By Debra Rorie, RPh, Pharmacy Implementation Supervisor, Deaconess Health System
This first-hand account of a real diversion scenario describes how it was uncovered and addressed by the healthcare institution.
2010 marked a year of change at Deaconess Health System in Evansville, Indiana. We had converted to an electronic medical record (EMR), added carousel technology, remodeled our pharmacy department, and were converting to Omnicell automated dispensing cabinets.
At this point we really believed we were doing everything we could to keep track of medications and prevent drug diversion. However, after hearing diversion expert Kim New speak at a National Association for Drug Diversion Investigators (NADDI) conference we felt there was opportunity for improvement to ensure our patients’ safety.
Changes to Diversion Monitoring Process
We had been using Pandora diversion analytics with our prior automated dispensing cabinets to analyze our dispenses. Our drug control officer evaluated and forwarded any concerns to the nurse manager, who would then do chart reviews and follow up as appropriate.
At the time, Pharmacy was seldom involved in this monitoring process, but after the NADDI conference we played a much more active role. We recognized that an unbiased investigator, with clinical knowledge, could find things that others could not—or were not—willing to see. We connected our Omnicell cabinets to Pandora and began to build reports to help with diversion monitoring.
First Investigation Pays Off
Immediately after running the first report, we saw an individual who fell above her peers for oxycodone and acetaminophen usage. I contacted our drug control officer and was informed that they had looked at this nurse (whom I’ll call “Sally”) multiple times in the past, but the nurse manager had found no issues with this individual.
Sally was known to pay great attention to pain management for her patients. I continued to search for information and found that the documentation regarding oxycodone and acetaminophen was indeed meticulous. Everything was documented in the allowed time frame, from the time it was withdrawn to the time it actually showed as administered.
But during my review I began to notice that many patients had orders for Vyvanse (lisdexamphetamine), a stimulant, which was not a medication I would expect to see on the neuro unit.
Using Another Nurse’s Login
I noticed that each of these patients had multiple orders for Vyvanse from multiple providers. Eventually, we realized that Sally was finding a computer left unattended but still logged in by someone else, and was placing orders for the patients. The order would come to pharmacy via the EMR, be verified, and sent through to the cabinet.
Sally would then use another nurse’s login ID and remove the medication from the cabinet, open the patient’s chart to document the administration of the medication, and then immediately discontinue the medication. That’s why Sally’s name was not showing up on our anomalous use report for the Vyvanse. Instead, another nurse was showing up for it.
Discontinuation of Orders
The discontinuation of the medication order was the only place Sally’s name appeared on the chart, other than administration which you would expect. Discontinuation of orders isn’t something we had monitored in the past; it didn’t seem like a risk point.
When we looked at the reasons given for why the medication was discontinued (which was also unusual because we did not require a reason to be entered), it stated things like patient’s preference, discontinued by physician, patient intolerance, and dose has been adjusted. However, the next day, the same medication would get ordered again in the same dosage form.
Video Surveillance Aids Investigation
It took effort to convince the nurse manager that Sally had engaged in a diversion scheme, because this nurse was not somebody she expected to do this. Consequently, she was willing to believe it was the person whose ID was being used, based on other characteristics. Until we provided evidence from the video camera, the manager was not willing to proceed with the case.
Eventually, after lengthy interviews with Sally, she did confess to diverting the Vyvanse. We would have been unable to detect that she had taken this drug based on her behaviors. She didn’t exhibit any of the signs we’re told to look for in someone who might be under the influence of a stimulant.
Need for Objectivity
We refer to this case as the “it can’t be my nurse” hurdle, and we learned that it’s much easier to spot diverters when you’re an unbiased observer removed from the situation.
We were able to finally open the eyes of our managers to the fact that a good, high ranking nurse can be just as capable of having an issue as someone who is a regular staff nurse. When I approach a nurse manager about a potential diverter, I would really prefer to hear that the suspect is a low performer or the manager has had concerns themselves. Unfortunately, it often turns out that the diverter is an excellent nurse who has a strong work ethic and is well liked.
So let’s fast forward five years. Where are we and what have we learned? We have strengthened our diversion surveillance with each investigation, as we have learned different ways that individuals have attempted to work around our system. We now monitor waste as much as we do the number of medication issues.
We actually look at unreconciled doses from our EMR. And we conduct education with our nursing staff, so they know what our expectations are and what we are going to be monitoring.
Analytics Tools Detect Diversion and Clinical Practice Concerns
We use the anomalous user reports in Pandora as a key to know where to start looking. We inevitably see some type of issue highlighted when we look at these reports.
The issue often times is not diversion. It may be inconsistency or sloppiness in practice, which can then be addressed with that individual. The reports also help us compare individuals—you can slice and dice the information and twist it in different ways.
Transparency and Teamwork
We have really emphasized transparency among departments in the last five years, which has helped. Having heightened awareness throughout the organization speeds our investigations. For example:
- A report writer sends us a report and mentions they noticed a name that stands out to them.
- Those who review our bedside bar code scanning reports call to say a particular person has a poor scanning record, and it always seems to be around controlled substances.
- A nurse manager looks at the Omnicell or anomalous use reports and notices that a part-time employee is removing more medications than a full-time nurse.
- Staff members call to state concerns about colleagues who are removing medications in off hours.
So, if people can be forthcoming with information, it is a much easier process and helps target your efforts. Having good communication is absolutely imperative to being able to move your organization to the next level in diversion management.
Debra Rorie, RPh, is the Pharmacy Implementation Supervisor at Deaconess Health System. She has extensive experience in administrative practice with focused expertise in hospital pharmacy operations and pharmacy information technology. She has broad experience leading drug diversion investigations and reducing organizational drug diversion risks through effective use of analytics. Her many areas of expertise include electronic medical record pharmacy build, implementation of controlled substance monitoring, and perpetual pharmacy inventory.
Want to Know More?
View our previous post on diversion here.
To access a podcast of a panel discussion on diversion that includes a talk by Debra Rorie and representatives of other healthcare institutions, click here.
Learn about Omnicell solutions for detecting diversion here.