April 25, 2019, 0424: Seven-hundred-fifty gallons of anhydrous ammonia gas were leaked from two towed, two-ton ammonia tanks in Beach Park, a suburb of Chicago. The Agency for Toxic Substances and Disease Registry determined that the resulting anhydrous ammonia plume created a 493-yard threat zone. The cool, humid air and calm winds around the plume prevented the plume from rapidly dissipating. (Morb Mortal Wkly Rep. 2020;69:109; https://bit.ly/3bqPqvH.)
0517: As I was driving into my single-coverage community ED, the night doctor texted me, “Just a heads-up what you're walking into.” There was only a three-minute delay between that text and the next, but that was plenty of time to let my mind wander. Down a nurse? Five ICU holds? Ten psych holds? What came next was not at all what I had imagined. “Mass casualty just called by Beach Park for tractor tanker of anhydrous ammonia spill. Lots exposed. Bringing us 6 so far. I told them to spread the love to other facilities as there are 12 total. Hazmat on scene.”
My greatest concern was the 58-year-old male driver of a vehicle thought to have been involved in a car fire. The initial 911 calls reported a stalled car with billowing smoke. The driver was speaking in full sentences with mild changes in phonation, significant uvular edema, and no hypoxemia.
0617: I spoke with the poison control physician after the night doctor was advised that there were no criteria to intubate the driver because he had been removed from the environment. I asked how this was different from any other burn with signs of airway compromise, and was told that it was inappropriate to intubate him because he was no longer in the environment; the swelling would only improve, he said.
0626: A 56-year-old female passenger in the vehicle abruptly progressed to upper airway obstruction. I immediately implemented rapid sequence intubation and placed a 7.0 tube between her swollen cords using a Macintosh 4. It was extremely difficult to advance the endotracheal tube, but I was able to advance it to 23 at the lip with slow steady pressure. (She was extubated after 28 hours, and discharged three days later.)
I immediately moved the driver to a ventilator-compatible room and chose to do rapid sequence intubation. In retrospect, doing an intubation without a paralytic would have been a better choice. Upon direct laryngoscopy, I called for a cric kit, and asked for someone to call anesthesia and turn on my GlideScope. When I inserted the blade, there was an audible gasp from all of the staff in the room as they saw the image appear on the monitor. I can only describe it as watching muffins rise in an oven in time-lapse fast-forward. Following the dark anterior area in the sea of rising edematous pink mucosa, I successfully passed the tube blindly into the trachea.
As a team, we became acutely aware of the gravity of the situation.
0646: I intubated a 46-year-old man with no preexisting conditions. I was able to use a Macintosh 4, but I was only able to pass a 7.0 tube. He went to the ICU.
0700: Several wonderful things happened. The nursing staff and techs who were scheduled to leave stayed as the new team came in. Administration implemented the disaster code, and ICU beds magically appeared. The assistant CEO said to me during the chaos, “Tell me everything you need.”
That unconditional support and rapid response saved lives. I asked for many bottles of propofol and to staff one-to-one nurses for my critical patients. I knew if we lost any of those tubes, I might never be able to get in them again. In retrospect, I should have asked again for an anesthesiologist. I should have requested that elective surgeries be cancelled so we had access to those doctors and that equipment. I could have stopped more incoming ambulances, following the examples of our Level I trauma center that accepted only two patients and our designated hazmat resource hospital that refused to take any patients at all.
0732: Medics brought in a 57-year-old man with a history of hypertension. I intubated him immediately because of a profound change in phonation.
0737: I was able to pass a 7.5 tube though his swollen cords using a Macintosh 4. (He was discharged two days later, on April 27, and was then readmitted to the ICU on May 1 for dyspnea secondary to pulmonary damage from anhydrous ammonia. He survived, and did not require reintubation.)
0803: A symptomatic 29-year-old diabetic patient who had been brought in by the paramedics at 0748 required emergency intubation. I was unable to visualize anything more than an edematous epiglottis using a Macintosh 4 and had to use the GlideScope. I was then able to visualize his swollen cords and intubate him with a 7.5 tube. He was bronched the next morning, and his entire airway, including the visible part of the trachea, main stems, and part of the lobar bronchi, were inflamed. (He was discharged two days later.)
0818: I intubated a 27-year-old man with no known risk factors. I was able to use a Macintosh 4, but the entire left half of his epiglottis was profoundly swollen, and the right half had no edema at all. I could not pass a 7.5 tube and had to use a 7.0. (He was discharged two days later.)
0833: Thirty minutes after being brought in by EMS, I intubated a healthy, symptomatic 24-year-old first responder. He was one of the first paramedics on the scene, and he had run into the plume unprotected to rescue people from a car fire. I was able to pass a 7.5 tube through his swollen cords using a Macintosh 4. (He too was discharged on April 27.)
1030: The medical director arrived and started seeing some of the patients who were backing up. Then the media trucks arrived, and we made a statement.
1200: A third midlevel provider arrived and started seeing patients immediately.
1245: I found out that the cuff of the 46-year-old man I had intubated at 0646 showed evidence of a leak while in the ICU, so anesthesiology was called to the bedside emergently. While preparing for a tube exchange, the tube dislodged, and the patient required reintubation with direct laryngoscopy. The anesthesiologist described the airway as having “copious secretions and swollen vocal cords.” (The patient failed his swallow study after extubation, but improved and was discharged on April 29.)
1300: The medical director, the administrators, and the media were all gone. I still had five hours left in my shift. It was a truly surreal feeling.
1810: The night physician showed up. “I heard you had a busy shift,” he said.
Our seven patients were the most critical of the 33 exposed to anhydrous ammonia gas, bringing our census that day to 183. Fortunately, I had two exceptional physician assistants running the room with me and a wonderful scribe. I cannot emphasize enough how important every single member of the team was—the housekeeping staff who turned over the rooms so quickly, the techs who anticipated my every need, the intensivists and internists who accepted patients with limited data, the administrative staff in labeled vests making things happen, and of course, as in every emergency department, the amazing nurses who make everything possible.
Our greatest shortcoming was that we did not strip down every patient and seal their clothing in bags; we wanted to be able to handle the volume and quickly turn over rooms. In fact, I had to care for some ICU staff who had upper airway symptoms from inhaling the fumes from the patients' contaminated clothing. No interventions were required. This lesson is critical as we face the COVID-19 pandemic: We must protect ourselves before we can take care of anyone else.
Anhydrous ammonia gas has long been known to cause severe respiratory sequelae. (Lancet. 1941;241:95; J Agromedicine. 2004;9:191; Arch Otolaryngol. 1980;106:151; Br J Ind Med. 1973;30:78; https://bit.ly/3age9kC; Thorax. 1992;47:755; https://bit.ly/2XHorYG.) The gas combines with water in the respiratory mucosa in a highly exothermic reaction to form ammonium ions.
This causes significant thermal injury, and the resulting alkaline solution leads to liquefaction necrosis. (J Trauma. 2009;67:93.) One published case report emphasizes the importance of early airway intervention. (Anaesthesia. 1983;38:1208.) My experience corroborates that recommendation.
There are three relevant takeaways from our experience. First, take care of yourself first. You cannot save lives if you are not protected. We put ourselves in danger by not removing patients' contaminated clothing. Secondly, tube early and tube often. This is important in anhydrous ammonia exposures, and the evolving data suggest it might also hold true for COVID-19. (Lancet Respir Med. 2020;8:e19; https://bit.ly/2VxbP3z.)
Finally, emergency medicine is a team sport. When we work together, we can conquer anything. We should celebrate all of the members of our teams and remember that we are all on the same team as a community of emergency providers across the globe. We will be able to persevere in this pandemic only if we are united. Stay safe out there!
Dr. Wrenis an emergency physician with more than 20 years' experience and is currently practicing in the Chicagoland area. Follow her on Twitter@KellyWrenMD.