Nuance: The Spoken Word

June 24, 2011
John Halamka, M.D. One of the biggest challenges for any software vendor is balancing the need to provide innovative,
John Halamka, M.D.
One of the biggest challenges for any software vendor is balancing the need to provide innovative, skate-to-where-the-puck-is-going solutions, with satisfying the customer needs of today. One company that has gone to extraordinary lengths to get it right is Burlington, Mass.-based Nuance Communications, Inc.

After spending $357 million to acquire Stratford, Conn.-based Dictaphone Corporation in February 2006, Nuance moved to integrate the two companies' offerings, expand its platform, and capture maximum market share. The core of its offerings remains its Dragon Naturally Speaking voice-recognition technology, which has gone through multiple releases and enhancements. That product alone has 100,000 users in North America, Nuance executives say. What's more, they estimate that over 20 percent of physicians in North America are using their solutions, while the company boasts more than 3,000 hospital customers.

Ben Rooks
And the company is looking relatively strong in what has been a jittery market for healthcare stocks. Nuance reported revenues of $216.8 million in the quarter ending Dec. 31, 2008, an 11 percent increase over revenues of $195 million for the same period last year.

What's particularly interesting about Nuance is how the company has created and enhanced a diverse platform of product categories focusing on several different groups of users. The three largest ones are radiology/diagnostic imaging; speech-enabled transcription for health information management in the acute-care (and increasingly ambulatory) sphere; and the EHR-focused line of business, looking at both acute-care and ambulatory physicians. Though the fundamental products are similar, specific features accommodate the different workflow patterns of different users. For example, radiologists, who are under pressure to produce more reports more quickly for their referring physicians, are self-editing 90 percent of the time; while attending and staff physicians using the other two products are only self-editing 10 percent of the time.

Still, that means given the 90 percent accuracy of voice-recognition capture, those health information management professionals who are editing the voice-dictated documents in the system and returning them to physicians are part of what the company claims is a huge cost savings to the industry. Indeed, many organizations are now referring to those people as “correctionists,” since they are no longer doing original transcription. As a result, Nuance executives believe they are saving hospitals $200-300 million a year across North America in traditional transcription costs.

Proof is in the pudding

Healthcare customers are looking at the practical application of technology to facilitate processes that simply must take place in an electronic world. John Halamka, M.D., vice president and CIO of Beth Israel Deaconess Medical Center in Boston, says, “We have to implement electronic health records for every physician in America by 2014. So how are you going to get the data into that? Nuance offers a server-based solution that doesn't require me to install tons of hardware and software. So it is a tool for me to foster adoption and use of electronic health records.”

For Halamka, the question is practical. “Doctors have 12 minutes to see the average patient” during the typical physician visit, he notes. “And they have no choice but to input certain kinds of structured data; but there are ways to speed up the process of getting unstructured data into the electronic record,” he says, voice recognition being one of them.

Halamka goes on to emphasize the need to implement tools early on in the automation process that improve productivity. As he puts it, “Planning for clinical IS implementations is like telling doctors, ‘You're going to lose 25 percent of your productivity, you won't have dinner with your spouse for six months, and it will cost you $40,000.’ And that's not an easy sell. So basically, what an electronic dictation or server-side voice recognition device does is to save you time and enhance productivity and save money, so it makes an electronic health record easier to sell.”

Making EMRs easier to sell is a goal for John Shagoury, president of Nuance's healthcare division, who says he feels good about the company's trajectory because of its role in the process. “At the end of the day, our goal is to produce improved care quality and cost savings and do it in a way that causes minimal disruption to physician behavior.” The key, he says, is to help physicians become more productive without forcing too much change. “They're willing to change their behavior to a degree, but only to a degree.”

What about keyboarding?

One of the key questions in all this is where the trajectory of practice patterns will take physicians and the industry. Ben Rooks, principal, ST Advisors, L.L.C., a healthcare strategic advisory firm based in Evanston, Ill., says, “Ultimately, I do think voice recognition will get less and less important. But if you asked me 10 years ago, would transcription or voice recognition be dead by 2009, I would have said yes. But we'll continue this way for a while, because physicians do not want to adopt new technology (such as EMRs which require them to type), and this is ubiquitous, fast, and easy.”

Rooks adds, “I think they've done a great job at aggregating these transcription assets; they've become the clear leader, especially by buying Dictaphone. That was surprising at the time, but it turned out to be a smart move. Many of us thought that dictation would go away among physicians. But clearly, that didn't happen. And if you want to get some of the intuitive benefits of EMRs, you need to turn all that text into real information; and Nuance does that really well.”

Physician groups, in particular, remain a very strong core group of dedicated users, and for good reason. Larry Garber, M.D., is the medical director for informatics at the 250-physician Fallon Clinic, based in Worcester, Mass. Garber and his colleagues conducted a small internal research study to gauge the potential effectiveness of using the Nuance product before they bought it outright (see sidebar).

And how did the purchase of the Nuance solution fit into the Fallon physicians' clinical IS strategy? Well, Garber explains, he and his colleagues were pleased with their implementation of a core EMR product from Verona, Wis.-based Epic Systems Corporation, except in one respect, and that is with regard to the elimination of transcription via the notes element in the Epic product.

“We had anticipated a reduction of 75 percent transcription” from implementation of that EMR, he says, “but we were only able to drop our dictation by about 30 percent.” As a result, Garber says, he came to realize it would be “unrealistic” to expect all his physicians to embrace the keyboard. In response, the company turned to Nuance for help. Garber and his colleagues are thoroughly satisfied with their purchase, and he is very pragmatic about what it means. “We've figured out that there are some things the EMR is very good for, and that is reusing information that's already been put into it, leveraging data that has already been entered, looking at the family history, past medical and surgical history, etc. EMRs are very good at that. What they're not as good at is putting together a history of present illness - those kinds of things will never fit will in the template. That's why there will always be a certain amount of dictation, and that's why we now use Dragon.”

Erica Drazen, Sc.D., partner in the Waltham, Mass.-based Emerging Practices Division of the Falls Church, Va.-based CSC Corporation, says the jury is still out on how all this will evolve. While Drazen does believe that youngest generation of incoming physicians, who spend a lot of time texting on mobile devices already, will ultimately move towards keyboarding, such is not yet the case.

“I do believe there's a window of time in which we're now operating when voice-recognition tools will remain important,” she says. “And if you can make things cheaper and faster by making 90 percent of voice recognition accurate, that's better. We want to move towards structured text, and I think it's really a wash as to whether people want to type or speak the structured text.”

Sidebar

Testing Out a Proposition

Larry Garber, M.D.
In the spring of 2008, Larry Garber, M.D., the medical director for informatics at the Fallon Clinic (Worcester, Mass.), called on the folks at the Burlington, Mass.-based Nuance Communications, Inc., with a proposition. The Fallon Clinic physicians had once tried an earlier version of Nuance's voice-recognition software, and found it unsatisfactory. But they knew that the software had advanced considerably, and believed it quite possible that the company's newest version would work well for them. So Garber and his colleagues initiated a small study, with the cooperation of the Nuance folks.

“We looked at the notes and practice patterns of 10 physicians for one week,” Garber reports. “Then we put them on Dragon Naturally Speaking and measured them again for another week to see how they were doing. And we didn't yank away their dictation devices if they felt they needed them.”

The result? A 95 percent reduction in the amount of dictation and transcription, as well as a dramatic improvement in availability of physician notes as facilitated by Nuance's voice-recognition software. The average time for availability of a document transcribed abroad, Garber says, had been two days, with another two days required to get a note finalized in the organization's EMR. With the Dragon product, that was reduced to 46 minutes. Not surprisingly, the Fallon organization quickly moved to implement the solution.

Of course, every patient care organization will come up with different results in terms of pre- and post-implementation of voice-recognition solutions, but one thing is clear, say industry observers: the pace of adoption of such solutions will only accelerate in the next few years, particularly because of the “meaningful use” requirements still to be fleshed out in the federal stimulus legislation passed earlier this year.

Healthcare Informatics 2009 June;26(6):68-70

Sponsored Recommendations

ASK THE EXPERT: ServiceNow’s Erin Smithouser on what C-suite healthcare executives need to know about artificial intelligence

Generative artificial intelligence, also known as GenAI, learns from vast amounts of existing data and large language models to help healthcare organizations improve hospital ...

TEST: Ask the Expert: Is Your Patients' Understanding Putting You at Risk?

Effective health literacy in healthcare is essential for ensuring informed consent, reducing medical malpractice risks, and enhancing patient-provider communication. Unfortunately...

From Strategy to Action: The Power of Enterprise Value-Based Care

Ever wonder why your meticulously planned value-based care model hasn't moved beyond the concept stage? You're not alone! Transition from theory to practice with enterprise value...

State of the Market: Transforming Healthcare; Strategies for Building a Resilient and Adaptive Workforce

The U.S. healthcare system is facing critical challenges, including workforce shortages, high turnover, and regulatory pressures. This guide highlights the vital role of technology...