Congratulations on your interest with using speech and dictation technology. You’ll find this exciting technology can save you vast amounts of time each day while increasing the quality of patient documentation.

You are embarking into this technology at a truly exciting time. We have helped hundred of veterinarians implement and become proficient quickly. To help you get the most out of your investment I’ve put together this summary of best practices and technology summaries.

First let’s discuss the technology. Today’s products perform with near real time speeds with accuracies in the upper 90% range. Mind you the average typist averages low 90% accuracy. So if when you make a mistake, chances are you are making less mistakes than had you types the same information.

We have helped hundreds of Veterinary practices implement and use these Technologies with outstanding results. Our best practices will help you get up and running almost immediately. First I think it’s important to understand how the technology works. Speech to text has been around for some time and is a mature technology. Today’s speech-to-text Solutions coupled with the processing speed of today’s computers give the illusion of pure magic. You speak, it types but what is really happening in the background?

First factoid; your computer is not really listening then processing the phonetics word by word. Instead it is gathering your speech pattern and matching it to prior examples. Most medical products use a 3 – 5 word speech pattern. So when speaking, it is always better to speaking full sentences or phrases. This is also why the technology itself has improved so much. Years ago these products required an hour or so of reading to “learn” your speaking patterns. Then the products were pre loaded with several hundred words of specific corpus. Today’s cloud based systems can search against hundreds of thousands speech examples. When we go live and coach a veterinarian the actual reading time is 5 minutes or less. While we typically spend an hour coaching, most the time in on semantics. Sound quality, microphone location, various automation commands.

Teaching a new word; When you make a mistake or experience a stubborn work, we can tell you one technique that never works; yelling, repeating, or over annunciating the word. Remember what I said above, the system is looking for speech patterns. If a new word must be added, add it. Speech or train the word as the application suggests. Then use the word in context. This will “teach” the system correctly.

Speed while talking; You can talk till you are out of breath. We have doctors that dictate 40, 50 even 100 + words before pausing. Even if you do not see the text on the screen while speaking there is no reason to stop and wait. Remember you’re dealing with a software. It is designed to cache and process data. Just continue to speak and when you stop, within a second or two you will see the text appear. This is why it’s very important when you’re speaking to speak in full sentences and or phrases. There’s no reason to hesitate, speak what is on your mind and don’t wait for the screen output.

Using speech to text will typically allow you to create notes 3 to 5 times fast than typing. The average person types 35 + – words per minute while speaking 130 – 150 words per minute.

Which product is best? There are many speech to text products today. Whether you are talking to Siri, Android, Dragon Natural, Dragon Medical, sayIt, or others it is important to understand the distinct differences. While all these products may use a similar speech engine, the output and quality of medical documentation is vastly different.

We typically see one or two customers each month who have purchased retail products and struggled for weeks or even months trying to train the product properly. We are glad to share testimonials of clients that made the investment in the right product and have enjoyed them since.

What are some of the major differences between retail off-the-shelf products vs medical version. Formatting is the #1 reason. The easiest example I can provide is dictating a blood pressure 120 over 80. You do not want 120 over 80 you want 120/80. The same can be said when dictating prescriptions frequencies, follow-ups, abbreviations and much much more. There are numerous abbreviations content structures that are unique to the veterinary industry. Products that specialize with medical formatting will do this correctly.

Another area these products shine in is workflow automation. This can take the form of templates, snippets of normal or abnormal conditions, dynamic worksheets, keyboard shortcuts and pc commands.

Templates: Templates is where you can leverage the technology to output common phrases and conditions with a few words. Thus you can speak a 2 or 3 word command and output 50 – 500 words. Utilizing templates can cut down on errors and increase efficiency 10 – 15 fold.

Template can be a simple normal that is outputted by saying something as simple as “ Normal Exam”. Here is an example.

Here is a template with a dynamic place hold. The [ ] is where the system knows to auto advance to and allow you to speak over the default variable or you can accept the default.

Thank you for taking the time to read and learn about these exciting technologies. If you have any questions feel free to email me at

Best regards,

Paul Messino