In today’s world, speech recognition has gained considerable popularity in everything from text messaging to medical dictation. And why shouldn’t it have? Accurate speech recognition saves time and allows for an improved, hands-free workflow.
But not all speech recognition solutions are created equally. Although there is a myriad assortment of solutions, not every solution is appropriate for every situation even if it’s produced by a tech giant.
Google recently announced a speech recognition solution that excels at recording medical conversations. While they take pride in a WER (word error rate) of around 20%, in the medical industry, that accuracy is not usually considered outstanding. Most doctors demand accuracy in the high 90’s or they’ll revert to typing.
In addition to that, Google’s newest ‘unobtrusive’ speech recognition tool is also designed to run in the background during an appointment in order to pick up natural conversations between patients and their doctors; while this may initially sound like a smart idea, the fact that it records everything could present potential confidentiality issues down the road.
Although Google produces some outstanding products and services, they aren’t (at least yet) delivering a commercial product that meets the privacy or performance bar demanded by medical providers. Having said that, the economics of consumer speech recognition solutions are certainly attractive – “free” or nearly free vs. a paid subscription service.
But if you or others in your organization are evaluating consumer speech recognition tools to document clinical findings, you’ll really want to consider two things:
- Using Siri® Speech Recognition, included with Apple operating systems, or Google Voice™, included with Google and Android operating systems to dictate a text message, a brief email, or perform a Google search is an efficient, simple and accurate way of completing a task. It is estimated that 50% of smart phone owners use these popular tools on a regular basis. The differences are that Siri and Google Voice do not include a robust medical vocabulary, understand medical context or support the medical syntax required for clinical documentation.
- More importantly, consider whether these tools really, provide the degree of protection you really need for Personal Health Information (PHI). They may provide the encryption necessary for HIPAA compliance, but they do not include a Business Associates Agreement(BAA). Under the U.S. Health Insurance Portability and Accountability Act of 1996, a HIPAA BAA is a contract between a HIPAA-covered entity and a HIPAA business associate (BA). The contract protects PHI in accordance with HIPAA guidelines.
Fortunately, there is a cloud-based speech-recognition solution that works with your iPhone or Android phone, conforms to HIPAA requirements of a BAA while providing data encryption, and all the medical vocabulary you need. SayIt™ from nVoq supports all major-medical specialties and is simple and easy to use. SayIt has an accuracy rate of 95% or greater and was designed with the medical community in mind, meaning it is designed to be both HIPAA compliant and highly secure.
One voice profile, encrypted in the cloud, is available to use wherever you have an Internet connection, on a Mac OSX or a Windows based PC.
SayIt is currently being used with Epic®, Cerner®, eClinicalWorks, Athenahealth®, MEDITECH®, GE®, McKesson® and many other EHRs. SayIt is as easy to use as any consumer speech recognition tool: simply place your cursor into a free-text field and dictate your notes, with the assurance that your patient’s data is secure.
Contact us to try SayIt now! firstname.lastname@example.org or call 720-562-4500.