Improvements on the Effectiveness and Scalibility of CMS Referral Analytics System

Referrals play an essential role in the healthcare system. It has been shown that 60%similar to 70% patients visiting hospitals are referrals. The Centers for Medicare and Medicaid (CMS) has released a large scale physician referral data set, which enables research on referral networks. Recently, we have developed a data-driven system to investigate referral networks based on the CMS open data. In this paper, we will introduce functionalities and discuss the system's effectiveness and scalability. Our system automatically identifies competitors for providers and find splitters with growth potential among physicians and physician groups. The system is designed to be user-friendly and scalable with advanced web development techniques and cluster computing framework.