Skip to Main Content

As an infectious disease doctor, Nasia Safdar is a detective of sorts at the University of Wisconsin Hospital in Madison. She tracks the patterns of infections, from the type of illness to the organism at its root, and in spring 2014, she noticed something odd: A cluster of bloodstream infections caused by an uncommon, and potentially deadly, bacteria.

The microbe, Serratia marcescens, can infect the lungs, bladder, blood, and skin, and usually causes a few infections per year at Safdar’s hospital; some studies estimate that about 1 out of 100,000 people fall prey to a blood infection from the bacteria annually. So it was strange that five cases had occurred in just five weeks, and Safdar did a double take.


“There were more cases than we had been used to,” Safdar said. “I thought that needed investigation.”

Because the bacteria were all the same strain, Safdar was confident the cluster stemmed from a single source. But what was it? To find out, Safdar scanned each patient’s chart, searching for a common thread.

For example, Safdar said, “Serratia is a moisture-loving organism, so we looked at all the medications and fluids the patients were getting.”


The patients infected by the bacteria had received the same drugs as those who hadn’t gotten sick. What differed was their location: All but one of the patients had spent time in the post-surgery recovery unit. What’s more, some of these patients reported pain after the operation that was poorly controlled by opioid medicines.

“Because we try to be diligent about pain control, that set off a red flag,” Safdar said.

Syringes with missing caps

Little did Safdar know, her colleagues in the pharmacy department at the hospital were conducting their own investigation. In mid-March 2014 — less than two weeks after the first Serratia case — a nurse had discovered that the tamper-resistant caps had been removed from 10 syringes of opioids. Less than a month later, three more were discovered. Lab testing revealed the tubes held next-to-no medicine; instead, they’d been filled with saline solution.

Subsequent testing of an additional 10 random opioid syringes in the hospital’s pharmacy and medication dispensers showed one contained only a tiny fraction of its original medicine.

The findings spurred an intensive investigation, including installing surveillance cameras. Over the next two months, investigators discovered more than 40 opioid syringes hospital-wide that had been emptied and refilled with salt water. The path led to a nurse working in the post-operative area; she had entered her password to remove medications from drug dispensers just before four of the patients who contracted Serratia received their medicine. The fifth patient who came down with the bloodstream infection lived with the nurse; he was her father.

Investigators believe the saline the nurse used to re-fill the syringes was contaminated by Serratia, although the containers were destroyed before they could be tested for the bacteria. They concluded she likely brought the liquid from home.

“It was very frustrating because it took awhile to link a person to all these cases, and we felt helpless,” said Safdar, who recently published the team’s findings in the journal Infection Control & Hospital Epidemiology. “We felt very relieved to find the culprit.”

The nurse, who reported struggling with opioid addiction, was swiftly fired for stealing the drugs. That ended the string of Serratia infections. She also lost her nursing license and was convicted on four counts of narcotics possession and placed on five years of probation; at the time of the trial, the investigation into the Serratia outbreak hadn’t yet linked her to the infections.

Still, for the five patients who endured severe pain after surgery, then got sick — one of whom died — the answer came too late.

An ongoing problem

In the midst of the opioid crisis, the risk of medication diversion — when a legal controlled substance is used or distributed illegally — is greater than ever. After all, opioids are the most commonly diverted drugs. Although drug theft occurs most often in outpatient facilities, such as clinics and urgent care centers, it can happen anywhere. Health care workers struggling with addiction face their own obvious dangers, but when a substance use disorder leads to events like those in Madison, patients’ lives also are on the line.

The last decade has seen numerous outbreaks of infections caused by health care employees siphoning medications intended for patients, from technologists to nurses to respiratory therapists, according to the Centers for Disease Control and Prevention. Yet many in the health care field remain unaware that drug diversion is a major problem.

“It’s become screamingly obvious that there are a lot of other people endangered” by a hospital employee’s addiction, said Dr. Keith Berge, an anesthesiologist at the Mayo Clinic in Rochester, Minn., who was not involved in the recent investigation but has written about controlled substance diversion. “It puts vulnerable patients at risk.”

Drug diversion is a complex problem that requires a multifaceted solution, starting with making it simple and acceptable for health care workers with addiction to seek help, according to Lauren Lollini, secretary of the International Health Facility Diversion Association, a nonprofit group dedicated to preventing drug diversion founded in 2015.

“Health care workers need to have a safe place to land if they are struggling,” said Lollini, who in 2009 was one of 19 patients infected with hepatitis C when a surgical technician stole opioids from a hospital in Colorado.

“It’s criminal to divert drugs, but somebody needs that drug to survive, so they are doing what they have to do,” Lollini said. “Breaking the law is separate from the addiction issue.”

Signs that an employee is struggling with addiction or is siphoning medicines can be subtle; vigilance for erratic behavior or other signs of substance use is vital, experts say, but red flags aren’t always obvious.

“They are often your best employees,” Berge said. “They may show up for work when they are supposed to be on vacation.”

When employees report suspicions that a colleague is stealing drugs, by law hospitals must relay these suspicions to the Drug Enforcement Agency within 24 hours. They may be reluctant to do so, fearing bad press and the financial and reputational losses that may follow. But Berge emphasized that it’s a crime to hide suspected diversion.

Along with health care institutions making it easy to report addiction or suspected theft, establishing a task force to investigate suspicious activity is key, said Berge, who helped create a 2012 road map to address drug diversion for the Minnesota Department of Health.

“The goal is to create a team that consistently investigates and learns from every diversion,” Berge said, adding that employees who divert drugs can be extremely clever.

That’s why tracing infections back to the theft that caused them is a team effort that requires persistence and vigilance.

“You have to follow the clues you’re confronted with,” said Safdar.

Allison Bond is a hospitalist at Massachusetts General Hospital. If you have dealt with a diagnostic puzzle that has been solved, either as a caregiver or a patient, please email Allison at [email protected]

  • PLEASE use the correct word for the singular of bacteria — it is bacterium. Bacteria is the plural of the Latin neuter noun. Would you say “the dogs is brown?” No? Then don’t say “a bacteria.” Or “a criteria,” either, while I think of it — the singular is criterion (in this case the word is Greek).

  • Yet I don’t know of a medical board that will allow a health care professional struggling with opioid addiction to be treated with buprehorphine/suboxone which is clearly the most effective treatment for opioid use disorder. Clearly if these people are “often your best employees” they are able to meet their requirements. So if we want to treat opioid addiction in health care professionals maybe we should start by opening up the most effective treatment as an option.

  • Interesting. We used Serratia marsescens in Micro teaching labs when I was an undergrad 40 years ago. If I remember correctly it forms red colonies, we used it track person to person cross contamination. Mouth pipetting back in those days, no gloves, no masks, somehow we all survived.

    • Yes, Serratia does form red colonies. I remember reading about an experiment (I believe in California) where Serratia was spread my airplane and then its spread was easily tracked because of the red colonies. Not a common pathogen, but then again it’s rarely injected directly IV. My question is why was her saline contaminated with it in the first place?

Comments are closed.