Skip to content Skip to navigation

Fighting the Fight Against Drug Diversion

February 22, 2013
by Rajiv Leventhal
| Reprints
A difficult issue to manage, Springhill Medical Center is tracking narcotic usage to help stop a growing problem in healthcare
Springhill Medical Center

Otherwise known as a criminal act involving a prescription drug, drug diversion in healthcare is a growing problem that can be difficult to identify; some studies say approximately 12-16 percent of healthcare professionals will be addicted to a prescription drug at some point during their career.  According to the Centers for Disease Control and Prevention (CDC), once they are prescribed and dispensed, prescription drugs are frequently diverted to people using them without prescriptions.

While organizations such as the National Association of Drug Diversion Investigators (NADDI) attempt to facilitate cooperation between law enforcement, healthcare professionals, state regulatory agencies and pharmaceutical manufacturers in the prevention and investigation of prescription drug diversion, Joe Adkins, Pharm.D., clinical pharmacist at Mobile, Ala.-based Springhill Medical Center—a 252-bed hospital serving the southwest area of the state—knows how tough an issue diversion can be to manage.


Historically, tracking diverters has always been a big problem, says Adkins, adding that in the past, paperwork has been the way to do it, but that becomes very labor intensive and can be especially challenging. While in recent years hospitals have been able to better track diversion by becoming more automated electronically, there was still the issue of doing analytics and figuring out where narcotic usage patterns may vary.

Springhill’s prior automated system tracked this data in spreadsheets, which were dumped into a file each day. In order to do any kind of analytics, those spreadsheets had to be recompiled one day at a time into a database that was homegrown.  And then custom queries and analytics had to be written on their own to do standard deviations, Adkins explained.

Now, Springhill is using Pandora business analytics solutions from Omnicell, a Mountain View, Calif.-based provider of technologically advanced automation. “When reviewing a report, you can dive into additional layers to a particular question area and find out more about a potential diversion case. It does all this in 30 seconds while I go get my coffee,” Adkins boasts. “When we upgraded to Omnicell last year, we were able to look at the narcotic usage patterns that were happening in the hospital and compare nurses to fellow nurses and compare patterns to users in the same area, and in doing so, it led us to finding outliers and tracking that way. We never had this capacity before.”

Specifically, Adkins says this is how the process works: the Pandora tool will look at the narcotic usage history of every user on that unit. The usage on that unit is then compared among nurses and anyone who has a disproportionally high number gets flagged by automated reports that are reviewed each month for each unit. Those reports are then shared with the nurse managers.

When usage numbers come back high, that is when the investigation begins. Questions start to arise, Adkins says. “Why does that nurse have so much usage? Is she taking extra shifts? Is she pocketing them? Is she selling them?” Although these questions don’t ever solve the problem themselves, they give you a narrow focus for the start of an investigation, he explains, adding that sometimes, people are forthright and honest, but other times they aren’t. “Careers and licenses are at stake here, so nurses know that with rehab and treatment, they have a chance to be reconciled with their license in time and come back into the workforce. But if they don’t cooperate, then it’s the worst case scenario.”


Springhill has engaged in five investigations of suspicious activity in the roughly six months that it has been using the tracking software. In three of those cases, Adkins says, diversion was taking place. And recently, the implementation of the system has seemed to spark significant changes in the hospital, including adjustments made by the nursing staff. For example, each time a nurse accesses a controlled substance, he or she is required to count the amount prior to giving out the dosage. But often, it is counted wrong, and when that happens, it creates an electronic discrepancy. Lately, as Springhill’s investigations have become more thorough, there have been far fewer narcotic discrepancies. Instead of paying little attention to those details, nurses are now asking how they can better clean up their discrepancies, Adkins says.

One challenge in the process is the meaning of the outliers that the software might pick up on. Some departments, such as orthopedics, might need to use more narcotics than other branches, so that unit’s usage would be higher than others. Another outlier may occur when a drug such as Methadone (used to relieve pain that has not been relieved by non-narcotic pain relievers, according to the National Institutes of Health) is used. Springhill doesn’t normally have many patients who would need Methadone, but when they are there, a good deal of the drug is used. So if a nurse has had one of these patients for two days, that narcotic usage will always come up as an outlier. This is why, Adkins warns, “you have to realize the outlier is a starting point, but it won’t solve the mystery for you.”

Overall, Adkins is highly satisfied with the new automation.  Prior to this, many problems were caught by nurses who would come in completely intoxicated or inebriated on whatever they were taking, he says. “That is how we found out. But [at that point] it was too late in the game, as you obviously can’t have those people taking care of patients. If the hospitals aren’t doing they type of data processing that we are now, it will be hard to find a good solution.”