Researchers in Brown University’s School of Public Health have launched an online tool that gives policymakers and the public an unprecedented look at how Medicare Advantage plans are contributing to billions in excess payments.
Called MediCode, the tool was created using data from the Centers for Medicare and Medicaid Services to analyze a tactic known as “upcoding.” In Medicare Advantage, the private alternative to traditional Medicare, upcoding happens when insurers or their contractors make patients appear sicker on paper to collect higher payments from the government.
For example, inflated coding happens when a doctor includes a diagnosis that isn’t actively being treated. A patient with a history of prostate cancer, who is in remission, could be coded as having active prostate cancer, even though they haven't needed treatment in years. This adds a high-risk diagnosis that boosts payment, even though it doesn’t reflect the patient’s current health condition or spending.
Another example is an insurer making a common condition sound more serious. A patient with uncomplicated diabetes might be labeled as having diabetes with chronic kidney problems based on a minor lab finding that wasn’t clinically significant or noted by the treating physician. This moves the patient into a higher-paying risk group.
It’s a quiet, technical process with billion-dollar consequences, said David Meyers, an associate professor of health services, policy and practice in Brown’s Center for Advancing Health Policy through Research (CAHPR) and lead researcher on the MediCode project.
“These are taxpayer dollars that are going into the pockets of insurance companies that aren't necessarily being translated into better benefits or care,” Meyers said. “It's money that could be used to support other programs like Medicaid or traditional Medicare.”
The Brown researchers who created the Medicode tool hope the data will help spark and inform discussions on issues in the Medicare Advantage program at a time when bipartisan scrutiny of the Medicare Advantage program is mounting in Washington. The rapidly growing program now costs taxpayers over $450 billion annually and is projected to receive $30 billion in additional payments due to coding differentials in 2025 alone.
Researchers and government watchdogs have long raised concerns about the billions of taxpayer dollars that are being spent because of coding practices. Every patient is coded, but the accuracy of those codes can vary. Currently, the Centers for Medicare and Medicaid Services applies a 5.9% cut to all Medicare Advantage plan payments to even things out and account for potential inflated codes. The measure, however, assumes that all insurers code the same way and follow the same coding practices, which is not the case, CAHPR researchers have found. Some companies code more aggressively than others, potentially listing more serious or extra medical conditions.
“The main finding we see is that there are just a couple of companies accounting for much higher spending—billions of dollars in additional payments due to how they code—while many other companies are not getting paid that much at all,” Meyers said.
UnitedHealth Group alone receives an estimated $12 billion annually in additional payments due to higher coding intensity, according to the data. Humana, which is now planning to revise how it adds diagnoses that boost payments in response to scrutiny over aggressive billing practices, follows with $6 billion. Smaller insurers like Anthem, Cigna and SCAN code even more aggressively, but their impact is smaller due to fewer enrollees.
Along with showing the average annual differential payments to the 10 largest Medicare Advantage parent companies, MediCode also breaks down coding differences at the state level. The hope is to give local stakeholders an understanding of what upcoding costs the federal government in their area. For example, health plans in Rhode Island may be overpaid approximately $53 million annually due to coding differences. The tool also includes an interactive section on potential policy solutions, including estimates on the cost of several alternative strategies that could be used to address inflated coding practices.
Medicode was designed with lawmakers in mind, but is free and open to the public. While the work was conducted independently by researchers at CAHPR, the project was funded by Arnold Ventures, a philanthropic organization focused on evidence-based policy.
The team at CAHPR plans to update MediCode regularly as new data becomes available. The estimates on the site are based on methods and peer-reviewed methodology and do not determine whether any coding is fraudulent or inaccurate. They instead aim to capture when plans code higher than their enrollment would predict. The documentation and data are open-source and available online.
“Our goal is to ultimately help policymakers and other stakeholders think about solutions to address these payment issues,” Meyers said. “By making it transparent how different companies engage in coding behavior, and the tools they use to capture codes, we may help to develop solutions that are more comprehensive than the current one-size-fits-all adjustments that CMS uses for these issues.”