SOAPware by DOCS, Inc. is very user friendly w/ Dragon NaturallySpeaking. With proper training, it is possible to dictate at speeds up to 160 words per minute. I have used Dragon NaturallySpeaking and SOAPware for several years and would not go back to traditional transcription or paper notes. As with learning to auscultate heart sounds with a stethoscope, equipment without proper training will not enable one to reap the greatest benefits from the equipment and technology.

Properly trained and corrected voice files generally do not have a problem with accents. Voice recognition will not work well for people who mumble and slur their words. Nevertheless, if one invests the time up front to learn to use the software, it can quickly pay for itself in productivity and elimnated transcription costs.

Disclaimer: I am a physician, a certified ScanSoft Gold Partner and a certified SOAPware dealer.

Regarding medical use of SR software =

Many doctors use SOAPware which is DNS enabled. Many of them have had great success with the product. One of the best sources for information is a podiatrist in Chicago, from whom we have had little contact of late, but who did a lot of lecturing and speaking to professional organizations about using this software in creating medical records. Perhaps you can search the archives of the group called medspeech.yahoogroups.com and find his many postings. Those postings provide a tremendous background of good information and advice from actual experience and testing. Look for postings by Larry Kosovo - he really knows his stuff in speaks from real experience. Another resource is the body of postings by Stan Swartz in the same discussion group. These guys have figured out how it is done and make use of if it all the time.

The key to making the software work is the ability of the physician to change the way he works and the way he dictates. You cannot speak too quickly, but you can speak in a normal, relaxed tone and pace. one must use particularly clear diction and complete each word carefully before sang the next word. This means, for example, saying the three distinct words, even if speaking rapidly, "This and that" - renouncing the "s" on "this" and the "d" on "and" and the "t" at the end of "that" - so that it does not, out all slurred together as we do in the vernacular, "this-n'-dat" - and this is a practice makes perfect kind of operation. Speaking too quickly, running words together, failing to enunciate each word clearly, all lead to low recognition accuracy. Many physicians are loath to slow down and work with the software and its requirements. It does not take much more time to speak clearly and distinctly than it does to speak in a hurry. This is analogous to driving a car fast through heavy traffic. Running at speed limit almost always gets you there as quickly as legally possible, whereas speeding between each stoplight will often cause you to get there at approximately the same time, but with greater risk of accident along the way. You might get there a few seconds earlier, but with greater chance for accident, higher gas consumption, more wear and tear on the automobile, etc.. Driving at the speed limit will usually get you there only a few seconds later, but much safer and with better accuracy. It works the same with speech recognition. Getting a physician to understand this can be difficult, as it is with any other professional who is used to using a hand-held recorder and having the transcriptionist sort out what he says when he or she does the typing. Lawyers, engineers, accountants, insurance adjusters, busy executives, -- these are all professionals who not want to change their behavior, especially if it feels like there slowing down. But slowing down and becoming more deliberate has its own rewards, just like slowing down and running the speed limit will give you there saver and on time.

One should speak at a normal, yet deliberate and careful pace. Speaking too slowly will be the same as driving too slowly -- it can be dangerous in itself. There is no reason to slow down to an abnormal speaking pace, just make sure that you are completing each word before moving on to the next. That takes just a little more care and deliberation -- but no more than that effort.

The beauty of this software is, of course, the immediacy of the result. The physician will have to slow down and proofread his work, or have someone else do that. This should happen whether or not he is using speech recognition software or a transcriptionist who might misunderstand what he spoke. Either way, the physician should proofread the resulting material. Either the typist or the software can make mistakes -- even though I suspect the typist will make fewer of them overall. Either way, the proofreading process is the same -- the physician must proofread the entire document to be sure it is correct.

The second advantage of using this software is the elimination of the middleman. The typist is entirely eliminated, so even though it might take little longer for the doctor to proofread, which he must do anyway, he will more than make up for any extra time (which is a loss of money) by offsetting the cost of the typist which has been eliminated. In my little practice, I found it took me just along you to use this software. Yet, that extra time was not so significant compared to the fact that I laid off a full-time employe. Yet, it can be a hard sell to get the busy executive or physicianto realize this advantageous offset.

Again, look for Larry Kosovo's previous postings about medical records and you might look for his Web site on the Internet by doing a search for him as a podiatrist in the Chicago, Ill., area and see what you find out. He is also posted many useful messages on the topic on the ScanSoft DNS discussion Forum. I would start there.

You can be assured, in any case, that the software works.

Just my take. Best Regards. ////// Richards //////
=========================================