Analytics, machine learning help spot drug diversion at Piedmont Athens

State and federal regulations can result in fines for hospitals with poor monitoring programs, says Russ Nix.


For 20 years, Russ Nix worked in the public safety and law enforcement arenas; while working as an undercover narcotics officer, he saw the impact of the growing opioid crisis and addictions to other medications.

Now, Nix works as a drug diversion specialist at Piedmont Athens (Ga.) Regional Medical Center, a job he pitched to the 360-bed hospital and got. He’s one of a growing number of specialists working in hospitals to tamp down on illicit diversions of drugs.

Drug diversion is the taking of medications outside of the hospital or pharmacy and it is primarily done by inside personnel, such as a nurse or physician and possibly a pharmacist. In general, however, pharmacy departments are put in charge of drug diversion programs given they usually have the most access, Nix explains.

Piedmont Athens became a pilot site with Invistics, a vendor of advanced healthcare inventory visibility and analytics software that uses machine learning technology and analytics to detect opioid and drug theft across the hospital—the program was funded via a grant to Invistics in 2017 from the National Institutes of Health.



“What was most impressive is the software not only detected known diversion cases faster than current methods, it also highlighted recent cases under investigation,” Nix says. “The software provided a historical baseline view that allowed us to clean up practices by partnering with all departments involved.”

Also See: Prescription drug monitoring program use continues to rise

Under the program, Piedmont Athens feeds a range of data to Invistics, which conducts analytics on the data. The hospital wanted analytics assistance, but it also wanted to control the actual program. Nix takes the analytics findings and runs them through machine learning technology to confirm that machine learning is showing the same detection of drug diversions that the hospital is seeing.

Data used to detect diversions include hospital information systems, medical record systems, employee time clocks, wholesale purchasing, inventory and dispensing cabinets. This level of enterprise data enables detection of drug diversions that would not be detected by traditional methods that focus on dispensing cabinets.

“Healthcare organizations need to embrace proactive drug diversion management,” Nix contents. “It’s not enough to just react when incidents occur. All organizations should be looking for more proactive ways to reduce diversion and develop best practices. Drug diversion is system-wide—you can’t point to one area as the main culprit, so we have active investigations at all times.”

Healthcare organizations that run occasional reports and analyses are not likely to be effective in reducing drug diversions, because these reports likely will be several weeks behind the present state of diversions, Nix advises. Doing a report every 24 hours, however, will enable identification of an incident that could have put the hospital at risk.

That’s important for safety and legal reasons, because the federal Drug Enforcement Agency and state Boards of Pharmacy can levy fines against hospitals that don’t adequately monitor diversions. “These agencies want to know what the facility has done to try to prevent diversions so you need to adopt best practices,” Nix warns. “We want to do everything we can to be the best there is at managing drug diversions.”

Pharmacy vendors, he adds, can help hospitals develop their best practice strategy. Invistics, for example, facilitates meetings to help providers develop best practices for a given region, so there is continuous continuity of surveillance and joint development of new ways to improve surveillance.

More for you

Loading data for hdm_tax_topic #better-outcomes...