Finding the Fraud 'Needles' in the Health-Claims Haystack

Fear Not the Software 

Health-care fraud detection historically has focused on finding the criminals. However, fraud prevention is best done by addressing the underlying waste and abuse that pervades America's health-benefits market, disguises fraud, and allows it to flourish.

Health-care claims are like a giant haystack riddled with needles of waste and abuse. However, some of those needles are bent -- they're fraudulent. Data mining acts like a giant magnet that pulls the needles out of the haystack so the fraud examiner can sort them, find the fraud, and reduce overall waste and abuse.

Of course, advances in computer technology and data-mining software have revolutionized fraud detection. Any interested health-plan fiduciary, manager, administrator, or internal auditor (as well as a fraud examiner) can now detect health-care fraud. Data mining narrows the field of suspects to a manageable number by categorizing them by standard criteria. This analysis (complemented by human judgment) detects potential fraud that would have gone unnoticed. Fraud examiners can then begin their examinations.

Health-care fraud takes many forms. Here we'll give some examples that illustrate how the tools of computerized data mining can be used to ferret out fraudulent behavior hiding in the midst of waste and abuse.