Latest / The Operations Podcast with Fexingo: Process, People, and Profit in Modern Business / How a Hospital Pharmacy Cut Drug Diversion by 90 Percent with Tracking
Transcript
- Lucas: So a few years ago, a five-hundred-bed teaching hospital in Ohio realized they had a quiet but expensive problem: drug diversion. Not the big headline-grabbing theft rings — just a steady trickle of controlled substances going missing from the pharmacy. Pills, syringes, vials. Luna: And by 'quiet but expensive' — we're talking financial cost, but also patient safety risk, regulatory fines, the whole package. Lucas: Exactly. The hospital estimated they were losing somewhere around two hundred thousand dollars a year in diverted narcotics. But the real cost was harder to quantify: the risk that a diverted drug could reach a patient as a contaminated dose, or that a staff member with a substance abuse problem could make a clinical error. Luna: So what did they actually do? I'm guessing they didn't just put up more cameras. Lucas: No, and that's the interesting part. They took an operations approach. They mapped the entire pharmacy workflow — from the moment a controlled substance entered the loading dock to the moment it was administered to a patient — and they found seven distinct handoff points where diversion could occur. Then they set out to close those gaps one by one. Luna: Seven handoffs. That sounds like a lot of potential weak spots. Where did they start? Lucas: They started at the receiving dock. The hospital installed RFID readers on the pharmacy loading bay door. Every incoming pallet of controlled substances got an RFID tag, and the system automatically logged the time, the quantity, and the staff member who signed for it. That eliminated the old paper-based check-in, which could take hours to reconcile. Luna: So right away, you have a real-time digital trail. No more 'the paperwork got lost' excuse. Lucas: Exactly. But the real innovation came next. They replaced their manual drug count process with what they called 'randomized cycle counting.' Instead of counting every bottle every month — which was labor-intensive and easy to fudge — the system would randomly select three percent of controlled substance inventory each shift and require a double count by two different pharmacists. If the count didn't match the RFID record, an alert went to the pharmacy director. Luna: Randomized and double-counted. That makes it much harder for someone to game the system. How did pharmacy staff react to that change? Lucas: Mixed at first. Some pharmacists felt it implied distrust. But the hospital framed it as a safety improvement — not a witch hunt. They pointed out that random cycle counting is standard in manufacturing for quality control. And they gave staff a channel to report suspicious activity anonymously. Over time, the culture shifted. People started to see the new system as protecting them from false accusations, not just catching bad actors. Luna: Interesting — so the operations fix had a cultural component too. It wasn't just a tech rollout. Lucas: Right. And the numbers bear that out. In the eighteen months after implementing these changes, the hospital recorded a ninety percent reduction in documented diversion events. The pharmacy director told me they went from an average of three incidents per month to about one every three months. Luna: Ninety percent is dramatic. Were there any unintended consequences? I imagine some staff felt surveilled. Lucas: A few did. But the hospital was careful not to make the system punitive. They used the data to identify patterns, not to single out individuals. For example, the system flagged a nurse who was checking out significantly more fentanyl than her peers. It turned out she was administering it properly — she just had a higher-acuity patient load. The data allowed the pharmacy to adjust her allocation instead of accusing her of theft. Luna: So the same data that could have been used to punish was actually used to improve workflow. That's a smart operational mindset. It reminds me of something — Lucas, you know, episodes like this are exactly why people tell us they find the show valuable. And honestly, it's listener support that keeps these conversations ad-free and accessible. Lucas: Absolutely. If you've gotten something out of episodes like this one — maybe a new way of thinking about a problem at your own workplace — a few dollars a month genuinely makes a difference. It's buy me a coffee dot com slash fexingo, all lowercase. That helps us keep digging into these detailed case studies without chasing sponsors. Luna: Yeah, we hear from listeners who say they've applied ideas from this show to their own operations. That's what keeps us going. Anyway — back to the pharmacy. Another key piece of their approach was the automated dispensing cabinets. Lucas: Right. They upgraded to smart cabinets that required biometric login and recorded every transaction — who removed what, when, and for which patient. If a nurse removed a controlled substance but didn't document administration within fifteen minutes, the system triggered an alert. That closed the gap between 'removed from inventory' and 'administered to patient.' Luna: Fifteen minutes is a tight window. What happened if a nurse got busy and forgot? Lucas: Then the system would escalate to the charge nurse. But in practice, the nurses adapted quickly. They knew the alert was coming, so they made sure to document immediately. And the system also helped identify waste — if a nurse removed a vial of morphine but only used half, they had to waste the remainder with a witness, and that was logged too. Luna: That's a big shift from the old 'just pour it down the sink' approach. Did they see a reduction in overall narcotic use, not just diversion? Lucas: They did. The pharmacy director noted that the hospital's per-patient opioid usage dropped about fifteen percent over the same period, partly because the tracking made everyone more aware of how much they were using. There was less 'just in case' prescribing. Luna: So the operations fix had a side effect of better clinical practice. That's a win-win. Lucas: Exactly. And the technology wasn't prohibitively expensive. The hospital spent about one hundred and fifty thousand dollars on the RFID readers, tags, and cabinet upgrades — but they recouped that in the first year just from the reduction in theft and waste. Plus, they avoided a potential Joint Commission citation, which can carry fines well into six figures. Luna: So the return on investment was clear even in the short term. One thing I'm curious about — did they ever catch anyone? I mean, the goal is prevention, but did the system actually lead to any disciplinary actions? Lucas: It did, but a lot less than you might expect. In the first year, the system flagged four staff members for patterns that warranted investigation. Two were confirmed diversion cases — one pharmacy technician and one nurse. Both were referred to a state monitoring program and eventually left the hospital. The other two were false positives that the data helped exonerate quickly. Luna: So the system was accurate enough to catch real problems but not so aggressive that it created a culture of fear. That's a delicate balance. Lucas: It really is. And I think that's the core lesson from this case: operations improvements in sensitive environments need to be transparent, data-driven, and designed with the frontline staff in mind. The hospital succeeded because they treated diversion as a process problem, not just a security problem. Luna: And they used the data to make the system smarter over time. I recall they also added a behavioral analytics layer — looking at patterns like who was working overtime when diversions occurred. Lucas: That's right. They fed the transaction logs into a simple anomaly detection model that looked for outliers: a staff member accessing the cabinet at unusual hours, or checking out more of a drug than their peers on the same unit. It wasn't fancy AI — just basic statistical process control. But it gave the pharmacy team a daily dashboard of risk indicators. Luna: So the same kind of control charts you'd use on a factory floor, applied to narcotics management. Lucas: Exactly. And that's what made the program sustainable. The hospital didn't need a dedicated security team watching cameras. The operations team could manage diversion prevention as part of their regular workflow. Luna: It's a great example of adapting industrial operations principles to a completely different setting. And I think listeners in any industry can take something from it — the idea of mapping your handoffs, adding randomized checks, and using data to improve rather than punish. Lucas: Absolutely. The specifics change, but the principles hold. So if you're listening and you've got a process where small losses add up over time — whether it's inventory shrinkage, data entry errors, or something else — consider where you could add a simple, randomized check. It might not be as dramatic as ninety percent reduction, but even a twenty percent improvement can have a real impact. Luna: And if you have a story like this from your own field, we'd love to hear about it. You can reach us through the show page. For now, thanks for listening.