians make good speech recognition dictators; it depends entirely on the individual. Rogers says the initial 10 licenses were “gobbled up quickly.” Brooks added an additional 18 licenses in 2004, and Rogers will be requesting an additional 5 licenses for capital purchase in 2006.