In 2010, the FBI organized an undercover sting of a Brooklyn medical clinic that was suspected of Medicare fraud. Agents installed a hidden camera in an air conditioning vent and watched employees pay kickbacks to patients in exchange for Medicare identification numbers, which they used to bill Medicare $50 million in fraudulent claims.

Agents eventually arrested 16 people in connection with the scheme and used the video evidence, along with audio and video from wired elderly clients, in their prosecution. However, it was data analytics that led them to the Brooklyn clinic in the first place, according to The Times.

Data analytics has helped investigators build cases and uncover fraud faster and easier, particularly in areas such as Detroit and Miami that have been hotspots for fraud schemes. In some cases, data mining has helped stop fraud even before criminal charges come to light.

“The idea of using real-time data to generate fraud cases is unique,” Leslie Caldwell, chief of the Department of Justice (DOJ) criminal division, told the newspaper.  “We have the ability to suspend–[when] there’s reasonable suspicion–[those] who are engaged in fraud even before they are prosecuted and indicted.”

The article points to the recruitment of Kirk Ogrosky, who spent time as a federal prosecutor in Miami. In 2006, the DOJ asked him to head the healthcare unit. Ogrosky accepted on the condition that the agency would “rethink the way they prosecute healthcare fraud, with an emphasis on real-time prosecutions.” Ogrosky began by searching for postal codes in which patient spending was three or four times the national average, and then employing old-school detective tactics to further the investigation.

“Most times, those zip codes would help generate a list of providers that had what I would call ‘medically impossible’ claims,” he told The Times. “[It was] like peeling an onion ring by ring–and yes, it always burnt my eyes at some point.”

Data analytics have since been used to uncover schemes related to chemotherapy drugs, home healthcare, and durable medical equipment. In Indiana, data-driven investigations have saved the state $85 million. FierceHealthPayer: AntiFraud previously reported on predictive models and algorithms such as the government’s Fraud Prevention System (FPS), which has led to more than $50 million in actual and projected savings in two years.

For more:
– read The Financial Times article