BOSTON – In this harsh economic climate, two Boston hospitals continue to reap millions in savings from voice recognition technology.

At Beth Israel Deaconess Medical Center, CIO John Halamka, MD says 1,000 physicians use the dictation system that was first deployed in 2002. The eScription software is now part of Nuance Communications’ product suite.

By making the patient documentation process easier for physicians, Beth Israel achieved a cost savings of $1.5 million in 2008, Halamka said. Since rolling out the technology, he figures the savings amounted to more than $5 million.

Also, Beth Israel physicians have decreased the time between when a patient is seen to the documentation of the visit by 90 percent – from five days to 12 hours. Radiology reports are delivered within 30 minutes.

Halamka, a self-described “Web guy,” who is more likely to e-mail than to talk on the phone, said he was surprised by the quick physician uptake of the technology.

“I didn’t recognize just how popular it would be,” he said.

When physicians dictate from their desktops, the voice is recognized in the Beth Israel data center, and it appears as an unsigned note in the electronic health record.

Children’s Hospital Boston is achieving similar outcomes with speech technology. Children’s, which uses outside transcriptionists, installed a platform from Philips Speech Recognition System, now part of Nuance, in 2007.

Today, up to 90 percent of the medical reports are produced in less than a day, said Mary Radley, director of medical record services. Productivity has increased by more than 50 percent.

Radley acknowledged financial savings, but she did not attach a number to it. She emphasized the satisfaction of clinicians.

As Laurie Fishman, MD, put it, “Often I get dictations within 12-24 hours. It makes a huge difference in caring for patients.”

While recognizing the value of speech recognition today, Nick van Terheyden, MD, said the technology must continue to improve.
Van Terheyden is chief medical officer for M*Modal, and formerly CMIO for Philips Speech Recognition Systems.

“It is the capture of structured clinical data that can automatically feed the EMR that is the real goal,” van Terheyden said. “It means the technology will not only be able to recognize the words, but also understand the meaning and context.”