BD, Microsoft create software to identify drug diversion
American medical technology company BD has unveiled software designed to help hospitals and healthcare systems detect the diversion of drugs by health professionals.
The software analyzes the dispensing patterns of clinicians to identify behaviors that suggest they may be diverting drugs for personal use or illegal distribution.
BD designed the software to contribute to efforts to bring the opioid crisis under control while protecting patients from harm and healthcare systems from financial losses.
The opioid crisis has created a surge in demand for narcotics. That demand creates a temptation for the clinicians in charge of dispensing opioids and other drugs to divert them from the legal to the illegal supply chains.
The threat is illustrated by a case at the University of Michigan's health system, where employees stole tens of thousands of hydrocodone pills. In August, the health system was hit with a $4.3 million fee to settle charges that its illegal distribution of narcotics from 15 locations was involved in two overdoses and one death.
While healthcare systems are under pressure to defend their supply chains against diversion, in BD's view the task is far from straightforward.
"Medication diversion is a growing and complex challenge for hospitals and health systems. Specific cases of diversion can be difficult to detect, and the impact can be devastating from a patient and healthcare worker safety standpoint," Ranjeet Banerjee, BD's worldwide president of Medication Management Solutions, said in a statement.
BD thinks connected technologies and accompanying analytics can help hospitals and healthcare systems tackle the drug diversion challenge. With that in mind, BD has created BD HealthSight, a cloud-based application designed to support investigations into drug diversion.
Working with Microsoft, BD developed machine learning algorithms that analyze data on a range of dispensing patterns such as overrides, canceled transactions, delays in dispensing, administering and wasting medications.
Those potential red flags represent a broader set of data than is traditionally used in diversion probes, which usually rely heavily on the amounts of drugs dispensed by clinicians. Use of these additional types of data has been limited by the time and effort involved in manually reviewing and reconciling anomalous dispense, administration and waste transactions.
By turning algorithms on the data, BD thinks it can efficiently analyze a broader range of dispensing activities and thereby more effectively identify clinicians who are engaging in behavior indicative of diversion.