Medicare Advantage Plan’s Use of AI Drives Rising Denials

Recent media reports have highlighted commercial insurers’ increasing use of algorithms to assess claims, a trend that seems to be leading to a surge in denied claims, especially within Medicare Advantage plans. There are plenty of stories on the negative effects on seniors as a result of these algorithms predicting the type and length of treatments patients should require. The Center for Medicare Advocacy published an article in April 2022, “When Artificial Intelligence in Medicare Advantage Impedes Access to Care: A Case Study,” highlighting many core concerns.

Commercial payors denying claims and giving patients the runaround to get needed therapies covered is nothing new. What’s new is the expanded use of predictive analytics to evaluate and make determinations on claims more quickly. Critics point out that many of the decisions appear to be more restrictive than Medicare coverage guidelines, leading to premature terminations of coverage for beneficiaries. In addition, while algorithm developers, such as NaviHealth (now owned by UnitedHealth Group), state that these algorithms are designed to “inform” decisions on coverage, it appears that in some cases payors are using the algorithms to make claim denial decisions outright.

Higher Denial Rates with Medicare Advantage Plans

According to a report published by the US Department of Health and Human Services Office of Inspector General issued in April 2022, federal investigators found that 13% of prior authorization requests and 18% of payment denials by private Medicare Advantage plans were wrongly denied and should have been approved under Medicare coverage rules. Patient advocates report seeing many instances where it appears that the private Medicare Advantage payor is taking the algorithm’s recommendations at face value, perhaps even in a fully automated process, and failing to adjust for a patient’s individual circumstances and against the basic rules on what Medicare plans must cover.

XiFin customer data shows that Medicare Advantage denial rates consistently exceed Medicare denial rates. According to XiFin’s analysis of 2021 claims, Medicare Advantage denials were 43% higher than Medicare denials.

Some XiFin customers, for example, report issues with Genetic Testing for Oncology Local Coverage Determinations (L39365) that provide limited coverage for tests based on specific clinical indications. While it is our understanding that Medicare Advantage plans should follow Medicare coverage—regardless of whether the provider is contracted or non-contracted—as long as the patient has met the clinical indications and LCD criteria, our customers indicate that even when patients meet the clinical guideline criteria based on the published LCD, they are receiving denials for the tests performed.

Hospitals and Regulators See Problems with Medicare Advantage Plans

According to Becker’s Healthcare, hospitals have been dropping Medicare Advantage plans due to higher claim and prior authorization denial rates since 2018, but it was fairly uncommon until recently. For example, Mayo Clinic warned Medicare-eligible patients in Florida and Arizona in October 2022 that it no longer accepts most Medicare Advantage plans.

Regulators also see problems with Medicare Advantage plans. In April 2023, CMS issued new regulations that state Medicare Advantage plans “must ensure that they are making medical necessity determinations based on the circumstances of the specific individual … as opposed to using an algorithm or software that doesn’t account for an individual’s circumstances.” In May 2023, senators warned the country’s largest Medicare Advantage insurers at a hearing that they must abide by Medicare’s coverage rules and cannot rely on algorithms to deny care that patients need.

Utilizing AI to Gain Operational Efficiency

Applying artificial intelligence (AI) to gain process efficiencies is not in and of itself bad. At XiFin, we use AI in a number of ways to streamline and automate the revenue cycle management (RCM) process and improve the patient experience. But it’s crucial to make sure that AI is being applied to the right processes and that the integrity of policy coverage guidelines is adhered to. As an industry, we need to embrace new technology that supports more efficient business processes, without letting it drive us off the cliff.

XiFin applies AI to identify issues, including changes in payor behavior, based on payor activity. Payors adapt and find new ways to increase their earnings by changing payor rules. Sometimes, payor policy updates are not announced. XiFin leverages AI to detect these changes early, identify the pattern created by the unpublished policy change, and find ways to streamline the reimbursement process and maximize payments. Trying to keep up with these changes manually is no longer a viable solution.

A Data-Driven Appeals Strategy

A data-driven appeals strategy combined with automated appeals processes and high visibility is essential. The value of this approach will be increasingly impactful as these denial trends continue, and put pressure on your bottom line.

Front-end edits and configurations can also help mitigate back-end denials. Capturing potential denial-related issues proactively is the most effective way to maintain a manageable AR and improve the propensity to pay. That said, some level of denials is unavoidable and not all known issues can be addressed at the front end of the process. This is where a robust and automated appeals process becomes critical. This is especially true for novel ancillary services, radiology, patient monitoring, and durable medical equipment (DME), where appeals are becoming increasingly critical to revenue generation. In molecular testing, for example, an analysis of XiFin customer data showed that in 2020, appeals accounted for 5% of total revenue generated by XiFin customers; in 2021, appeals-generated revenue had increased 30%, to 6.5% of total revenue.

XiFin provides visibility into the data that helps our customers understand their organizations’ exposure, denial trends, and appeals success by payor, which informs their revenue cycle management strategy. XiFin helps organizations understand the cost and time impact of multiple appeals, and helps organizations prioritize their appeal efforts based on the likelihood that the appeal will be successful.

XiFin customers have first-hand experience with payors incorrectly applying commercial edits to Medicare Advantage plans. One private Medicare Advantage payor, for example, routinely denied claims despite documented coverage policies. After working with this private Medicare Advantage payor for more than two and a half years, including public advocacy efforts, defending coverage, and getting legal representation involved, the payor agreed that commercial edits were being applied to Medicare Advantage plans in error, and confirmed that claims were erroneously denied.

When it appears that a private Medicare Advantage payor is not adhering to the required Medicare coverage policies, XiFin’s payor relations team supports our customers in addressing these irregularities. We work collaboratively with the payors and our customers to find solutions. That includes taking a strategic approach to appeals automation and tracking what works. We understand how to maximize success with both procedural-related appeals and clinical/medical-related appeals. We work to ensure that all payors, including private Medicare Advantage payors, meet coverage policies.

For more information on how XiFin is utilizing AI watch the free on-demand webinar, Unlocking RCM Potential: How AI Is Shaping Revenue Cycle Management Workflow.