Assessing Risk When Everyone's Afraid: The Challenge of Seeing Health Care Workers as People When Our Need for Them Is So Great

I was a second year resident when I witnessed, from across a hallway, a failed resuscitation of a child hit by a car. Her image on that table—looking so much like my own young daughter—and her mother's screams haunted me for many weeks. I couldn't understand why: "She wasn't even my patient!" I told my mother in disgust. I was just there. It seems strange to me now, but in the moment, my mother's response came as a shock: "You may be a doctor now, but you are still a human being who witnessed a child die." I was still early in my training, but already it had become a fixed, unquestioned belief: a physician is absolutely invested in the care they provide, but ordinary human emotions should no longer affect them.

The idea that traumas you experience as part of your job do not "count" is widespread. I think it is likely at the core of a number of the challenges I experienced as a researcher attempting to address the traumatic stressors experienced by health care workers and first responders working during the COVID-19 pandemic. I am a VA psychiatrist working in a PTSD specialty clinic, and I run a research program focused on understanding how particular combinations of chronic and acute traumatic stressors can lead to the long-term changes we associate with PTSD. I also study how to treat, and, hopefully, prevent these changes.

When the pandemic first hit New York City, I began to hear my friends and colleagues in medicine describe not just long, harrowing work shifts, but also insomnia, nightmares, intrusive memories of the horrors they were seeing, and a sense of always being on edge. Although these symptoms can't count as symptoms of "PTSD" until they have been present for more than one month, what we know so far of the pathophysiology of both acute stress symptoms and PTSD suggests the underlying mechanisms are fundamentally the same. If treatment is indicated earlier than one month after trauma, in practice, most of the medication options are the same as well. The biggest difference is that it's so hard to study acute stress disorder. Here, we are always using these treatments "off label," relying on what we know of PTSD treatment, pragmatic experience, and rare case series, rather than large, organized clinical trials.

This gap in evidence base is most frustrating in the area of long term outcomes. Our theoretical and preclinical models would suggest that treating acute stress symptoms with medications that block the noradrenaline response to stress, such as the common PTSD medication prazosin, will also decrease the likelihood of these symptoms becoming the chronic symptoms seen in PTSD. However, there is no good clinical evidence for or against this hypothesis.

When I began to hear all the symptoms of acute stress that were emerging from frontline clinicians working during the COVID-19 pandemic, the right research move seemed obvious and urgent: if we could treat frontline clinicians experiencing such symptoms with prazosin, we would be providing the best treatment option I know of to a population that needed immediate intervention. We would be generating the first structured clinical trial data to address the efficacy of this intervention for, in particular, the acute sleep-related symptoms of acute stress disorder. We would also provide the first direct test of whether treatment with bedtime prazosin during or immediately after a traumatic stressor could decrease the risk of PTSD at 6 months. Although we would need to conduct the trial virtually to reduce the risk of COVID-19 transmission, we could do this using the same adaptations we were using in our clinical practices, which had rapidly converted to entirely virtual care. As the PI of a research team already running randomized clinical trials using the drug prazosin, this was a trial we could begin within weeks, with resources we had on hand. My team and I poured ourselves into the most rapid start-up of a...