As the pandemic heads toward a second year with no further financial stimulus guaranteed, hospitals and health systems are seeking ways to reduce costs and improve revenue cycle performance. Intelligent process automation (IPA) is an emerging solution designed to optimize operations and increase productivity through a combination of process modeling, process automation, and artificial intelligence.

IPA in the revenue cycle enables healthcare organizations to shift manual, repetitive work to automated processes that improve efficiency, accuracy, and financial outcomes. These benefits are particularly important in the healthcare revenue cycle where a maze of confusing payer requirements, redundant workflows and siloed administrative functions push up operational costs and departmental overhead.

Connecticut’s leading healthcare system, Yale-New Haven Health, is breaking the pattern of costly revenue cycle operations—one function at a time. This article explores how the health system’s 1,200-employee Corporate Business Services organization uses IPA in the revenue cycle to tackle inefficiency.

Revenue Cycle Automation at Yale-New Haven Health
Yale-New Haven Health began using IPA to streamline revenue cycle operations in 2019. The organization first analyzed all their high-volume, repetitive tasks that required no human intervention until there was an exception in the case or workflow. Their assessment process involved four steps:

– Evaluate each revenue cycle function for high levels of repetitive, redundant tasks, or work overlaps.

– Step back and perform process mapping. Look at EHR and other existing vendors to ensure efficient uses of all current application capabilities. Implement any capabilities not currently being used.

– Identify any remaining gaps and determine if revenue cycle automation using an IPA platform could fill the gaps for that specific revenue cycle function.

– Work with internal staff and IPA vendors to create a comprehensive physical map of the entire process, new workflow changes, and a timeline for implementation.

In addition to choosing the right revenue cycle process to automate, it is critical to re-engineer those functions to achieve the greatest impact and value to the healthcare organization. “We needed to use all of our existing systems before bringing in new revenue cycle automation,” says Melisa Brereton-Esposito, Director, Systems, Training and Development, Corporate Business Services at Yale-New Haven Health. “We first focused on cash reconciliation and posting, which provided a valuable learning experience for future projects.”

The four-step approach takes time, but yields dramatic results in cost reduction and staff adoption. “If our team doesn’t use the recommended assessment process, the introduction of IPA is of little value,” adds Brereton-Esposito.

Overcoming Adoption Challenges

Initially, there was general distrust among staff regarding how automation would improve or replace their manual work. Concerned about job security, many were reluctant to turn over tasks to the computer. Revenue cycle staff tend to be long-term employees who are cautious by nature. Brereton-Esposito’s department implemented three managerial guidelines with regard to staffing:

– Keep staff whose jobs are replaced by technology—never let them go based on automation.

– Reassign and retrain to jobs that require more analytical thinking. Encourage staff to focus on the next “better” job

– Redistribute staff or wait for attrition in areas that have been automated.

Example of a task currently automated: Correspondence Workflow—Applies to mail that comes into the revenue cycle department, centralized across five hospitals.

Before automation: All letters are received from a lockbox in random order in batches. Staff are assigned to read, sort, and process the letters to different work queues such as an explanation of benefits (EOB), financial assistance applications, approval, and denial letters in the EMR system. This is a highly manual effort and delays in this process may sometimes lead to missing time-sensitive correspondence from the payors and other external entities.

After automation: The technology uses OCR and machine learning to categorize each piece of correspondence based on the content and then moves it to the correct person or place. For all types of letters, the system takes steps to sort and send to the right category. The technology is expected to read approximately 70% and send 30% to the human in the loop. Percentages should improve with ongoing testing, validation, tracking, and working on the exceptions in incoming correspondence.

Checklist for Evaluating Solutions

Automation platforms should use a combination of AI tools along with RPA (robotic process automation) to enable automated workflows, specifically processes like document classification. Solution providers who have an enterprise approach and multi-tenant automation technology platforms can help with long term organizational goals.

Organizations should look for vendors with the knowledge and experience of healthcare processes and have deep technology capabilities beyond RPA, like the capability to handle large amounts of structured and unstructured data, to drive automation. Evaluate vendors beyond a point solution on how the automation platform can scale across various functions and their ability to partner with you to maximize value.

Finally, these systems learn as they go. Vendors should have the ability to scale with reusable components and continuous learning for enterprise-wide automation.

Feedback and Outcomes

Achieving positive outcomes with revenue cycle automation depends on staff trust in the technology and new processes. Partnering with a reputable IPA vendor will allow the management to build trust with the staff and get staff involved in the process. Accuracy is one of the key determinants of success and must be measured consistently since intelligent systems learn and improve over time. When staff and leadership agree that the implementation has been successful, then they can rely on IPA to address the next costly and inefficient revenue cycle function.

About Albert Porco

Albert Porco, Chief Solutions Architect at Cognitive Health Technologies

Albert Porco serves as Chief Solutions Architect at Cognitive Health Technologies. Albert has served as CIO for several New York metropolitan area hospitals and health systems. Prior to joining Cognitive Health Technologies, he also served as the Chief Technology Officer for the New York Department of Health. He can be reached at