MaineHealth gives docs voice with EHR rollout

August 30, 2012 in Medical Technology

PORTLAND, ME – In the midst of rolling out a new electronic health record system across eight hospitals, MaineHealth will also deploy speech recognition technology to make it easier and quicker to fill in the patient chart.

MaineHealth includes Maine Medical Center in Portland, seven other hospitals across the state, as well as physician practices.

Maine Health will deploy Nuance Communications’ Dragon Medical 360 | Network Edition across health system. It already uses Dragon Medical 360 | eScription, Nuance’s hosted transcription service. eScription is what Maine Health CIO Barry Blumenfeld, MD describes as a backend system. The physicians dictate as they would in traditional transcription. The words run through an engine and are converted to text that is then reviewed by a transcriptionist.

[See also: Maine docs move to one record]

The network edition of Dragon will populate the chart as the physician talks into a microphone.

“This technology would recognize their speech and would fill it in as they are speaking – in near real-time,” said Blumenfeld.
As part of its “One Patient, One Record” initiative, MaineHealth is in the process of deploying its Epic’s EHR across its eight hospitals, with Maine Medical Center slated to go live Dec. 1. The health system will go enterprise-wide with both the new Epic EHR and the speech recognition technology, Blumenfeld said.

[See also: Maine docs move to one record.]

“At this very moment, we’re doing our testing and our user acceptance testing,” he said. “We have people that are about to begin their training, and we’re finalizing our go-live plan. We will have all eight hospitals up on Epic by Dec. 6, 2013.”

[See also: UPMC, Nuance aim to take speech recognition to new level]

That means going live with Dragon, too. Blumenfeld sees it as an integral piece of improving accuracy and efficiency within the electronic health record – and also a way to lower the costs of clinical documentation and related traditional medical transcription.

But, while Maine Health purchased an enterprise-wide license, he said, it will not be forcing doctors to use it.

“We’re not getting rid of transcription altogether,” Blumenfeld said. “There are certain specialties and there are certain areas where we may not be able to eliminate it today or, in some cases,  ever. There are areas that just are resistant to that sort of change. On the other hand, there are areas that are very open to it.

One area primed for adopting speech recognition technology is the emergency department, he said. It’s also an area ripe for reducing costs.

“We spent hundreds of thousands of dollars on transcription each year, he said, and we hope as we introduce Dragon in the ED to eliminate those transcription costs.”

Maine Medical Center has saved more than $5 million to date and doubled transcription productivity as a result of leveraging Nuance’s hosted transcription service, eScription.

[See also: UPMC, Nuance aim to take speech recognition to new level.]

Blumenfeld appreciates the savings, but as sees it, there are benefits that trump even that level of cost-cutting.

“Transcription – even the best transcription – there’s a lag time until it comes back,” he said. “Obviously anytime a physician is done their documentation using speech and has got that text in there immediately, it’s available immediately. So, one big benefit is faster turnaround time on documents, greater availability and as such better quality and safety for patient care because you’ve got the information right away. All of that information is available very, very quickly, with no lag time.”

Perhaps dictating is a chore for physicians, but for many it is a chore they prefer to typing.

“You find sometimes when you force physicians to use direct entry when they’re just not real good typists, or they’re not real wild about it — whatever their reasons, they tend to be a little more terse than you would like,” Blumenfeld said. “With speech recognition, they tend to say more and again there’s value in terms of quality and safety when a physician tends to give a more complete description of the problem at hand.”

Janet Dillione, executive vice president and general manager of Nuance Healthcare, says Maine Health is doing the right thing by introducing speech recognition. 

“MaineHealth is ahead of the curve in embracing these changes and it’s leading the charge to create a more universal approach to patient care across their IDN,” she said. “This endeavor will undoubtedly benefit MaineHealth’s patient community and the hospital’s bottom-line in the coming years.”

Blumenfeld agrees. The adoption of Dragon technology fits with Maine Health’s mantra: “One Patient, One Record.”

But, he recognizes there will be difficulties.

“The challenge with it is that even with the best technology — and I believe that Nuance probably has the most developed speech engine out there – there still will be occasional defects in what’s transcribed,” he said. “So, the physician has to be willing to go in and correct — or at least review then correct any mistakes that are in the text. And, there is also a learning curve associated with it – just learning to navigate and learning the various commands, and so forth…”

The physician informaticist in me always wants to add a caveat,” Blumenfeld adds. “It’s text. It’s not as valuable as discreet data. But we are requiring certain discreet data to be entered and there are often other individuals that enter it. So problem lists, labs, vitals, meds — those things are all entered directly as discreet data. The combination of the text from the voice recognition and the discreet data makes for a much more complete chart that’s available to many more people.”

Be the first to like.
VN:F [1.9.22_1171]
Rating: 0.0/5 (0 votes cast)

Article source:

Be Sociable, Share!
Bookmark and Share

Leave a reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>