Industry News & Events

Speech Rec Software: Ready for prime time?

May 2, 2011 - Over the last few years, medical transcription companies have been giving some serious consideration to Speech Recognition Technology (SRT), and the implications it has on the industry as a whole. Reactions and experiences have ranged across the board, making it hard to reach a definitive verdict and conclude what role it will play in the future. The simplest way to evaluate this trend is to focus on some basic questions.

What is SRT, and how does it work?

Speech recognition is being applied to a growing number of services and applications. The technology boils down to software that converts voice audio into text with an electronic format. It powers everything from simple voice-dialing on cell phones to advanced robotics and translation. Within the healthcare environment, there are models that produce text as the physicians speak, allowing them to correct and sign off on a document immediately. Other models route the text to an editor or MT for completion, and deliver a final version back to the provider. In either of these options (Front-End/Back-end), there is always a need for a qualified professional to verify that the transcription is accurate.

How does it affect medical transcription service providers?

While some people see SRT as a threat, others have embraced it as an opportunity to apply technology and get ahead of the curve. MTīs have kept their eye on the technology as a potential source of competition for years, assuming that it will eventually match their transcription standards and put them out of business. Others have acted on this assumption by making massive investments to embed SRT into their existing workflow, and results have been mixed at best.

Technologies in their early stages have a tendency to be expensive and unreliable. Many actors in different industries jump into these solutions too early, footing the bill for research, integrations, developers, and other loose ends. A good example in the medical field is found in EHR adoption, where healthcare services that switched a few years ago paid exponentially higher prices than options available in the market today.

What's the bottom line; is it right for me?

Evaluating the pros and cons of SRT boils down to identifying priorities. Some medical transcription companies focus on quick turnaround, while others have a higher emphasis on quality. At this point, the technology does not have good solutions for quality issues related to thick accents, drugs that sound similar, background noise, grammar, and other concerns that MT.s have learned to handle through experience. This means that transcriptionists must pore through every single line of documents to check for mistakes, or worse, that doctors have to spend their valuable time correcting reports.

If SRT is not reliable enough for sensitive operations like banking transactions or legal transcriptions, is it ready for prime-time in the healthcare environment? In our case, we will only consider the implementation of SRT if it offers a considerable reduction in costs to our clients, while maintaining the superior quality standards that set us apart. For now, that combination of factors is still not clearly available.

- VIVA Transcription News Team