Claims denial remains one of the most persistent and costly challenges facing health systems.according to Research Published last year by the Healthcare Financial Management Association, “of the $3 trillion in total claims filed by healthcare organizations, $262 billion was denied, with an average of nearly $5 million denied per provider.”
The outbreak of the Covid-19 pandemic has exacerbated the problem, with national denial rates rising 11% Between July 2019 and June 2020.
The fact that most rejected claims are preventable is just one aspect of a deeper problem. Reprocessing claims takes time, money and resources.
Health system claims quagmire often stems from several potential internal and external issues, including:
- Complex and ever-changing rules for undocumented adjudication across payers make effective claims management incredibly difficult.
- Errors and inefficiencies in one area, such as benefit eligibility, can cause problems downstream in the revenue cycle.
- Inefficiencies in handling the increasing number of prior authorizations can easily drive up a provider’s rejection rate.
- Relying on manual processes is expensive and laborious.
Solutions such as robotic process automation can handle simple, repetitive tasks well, but to truly optimize reimbursement, health systems should investigate the potential of artificial intelligence (AI).
AI provides the power of automation, but takes it to a whole new level by adding predictive capabilities, continuous learning and insights to proactively prevent claims from being rejected before submission, expediting and prioritizing manual rework, and leveraging advanced analytics from Discover actionable insights in the health system’s own data to optimize claims management across the revenue cycle.
level playing field
Most claims management techniques are designed to prevent service of claims that might be rejected by payers under well-documented payer rules.
The billing department will apply standard or custom edits to claims based on what the payer publicly states they will use to adjudicate the claim. After editing, send the clean claim to the clearinghouse or payer for adjudication.
Unfortunately, the rules that payers use to adjudicate claims are constantly changing, and often without notice. This has left suppliers playing a constant game of catching up.
AI has the potential to level the playing field by making highly accurate computational predictions of the likelihood of claims being denied, including denying undocumented payer changes. Thus, empowering providers to monitor capabilities in real-time and, if needed, take immediate action to correct a claim before submitting it to maximize compensation. The key to doing this successfully requires AI modeling based on the provider’s own claims and payer payments, rather than from a generic pool of mixed provider claims and payer reimbursements.
Focus on recoverable claims
The effort and resources devoted to appealing denied claims are often based on amount rather than likelihood of recovery.
It seems intuitive to focus on recovering a $100,000 claim rather than a $5,000 claim. But if the former has little chance of appeal, the claims department is just wasting time and money trying to collect potentially unrecoverable revenue.
This is where AI can really make a difference through modeling complex claims rejections and successful resubmissions/appeals. By searching the provider’s own system of record, the AI platform can analyze the history of related claims and similar claims to reveal insights such as past appeal success rates and reimbursement percentages recovered.
This level of granularity enables providers to score denials based on their likelihood of being reimbursed, an approach that yields more results than focusing on claims simply because they are premium.
End-to-end claims management
AI platforms can do more than just claims management. Properly scaled, AI can address many of the problems and inefficiencies that have significant downstream impacts on the revenue cycle. The two most obvious applications relate to patient eligibility and prior authorization.
Patient visits can be a major source of error, especially when it comes to eligibility. AI can bring intelligence to real-time qualification checks, providing a higher level of automation, accuracy and efficiency.
Prior authorization and medical necessity are also common sources of claims problems. As the number of prior authorizations increases, the need for improvements in the claims management process becomes even more important. The potential of AI in this area includes automating key aspects of the previous authentication process, including determining the need for pre-authorization, displaying and attaching relevant medical documents and images, and monitoring authorization status.
Another benefit of AI is the ability to apply learning to predictions. The intelligent platform can then flag current and future claims risk from rejection based on the insights it emerges from the sea of historical remittance data. In addition to flagging potential roadblocks, AI programs can use their analysis to identify root causes in the claims workflow.
Employee Influence
One of the downsides of AI technology is the widespread perception that the technology poses a threat to human job security.
These concerns are valid, but the intent of AI deployments can be misinterpreted. In today’s hospital billing environment, employees are increasingly burdened with doing more with less. Too thin employees can lead to wear and tear on basic efficiency.
In the case of claims management, AI is best viewed as human augmentation, handling repetitive tasks and performing complex analytics at scale, while employees focus on reprocessing claims with the greatest potential for reimbursement.
Provider organizations may never fully eliminate their denied claims. However, the power of AI can help suppliers recover more of what they owe, better predict at-risk claims, and identify claims that are worth the effort to appeal.
Photo: Michail-Petrov-96, Getty Images
