In light of the most recent final rule and the increased risk for non-compliance penalties, what can payers and providers do to ensure transparency in their risk adjustment?
The increasing scrutiny by the Centers for Medicare & Medicaid Services (CMS) around risk adjustment coding in Medicare Advantage (MA) plans has been garnering headlines for over a year. The strong signals from CMS to MA plans is “get it right, or get penalized.”
In light of CMS’ most recent final rule and the increased risk for non-compliance penalties, what can payers and providers do to ensure transparency in their risk adjustment?
Medicare is big business
Medicare is big business for everyone. In 2020, the program spent an estimated $861.9 billion providing health care services for approximately 63 million beneficiaries—a figure that translates to approximately 13 percent of federal spending. Indeed, the U.S. Government Accountability Office (GAO) first designated Medicare as a “high risk” program in 1990.
With lower monthly premiums and additional benefits, MA plans have quickly emerged as a popular choice for seniors. Almost half of all Medicare beneficiaries were enrolled in MA plans in 2022, according to the Kaiser Family Foundation. Today, some 3,998 Medicare Advantage plans are available, representing a 6 percent jump over the number available in 2022. By 2032, 61 percent of all Medicare beneficiaries will be enrolled in MA, according to the Congressional Budget Office (CBO).
Given the scope of the expenditure and the program’s projected growth trajectory, it’s no wonder CMS is keeping a prudent eye on expenses.
Margin for error
As opposed to traditional Medicare fee-for-service plans, Medicare Advantage Organizations (MAOs) receive payment for individual beneficiaries based on a risk score, which CMS calculates using its Hierarchical Condition Category (HCC) model. The scores are used to adjust payments to health plans based on the estimated annual cost of care for members. Health plans receive higher monthly rates for members with more chronic comorbid conditions, and less for the healthier members. The difference between accurately capturing a member’s comorbidities versus under-presenting their conditions can be thousands of dollars per member per month in reimbursement from CMS. This threat of underpayment leads to payers expending large efforts to ensure each diagnosis is captured to its maximum specificity.
Upcoding under scrutiny
Within this model, there is of course the risk of “upcoding,” or the systematic exaggeration of members’ health problems to secure higher payments.
How pervasive is the problem? In late January, Kaiser Health News (KHN) released details of 90 government audits of medical records, which revealed millions of dollars in overpayments to MA plans. The audits, which cover billings from 2011 through 2013, are the most recent financial reviews available, despite the dramatic increase in enrollment in MA plans.
CMS currently imposes a 5.9 percent reduction (the coding pattern adjustment) to MA plan payments to counter the effect of different coding intensity across plans. But even with the reduction, excess payments to MA plans came to nearly $12 billion in 2020, the Medicare Payment Advisory Commission (MedPAC) reported to Congress last year.
There are specific guidelines around which diagnoses can be submitted to CMS under MA. Each submitted diagnosis is required to have clinical documentation evidencing that the given condition has been actively Monitored, Evaluated, Assessed and Treated (MEAT). In other words, a diagnosis in a problem list alone is not enough evidence that the patient is currently under management for that condition–and claiming for it would be inappropriate.
CMS RADV Final Rule ups the ante
On February 1, CMS published details regarding the MA Risk Adjustment Data Validation (RADV) program that it uses to recover improper payments made to MA plans. And although the final rule includes a decision to not issue penalties on audits prior to 2018, it excludes the fee-for-service adjuster, which allowed MAOs a limited number of payment errors.
The heightened oversight of MAOs is expected to net the administration as much as $4.7 billion over the next ten years. Meanwhile, there is speculation that the elimination of the fee-for-service adjuster could result in litigation on the part of payers.
To protect themselves, how can payers and providers ensure transparency in their risk adjustment?
Getting risk adjustment right the first time
Accurate and audit-proof risk adjustment requires transparency and ready access to a verifiable audit trail. With as much as 80 percent of clinical data in unstructured or semi-structured form within electronic health records (EHRs), organizations are often challenged to access and validate clinical encounters.
Fortunately, there are technologies that can help streamline risk adjustment coding and workflows and provide a paper trail between clinical documentation and submitted claims.
Artificial Intelligence (AI), and specifically Natural Language Processing (NLP) are now being used to actively help payers and providers increase transparency and reduce the administrative burden of risk adjustment. NLP can help improve coding accuracy by surfacing key clinical information from member medical records to identify risk adjustable diagnoses, and the required supporting documentation as per the MEAT framework. The technology can also identify potential documentation improvement opportunities, as well as flag where there is no supporting evidence for certain conditions. Health care specific NLP understands clinical context such as negation, family history, synonyms, and abbreviations.
Consider, for example, one payer that used NLP to improve the effectiveness of chart reviews and the accuracy of HCC code capture. The NLP tool processed documents between 45 and 100 pages long per patient and identified features for HCC codes with over 90 percent accuracy. The technology has enabled the organization to process millions of documents per hour, a significant improvement over manual chart review.
The number of payers and providers challenged with accurately identifying and documenting risk-adjustable conditions is growing exponentially. With escalating regulatory changes and potential penalties, stakeholders should strongly consider technologies such as NLP to help them readily maintain compliance and promote ongoing fiscal fitness.
To learn more, visit the IQVIA team at RISE National or visit https://www.iqvia.com/riskadjustment.
About the author
Dr. Calum Yacoubian is the director for Healthcare Strategy at IQVIA NLP. He has 10 years’ experience in the health care industry, starting off as a clinician and moving into medical technology in 2015.
His passion is the intersection of innovative technologies and healthcare, particularly where Artificial Intelligence can augment human expertise across a broad range of use cases, from risk adjustment to precision medicine.