Weighted risk models for dynamic healthcare fraud detection

摘要

Despite efforts to prevent it, fraud in the United States healthcare system remains a serious and pressing issue. Since healthcare fraud is a complex and multi-faceted problem, fraud-fighting solutions must be flexible enough to address the ever-evolving nature of the crime. Here, we present a method to identify healthcare fraud in such a manner that incorporates both potential fraud as well as risky provider behavior. The proposed weighted risk model provides a framework for creating a dynamic fraud detection database that can be easily scaled up to incorporate emerging fraud schemes.