Last week, the Centers for Medicare & Medicaid Services (CMS) reported $1.5 billion in savings due to “big data” initiatives with the Fraud Prevention System (FPS), which started in June 2011. Over the past five years, CMS used data and predictive analytics to quickly identify and take action on cases of fraud, waste, and abuse in the Medicare program. Working closely with public and private predictive analytics experts, data scientists, and law enforcement, the FPS “has had a profound impact on fraudulent providers and illegitimate payments,” according to CMS.

Nationwide, CMS’ predictive analytics resulted in over $1 billion in savings during 2014 and 2015 by working to prevent illegitimate payments rather than make payments and follow-up on potential fraud afterwards. CMS says it analyzes 4.5 million Medicare claims every day, identifying considerable savings on claims before payment is initiated.

To illustrate on-the-ground results, CMS provides a few practical case examples:[x_blockquote cite=”” type=”left”]The FPS identified a home health agency in Florida that billed for services that were never rendered. Due to the FPS, CMS placed the home health agency on prepayment review and payment suspension, referred the agency to law enforcement, and ultimately revoked the agency’s Medicare enrollment. [/x_blockquote][x_blockquote cite=”” type=”left”]In Texas, FPS identified an ambulance company submitting claims for non-covered services and services that were not rendered. Medicare revoked the ambulance company’s Medicare enrollment. [/x_blockquote][x_blockquote cite=”” type=”left”]Likewise, in Arizona, FPS identified that a medical clinic had questionable billing practices, such as billing excessive units of services per beneficiary per visit. A review of medical records showed that physicians were delivering repeated and unnecessary neuropathy treatments to beneficiaries. The medical clinic’s Medicare enrollment was subsequently revoked. [/x_blockquote]

Learn more about the Fraud Prevention System on the CMS blog.