In 2017, fentanyl surpassed heroin as the leading drug involved in overdose deaths, increasing from 57 percent of opioid-related deaths in 2016 to 84 percent in 2017.
On a busy summer evening, six patients were brought by EMS transport to the Emergency Department at Penn Presbyterian Medical Center for suspected overdose of crack cocaine, which, unbeknownst to the patients, had been mixed with fentanyl. Within four days, ED providers treated 12 more patients with the same life-threatening overdose symptoms. For senior resident Utsha Khatri, MD, it was clear something was wrong. Khatri’s actions set off a chain reaction among ED leadership, Philadelphia’s Poison Center, and the Department of Health, and resulted in a coordinated rapid response that helped to stem the influx of overdose patients in just a few days.
Khatri and Jeanmarie Perrone, MD, a professor of Emergency Medicine and director of Medical Toxicology, share how things unfolded that weekend, lessons learned, and how their actions can serve as a model for other hospitals on the front lines of fighting the opioid epidemic.
Q: Let’s start at the beginning. What was going on in the Emergency Department that night?
UK: Every day we see EMS teams bring in patients who are suffering from an overdose. But this cluster was different. It was a Saturday evening, so the ED was pretty chaotic, with doctors and nurses trying to keep up with an influx of patients. As I arrived, the resident who was leaving mentioned that there had been at least four patients who’d come in around the same time with overdoses. Two had been successfully reversed with naloxone [Narcan], but it wasn’t long before more patients were brought in by paramedics.
Q: Why was this particular cluster worrying to you?
UK: Those who were able to tell us what happened – where they’d been and what they’d taken – were all adamant that they had smoked crack cocaine. They didn’t hide it at all, but they insisted they’d never used opioids and had no intention of using fentanyl that night. In fact, they were actually pretty surprised when we informed them that their response to naloxone suggested their overdose was due to an opioid.
Most people who struggle with opioid use disorder and come to the ED don’t deny their history; they often tell you what they used and how much – and many know how to titrate their doses to a desired effect. If you haven’t used opioids before, any “usual dose” can be deadly, which is what happened in this case. We then realized that all of these patients were coming from generally the same neighborhood, and the alarms really started going off in my head.
Q: When you noticed the unusually high volume of overdoses, the patients' unawareness that they’d taken fentanyl, and the geographic similarities, what did you decide to do?
UK: I called poison control to report the cases, as is standard protocol, and to see if any other EDs in the area had reported anything similar. They had not. Then, I texted Dr. Perrone to tell her what was going on. The next day, we had another round of patients in the same situation. It just seemed very off. Dr. Perrone was on board right away.
JM: I agreed that something seemed really strange. Even for our ED, where we regularly see overdoses, the number of patients was pretty worrying. I recommended we start testing all overdose patients for fentanyl, which is not in our practice protocol. Standard urine drug screens don’t include fentanyl, so initially the drug tests came back negative.
I contacted my colleagues at the Philadelphia Department of Health, and we set to work drafting a warning that the DOH then sent out to participating ED providers within 24 hours. The warning talked about the escalation of overdose cases, what health care providers should be on the lookout for, and special discharge plans for these patients. The news media was also alerted. The goal was to get the word out that there had been a batch of crack cocaine going around that had either been tampered with or had been adulterated with fentanyl. We wanted to warn people in the community.
Q: Speaking of reaching the community, part of the article you recently published in the New England Journal of Medicine about this cluster focuses on the need for increased access to and use of fentanyl test strips. How do those test strips work, and why is there a need for more of them?
UK: Rapid test strips are pieces of paper that individuals can mix with water to test the drug they’re going to take. The strip will indicate whether there’s fentanyl in the product. Most overdoses come from fentanyl, so knowing if a product has fentanyl in it can prompt the user to use less, inject more slowly, use with a friend, keep naloxone on hand, or possibly not use at all.
Q: Were there any other efforts made to reach members of the community?
UK: The DOH advised all EDs to do urine tests for fentanyl for these patients. They also notified people in the community of the uptick in opioid naive patients and recommended counseling and connections to addiction services. It was also recommended that naloxone prescriptions be made available as a standard part of discharge instructions.
Q: What was the response from these efforts?
UK: To have 18 overdoses in four days, and especially ones with such severe complications, is pretty drastic. But within four or five days, word got out among the community that whatever had been sold that weekend was a bad batch. We can’t say that people stopped using it because some overdose patients don’t come to the hospital, but we can say that the number of patients we saw with these symptoms dropped dramatically soon after the communications went out to the other hospitals and the communities.
Q: How did this experience change communication among the teams in Philadelphia?
UK: We’re really happy that the effort to better communicate between EMS providers and EDs in real-time was effective. The coordination among our teams started conversations about having more real-time alerting systems. We also have to be able to quickly communicate with the Poison Center and the DOH, and following the outbreak, we convened a meeting and created a group messaging system with the DOH, Poison Center, key toxicologists, and Emergency Medicine providers so we can communicate rapidly as cases come in.
JM: The city has an elaborate surveillance system that tracks ED visits for overdoses, but correlating the sentinel signals from that system with real time ED providers helps to identify any significant change in volume and determine whether an alert should be issued.
Q: What are the lessons learned from this cluster, and how might the rapid response coordination efforts be applied in other cities?
UK: What this cluster highlights and what other cities have observed is that the exposure of fentanyl is expanding to people who aren’t intending to use it. That’s what’s really scary. We have convincing patient and laboratory evidence to show that this was definitely multiple cases of unintentional fentanyl exposure.
JM: In the same week that the article was published, the city issued recommendations to all hospital EDs to begin rapid fentanyl testing of all overdose patients. Despite not being part of our current routine, or hospital and many others are exploring implementing rapid fentanyl testing soon, which will greatly help confirm the drugs of exposure for both patients and providers. I don’t think this cluster is unique to Philadelphia, but by quantifying it the way we have and describing the response, our hope is that other cities can feel empowered to reach out to groups beyond physicians who are also trying to fight this epidemic.