Collecting Earned Revenue: How AI Maximizes Radiology Payments

Today’s most pioneering radiology groups are using artificial intelligence to chase down payments, increase patient satisfaction, relieve front and back office workload, and recoup earned revenues.

“A large chunk of uncollected revenues is sitting with the payers,” explains Navaneeth Nair, chief product officer at Infinx Healthcare.

Artificial intelligence along with machine learning and automation empowers radiology groups and imaging centers to go and get those revenues.

An AI-driven revenue cycle management solution combines the impartiality of data and analytics with the insights of experienced revenue cycle managers and billing teams. It takes both to resolve substantial backlogs of claim denials and rejections. The American Medical Association (AMA) calls the combination of artificial intelligence, automation, and human insight, “augmented intelligence.

And as Nair explains, “The goal is to omit as many redundant and repetitive tasks as we can from cold-case revenue generation. This allows the provider’s staff to concentrate on knowledge-based and value-added tasks.” Together, technology and billing experts can help providers bring in earned revenue, improve patient satisfaction, and lend healthcare the transparency required to end the adversarial provider-payer relationship once and for all.  

Radiology big data, mined and understood

Starting about 2014, payers and other large business entities began leveraging AI-driven solutions to provide clarity, minimize costs and improve revenue. In 2017, Infinx Chair and Cofounder Jaideep Tandon realized that the company could better serve providers by expanding past automation to understandable AI.

Infinx led the push to develop artificial intelligence that can learn from a large amount of provider data about recovery experience. The data includes information on which claims and denials were reimbursed, which got reimbursed at the highest levels, and which were paid in a timely way. Based on this data, AI and machine learning learn patterns that predict which new claims and denials providers are most likely to recoup. The algorithm learns continuously from all new claim recoveries and recalibrates as the payers change rules or claim approval behavior. But AI has its limitations. In complex cases, specialists intervene to evaluate and resolve.

When Infinx took its revenue cycle management solution into the C-suite, “the CFOs got it right away,” Nair says. “Financial leaders in healthcare are all about bringing in every dollar that could and should be coming through the door. If you can promise them a 5 to 10% lift in revenues received, they recognize that as worth investing in.”

Directors and managers of RCM operations initially expressed a kind of intrigued skepticism. “They’d heard many such pitches before,” Nair explains. “They knew what it was like to get sold on revenue software that, in day-to-day use, turns out to be nothing more than glorified analytics of some kind.”

Infinx set out to prove AI-driven solution could return a positive ROI without the need for additional staff.

AI exceeding expectations

To prove that AI could solve AR backlogs, Infinx lined up one of its largest RCM customers as a design partner. This opportunity gave both sides real-world, real-time experience.   

This Los Angeles-based provider of outpatient imaging services has 330 sites and a staff of more than 8,500. Its robust data served well to challenge and prove Infinx’s solution. 

In six months, the two companies produced an alpha product and then beta-tested it for performance. Infinx also began internally beta-testing it with some of its full-cycle billing customers.

The results exceeded Nair’s expectations. Among the highlights, as documented in a recently published case study, the product’s first year with this radiology partner produced: 

●      28% increase in collections from denials and aged AR;

●      90% of denials addressed in less than five days, bringing in faster cash flows;

●      120-day aging AR cut by 60%; and

●      90+ days aging AR reduction by 20% in just 2 months.

The striking successes owed much to Infinx’s AI-powered technology which uses impartial, more complex criteria to determine the A/R most likely to win reimbursement.  Thanks to these early successes, AI has now taken root as a reliable tool to help recoup earned revenue from accounts 60, 90, and 120 days old.

Boosting patient satisfaction for the bottom line

An AI-driven solution reinforces revenue in more ways than A/R recovery, however. Patient satisfaction makes for loyal customers who return to the provider for additional services. It also underlies provider reputation, a factor credited for bringing in new patients. Because of the short- and long-term benefits, most providers today are laser-focused on optimizing patient satisfaction. 

While, of course, clinician skill and personality count for a great deal, patient satisfaction stems from more than clinician interaction. The whole benefits and billing aspect of the patient experience also arouse either appreciation or frustration. Most often, it’s the latter. Even patients with good coverage spend much more time with billing and admin than with their clinicians.  

Patients’ greatest source of frustration is the confusion over the lack of alignment of statements coming in from the payer and possibly multiple providers post-service. The solution for both provider and patient billing frustration is the “patient payment estimation.” AI-driven patient access technology unleashes deep data analysis, artificial intelligence, and machine learning to calculate an accurate estimate for treatment. When the front office shares this statement before treatment, patients feel more in control and more clear on their financial responsibility. 

Providers have been pleasantly surprised that patients actually appreciate the pre-service payment. One survey of 1,000 patients found that 90% actually want to see a pre-service price estimate.

The patient pay estimate also has benefits for the provider. Patients tend to pay what they owe upfront when presented with a clear estimate and payment options. Given the confusion and frustration medical statements cause, Nair encourages, “if the provider can help the patient have a better experience with the financial part, they’re going to remember a better overall patient experience.” 

Another source of patient frustration is the time it takes for providers to determine eligibility and win prior authorizations. It’s well known that the administrative burden during patient access is enormous, particularly when staff must execute all tasks manually. An AI-driven system speeds insurance discovery and helps to guide staff in inquiring about coverage, winning prior authorizations, and submitting claims. By taking on the manual tasks, an augmented intelligence solution frees staff to spend more time with patients and get answers from payers on the more complex cases faster.

Of course A/I needs a human assist

AI and automation alone sometimes aren’t enough to resolve the mountain of work practices face when it comes to prior authorizations and benefits verifications. Providers with a huge backlog of A/R can get ahead when they bring in a third party of billing experts to clear as much as possible. These experts help in-house billing specialists learn how to prioritize work and reduce errors as well — training proven to boost revenue long-term.

Complex cases escape the capabilities of AI and automation. Cases considered “exceptions” are best handled by human specialists. These specialists rely on their past experiences as well as discussions with payer representatives to resolve complicated A/R accounts. When only an experienced billing specialist can resolve a complex case, providers need to rely on their team or that of a third party. We refer to AI-driven revenue cycle management solutions as “augmented intelligence” because, while it does carry out many important tasks, it’s primary role is to assist the human providers and their billing specialists.  

Technology’s role in radiology

Ever since a prominent computer scientist declared, “We should stop training radiologists now; it is just completely obvious deep learning is going to do better than radiologists,” radiologists have been watching technology as it approaches in the rearview mirror.  

Radiology administration’s use of AI has nothing to do with the subtle skills of diagnosis on the clinical side, however. Revenue cycle management is in the domain of numbers, revenues, and dates. When incorporating AI for revenue cycle management, a group moves from sifting through endless spreadsheets and balancing stakeholder opinions to data-driven, intelligent insights. No “bots” are reviewing films.

As American healthcare continues to adopt more and more sophisticated technology, Nair — who has worked for Aetna and Anthem as well as on IBM’s Watson project — is confident that AI-powered revenue cycle management will improve its quality and cost-effectiveness. After working with entities from small provider organizations to those with billion-dollar revenues, he’s seen firsthand that all healthcare suffers from revenue write-offs. That just means that all can leverage automated intelligence to reinforce cash flow, relieve staff workload and increase patient satisfaction. 

Dave Pearson

Dave P. has worked in journalism, marketing and public relations for more than 30 years, frequently concentrating on hospitals, healthcare technology and Catholic communications. He has also specialized in fundraising communications, ghostwriting for CEOs of local, national and global charities, nonprofits and foundations.

Around the web

CMS has more than doubled the CCTA payment rate from $175 to $357.13. The move, expected to have a significant impact on the utilization of cardiac CT, received immediate praise from imaging specialists.

The all-in-one Omni Legend PET/CT scanner is now being manufactured in a new production facility in Waukesha, Wisconsin.

After reviewing years of data from its clinic, one institution discovered that issues with implant data integrity frequently put patients at risk.