HHS Data Gurus Are Saving Taxpayers Hundreds of Millions of Dollars

Last year, a team of data geeks, statisticians and modelers teamed with law enforcement agents after unearthing curious analytic patterns in Medicare data. They suggested a number of health care providers were submitting claims for services they weren’t providing.

With access to petabytes’ worth of data housed within the Centers for Medicare and Medicaid and newfound analytic tools to delve into it, officials from the Department of Health and Human Services were able to team with the FBI and Justice Department in a $1 billion fraud takedown—the largest in Medicare’s history.

The case resulted in criminal and civil charges for 301 people allegedly involved in fraud schemes, and upped the tally to $1.6 billion returned to the Treasury Department in just two fraud takedowns; a 2015 sting led to the return of $700 million in defrauded Medicare payments.

The secret, according to Caryl Brzymialkiewicz, chief data officer and assistant inspector general at HHS, stems from the department’s focus on data. Her growing office is the backbone of many of HHS efforts to combat Medicare fraud.

The 301 alleged fraudsters busted in 2016 “had figured out how to hide in the data,” said Brzymialkiewicz, speaking Tuesday at the Cloudera Federal Forum. The forum was produced by the events division of Government Executive Media Group, Nextgov's parent company.

“For us, it was about using data analytics and partnering with the DOJ and FBI to understand what was happening,” Brzymialkiewicz said.

Brzymialkiewicz’s team, which includes data modelers and statistics experts, poured over internal Medicare data and compared it with external intelligence derived from field agents. The team sought data that might indicate providers that receive Medicare payments for services they didn’t perform or for providers that might harm patients.

The complexity of the data and sheer size—Medicare spending accounted for 14 percent of the federal budget in in 2014—makes Medicare a common target for fraud.

Sometimes, Brzymialkiewicz said, fraudsters are brazen, as was the case with Dr. Farid Fata, a Detroit-area doctor who used cancer treatment drugs on hundreds of patients who didn’t need them. Fata lined his pockets with nearly $18 million he collected from Medicare or private insurance companies—money he forfeited back when he was caught and later sentenced to 45 years in prison.

The criminal acts allegedly perpetrated by the fraudsters in 2016’s Medicare fraud takedown were more orchestrated, requiring “robust analysis” and personnel “armed with the right tools and visualizations.” Data also has to be presented “in ways your customer understands.”

Lastly, the team had to be persistent, following up with law enforcement to ensure “folks take action” on data and “wrongdoers are held accountable.”

“For us, it’s about figuring out the part we have to play,” Brzymialkiewicz said. “Are we doing as much as we can?”