Applying Business Intelligence Concepts to Medicaid Claim Fraud Detection
Leanndra Copeland, Dana T. Edberg, Anna K. Panorska, Jeanne Wendel · 2011
U.S. governmental agencies are striving to do more with less. Controlling the costs of delivering healthcare services such as Medicaid is especially critical at a time of increasing program enrollment and decreasing state budgets. Fraud is estimated to steal up to ten percent of the taxpayer dollars used to fund governmentally supported healthcare, making it critical for government authorities to find cost effective methods to detect fraudulent transactions. This paper explores the use of a business intelligence system relying on statistical methods to detect fraud in one state’s existing Medicaid claim payment data. This study shows that Medicaid claim transactions that have been collected for payment purposes can be reformatted and analyzed to detect fraud and provide input for decision makers charged with making the best use of available funding. The results illustrate the efficacy of using unsupervised statistical methods to detect fraud in healthcare-related data.