Medicare has a notable issue with overpayments, especially through the Medicare Advantage program. The Centers for Medicare and Medicaid Services (CMS) currently uses a method called Hierarchical Condition Category (HCC) coding to determine payment adjustments. However, this approach can be inaccurate, leading to significant waste of taxpayer money—estimates suggest overpayments could reach $1.2 trillion over the next ten years.
To tackle this problem, a team of researchers created an AI model named Franklin, which aims to improve payment accuracy. Ezekiel Emanuel, one of the senior researchers, shares that Franklin is three times more accurate than HCC in predicting care costs. This accuracy stems from using the same data as HCC but applying advanced machine learning techniques to better assess patient risk.
Franklin’s design focuses on identifying cost patterns associated with various medical diagnoses, including heart disease and cancer. This approach helps target lower-cost groups, where many overpayments arise.
Experts argue that improving risk adjustment accuracy from models like Franklin is crucial to reduce incentivization for inaccurate coding and patient selection. Daniel Shenfeld, another researcher, emphasized that current systems often overlook the true complexities of patient care, allowing for financial gains through “upcoding” — a practice where healthcare providers list more severe diagnoses than necessary.
The need for effective risk adjustment is underscored by a recent simulation, which found that Franklin could potentially cut overpayments by up to $3.25 billion annually. This result could lead to better allocation of Medicare funds and ultimately more efficient healthcare delivery.
While Franklin offers a promising step forward, experts warn that it cannot entirely eliminate the potential for gaming the system. Continuous updates and monitoring will be necessary to ensure its effectiveness. Emanuel notes that improving the risk-adjustment process isn’t a partisan issue; there’s broad agreement on the necessity for effective reforms.
In conclusion, as Medicare continues to grapple with challenges related to overpayments and efficiency, innovative approaches like Franklin may offer the solutions needed to ensure fair and accurate healthcare reimbursement for millions of Americans.
Source link

