Drug Diversion – Can AI Monitoring Solve This Growing Issue?

The following is a guest article by Claire Reilly, Director of Clinical Operations at Imprivata and former Emergency Room Charge Nurse.

Although often overshadowed by major news and events, drug diversion – the rerouting of medications intended for patients by healthcare staff and is actually theft – has been a persistent problem in all healthcare settings. Pharmacies, doctors’ offices, hospitals, and care homes are all places drug diversion can occur if prescribing is not carefully monitored and managed.   

In recent years, multiple factors have combined to create a perfect storm for drug diversion. Healthcare workers were already overworked and struggling to reduce diversion when the COVID-19 pandemic first hit almost three years ago. With hospitals overwhelmed and under resource constraints, less focus was directed toward improving workflows and prescribing processes as clinicians and nurses struggled to manage and care for the influx of COVID-19 patients. In addition, increased stress and burnout has led an increased number of healthcare staff to divert drugs as a coping mechanism. Considering the pressures they face doing such challenging and taxing work, this isn’t unrealistic. In fact, according to some U.S. studies, it’s estimated that 10-15% of healthcare workers will abuse drugs or alcohol at some point in their careers. 

But outside hospital walls, demand for prescription drugs is growing too. Drug overdoses were on a steady decline up until 2020. Once the pandemic hit, overdose deaths increased 30% from the previous year. Drug overdoses continued to climb in 2021, resulting in over 107,000 deaths. Although many overdoses are caused by synthetic substances like fentanyl and heroin – not to mention the rise in counterfeit pills – prescription drug abuse is often a factor. Nearly 80% of individuals reported using a prescription opioid before heroin, according to the National Institute on Drug Abuse (NIDA).

And opioids are not the only type of medication being diverted from healthcare organizations. Recently, incidents of HIV medication, Adderall, COVID-19 drugs, and even insulin being stolen have made headlines.

Despite the current situation, many healthcare organizations continue to treat drug diversion as an isolated problem, but it is not. It’s a systemic issue that impacts all healthcare settings. The shockwaves of each instance extend far beyond the original diverter, impacting coworkers, system administrators, patients, and their families.

How Often Does Drug Diversion Happen – and Why Don’t We Know More? 

The first step to reducing drug diversion is understanding the scale of the problem. But there are few resources dedicated to tracking and monitoring how medications are dispensed and administered. Although many health systems are required to do audits, these are often done manually and can be error-prone and time consuming. So, all we know for certain is that drug diversion often goes undetected.  

Take the example of a recent study of assisted-living facilities and nursing homes in Minnesota, which investigated 107 substantiated reports of drug diversions. On average, each incident took place over 56 days, with over 30 doses stolen per patient. But in some cases, incidents went undetected for months, resulting in a dozen or more victims at a single site, according to the Star Tribune.

Can AI Help Healthcare Providers Prevent Incidents Proactively? 

Currently, most healthcare organizations measure drug diversion through randomized audits that focus on a small subset of transactions. Others do long-winded manual checks that involve combing through hundreds of pages of reports and cross-referencing them with electronic health records.  

This reactive approach typically uncovers incidents long after they’ve taken place. The negative effects of this include: 

• Multiple patients being denied pain relief or exposed to blood-borne pathogens through stolen intravenous drugs
• Workers diverting drugs may develop addictions that spiral out of control, impacting the ability for remediation or recovery 
• Significant monetary losses and exposure to regulatory liability for healthcare organizations – like the well-known case of Massachusetts General Hospital’s $2.3 million fine after it was found to have 20,000 missing or incomplete medication inventory records and 16,000 oxycodone pills stolen 

So, to detect drug diversion before the problem escalates, many healthcare organizations are turning to AI and big data technologies to help. AI can be used to turn huge quantities of raw data – patient pain scales, clinical practice notes, staff reports, automated dispensing cabinet reports, and more – into actionable insights at a speed and scale that human auditors simply can’t achieve. The right solution can detect a pattern of diversion within days or weeks, rather than months or years. 

Take the example of monitoring unusual medication waste. Wasting drugs is a normal part of hospital operations, as vials often contain more medication than patients need. But what if over a month, one nurse wasted medicine at an average rate three times higher than her colleagues? This type of incident would warrant further investigation. But without the help of AI – which can immediately spot the employee in question as an outlier – it could take months for anyone to notice. 

A Holistic Approach to Reducing Drug Diversion 

Preventing drug diversion is a continuous effort, the success of which depends on having a robust strategy to address diversion from all angles: technology, task forces, and education. Alongside behavioral analytics and AI, this requires executive-level support for combating the issue and a specialist task force dedicated to preventing drug diversion. 

On a grassroots level, all staff need training in how to detect drug diversion.  Employees need a channel to report their suspicions anonymously. Above all, when incidents are identified, organizations should offer rehabilitation resources to help addicted diverters recover and get back to work. With all the above measures in place – with the help of cutting-edge technology – healthcare organizations can start to turn the tide on this endemic problem.