How Using AI Automation for Outstanding Claims Accelerated Cash Flow by $20.6M

Using AI-powered automation to follow up on the status of outstanding claims resulted in increased revenues and cost savings for one health system.

The revenue cycle staff at South Dakota–based health system Avera Health worked tirelessly to manually follow up on outstanding claims, says Mary Wickersham, vice president of central business office services for the organization.

But no matter how doggedly they worked, they "hardly made a dent in the ever-increasing backlog," says Wickersham.

And Wickersham couldn't keep hiring staff to solve the problem. "That gets expensive," she says.

That's why Avera Health is now using artificial intelligence (AI)-powered automation to follow up on the approved/denied status of outstanding claims. Doing so has resulted in several improvements, including $20.6 million in accelerated cash flow within 18 months of implementation.

At the HealthLeaders Revenue Cycle Exchange in March, AI and automation came up frequently as a topic of discussion among attendees, including one who described Mayo Clinic's recently launched pilot project to use AI-powered predictive models for denials.

"This is revenue that would otherwise have been delayed if received at all," Wickersham says. "By automating claim status follow up, we’ve been able to discover and resolve claims set for denial more quickly and accurately."

They've also seen a $1.1 million net impact in timely filing improvement and $260,000 in FTE annual savings, based on what it would have cost Avera to add additional staff to process the growth in claims instead of automating claim status follow up.

A vital process
 

The claims follow-up process is an important one, Wickersham says.

"The sooner you know a claim is going to be denied, the better chance you have at adjudicating that denial," she says. "You have fresher information to work with and there is less investigating and digging you have to do. On the other hand, the longer a claim goes unresolved, the more … likely that you won’t be able to resolve it."

Following up on claims also reveals important revenue cycle patterns that can help determine root causes of denials, uncover which insurance companies reimburse more slowly, and understand which claims are reimbursed more slowly.

Wickersham adds that following up on claims also helps patients.

"You certainly don’t want to create added stress for recovering patients by involving them in a convoluted claim issue. That is something that all healthcare financial leaders should heed," she says. "We may be working behind the scenes on fairly dry issues of billing and administration, but our roles—and how we serve in them—are meaningful to the patient’s health and well-being."

Keeping pace with growth and productivity
 

Wickersham says that Avera’s growth and employee productivity were the catalysts for switching to automated follow ups. The volume of claims was always increasing, and Wickersham knew she didn’t have enough people to do the work.

"My existing staff was already discouraged by the volume of claims that never seemed to abate," she says.

The manual process of following up on the status of outstanding claims involved contacting the insurer, inputting the claim status into different MEDITECH EHR workflows, and forwarding denied claims for specialized attention to others within the department.

The work was repetitive, slow, and frustrating.

Every day, staffers had to either call insurance companies or visit insurers' websites to pull up the status of claims. Each method had its own frustrations. For instance, Wickersham notes that most insurance companies will only discuss three claims per phone call.

"If you have more than three claims to work through, you literally have to hang up the phone and call back," she says.

Online follow ups were hardly better, since it required logging in, inputting the claim number, and transcribing the status and any supporting details into the workflows.

Plus, the staff members' specialized skills were going to waste.

"Essentially, you have even your most experienced staff spending time on the phone and logging into websites to chase after claims," she says.

Automating a manual process
 

Now Avera Health is using the ClaimStatusPlus robotic process automation tool from Recondo Technology, which integrated its rules engine into Avera's existing MEDITECH system and claims-related workflows.

After a claim goes to a commercial payer or Medicare, the tool queries the payer’s website, looks up the claim status, brings the answers back to the MEDITECH system, normalizes the data, and assigns it to specialized work queues.

Wickersham says the tool has cut in half the number of claims needing follow up. It also "routes the claims set for denial to specialized staff, according to reason for denial."

Metrics of success
 

In addition to the accelerated cash flow, costs savings, and increasing revenue, the automation process has resulted in several additional improvements:

• Average days in AR reduced by 7 days
• Average AR over 90 days reduced to 8%
• Timely filing write-offs as compared to net revenue down from 0.9% to 0.4%
• Number of accounts needing follow up per FTE down from 719 to 339

Plus, Wickersham notes another important improvement: staff morale.

"[C]laim status queues are shrinking, work goals are being met, staff is working on more challenging and interesting claims instead of going through repetitive hoops," she says.

Wickersham says her staff's experience trying to manually follow up on thousands of outstanding claims is only a small glimpse of the outmoded methods that have bogged down revenue cycles everywhere.

"[This is] just one health system; this same scenario is playing out in almost every hospital across the country," she says. "We have to break out of this expensive status quo that has taken root in the healthcare revenue cycle. Automation has proved to be a successful method for doing so."