Latest / The Operations Podcast with Fexingo: Process, People, and Profit in Modern Business / How a Hospital Pharmacy Cut Drug Diversion by 90 Percent with Operations
Transcript
- Lucas: So let's start with a number that still catches my attention: It is estimated that somewhere between ten and fifteen percent of healthcare workers will divert controlled substances at some point in their career. Luna: That is a remarkably specific and unsettling statistic. Ten to fifteen percent—across doctors, nurses, pharmacists, technicians? Lucas: All of the above. And the data comes from a mix of DEA reports, academic surveys, and hospital internal investigations. The real number is probably higher, because most diversion goes undetected for months or years. Luna: And today's story is about a hospital that actually did something about it. Which hospital are we looking at? Lucas: A large academic medical center in the Midwest—they asked not to be named publicly because they're still under some regulatory scrutiny—but their pharmacy director shared the details on condition of anonymity. They had a problem: between 2020 and 2023, they had five confirmed diversion cases involving fentanyl, hydromorphone, and oxycodone. Two were pharmacy technicians, one was a nurse, one was a pharmacist, and one was an anesthesiologist. Luna: Five confirmed cases in three years—that sounds like a lot, but I imagine most hospitals would say that's just the tip of the iceberg. Lucas: Exactly. The pharmacy director told me that after each case, they'd do a root-cause analysis, update a policy, and move on. But the pattern kept repeating. So in early 2024, they decided to treat diversion not as a compliance issue, but as an operations problem. Luna: I love that framing. What does that mean concretely? Treating it as operations versus compliance? Lucas: It means you stop asking 'who did this?' and start asking 'what about our process made this easy?' They mapped the entire medication flow—from the loading dock where the pharmacy receives the order, to the automated dispensing cabinets on the nursing floors, to the waste disposal process in the OR. Luna: And what did they find? I'm guessing the loading dock wasn't the weak link. Lucas: Actually, it was part of it. They found that the pharmacy vault—where the bulk controlled substances are stored—had a single key that floated among staff. There was no log of who had it at any given time. The vault door was propped open during busy hours because the airflow was bad. So anyone could walk in and grab a vial. Luna: That feels almost negligent. But I know from experience that many older hospital pharmacies were built in the 1970s and never updated. Lucas: Right. The fixes they implemented were not high-tech. First, they replaced the single key with a biometric lock. Second, they installed a simple RFID reader at the vault door, so every entry was logged to an individual. That alone eliminated the anonymous access loophole. Luna: But the bigger problem is probably after the medication leaves the vault. On the nursing floors, in the OR—that's where the real diversion happens. Lucas: Exactly. The second major finding was that the automated dispensing cabinets—the Pyxis machines—were not being used correctly. Nurses were overriding the cabinet to remove medications without a patient order. They called it 'stat removal' or 'emergency override.' In many cases, it was legitimate—a crashing patient—but in others, it was a workaround to bypass the audit trail. Luna: How did they fix that without slowing down emergency care? Lucas: They didn't eliminate overrides entirely. Instead, they required a second nurse to co-sign any override within fifteen minutes. That was already the policy on paper, but it wasn't enforced. So they added a pop-up in the system that locked the nurse's account if the co-sign wasn't done by the end of the shift. That drove compliance from about sixty percent to ninety-eight percent in three months. Luna: So process enforcement through technology, not just policy. That's a classic operations move. Lucas: Exactly. The third big change was in the operating room. Anesthesiologists and CRNAs often waste unused portions of controlled substances after a procedure. The old process: they'd waste it in a sink, with a witness, but the witness was often a nurse who was busy and not really watching. So they switched to a 'waste kit'—a sealed container that uses a chemical neutralizer and a RFID tag. The waste is scanned before disposal, and the container can only be opened with a unique code. Luna: That sounds expensive. How much did all of this cost? Lucas: The pharmacy director estimated about two hundred and fifty thousand dollars in capital—mostly for the RFID readers, the biometric locks, and the waste kits. But they calculated that each confirmed diversion case costs the hospital roughly one point two million dollars in investigation, legal fees, fines, and reputational damage. So if they prevent just one case a year, the investment pays for itself in three months. Luna: And did it work? The title of the episode promises a ninety percent reduction. Is that real? Lucas: The hospital says that in the eighteen months since the changes, they've had zero confirmed diversions. But they also changed how they detect diversion. Before, they relied on tip-offs and occasional audits. Now they run daily analytics on every controlled substance transaction—flagging unusual patterns like a nurse wasting more than ten percent of their dispensed doses, or a pharmacy tech accessing the vault at odd hours. Luna: So the ninety percent figure might be partly a measurement effect—you catch more, so you think you have less, but actually you're just seeing it. But still, zero is zero. Lucas: That's a fair point. The director acknowledged that some diversion is probably still happening, but the low-hanging fruit is gone. The casual opportunities—the unlocked vault, the unmonitored waste—those are closed. And the staff knows it. Luna: I want to push a little on the cultural side. You mentioned a culture that looked the other way. Did the operations changes shift the culture, or did the culture have to shift first for the operations to work? Lucas: The director said it was both. They made a deliberate effort to communicate that these changes weren't about distrusting staff, but about protecting patients and protecting the staff themselves from temptation. They framed it as 'we are removing the easy paths to making a terrible mistake.' And that message actually resonated. The union initially opposed the biometric vault locks, but after the director explained the rationale—and after a couple of staff members admitted that the old system was too tempting—they came around. Luna: So it's almost like an operations version of 'design out the error.' In aviation, they design cockpits so that a pilot can't accidentally flip the wrong switch. Here, they designed the pharmacy so that a well-meaning person can't accidentally become a diverter. Lucas: That's exactly the analogy. And I think that's the real lesson for any operation that deals with high-value or high-risk inventory. Whether it's opioids in a hospital or expensive components in a manufacturing plant, the question is: Are you designing for the best-case behavior or for the reality of human frailty? Luna: It's interesting that you mention manufacturing. Because I think a lot of listeners might hear this and think, 'Well, my business doesn't deal with opioids.' But the principle applies anywhere you have valuable stuff and people who handle it. Lucas: Absolutely. And one last thing the director mentioned: They now do random, unannounced cycle counts of controlled substances. Not just at the vault, but on the floors. If a nurse's count is off by even one vial, there's an immediate investigation. That kind of operational rigor—the same kind you'd see in a high-reliability warehouse—is what drives the culture change over time. Luna: Speaking of operational rigor and making things work better for people, if these conversations have moved your work forward in some small way, we want to mention that listener support is what keeps this show ad-free and independent. It's a simple way to say 'this is valuable to me' without a big commitment. We're at buy me a coffee dot com slash fexingo. And we mean that genuinely—no pressure, just an option for those who find value here. Lucas: Yeah, and it's a nice way to keep the conversation going. So, back to the pharmacy: One thing I didn't mention is that they also changed the way they handle returns from the OR. Previously, unused opioids were often left in a bin and someone would take them back to the pharmacy hours later. Now they have a dedicated return process with a locked cart that gets picked up within thirty minutes. Luna: That closes another window. So in the end, it's a series of small, boring, unglamorous fixes that add up to a big change. No AI, no blockchain, just good process design. Lucas: Exactly. And I think that's the takeaway for any operations leader. You don't need a shiny new technology. You need to map your flow, find the points where the system makes it easy to do the wrong thing, and then redesign those points. The hospital did that, and it worked. Luna: Alright, so next episode we're going to look at a completely different industry—one where the 'valuable inventory' is data, and the 'diversion' is insider data theft. Stay tuned. Lucas: That's a good one. Thanks for listening.