Generative AI delivers new revenue cycle value for Cleveland Clinic
Health systems are increasingly turning to generative AI to tackle persistent pain points in RCM, especially in the labor-intensive mid-cycle. During a recent conversation hosted by Becker’s Healthcare and AKASA, leaders from Cleveland Clinic and AKASA shared how generative AI is driving measurable improvements in inpatient coding, documentation integrity and claims accuracy.
The panel featured Bob Gross, executive director of financial decision support and analysis at Cleveland Clinic and Benjamin Beadle-Ryby, senior vice president and co-founder of AKASA.
Here are four key takeaways from the conversation.
1. GenAI advances traditional automation
Legacy solutions like robotic process automation and natural language processing helped health systems chip away at revenue cycle inefficiencies during the advent of digital transformation but failed to solve the mid-cycle, where coding and documentation tasks require significant cognitive effort.
“Computer-assisted coding was the big innovation of the 2010s,” Mr. Gross said. “For the first time we were using natural language processing and word matching to help optimize the encoding process of inpatient coding. However, it’s still a very manual process that’s entirely a human activity in its current state.”
Generative AI changes the equation. By reading and interpreting the unstructured data in medical records, large language models can now support coders and documentation specialists at scale. This allows health systems to extract accurate, auditable insights without relying solely on human effort.
2. Proving return on investment
Cleveland Clinic partnered with AKASA to deploy a generative AI tool across its enterprise with a focus on coding and clinical documentation improvement (CDI). In the first 60 days, the health system saw notable gains.
“When it comes to something like CDI, that is something that no health system across the country today has the luxury of reviewing 100%, just because it is cost and resource prohibitive,” Mr. Beadle-Ryby said. “AI’s ability to understand and reason with complex language opens the door for so much more efficiency.”
About 15 percent of inpatient cases reviewed by the tool revealed missed diagnosis coding opportunities. Human coders accepted half of those recommendations, leading to an improvement in overall case mix and the capture of comorbidities, severity of illness and other quality indicators.
“We are unlocking revenue that couldn’t be captured,” Mr. Gross said. “Generative AI is consistent Sunday to Saturday, it is always on, it is always reviewing and the documentation that we’re assembling is fully auditable and justifiable.”
3. Better compliance and audit results
A major concern among revenue cycle leaders is how AI-driven coding might affect audit risk or payer relationships. Mr. Gross noted that Cleveland Clinic’s implementation was designed to increase, not reduce, compliance.
Since implementation, Cleveland Clinic has not observed an increase in denials linked to coding changes. If denials do arise, the AI-generated documentation gives Mr. Gross’ team stronger grounds for appeals.
“The audit packet we get with each AI-assisted claim clearly documents line-item evidence for every diagnosis code,” Mr. Gross said. “That level of documentation and evidence simply does not exist within the manual coding process. We’re producing a better result from an audit and a compliance perspective.”
4. The right partner matters
One reason Cleveland Clinic prioritized generative AI in coding and CDI is that the process is relatively governed, rule-based and high-volume. But leaders were also strategic about how they deployed the technology. Prioritizing people over technology and making sure AI “supercharges” their team members was key in deployment success.
Equally important is the selection of the right partner. Mr. Gross advises health systems to look for partners who understand both generative AI and revenue cycle operations.
“When we look for partners, we’re looking for technology excellence and domain expertise,” Mr. Gross said. “This is definitely a domain where you want to partner, you want to buy it. It’s not going to be something that you’re going to be able to build and maintain at scale on your own.”
The post Generative AI delivers new revenue cycle value for Cleveland Clinic appeared first on Becker’s Hospital Review | Healthcare News & Analysis.