Medical conversation startup Abridge raises $12.5M

The company also unveiled its enterprise product, summarizing reports for clinicians

When Abridge launched in 2020, it introduced technology for patients to record their doctor appointments, using machine learning to transcribe and highlight the most important parts of the conversation. The idea was helps patient better understand the details of what their doctors had told them by distilling it down and making it easy to understand.

Of course, there's the other side of that conversation: the physician, who is becoming increasingly burned out. More than 50 percent of providers experience burnout-related symptoms, and systems are anticipating healthcare staffing shortages that will cost $4.6 billion every year. They would also benefit from a service that could accurately summarize their reports for them.  

That's what Abridge is now offering, announcing a new enterprise solution on Thursday, along with an oversubscribed $12.5 million round of funding.

Healthcare is about people

Founded in 2018, Abridge's initial product was aimed at patients, using machine learning to give users a transcript of the medical parts of their appointments, highlighting important definitions and takeaways, such as what next steps a patient is supposed to take, or instructions on how to properly take their medication. 

The challenge on the patient side is that people forget up to 80% of what they've heard, Dr. Shivdev Rao, CEO and co-founder of Abridge, explained in an interview.

"When we started this company our thesis really was, and continues to be, that healthcare is about people. When you think about healthcare delivery, there's a doctor like me, and then there's a patient, and when you think about what they're doing, they're having conversations," he said.

"Increasingly, those conversations are happening over telemedicine, especially since the pandemic and that's just a part of healthcare now. They're also always going to happen over the phone, and they're certainly always going to happen in person, as well," he said. 

All the patient has to do is hit the record button during their appointment; the recording is then uploaded to a HIPAA-compliant server, and Abridge’s machine learning algorithms create a selective transcript of the medical parts of the conversation. It will also highlight takeaways from the visit, such as what next steps should be taken and prescription instructions. The app, can be used in routine annual exams, as well as in urgent care appointments and specialist conversations.

"We set out to build technology that can be a part of all of those conversations. The thesis really is that healthcare is about dialogue, it's about conversations, and so we've been building technology, since day zero, that can ambiently be a part of them, and then automatically create draft summaries of those conversations, and also structure information from them in a format that can help everyone involved," said Rao. 

Launching an enterprise solution

What Abridge hadn't yet been solving for where the challenges on the provider side, most notably what Rao calls "pajama time," aka the hours that physicians spend at night, after work, writing up notes about all the patients they've seen that day.

"There's nothing more soul crushing than at night to be in front of your computer spending hours writing notes about the patients that you saw earlier that day. And now, more than ever, that burnout challenge for doctors and patients is top of mind for health systems across the country," said Rao.

Abridge's enterprise solution takes the same technology that it has used for its patient-centered product, transcribing the call and pulling out highlights, but then goes a step beyond that by automatically creating a draft note of the conversation in the standardized format that  doctors in the United States, and around the world, use daily.

That means breaking down the conversation into different headings including social history, which can include the patient's alcohol usage and smoking; diagnostics and appointments, including when and where the patient should be going for follow-ups; and medications that the physician prescribed during the visit.

This is all sent to the doctor in near real-time: it currently takes around a minute for the report to be sent to the physician, and Abridge plans to make that even faster going forward. The doctor can also then share this summary with the patient, along with the original transcript that they get from using the app.

"We're automatically creating a note draft here that shows up in the medical record for the provider to trust and verify. Providers currently can't afford these sorts of solutions at scale; they spend incredible amounts of money for humans to help them do this. Now we've got technology, finally, after four years of R&D, that's above and beyond anything in this space. We've got technology that can scale to all the conversations that are happening across the system," Rao said.

So far, Abridge's enterprise solution has been used by 2,000 physicians and the ROI for them has been two-fold: first, it takes care of that aforementioned pajama time, meaning the doctor no longer has to spend hours at night creating these notes and summaries. Second, is that it  structures data for them, which can then be piped into the electronic medical record, and can help with other clerical tasks.

On top of that, it can also help with physician recall: Abridge's solution recorded over 90% in recall, compared to around 70% for a human. That helps create a more complete record, which Abridge believes doesn't just help the bottom line for health systems, but that it also helps patients, because now the patients have a better care plan.

"That makes a huge difference on the value that the provider can capture and can create from that note for the system, but it also makes a huge difference in terms of the value that the patient could potentially get later, because they're going to see those visit summaries, they're going to read their care plans and the notes that we create are a more complete and accurate depiction of what was actually discussed," said Rao.

In addition to its new enterprise offering, Abridge also revealed that it is now making its machine learning technology available via API, which will help care delivery integrate with their existing telemedicine and call center platforms. The company so far has technology partnerships in place with Twilio, Genesys, and Vonage.

New funding round

The new funding, Abridge's Series A-1 round, was led by Wittington Ventures, along with participation from all existing investors including Union Square Ventures, Bessemer Venture Partners, Pillar Venture Capital, and UPMC Enterprises, as well as new investors Yoshua Bengio, and Whistler Capital. This bring Abridge’s total raised to $27 million.

The company says it will use the funding to grow its team, and to expand its R&D to further improve and automate documentation workflows, with the goal of getting its technology everywhere as fast as it can. That mwans spending time and effort on the go-to-market, partnerships, and also on sales to get its software distributed.

"At a high level, what we're looking for is to improve the healthcare delivery experience for everyone involved, most importantly for patients, and for those providers that are caring for them. And we're really trying to improve the quality of those conversations and we’ve built technology that we believe can scale to all of them," said Rao.

"We'll always have a focus in this company as a product-driven company, which is also a rarity in healthcare, we'll always have a focus on improving the end user experience, on improving the technology, but also making it as easy as possible for people to be the best versions of themselves, whether it's as a provider or the healthiest versions of themselves as a patient."