The U.S. Postal Service found more than it expected when it began using big data and analytics to detect mail theft.
The postal service’s office of inspector general used analytics to set up trip wires, which set parameters around data to help detect anomalous activity. That allowed postal inspectors to determine that a carrier was disposing of mail rather than delivering the mail.
When an agent arrived to serve the carrier with a warrant, the agent discovered the carrier also was cultivating marijuana in his home.
“That was a fairly interesting case,” Michael Mashburn, director of the data analytics lab at the U.S. Postal Service’s Office of the Inspector General, said at MeriTalk’s Second Annual Big Data Brainstorm on Thursday.
Agencies and other organizations are relying on analytics more and more to detect and stop fraud, but agencies should start with small projects and small data sets to pinpoint risk.
“Working in increments is the way to go,” said Alex Habershon, program coordinator at the World Bank. “Pick your sectors based on where you know the risks to be and use the data analytics to help focus and target your investigations.”
Medicare and Medicaid remain heavily targeted by fraudsters, said Tony Trenkle, chief health information officer for U.S. Federal software at IBM.
Two health care programs alone accounted for nearly $80 billion in improper payments identified last year:
• Improper payments by Medicare amounted to $60 billion in 2014, according to a 2014 report from the Government Accountability Office (GAO).
• Improper payments by Medicaid reached $17.5 billion in 2014, according to the same report.
The Health Care Fraud and Abuse Control (HCFAC) Program, a joint effort of the Department of Health and Human Service’s Office of Inspector General and the Department of Justice, has recouped $14.8 billion over the past three years by relying on analytics.
Analytics are helping the agencies detect anomalies in data, Trenkle said, and helping agencies move from a reactive mode to a predictive mode.
An anti-fraud system using predictive analytics at the Department of Health and Human Services stopped $133.2 million in Medicare waste, fraud, and abuse in 2014. The program recouped $2.84 for every $1 invested.