October 8, 2024 | Jason Burke and Deanna Berkowitz
I sat down with Jason Burke, vice president of revenue cycle solutions at Solventum, to uncover the true benefit of autonomous coding and how it differs from computer-assisted coding. I also got a sneak peek at the booth activities for AHIMA24.
What is autonomous coding technology solving for revenue cycle leaders and what are the most important elements they should look for in automated solutions?
Revenue cycle leaders are facing a gap in qualified coding resources in the market, and the cost of those coding resources continues to increase. Autonomous coding technology helps fill this gap by addressing the less complex patient records that can be auto coded, allowing coding staff to focus on more complex patients. It can also help organizations minimize the need for outsourced coding to supplement their coding staff.
Accuracy and compliance are the foremost priorities for organizations thinking about adopting autonomous coding solutions. Being confident that the technology is not only coding correctly and adhering to the latest coding guidelines but can also identify inaccurate coding is essential to trusting the results. The next most important priority is that the tools can work with their electronic health record (EHR) and coding staff for a more efficient way of ensuring up-to-date documentation funnels through the autonomous process and final code sets are sent back to the EHR, should the healthcare system’s process require it. Having an integrated tool also means being able to address those records that cannot be auto-coded in a semi-autonomous coding environment.
How is Solventum approaching coding automation today, and how does that differ from what we currently do with computer-assisted coding (CAC)?
We started with CAC as our first solution to address coding challenges, but the models that exist today are so much more sophisticated than they were 7-10 years ago. The evolution of generative artificial intelligence (AI), machine learning, large language models (LLMs), etc. has changed the game and allowed for more truly autonomous coding technology. As we deploy those models, we marry AI with rules to ensure we validate that the information coming from the AI model is correct. I will stress that AI alone is not enough; it gets you very far in producing codes, but must also have underlying rules and other knowledge-based intelligence pieces to be able to supplement and validate it to make sure it is compliant.
What can AHIMA 2024 attendees who are interested in this technology expect to discover at the event?
Honestly, the difficult thing about autonomous coding is that people are used to seeing a demo, but with autonomous coding there isn’t really a “demo” per se because it's automated; it all happens behind the scenes. Sort of a catch 22, right? The key is the trust factor in the coding results and AHIMA attendees who stop by our booth (1118) can view results from our engines. We will show our coder workflow tool, which has been enhanced, and our dashboards that track the success and failure of the models and how we're improving on the models, providing full transparency to the user. We’re also bringing back our caricature artist, just one more reason to say hello.
Jason Burke is the vice president of revenue cycle solutions at Solventum.
Deanna Berkowitz is a marketing communications specialist at Solventum.