Getty Images

Epic Taps Ambient Intelligence to Streamline Clinical Documentation

A new Epic EHR integration uses ambient intelligence to draft clinical documentation based on patient-provider conversations in a matter of seconds.

EHR documentation requirements have led to clinician burnout, with the average clinician spending over 16 minutes per patient encounter using the EHR. However, advancements in ambient intelligence are transforming clinical documentation, according to Garrett Adams, product lead for the ambulatory research and development division at EHR vendor Epic.

Ambient intelligence refers to physical spaces that are sensitive and responsive to the presence of humans, according to an article published by the National Center for Biotechnology Information (NCBI).

The technology uses sensors and processors embedded into everyday objects to collect data and utilizes machine learning algorithms for data analytics. For instance, Google Assistant and Amazon Alexa—devices that automatically respond to a person's voice—use ambient intelligence.

In the context of the EHR, ambient intelligence can draft clinician documentation based on patient-provider conversations.

Earlier this year, Epic announced an ambient clinical documentation integration in partnership with Nuance Communications, a Microsoft company.

The integration of the Nuance Dragon Ambient eXperience Copilot (DAX Copilot) allows providers to record in-office and telehealth patient visits with patient consent directly in Epic's mobile application.

Then, the technology produces a draft note within seconds for physician review. Physicians must sign off on the note and make any edits as needed before the system locks documentation into a patient's chart.

Adams, who is also a member of Epic's cognitive computing platform team, noted that ambient intelligence has benefits for clinicians and patients alike.

He explained that traditionally, there are three entities in the room during a patient encounter: the provider, the patient, and the computer where the provider is completing documentation. 

However, with ambient clinical intelligence, the computer takes a back seat during patient encounters.  

"In a scenario where you have ambient, providers can place their phone down on the table or in their pocket, hit record, and just focus on the conversation with the patient," Adams told EHRIntelligence in an interview.

In addition to an improved patient experience through more eye contact with their provider, ambient documentation has carry-over benefits for clinicians even after a visit is over, Adams said.

Often, clinicians must complete EHR work after hours, also known as "pajama time." In a recent nationwide survey, physicians reported spending an average of 15 hours per week working outside of their scheduled hours.

"Clinicians have a responsibility to complete documentation and close the visit at a certain point," he explained. "In a packed schedule, that work gets pushed later if they're not able to do it in the room with the patient."

However, with ambient intelligence, providers have a summarized note that captures relevant medical information within five to 10 seconds after the patient-provider conversation is complete.

"It has major potential and already proven ROI on reducing that additional time that the provider needs to spend later to complete documentation," said Adams. "I think that's one of the strongest value props for clinicians specifically."

Pilot users of the DAX Copilot Epic EHR integration have reported decreased pajama time, reduced feelings of burnout, and the ability to see more patients in a day, Adams said.

Additionally, some providers have reported an increase in documentation quality.

Adams saw this firsthand when he shadowed one of Epic's first ambient go-lives.

When a physician got his first AI-generated clinical note back, the provider remarked on the level of detail the system captured within the patient history section of the note. He even said that he didn't think he could have captured the information as effectively on his own, Adams pointed out.

"Folks don't want to go back once they use it, so that's been really rewarding," Adams said.

"This is something that Epic and the folks we work with have viewed as transformative for the patient-clinician experience for quite a few years," he noted. "I'm incredibly excited that the time is now and that the feedback that we're getting from organizations, from healthcare providers, from patients is exactly what we could hope that it would be."

Lifespan Health and UNC Health are among the over 150 health systems, hospitals, and medical centers deploying the DAX Copilot EHR integration.

"I think that this is something that will likely become ubiquitous and expected in workflow in taking care of patients and entering documentation," Adams said.

Epic and its partners are working to develop the ambient intelligence integration further. For instance, he said they are working on using ambient to provide suggestions for next actions within the EHR, such as updating a patient's allergy list.

"We can pull that out of the conversation and tee it up as a discreet action for them to enter into the chart," Adams said.

Just like physicians must review ambient clinical documentation before the system logs the note, Adams emphasized that the system will also require providers to confirm any recommended actions derived from patient-provider conversations.

"The human is always in the loop with each of these interactions," Adams said. "Rather than them going and navigating to an activity or clicking and typing in the allergy that they're going to add, they'll see it teed up, and they can accept it and go from there."

As AI technology continues to advance, Adams predicts ambient intelligence will become ubiquitous in clinical documentation workflows.

"Just like you don't think about AI being your auto-complete as you're sending text messages, this is another thing where I think, in a long enough timeframe at least, it'll just be normalized and something we expect," Adams said.

Next Steps

Dig Deeper on Health IT optimization

Cloud Computing
Mobile Computing