How to reduce claim denials with AI and automation

Claim denials continue to be a costly problem for health systems of all sizes, and the claims denial rate has been steadily increasing since 2016. The pandemic only accelerated the trend, and by the third quarter of 2020, the claims denial rate upon initial submission was up to 11 percent. That means more than 1 in 10 claims submitted are denied. The vast majority of these denials are preventable or recoverable, meaning that hospitals are leaving valuable revenue on the table.

Why is this problem still growing? There are many complexities in the processes and rules that are making claims management increasingly difficult, and most health systems rely heavily on human resources instead of technology. But human workforces are stretched thin, and continuing to add employees doesn’t solve the underlying problems. Here’s the good news: Today, artificial intelligence and automation power proven solutions to reduce claim denials and provide a positive ROI. 

At each step in the hospital revenue cycle, there are opportunities to deploy artificial intelligence to avoid costly denials. AI can eliminate simple errors, increase staff time to work claims, and provide valuable insights into denials management that reduce revenue loss and improve cash management and cash flow for the hospital. Here’s how:

1. Ensuring accurate eligibility

From the very first patient interaction, errors and inefficiencies can occur that lead to downstream problems. Ensuring accurate eligibility by automating benefit verifications with AI has positive ripple effects down the entire revenue cycle, including reduced denials. In fact, 23.9% of claims are denied because of registration and eligibility issues. 

Many of these stem from changes in coverage between initial scheduling and the appointment or inaccurate benefit information being pulled into the EHR. Intelligent automation can augment your existing EDI Real-Time Eligibility checks by checking coverage more frequently and reducing errors in benefit pulls, driving down the number of denials caused by inaccurate or out-of-date benefit information.

2. Automated prior authorization

A large percentage of denials are due to prior authorization and medical necessity issues. And since the number of prior authorizations is increasing, the amount of denials is likely to increase as well — unless steps are taken toward significant process improvement. Currently, the prior authorization process is one of the most high-touch and labor-intensive steps in the hospital revenue cycle. 

An end-to-end AI-powered prior authorization solution reduces denials by automating multiple steps in the process, from determining if an authorization is required, to helping submit prior auth requests with data from the EHR, to continually checking prior authorization statuses. This reduces errors that lead to future denials. 

3. Claims status checks

Claims status checks are a necessary step in healthcare claims management, but not one that is always a value-add. Often, a claim status is checked and no additional action needs to be taken. But checking a claim status manually takes an average of 14 minutes, and when you consider the number of claims and the frequency at which they would optimally be checked, it becomes apparent that this is impossible for humans to perform. As a simple, repetitive task, it is a perfect candidate for automation. But how does this reduce denials? 

The majority of revenue cycle departments are chronically behind in claims, without time to rework denials that could be recovered and without time to improve other processes in the revenue cycle that would prevent denials in the first place. Currently, 48% of claim rejections and denials go unappealed or unworked. So, in itself, automating claims status checks may not reduce denials, but it saves such a significant amount of staff time that can be redirected to other steps in the revenue cycle process, including reworking denials. 

4. AI-powered denials management 

Once a claim has been denied, artificial intelligence can help health systems recover more of these claims. Automated claims status checks as described above are actually the first step in an AI-powered denials management system. When the claim is flagged as “denied,” artificial intelligence can automatically address any simple errors and then resubmit the claim. For more complex errors, AI can pass the denial to a human, but will provide detailed patient information and denial information, greatly speeding the rework time. An AI-powered denials management system means that denials will actually be worked and all available reimbursements will be collected, increasing revenue and reducing days in A/R.

5. Deep learning insights into denials 

Artificial intelligence, unlike simple RPA solutions, also has the power to analyze the data behind the claims processes, uncovering actionable insights. For example, at one health system, it was found that one specific drug denial was due to missing prior authorizations and medical necessity. Knowing this information, the hospital was able to target and solve this recurring issue. Opportunities for process improvement are hidden in the data your organization works with every day — you just need AI to find them.

Claim denials are costing you more than you think — and there is a proven solution available today. 

Claim denials are costly, from revenue leakage and write-offs to the vast administrative burden. Artificial intelligence and automation can reduce these denials and optimize your claims management at multiple points, from the moment a patient is scheduled to providing intelligence that proactively offers opportunities for improvement. 

Learn more about our AI and automation solution, Olive Works, for claim status checks and denials actioning.  


  1.  Change Healthcare 2020 Denials Index
  2.  Change Healthcare Healthy Hospital Revenue Cycle Index
  3.  2018 CAQH Index
  4.  HIMSS – “How to Improve Your Clean Claims Rates”

Investing for Enterprise Impact Whitepaper

A Healthcare Executive's Guide to AI Workforce Models

Read this white paper to understand the total cost of ownership of various purchasing models for AI and automation, so you can chart the right course as you invest.

Get the White Paper

Related Articles

Learn more