This year at the Healthcare Financial Management Association’s 2016 Annual National Institute (ANI), Brian Sanderson, managing principal of the Crowe healthcare services group, presented a session with Scott Hawig of Froedtert Health. Hawig is the senior vice president of finance, CFO, and treasurer at Froedtert. Sanderson and Hawig presented to nearly 200 ANI attendees in Las Vegas.
The session, “Exceptions Resolution: How Machine Learning Can Transform Your Revenue Cycle,” focused on how integrating machine learning algorithms can continuously improve an organization’s revenue cycle performance and quickly resolve account exceptions. The pair discussed the results of Froedtert’s experience with automation and machine learning.
“To create a parallel within our country’s history, the greatest impact to the industrial revolution was the introduction of automation within the manufacturing environment,” said Sanderson. “And automation will have the same effect on revenue cycle operations now. This will fundamentally change the way we do business going forward. It’s going to happen.”
Using artificial intelligence and automation, organizations such as Froedtert Health have applied proprietary Crowe machine learning technology to track account exceptions, discover trends among exceptions, and resolve exceptions through artificial intelligence. By finding root cause patterns using historical data, the technology can provide transparency on follow-up activities at the account level, increase the reliability of resolving accounts in a consistent manner, and resolve a high volume of accounts with minimal human intervention.
Sanderson and Hawig described how uncovering exceptions via machine learning can help healthcare systems achieve improved and standardized performance in resolving account anomalies. They discussed studies completed to determine key exception areas of focus, statistical validity findings associated with data scientist reviews, and process and portfolio changes for credit balance and denials follow-up responsibilities. Early study results related to credits included identification of 41 percent of accounts and $2.2 million in outstanding liability (in a total credit balance population of $5.4 million aged 0-30 days) that could be resolved without human intervention, saving organizations 15 minutes per account and providing greater standardization.
Attendees left with a greater knowledge of the application, methodology, and benefits of machine learning, which will allow organizations to maintain control, increase reliability, and reduce cost. In addition, machine learning will help revenue cycle leaders stay ahead of the industry and control risk.
Learn more about Exceptions Resolution and how Crowe is using machine learning to help organizations improve revenue cycle performance.