Artificial intelligence (AI) might be one of the most misused words in the technology vernacular. It’s often erroneously interchanged with machine learning (a subset of AI) and has become an umbrella term that either incorrectly groups capabilities of dissimilar technologies or synonymizes related technologies that are actually different.
As the healthcare industry rapidly invests in new technology capabilities, the confusion of what “counts” as AI has spilled over into the world of intelligent automation (IA) in revenue cycle management (RCM). To clarify, think about AI as a subject, and intelligent automation as a tangible application of that subject.
When people think of AI in RCM, the first thing that comes to mind is often robotic process automation (RPA). Having a few bots, however, does not actually equate to intelligent automation. RPA is a primary component of intelligent automation but on its own it lacks true “intelligence.” Wrapping RPA with other disciplines of AI results in intelligent automation — for example, creating a digital worker that leverages RPA to take actions informed by a machine learning model to direct workflow.
Additionally, given the vast amount of paper that still powers much of the healthcare system, computer vision/optical character recognition and natural language processing (NLP) are foundational tools for IA. They bring more intelligence to expand the kind of work digital workers can automatically perform.
For example, can a bot read a paper document? No, but a digital worker with computer vision can. It operates at the transaction layer, automatically filling in required fields extracted and validated from paper documents. And by leveraging workflow orchestration, digital workers can initiate seamless handoffs between the digital and human workforce to enable meaningful operational transformation.
As a general expectation, the point of automation in RCM is not to automate everything. Detailed process review and domain expertise within revenue cycle functions are necessary to avoid automating high-risk or overly complex tasks. It’s possible to automate 80% to 90% of the volume that flows through certain paths, but at some point, the effort to automate edge cases outweighs the return.
That’s why digital workers take on the redundant, error-prone processes at scale, while people on the front and back ends manage complex items. Both are needed to realize the full value of the automation ecosystem and codify the steps to a healthcare system’s particular needs. The human worker then takes on the exceptions, parts of the process subject to complex variabilities.
Ultimately, there’s no intelligent automation without artificial intelligence. But AI on its own, does not transform RCM in a meaningful, sustainable way. It takes layers of technology, a combination of digital and human workers and deep expertise to create a modern, scalable revenue cycle.
You can find out more about how to apply automation for the best results in this white paper: The Case for Intelligent Automation in Revenue Cycle Management as Part of Your System-wide Technology Upgrade.
Sean Barrett is the Senior Vice President of Digital Transformation at R1 RCM. He joined R1 in 2018 and currently oversees R1’s core product management, automation, and machine learning functions. Prior to R1, Sean spent 14 years at Deloitte Consulting focusing on serving clients primarily in the healthcare provider segment-leading operational performance improvement and technology-driven transformation engagements at many of the largest health systems in the country.