Pharmacy benefit manager slashes fraud, waste and abuse using artificial intelligence

Prime Therapeutics saves its clients $355 million in 18 months with AI-powered SAS^® Detection and Investigation for Health Care

Fraud, waste and abuse (FWA) ravages the US health care system. Estimates from the National Health Care Anti-Fraud Association show fraud costs health care organizations $70 billion to $230 billion each year. The precise figure is unknowable because only 10% of such fraud is ever detected.

Of course, health care organizations work tirelessly to thwart FWA. Pharmacy benefit managers (PBMs), which manage prescription drug benefits on behalf of health insurers by negotiating with drug manufacturers and pharmacies, are particularly adept at detecting FWA in drug dispensing, but typically lack visibility on the medical benefit side.

“I could tell you all the drugs a patient received from a hospital, a retail pharmacy or a mail order pharmacy. But I couldn’t show you the diagnosis attached to that claim,” explains Jo-Ellen Abou Nader, Vice President of FWA and Supply Chain Optimization at Prime Therapeutics (Prime).

This inability to connect and analyze all relevant data sources inhibited Prime from stopping fraud committed by prescribers and members – a bountiful source of FWA in the health care industry.

A holistic FWA solution

Based in Minnesota, Prime serves 30 million members across the US through its relationships with health plans, employers and government programs including Medicare and Medicaid. The PBM serves 23 Blue Cross plans, which typically conduct their own FWA investigations on pharmacy claims paid under the medical benefit.

But this siloed approach causes too many fraudulent claims to fall through the cracks, according to Abou Nader. She saw an opportunity to save her clients’ money – and lives – by looking more holistically at FWA.

Prime invested in SAS Detection and Investigation for Health Care in the cloud. This advanced analytics platform allows Prime to consolidate data from pharmacies, medical drug claims and medical services. By applying the platform’s artificial intelligence (AI) and machine learning capabilities to this comprehensive data set, Prime can now detect and prevent fraud regardless of the source. Because Prime opted to deploy the solution in the cloud, it also gained numerous benefits, like expedited access to the software, the ability to easily share results with others and savings on infrastructure expense.

“Before SAS, we had data in multiple databases and data stores,” says Korisa Matezevich, Senior Director of Supply Chain Shared Services at Prime. “We would have to take disparate tools and create temporary tables to blend that information and make it consumable for our analysts to query and model for what we wanted to identify. With our SAS fraud solution, we’ve removed a lot of those intermediary steps by sending SAS that information and storing it, consolidating it and cleansing it so that it’s ready and prepared for modeling. Now we have confidence in the outputs to triage the most valuable cases of fraud to our special investigations team.”

SAS Detection and Investigation for Health Care automatically scans through thousands of claims in real time, searching for patterns and flagging outliers, helping Prime analyze billions of claims annually.

For example, Prime can identify members who use deception to get prescriptions, whether it be through insurance fraud, collusion with a pharmacy or prescriber, or identity theft. The platform can also find prescribers who falsely diagnose patients, receive kickbacks for certain medications or submit duplicate claims.

Visual link analysis helps prosecutors

Prime uses a capability in the software called visual link analysis to show connections between providers, members and pharmacies. The tool visually reveals the shifting of claims or hiding of money and can show multiple pharmacy ownerships and transfers of ownership.

This makes it easier for Prime investigators to share evidence with law enforcement. “Once fraud is detected, we need to be able to show a visual snapshot of exactly what happened,” Abou Nader says. “Visual link analysis makes it easier to tell that story.”

She adds that visual link analysis was a key factor in selecting SAS. “Our teams used to spend months recreating link analysis manually … SAS lets us do it on the fly.”

Curbing the opioid epidemic

SAS Detection and Investigation for Health Care enables Prime to harness the power of embedded AI and machine learning to continuously enhance fraud detection models. By staying ahead of new fraud schemes, Prime has the power to save lives.

“Patient safety is a key factor for us in the fraud, waste and abuse area,” Abou Nader says. “We’re always searching the data for things that could potentially harm patients.”

Opioid abuse is a prime example. In 2018, nearly 50,000 people in the US died from opioid overdoses, exacerbating the public health crisis. Using SAS, Prime pioneered a new predictive modeling process that uses more than 170 potential predictors to identify patients most at risk to receive high-dose opioids – a factor the US Centers for Disease Control and Prevention has linked to increased harm.

With SAS, Prime can more easily identify and stop opioid FWA, such as when pharmacies or patients create or alter prescription orders, or when prescribers overprescribe drugs beyond guidelines. It can also reveal “doctor shopping,” where patients get more medicines than appropriate by seeing multiple doctors.

In one case, Prime discovered a member went to 10 prescribers and 10 pharmacies over 12 months, collecting 48 separate prescriptions for a total of 1,870 narcotic tablets in one year. Spotting this sort of illicit – and potentially harmful – activity had previously been difficult. “Finding these disturbing patterns and connections among millions of claims had been a manual process in the past,” says Anne Mack, Senior Director of Pharmacy Audit and FWA at Prime. “Now it’s easy with SAS because of the ability to integrate pharmacy and medical side data. We can combine controlled substance claims from the retail pharmacy, along with the emergency room or physician.”

Market-leading innovation

By broadening its FWA detection program to include prescribers and members, Prime saved its health plan clients $279 million in the first year: $51 million in recovered payments and $228 million in cost avoidance, where Prime detected signs of FWA before claims were paid. The PBM also increased the speed and accuracy of investigations and has now exceeded $355 million in recovered payments and cost avoidance over the 18 months the solution has been in place.

These results helped Prime make Fast Company’s 2020 Most Innovative Companies list. The company ranked second in the health category, acknowledged for its unique use of AI and advanced analytics to identify more areas prone to FWA activity and recover overpayments due to fraudulent claims.

“The recognition is amazing,” Abou Nader says. “We’ve worked so hard over the years to fight fraud, and this new capability with SAS has helped us take it to the next level to safeguard our members and save our clients millions of dollars.”