Healthcare organizations have a tremendous opportunity to harness artificial intelligence in an area that may not seem flashy but is important nonetheless: revenue cycle management.
Although 94 percent of companies firmly believe AI can give healthcare organizations a “massive competitive advantage,” according to Change Healthcare’s Adam Sullivan, PhD, less than 5 percent have extensively incorporated the technology.
Few organizations are incorporating AI into their operations because there are considerable barriers to entry, but the low utilization could also indicate that they simply don’t know where to start. During a June 12 webinar sponsored by Change Healthcare and hosted by Becker’s Hospital Review, four panelists explained how provider organizations are looking to take the next step and integrate AI in meaningful ways.
The panelists from Change Healthcare were:
- Adam Sullivan, senior director of artificial intelligence
- Brian Andrews, senior vice president of product and hospital revenue cycle management
- Vindali Vartak, senior director of advanced analytics
- Greg Arnold, senior vice president and general manager of physician services
Four takeaways from the presentation:
AI Is The Key To Reducing Costs And Eliminating Inefficiencies
While AI has contributed to significant advancements in imaging and other fascinating areas, the most meaningful use of AI right now isn’t necessarily the most groundbreaking, Dr. Sullivan said.
Health systems should understand that AI tools can help improve billing, collections and other administrative processes that are historically inefficient and very costly if not done properly, according to Dr. Sullivan. He said as much as one-third of the country’s healthcare spending is wasted on problems such as inefficient care delivery, fraud and preventable complications.
“If you have the data, and you have a way to effectively identify the pathway to make change, then some of these problems become easier to solve — unnecessary services, inefficient care delivery,” he said. “Whenever they’re framed in this way, we can state problems that we want to solve using AI and tackle these particular areas.”
AI Can Predict Denials With A High Degree of Accuracy And Precision
To ensure the greatest value for customers when it comes to denial prevention, Change Healthcare’s model carefully measures how accurately it can predict the reasons for claim denials. For example, while a modern breast cancer diagnosis needs to have an AUC (area under curve) score of at least 0.7 to be considered clinically valid, Change Healthcare’s model can predict medical necessity with an AUC of 0.95, indicating high accuracy, Dr. Sullivan said.
By predicting whether a claim will be denied and building that intelligence into workflow prior to claim submission — with tools like the ones Change Healthcare has developed —- organizations can significantly reduce denials and appeals costs on the back end, according to Mr. Andrews.
“Our customers now, with our claims workflow solution, can actually see that a claim is going to deny, what that likelihood is, and an indication of the reason,” he said, “so we can help point them toward action.”
AI Tools Can Prevent Losses And Missed Opportunities
Hospital systems and providers can lose up to 1 percent of revenue by missing charges, creating a perfect opportunity for meaningful AI implementation, according to Ms. Vartak.
Change Healthcare’s Charge Capture Advisor solution develops predictions about missing charges, which is beneficial whether the user is a provider, clinician, coder or revenue integrity nurse, Ms. Vartak said. The solution, which is based on a robust data set that enhances reliability, integrates predictions into the workflow as early as possible. This enables users to take proactive measures instead of devoting time and resources to fixing errors on the back end.
Change Healthcare’s charge capture tool demonstrates AI’s clear potential to reduce missed opportunities, in turn decreasing tedious, labor-intensive work and improving revenue streams, according to Ms. Vartak.
“The earlier in the process we use AI — in any type of AI application — the better our ability to actively respond rather than react to the problem you’re working on,” Ms. Vartak said. “The idea for charge capture is similar: The charge capture predictions are available to the user in the pre-build stage.”
AI Implementation Benefits All Parts of The Revenue Cycle
While AI has applications in many different areas of healthcare, it has dozens of use cases in revenue cycle management alone, according to Mr. Arnold.
On the front end of the revenue cycle, AI not only helps confirm patient coverage through preauthorization, but it can also improve payment collection at the time of service, Mr. Arnold said. Mid-cycle, its applications include ensuring codes are appropriate.
And, on the back end, organizations can harness AI to prevent denials and predict reimbursement — before a claim is even sent to the payer.
“Doing all of this right — and the use of these innovative tools in the front, middle and back end of the revenue cycle — achieves our goals of improving cash collections,” Mr. Arnold said.