Healthcare News & Insights

Does speech recognition software save time, or create too many new problems?

Will speech recognition software help make clinical documentation easier, or will it open the door to new problems and dangerous mistakes? The answer depends on who you ask. 

134931257One of doctors’ biggest complaints about EHR systems and other health IT is that clinicians now have to spend more time entering data into a system, which takes away time that could be spent with patients.

That’s why some hospitals and other providers have turned to speech recognition software. With those tools, doctors speak into a device and the software converts the speech into text and imports it into the chart.

But while there’s a potential to save time, speech recognition systems can create complications of their own, according to a recent debate among doctors and researchers in the magazine For the Record.

The biggest issue: Those automated transcriptions aren’t always accurate. Therefore, doctors either have to spend time carefully reading notes to make sure the right information is included, or risk having missing or inaccurate data attached to a patient’s record.

In addition, systems often need time to adapt to a particular clinician’s voice, which requires more time on the part of the doctor, at least at the beginning. And even if the technology itself is perfect, there may be background noise that interferes with the transcription.

That’s why many hospitals choose instead to rely on back-end transcription services in which a service provider handles the double-checking.

Potential cost savings

However, many hospitals report positive results after switching to front-end speech recognition. One hospital mentioned in the article was able to trade $1.5 million spent each year on transcription services for a much more cost-effective software system.

Also, after an initial training phases, officials say doctors have become more efficient because they’re able to update a patient’s EHR at the point of care, rather than waiting for notes to be transcribed.

Supporters of the software offer these tips for a successful implementation:

  • Choose software that adjusts itself based on individual speakers, rather than those that claim to be speaker-independent. Everyone talks differently, and using an adaptive system will pay off in the long run.
  • Invest in high quality microphones and other necessary hardware to make sure the software has good data to work with.
  • Make sure the system is compliant with HIPAA and other regulations.
  • Train users on the best ways to edit their notes and spot mistakes.

Subscribe Today

Get the latest and greatest healthcare news and insights delivered to your inbox.