Athenahealth revamps revenue cycle management in the era of AI
Athenahealth is adding AI features to its practice and revenue cycle management capabilities, at no additional cost, as the health IT company leans toward becoming AI native.
What if your revenue cycle management system were developed in this new age of artificial intelligence? That's what athenahealth is pondering as the company revamps its practice and revenue cycle management capabilities with emerging technologies, like agentic and voice AI.
Athenahealth recently announced a slew of new AI capabilities being built into athenaOne, an integrated, cloud-based health IT system that combines EHR, medical billing and practice management and patient engagement. The system is popular among small and mid-sized physician practices, which tend to struggle with AI adoption due to high initial costs, lack of staff skills and training and difficulties with legacy system integration.
The new AI-powered capabilities include automated insurance selection, patient liability estimation, waitlist scheduling, express coding with real-time clinical documentation improvement, payer portal agents and voice AI and automated denial advice, at no extra cost to users. These capabilities, according to athenahealth, are going to produce "revenue cycle results that were unimaginable only a year ago."
"We're thinking about, in some cases, orders of magnitude improvement that we can bring in terms of reducing rejections, improving posting quality and getting rid of the exceptions because revenue cycle management is almost all about exception management," Paul Brient, chief product and operations officer at athenahealth, told Rev Cycle Management.
"We're in this incredibly cool place where these new gen models show up that can do things that previous iterations of AI could not, and we're already structured to be able to take advantage of those things on behalf of our customers," Brient continued.
Practice and revenue cycle management are far from AI native, but that is the goal for athenahealth.
"AI native to us is different; it's the next logical extreme, which is: What if we wrote athenaOne when AI was present in its current form? So, it is a pretty fundamental rethink now on revenue cycle," Brient explained.
Being AI native is starkly different from the current reality for most practices. The majority of providers use multiple, standalone solutions to manage their revenue cycles. Oftentimes, those solutions merely bolt on to practice or revenue cycle management systems. Other practices may also utilize solutions that incorporate AI, but it is not part of the system's core functionality.
The bolt-on ecosystem of solutions can be difficult to manage for resource-strapped practices, even as the cost of AI comes down. When practices have to navigate multiple vendors and their contracts, the costs add up.
Native AI is distinctive because it describes AI as a core, fundamental part of an IT system from the beginning, in which AI capabilities drive the system's functionalities. For athenahealth, being AI native allows users to access AI capabilities while being an IT architecture that enables rapid deployment of new features, resulting in faster innovation within revenue cycle management.
"For the most part, we hear practices being excited about AI and they feel they need to adopt it, but we are inundated by a whole set of startup companies that are promoting all kinds of cool AI," Brient said. "It's exciting, but it's overwhelming, and frankly, as some startups do, sometimes the marketing vision and on-the-ground reality aren't as well correlated as people would like."
Becoming AI-native can help athenahealth's users tap into the world of AI with a trusted partner. The company wants to "provide customers with the ability to take advantage of all this stuff without having to have an IT shop that has to evaluate all kinds of technologies and run trials and whatnot," Brient explained.
AI to take revenue cycle to the next level
Revenue cycle management is one of the top use cases for AI in healthcare; its repetitive nature lends itself to the application of analytics and AI. That's why athenahealth is weaving AI capabilities into its system to achieve this smart, smoother revenue cycle.
"When you send a claim in and it gets paid, that's fine and easy, but that happens just a percentage of the time, not a hundred percent," Brient said. "Revenue cycle is really all about how we deal with those exceptions, and with the current gen of AI built into athenaOne, we can manage more and more of those exceptions and take away a ton of the work that our clients do in the cosourcing model."
Speaking of those exceptions, Brient called out claim rejections as an area AI can really make a difference for practices. He envisions rejection rates being in the single digits as a percentage of claims, even nearing 0%, with the help of AI.
Revenue cycle is really all about how we deal with those exceptions, and with the current gen of AI built into athenaOne, we can manage more and more of those exceptions.
Paul Brient, chief product & operations officer, athenahealth
"I'm not certain exactly how low we can go, but we have AI out scrubbing payer websites to see what's changed," Brient said, speaking of claim payment policies and requirements. "If payers actually tell us what's going on, we can react to it proactively many times. Then, if we see a rejection or receive a denial, AI is able to identify the anomaly, react to it and get a rule in place quickly."
Implementing a new rule in the revenue cycle management system can take months, as humans must first identify a pattern in rejections or denials, then develop the code for the rules-based engine. AI can reduce that reaction to days or even hours, according to Brient.
With the maturation of agentic and voice AI, technology is poised to significantly improve revenue cycle outcomes significantly and at scale.
For example, athenahealth is using agentic and voice AI to support capture of claim status, remit downloading and determining prior authorization status to provide faster and more accurate results for practices. AI is also generating real-time claim correction suggestions for coding-related denials, which athenahealth reported leads to a 26.4% increase in payment recovery versus manual corrections.
Improving the front-end, patient experience
Additionally, AI can optimize the front-end experience by selecting the appropriate coverage from an insurance card image and automatically identifying open time slots to get patients off the waitlist.
In pre-Alpha development, athenahealth is also testing AI to estimate patient financial responsibility for commercially insured patients based on payer contracts and services likely to be provided during a scheduled encounter.
"One of the most frustrating parts about the U.S. healthcare system is that patients show up, get a service and they don't know how much they're going to have to pay until, in many cases, months after the service," Brient said.
Athenahealth is leveraging data from approximately 160,000 providers to analyze historical payments. AI can then estimate what a patient's liability is for similar services performed on patients with similar insurance coverage.
"This is patient liability estimation at a level we haven't been able to do in the past," Brient stated. "And if we can't give you an estimate, we can at least give you ranges. So, this helps practices collect more money upfront and helps patients understand what the expense is going to be."
Applying AI to the front-end of the revenue cycle can get healthcare closer to that retail-like experience that providers and patients have been hoping for.
"On the administrative side, the holy grail is having a patient come in, pay upfront or at checkout, if appropriate, and they know exactly how much they owe," Brient said. "This transaction occurs, then it goes to the insurance company, and it comes back paid. In the rest of the economy, for the most part, that's the way transactions work. I think AI might be able to get the U.S. healthcare system there; it would take so much friction out. So, that's where we're trying to go."
AI to put the human back at the center of healthcare
Healthcare is going through a digital transformation, but it hasn't always been the easiest transition for providers. The introduction of technology in healthcare has shifted the very human interaction at the center of care, Brient explained. But AI has an opportunity to reprioritize that relationship.
"My vision [for AI] is it being like a resident that has talked to the patient, studied the chart and knows you, as a provider, to serve up the right information to you," he said. In this world, providers may never even have to touch a computer to get all the necessary information for a high-quality patient visit, he continued.
Evolving AI capabilities are bringing about a "quantum change to the healthcare workers' experience," whether they are physicians or practice staff, he added
As health IT systems, including revenue cycle management solutions, mature, AI can remove administrative barriers between providers and patients, restore personal connections from earlier decades of healthcare and enable providers to focus on what matters most: human-centered healthcare.
Jacqueline LaPointe is a graduate of Brandeis University and King's College London. She has been writing about healthcare finance and revenue cycle management since 2016.
