Dr. Phillips is a Professor in the Departments of Radiology, Neurosurgery, and Otolaryngology-Head and Neck Surgery, and the Director of the Division of Neuroradiology in the Department of Radiology, University of Virginia Health Systems, Charlottesville, VA. He is also a member of the editorial board of this journal.
I work in what I believe to be an enlightened department. We have PACS and we have voice recognition. One works spectacularly. The verdict is out on the other. I'm going to use this little editorial opportunity to express opinions on this wonderful, breathtaking (literally), and time-saving (that's what they say) technology. I was a resident during the time of the "transcription pool." Transcription ("the conversion of speech sounds to phonetic symbols") did indeed occur there, but it wasn't a pool, at least not in the sense I understand pools. I used to be a lifeguard, just for the record. Some of the folks that worked in transcription were FAST; but there were only so many of them; as the work expanded, you can guess the outcome. There were delays, progressive delays, mounting, cascading, ever-expanding, brain-twisting delays. If you were in the ER at night, you got your transcriptions back the next morning, but the routine work could drag on for weeks. The problems were in getting the clinical staff their reports, and also in remembering if the dictation you were reading 3 weeks later related in any way, shape, or form to what you had seen. Was it on the left side, or the right side? Was it that big? Usually, you'd look for typos, and just sign the stupid things.
Fast forward a few years. No one was happy with the delays. You couldn't send a bill out on time, you couldn't keep the clinical staff happy, and the patients deserved better. Some brilliant person decided the time was ripe to marry fast computers, high-quality microphones, and some software that can make words out of the digital conversion of sound waves. Being the technological gimps we are, we decided it was going to make everything well.
There are too many sides to this issue to cover in a short editorial. The wonderful thing is that the system can speed you up. Way, way up. Macros rule. If you want to say the same thing you've said over and over again about a particular study, or just need to fill in some minor details, you can get out a signed report in seconds. And, it is FINAL. A billing document, there on the system for your clinicians. It is great for repetitively evaluating normal studies. Hmm...unfortunately, I don't see a whole lot of those. I have to enunciate carefully, watch the screen, and hope it gets most of it right. If it doesn't, well, therein lies the rub. No longer does a trained typist put the necessary words to paper. The transcriptionists are now (drum roll, please)….US. Some of us type relatively well. It is a computer age, after all. Others get great voice recognition and need to do very little. And many of the others are ready to bash up keyboards. Destroy those things. Wrap the cords around someone's neck. I cannot write what we have contemplated doing with those expensive digital quality microphones. Although, I must admit I'm amused from time to time about what it has transcribed when I curse a blue streak at it.
The short story is that we have gotten our report turn-around time down considerably. Is this valuable? Absolutely it is. The costs, however, are in the radiologists' sanity and time.
Rule number one in the universe is that computers will get faster. The next software release and computer upgrade might make voice recognition so good that you can't even joke about it anymore. God, please hurry that day. In the meantime, I find myself typing words and visualizing the corresponding finger movements when I jog. I'm just trying to keep my fingers limber for the next session of voice transcription.Back To Top
Editorial: Voice recognition: Don’t you recognize my voice?. Appl Radiol.