Speech or Voice Recognition Software in Medicine
Experts share ways physicians can implement and use speech recognition software effectively.
When Scott Sachatello, M.D., of Baptist Medical Center in Florida, first used speech recognition software, he wasn’t sure how well he would like it.
“Just within a few sessions, I was overwhelmed with how good it was and how high the accuracy was,” Dr. Sachatello says. “I began to realize how complete the charting process could become. My thought process was translating into text right before my very eyes.”
Still, it takes some product research and physician training to translate speech recognition into workplace efficiency.
Before You Begin, Ask an Expert
Eric Fishman, M.D., is the founder and CEO of 1450, a distributor of speech recognition software. In 17 years in the speech recognition industry, he has seen a lot of obstacles to adopting speech recognition software. Dr. Fishman recommends practices work with a value-added reseller when they buy speech software.
“The manufacturers of speech recognition software understand this is complex software,” Dr. Fishman says. “In a professional environment, the increase in accuracy that a physician can [gain] by spending a few hours with a value-added reseller will pay tremendous dividends.”
Some specialists are more likely to need a hands-free microphone, while others want to work in conjunction with a graphical user interface of some kind. Still others will need a headset that travels with them.
Often, a value-added reseller may be familiar with EHR technology and even have worked with a client’s particular EHR.
Cloud or Ground?
Practices will need to choose between cloud-based and local voice profiles. Each kind has its proponents.
For instance, Dragon Medical offers portability and ease of access. Machines don’t have to be connected to the Internet to use the technology. However, users may have to “train” each instance of Dragon Medical on their voices.
Other products, such as M*Modal, offer voice profiles that can be accessed by any machine with Internet capability, meaning the user doesn’t have to re-train multiple machines on the same software. The downside? Standalone machines may be left out in the cold.
No matter how physicians first encounter speech recognition software or where they use it, it has to work smoothly and feel right for the individual user.
“For me personally,” says Dr. Sachatello, “[I appreciated] that the technology was as good as I thought it needed to be to remain efficient in the emergency room.”