The future of dictation and transcription technologies

With Dictran’s Bill Lupinacci

EHM. Cost pressures, aging dictation systems, and a shrinking pool of medical transcriptionists means that companies are looking for alternate dictation and transcription technologies. What are the various technologies that you offer?
BL. For decades our competitors have always thought of dictation and transcription as separate entities that are created and stored in separate servers in separate databases. The problem is that, from a hospital or clinical information viewpoint, they aren’t really separate. They are both the same medical report, just in different formats: voice versus text. So we process and store both voice and text in one server, in one database, and make it available to all the other hospital applications that need to access those reports. And again we’re talking about both voice and text. We call it our 2-4-1 architecture. We take voice from anywhere: phone calls, digital recorders, PDA’s PC microphones, other dictation systems, etc. We are the only major brand supporting all brands of hand-held digital recorders. We can handle over 150 different digital audio and digital video formats. And we send text anywhere, in any format needed for all the other hospital applications. We also provide state-of-the-art server-side speech recognition at less than half the price of our competitors. We make speech recognition non-intrusive, so neither dictators nor quality assurance editors have to change their work behaviors to accommodate its implementation.

EHM. How are these new technologies improving productivity?
BL. Our designers are our customers. And our designers are obsessive about reducing costs and eliminating errors. Our job, as our customers see us, is to enable the flow of information from a physicians’ thoughts, through a complex workflow, that eventually is returned to the physician as codified medical information being examined and approved – signed – by the physician. Throughout that entire workflow loop, we provide capabilities that eliminate the possibility of human error wherever possible. And our unique 2-4-1 architecture reduces the costs of that workflow in many areas of operations.

EHM. What are the main challenges facing dictation and transcription technology?
BL. This is just a sample of common problems within transcription departments that the Dictran systems faces head-on and provides solutions to reduce or eliminate:

• The physician specified the wrong patient ID for the dictation.

• The physician’s voice was too garbled for transcription to understand it.

• The database identified only one patient record, but because the physician forgot to hit ‘Next Job’, the database record really contains multiple different patient records.

• Medical records can not find the transcription associated with that dictation.

• Medical records has no record of a physician dictation, but the physician is certain that he/she dictated the report.

• The report was delivered to the wrong physician because the physician used the wrong ID to dictate.

And one of the worst problems of all:

• The wrong operation was performed because of an error in transcription.

Our customers have designed features and functions into our systems that help reduce or eliminate all of these challenges and many more.

EHM. A single physician can spend upwards of $25,000 per year to digitize patient information. How do you reduce this spending?
BL. We were the first major brand to put a footpedal on a PC for transcription. Before that invention, transcription was performed on a separate transcriber hooked up to a phone line and electrical outlet. Any medical report going from a dictation system to a transcription system could only get there through a human being using both systems simultaneously. We were also the first major dictation brand to offer Internet-based transcription. Before that invention the digital movement of voice dictation and transcription was much more costly. Last year we announced the availability of Voice-Over-IP dictation which is cutting long-distance costs for our customers to as low as one-tenth the price of previous levels. Our proprietary lossless voice compression yields WAV files half the size of our competitors, while, at the same time, increasing the quality of the average recording. Our competitors all use a technology called lossy compression, which degrades the quality of the telephone recording. And our speech recognition solutions hold the promise of providing another quantum leap in reducing the cost of digitizing patient information

EHM. How has speech recognition changed the transcription and dictation market?
BL. Not as much as vendors like us would have you believe. Speech recognition holds great promise. And the past year has yielded significant advancements in the quality of speech recognition.

However, speech recognition works best where the number of words being used is limited. That’s why radiology, for example, has a much higher success rate with speech recognition than medical records in general. A radiologist uses a much smaller number of different words than a general practitioner, so it’s easier to predict what a radiologist is saying.
A few institutions have even demonstrated cost savings outside of radiology, while maintaining quality with this technology. But there are still far more horror stories out there than great success stories. Speech recognition is still a mine field that you shouldn’t venture into without good roadmaps, competent advisors, realistic expectations, and only the best technology available. Many major respected institutions still consider the technology too risky to jump into head first. With all those caveats, we believe we have the best speech recognition platform available and lots of great customer references.

EHM. How do you make sure that you maintain maximum security and confidentiality, in accordance with HIPAA regulations?
BL. Before HIPAA even existed, Dictran systems were in use in military, federal law enforcement and other secure environments. When HIPAA was first released, our system already satisfied its specifications. Additionally, our main dictation server avoids the use of all commonly hacked Internet protocols. Even though a Dictran VoiceWare Server 241 system is Internet-enabled, we do not use http (the standard web protocol), ftp (the standard Internet file transfer protocol), or SMTP (the standard e-mail protocol). While our competitors are trying to build the best locks they can to secure their systems, we follow a policy of “the most effective locks are the ones that no one knows exist.”

EHM. What are your predictions for the future of dictation and transcription?
BL. I’d like to give a short, simple, candid answer and just say ‘us, we are the future of dictation and transcription’ but I don’t want to sound too smug. But at the rate we are growing, we are trying our best to make that answer a reality. We see no future for companies who limit themselves to lossy compression techniques. We believe companies with closed architectures that make you choose from a very limited list of digital recording devices have a very limited future. We don’t believe that speech recognition is a panacea and we don’t believe transcription will be a dead profession in five years. We believe some of our competitors have gotten into a lot of trouble over-promising and over-selling its benefits. But speech recognition will continue to improve and play a larger and larger role in medical transcription over the coming years.