VoiceIt: “Speech recognition can bring standardization”

(HealthTech Wire / VoiceIt) - From a tool that was once predominantly used by radiologists, speech recognition is now enjoying much wider usage. All over Europe, hospitals and even regional networks are introducing the technology in order to improve the documentation workflow. According to Marcel Wassink, Managing Director of Philips Speech Recognition Systems, and Robert Thornton, the company’s Commercial Director, speech recognition could also be used as a vehicle to support cross-border interoperability in Europe.

Published: 09/24

Mr. Wassink, could you give us a brief overview of the adoption of speech recognition by the European healthcare systems?

MW: Currently, more than half of all radiologists in Europe use SpeechMagic. Around a third does not use any speech recognition at all. This is in radiology. When it comes to other medical specialties, or to hospital-wide implementations, numbers are different. Europe-wide, the usage of speech recognition is still in the lower one-figure range for hospital-wide implementations. But things are changing. There are some countries that are really pressing forward. Take Spain, for example: they have several regional implementations of speech recognition, with a number of hospitals in each region that use speech recognition hospital-wide. Norway has even signed contracts to implement speech recognition in all of its public hospitals. And in France, all 39 public hospitals in Paris are in the process of implementing SpeechMagic. I could go on with examples…

So there is a clear European trend for speech recognition to be used beyond just radiology?

RT: Absolutely. We can see that everywhere. There is a hospital in Denmark where 1,400 physicians, nurses and secretaries are using speech recognition. In Germany, the majority of university clinics are moving ahead following the success of their radiology implementations. And in Norway, as Marcel has already mentioned, I would be so bold to say that in two to three years from now, we will have an 80% to 90% penetration of all public hospitals – and not only in radiology.

Is it correct to say that more and more regional networks are using speech recognition, albeit in an early stage?

RT: This is certainly true. In Spain for instance, the regions of Valencia, Castilla La Mancha, Navarra, Galicia and Pais Vasco all have large-scale speech recognition projects with a regional scope.

Why is it interesting for regional care networks to work with speech recognition, and to make it accessible to doctors in private practice?

RT: One aspect certainly is that speech recognition is highly scalable. One installation in North America has as many as 90,000 physicians. A central installation makes it far more efficient to manage user profiles, for example. From the point of view of a care network, there is another aspect as well: speech recognition can be used as a tool to help standardize a heterogeneous IT environment. Technical standards are being established; speech recognition can help to introduce common documents, language standards, etc.

MW: There is another point: cooperative care networks are increasingly using electronic patient records, in whichever country you look. But an electronic patient record will only be a success if healthcare staff use it on a regular basis; and speech recognition can support this. The operation of electronic patient records is facilitated when voice is used as an input method to supplement the keyboard and mouse. In fact, it is the most accurate, convenient and efficient way to capture medical information at any point it might be needed. We know that many doctors feel pretty frustrated by electronic patient records, and some of them have even stopped using them because they are so inconvenient.

The European Commission is currently pushing for greater interoperability among healthcare systems. The aim is to make basic patient information available everywhere in Europe. Is there a role for speech recognition in this context?

RT: Think of coding: coding systems like SNOMED or diagnostic coding systems in radiology are increasingly used internationally. So if you standardize coding and the coding engines in the applications, this will automatically help interoperability. When a doctor uses speech recognition to capture a diagnosis or a radiological finding, this information could automatically be transformed into code. If internationalized, this code could be used and understood by medical information systems anywhere in Europe. And it could even be transformed into the respective language. So, yes, speech recognition can help cross-border interoperability.

MW: But it is not speech recognition alone; there must be a joint effort from the relevant companies and standardization bodies in this field to achieve the goals set by the European Commission. The recent technology advancements of SpeechMagic and our partners' technologies are clearly pointing in this direction. We are on the right track.

Mr. Wassink, Mr. Thornton, thank you for your time and answers. (HTW)

###

This is a HealthTech Wire VoiceIt interview. You are free to use parts of it for journalistic, non-commercial purposes. You must, however, always quote or link to the source: HealthTech Wire (www.healthtechwire.com).

© 2008, so2say communications