The following is a guest article by Yair Saperstein, MD MPH, CEO and Co-Founder of Avo
As many anticipated, Epic announced the development of their native AI-Scribe this week at their user conference, UGM. As the founder and CEO of an AI healthcare startup that includes AI scribing as a feature and that integrates with Epic, here is my perspective on the announcement and what it means for healthcare organizations and startups serving this market.
Health systems no longer need third party vendors for AI scribes.
Epic’s investment in native workflows isn’t new. The company has a pattern of waiting for market proof, and then scaling through their distribution moat as the dominant EHR in the country. With AI scribes among the fastest adopted technologies ever in healthcare, this addition is not surprising given a large portion of a clinician’s time is spent on note writing.
Epic (and most other EHRs) began as the digital medical record, a single database of patient data, and grew by building workflows on top of that database. Epic continually enhances its applications with features to enhance those workflows natively – now adding ambient listening.
This has always been part of their strategy, maybe controversially. There have been recent claims that “Epic allegedly mandates that clients…must also use Epic products in related software categories, even if third-party alternatives are superior or already in use.” Regardless, health systems will make decisions based on what’s simple to integrate, cost effective, and driving ROI via true adoption. Epic will eventually get that right for scribing. Interestingly, in the announcement today, Epic shared that they are working with Microsoft to provide that function natively.
Scribe-first vendors will struggle.
Large AI scribe startups have been preparing for this moment by expanding horizontally with additional functionality. Most notably, Abridge announced their entrance into prior authorization. Will this additional functionality beyond scribing be enough to convince health systems to continue to use Abridge in place of just adding the feature in Epic?
I don’t think so. Think about Amazon’s Alexa. Alexa was judged to be a “colossal failure,” because it never generated revenue for Amazon (in fact, it led to $10 billion in losses). Amazon hoped Alexa would be the new de facto workflow for shopping on Amazon. Instead, people play music on Alexa and ask for the weather forecast. Similarly, a lot of money has gone into voice-first EHR workflows (from Oracle Health’s (Cerner’s) new EHR to the multiple billion dollar scribe valuations), but as Epic has just shown, scribing is just a feature. Furthermore, Epic is just starting with scribing. They shared from the stage today that they are making a foray into prior authorization as well.
No, this isn’t the end of AI startups.
Health systems will always experience workflow gaps in Epic. Startups who are built for AI scalability and customizability will survive; others will die.
Epic has become a massively complex database. It will never be capable of deploying AI to streamline every single workflow. For example, many high-ROI clinical workflows rely on sourced medical content and proprietary guidelines, and content owners are hesitant to partner with Epic, fearing intellectual property risks. Health systems also have unique needs for customization. Going through Epic means waiting for their often lengthy IT timelines.
How do startups compete? They must build scalable AI engines (not point applications) that are able to ingest patient EHR data, process that data alongside trusted clinical content, and provide integrated clinical workflows. They must go a step further – ensuring those workflows are easily customizable so health systems turn to them first because their time to value will be faster than going through Epic.
As in every industry, startups exist for a reason. Incumbents never move fast enough, and they never will. In healthcare, the challenge for startups is not only to move fast – but to deliver value so great and so seamless that health systems are willing to utilize them in addition to their trusted EHR partners.
About Dr. Yair Saperstein
Yair Saperstein, MD MPH is CEO and Co-Founder of Avo, the AI-engine used by healthcare organizations to improve care and operational outcomes in a way clinicians love. He is also a dual board-certified physician in internal medicine and clinical informatics. He is a member of Alpha Omega Alpha (AOA) and is a hospitalist at Mount Sinai Hospital. Dr. Saperstein graduated from Albert Einstein College of Medicine with distinction in research in global health and from SUNY Downstate with a Masters in Public Health in hospital policy and management.
Before founding Avo, Dr. Saperstein served on the implementation team for multiple Epic go-lives and was inpatient medicine Chief Resident at HHC’s Kings County Hospital, leading over 400 providers through New York’s first wave of COVID. As a dual board-certified physician and clinical informatician, Dr. Saperstein co-founded Avo to meet an unmet need to standardize care in a way that would also reduce clinician burnout.