According to estimates from the federal government and issues-based groups such as the National Health Care Anti-Fraud Association (NHCAA), as much as 10 percent of all healthcare expenditures in the United States, or $100 billion dollars, may be lost each year to fraud, waste and abuse. Prevention and recovery of only a fraction of this $100 billion represents both a significant ROI opportunity and a competitive advantage.

Since health care payments are among the largest government payments to citizens, they are also often the domain for fraud. While Medicare fraud prevention is to some degree a focus for the federal government focus while prevention and reduction of Medicaid fraud is a much greater focus at the state level. There are increased penalties for Medicaid fraud and organizations receiving substantial Medicaid payments needs to describe their policies for preventing fraud particularly at the state level, where Medicaid programs are administered. Many states have fraud prevention initiatives already under way, and advanced analytics is a key tool for identifying payments that may be fraudulent.

As a prominent example, New York State is the largest provider in the U.S. of Medicaid services, at $44 billion per year, and has a strong focus on analytics for fraud prevention. The New York State Office of the State Comptroller identified more than $150 million in Medicaid claim overpayments in 2005 and 2006 after analyzing historical claims data in the eMedNY data warehouse. These analyses identified duplicate payments, overpayments to health care providers, non-billing to Medicare, and miscoding of diseases and payments. Because Medicaid payments in New York State are distributed through county governments, particular counties also have analytical fraud prevention initiatives under way. County officials believe that they have saved millions of dollars in savings since 2003, when they began to use the software tools to analyze Medicaid claims.

InferX Fraud System supports the various aspects of fraud investigation and management, including identification, prevention, decision aid in investigation, and decision aid in detection. Using a unique combination of data mining capabilities, visualization techniques and reporting tools, the system can identify potentially fraudulent and abusive behavior before a claim is paid, or retrospectively analyze providers' past behaviors to flag suspicious patterns within distributed data and then rank providers as to their degree of potentially abusive/questionable behavior.