Abridge launches generative AI tool for emergency medicine with Emory, Johns Hopkins as early adopters

Abridge has spent the last six years building generative AI tools to help doctors with medical documentation, and the company continues to rapidly roll out new capabilities and features.

The company launched a new gen AI product for emergency care, and it's currently in use at several health systems including Deaconess Health System, Emory Healthcare, Johns Hopkins Medicine and UChicago Medicine. 

Abridge Inside for Emergency Medicine was developed as part of Abridge's participation in the Epic Workshop program. The program features third-party vendors that are co-developing technology with Epic. 

Abridge's new gen AI product was developed to suit the needs of emergency medicine as it integrates with Epic’s note drafting workflows for emergency department clinicians and works with Epic’s ASAP module in Haiku and Hyperspace, according to the company.

Clinicians can select patients from the department’s Track Board and begin an ambient recording immediately with Haiku, Abridge executives said.

Abridge’s automatic speech recognition models detect specialty, language and multiple speakers, without manual setting adjustments, to create an accurate transcript and clinical documentation.

“Abridge Inside has reduced burnout and increased job satisfaction in the Emergency Department without changing anything else,” Tricia Smith, M.D., an attending emergency physician at Emory University Hospital Midtown in Atlanta, said in a statement. “We have adopted one technology that has really impacted our work and, therefore, our patients—and I can’t think of a single person who uses Abridge who would go back.”

As Abridge rapidly builds gen AI tools for clinicians, physicians and nurses, there was an opportunity to tackle the documentation burdens facing ED clinicians, said Shiv Rao, M.D., a cardiologist who launched Abridge in 2018 and now serves as CEO.

Emergency medicine clinicians face the highest rates of burnout of any specialty at 63%, according to a 2024 Medscape survey.

"There is a lot of burnout for emergency medicine clinicians, and that's a big part of our why we want to be able to unburden them and help them focus on the most important person, their patient, and help them get more fulfillment from their medical practice," Rao said. "When you think about emergency medicine, it's one of those specialties that is very singular in terms of the style of the note they actually create. When you're a doctor in emergency medicine seeing a patient for shortness of breath, you need to write a note that takes into account the differential diagnosis in a very complete and thorough way."

Emergency medicine also has a different workflow compared to other medical specialties, he noted.

"It's a chaotic atmosphere. Sometimes you're walking into a room and you're paged to another room and you've got to go to some other emergency and then come back to that first room and continue the conversation. Also, sometimes you're seeing a patient and then they're getting an X-ray or CT scan or getting some labs drawn and coming back later to review. We also had to build features that support that kind of discontinuous workflow with any given patient, that allows them to stitch together all of their conversations or encounter interactions with that patient to create one coherent note at the end," he said.

Abridge developed its emergency medicine gen AI tool in collaboration with ED clinicians.

"We had to build a note that checks off their stylistic preferences and medical-legal compliance-related objectives and that has been part of the challenge. The last piece is being able to integrate it deeply into their medical record," he said.

Working alongside electronic health record giant Epic as a co-development partner enabled Abridge to integrate it seamlessly into one of the most widely used EHR systems, he noted.

Pittsburgh-based Abridge, one of Fierce Healthcare's Fierce 15 of 2024 honorees, uses AI to increase the speed and accuracy of medical note-taking, leveraging a proprietary data set derived from more than 1.5 million medical encounters. The company's AI converts a patient-clinician conversation into a structured clinical note draft in real time and integrates it seamlessly into the EHR system.

The company has rapidly grown to be one of the top players in the medical gen AI space and has raised $212.5 million to date, banking a $150 million series C funding round last February.

The company cites data that show clinicians who use Abridge expend 86% less effort writing notes, do 60% less after-hours documentation work and report a 55% reduction in burnout, the leading cause fueling the U.S. doctor and nurse shortage. 

2024 marked a year of rapid growth and expansion for Abridge as the healthcare industry increasingly adopts gen AI technology for medical documentation.

“We scaled across the country very quickly and got incredible amounts of feedback from tens of thousands of users,” Rao said. “That feedback allowed us to very quickly iterate and meet the scale of the opportunity here. I think one of our most important mantras in this company is to deliver a truly enterprise healthcare-grade solution, and, to us, that means being able to scale across all the different specialties in all the different settings and even in all the different spoken languages inside of a healthcare system.”

He added, “It sounds easier than it actually is, because, as I can attest, I was on call last night in the hospital, and the cardiology encounter can look and feel very, very different than a primary care encounter, than a gastroenterology encounter. The notes that I end up creating also have a very specific style and structure to them that is idiosyncratic.”

In the past year, Abridge has focused on scaling up gen AI technology that can “lean into” provider preferences and pull in patients' medical history to create a “clinically useful and billable note,” Rao said.

In August, Abridge inked its largest partnership to date with Kaiser Permanente, making the company's AI-powered medical note-taking app available to more than 24,000 doctors across its system.

Its gen-AI-powered platform has been deployed at the University of Vermont Health System, Christus Health, UChicago Medicine, Sutter Health, the Yale New Haven Health System, UCI Health, Emory Healthcare, the University of Kansas Health System, UPMC and dozens of other health systems.

The company is also collaborating with Mayo Clinic and health IT giant Epic to develop a gen AI ambient documentation workflow for nurses. 

Many health systems are ready to shift from gen AI pilot projects that focus on primary care or particular specialties to scaling the technology enterprisewide across the organization.

“It can quickly become a case of the ‘haves’ and the ‘have nots’ inside of a system where, if you're a doctor, and you're looking over your shoulder and you see someone else using this tool and leaving to go home on time, then you want to also benefit from this technology. Increasingly, what we saw over the last couple quarters, especially, were health systems opting for an enterprise agreement with us,” Rao said.

“That allow us to innovate in unexpected ways because we might learn something we never anticipated, that a certain specialty was going to use this for a very specific type of encounter that wasn't even on our list. That kind of unexpected innovation and unexpected ideation is always something that we look forward to with these enterprise partnerships,” he said.

As evidence of Abridge’s rapid adoption in healthcare, the company’s name has now become synonymous with ambient note-taking.

“We've heard clinicians start to use Abridge as a verb, which is delightful and unanticipated,” Rao said. “They’re going to ‘Abridge their conversation' or ‘Abridge a medical encounter.’”

It’s similar to how people use the company name Google to refer to searching for something online: “Google it.”

“What that means to us is that they're recognizing that we're not only trying to summarize and distill, but we're also trying to help them build a bridge between themselves and all the work they have to do and all the different stakeholders they serve,” Rao said. “We’re helping them create a clinically useful, specialty-specific note, and we're also helping them create a billable note that meets criteria for coding completeness, whether it's fee-for- service or value-based care. And, we’re also helping them create a more patient-readable and understandable summary or note that can meet their patients wherever they are.”

On the the patient-facing side, Abridge is partnering with OpenNotes, a health transparency research group that is part of Beth Israel Deaconess Medical Center in Boston, to evaluate AI-generated patient visit summaries, but from a patient-centric lens.

Artificial technology is quickly evolving, and, as a vertical AI company, Abridge can build capabilities by leveraging new discoveries and innovations both inside and outside the company, Rao noted.

He pointed to DeepSeek, a Chinese AI startup that shook up the tech industry last week when it released an AI model that appeared to perform on par with OpenAI’s reasoning model but with far less computing power required, which means lower costs, as The Wall Street Journal reported.

“It’s one of the most exciting things I've read about or learned about in the last two weeks,” Rao said.

“We've built this company wherein we have a scientific center that understands that our priorities are to scale the best-in-class note across the country and we have built a modular stack of technologies that allows us to very quickly ‘mix and match’ ingredients from inside and outside to deliver on that,” he noted.