As the COVID-19 pandemic rages across the US and the world, health care workers have confronted profound ethical dilemmas that strike at the core of the medical profession. In some cities, shortages of the personal protective equipment (PPE) that usually defends front-line health care workers from virus-containing droplets have led to critical conversations about what risks health care workers can be reasonably expected to assume in the care of patients.^{Reference Ranney, Griffeth, Jha and Jauhar1} The narrative review by Giwa and colleagues now published in this issue of the Journal of Law, Medicine and Ethics is a striking reminder that this question, this struggle that forces health care workers to balance patients' safety with their own, is not new.^{Reference Giwa, Milsten, Viera, Ogedegbe, Kelly and Schwab2} In an examination of the literature surrounding a very different epidemic — gun violence in America — the authors tackle an uncomfortable question that all health care workers hope they never have to face: what is a health care professional's moral obligation to protect patients during an active shooter incident in the hospital?

The investigators intended to conduct a systematic review, but converted the format to a narrative review when a search of the literature revealed, in their words, “a large number of editorials, commentaries, opinion reviews and very few cross-sectional studies.” Their thorough analysis of 32 publications yields no easy answers should this nightmare scenario become a health care worker's reality. In the end, the authors conclude that whatever framework is accepted for the individual (and for that they endorse a “Secure, Preserve, Fight” approach), the onus must fall not on a single person, but rather on the institution to priori-tize safety measures. As examples of these, the authors focus on the need to make hospital areas securable in the event of an active shooter. They go further, recommending that every “hospital's Emergency Management Committee needs to regularly update and educate hospital staff about active shooter response plans. Many authors have recommended that hospital administrators consider training key staff in the use of antiballistic armor or employing properly trained and equipped hospital security guards until law enforcement arrives …[health care professionals] need to all be aware of their environment and those in it at every moment and in every location.”

Anticipatory safety measures that are thought through and implemented by trained security professionals are critically important. Still, the potential harm of policies that seek to turn all health care workers into trained active shooter incident responders warrants careful consideration not only within health care, but also in other environments across the US, where similar discussions are taking place. Interventions that turn lay people into mass casualty responders, perhaps without fully considering the potential for harm of such approaches, are becoming widespread.

According to the National Center for Education Statistics, approximately 95% of public schools in the US now conduct active shooter drills. Despite a dearth of evidence about the effectiveness of such trainings, uptake has been swift, with too little recognition or debate about the possible harms and trauma such trainings may be imparting to students. More recently, school nurses and teachers unions have sounded the alarm, reporting that the fear of these drills, school absences when they are announced ahead of time, and anxiety that results is real.^{Reference Cogan, Nickitas, Mazyck and Kamenetz3} Without clear evidence that such drills are helpful if the almost unimaginable terror of an active shooter is unleashed in a classroom, how can we blindly accept the risks of militarizing schools?

Such efforts have moved beyond schools, reaching civilians in a range of locations. For example, the American College of Surgeons launched the “Stop the Bleed” campaign, with a goal of training Americans in bleeding control techniques (such as use of tourniquets). As of March 2020, nearly 1.5 million people have participated in these trainings.⁴ Of course, such trainings may empower those who attend with knowledge and a sense of security in the event of an exsanguinating extremity wound. However, fewer than 2% of firearm-related injuries are the types of vascular extremity injuries that might benefit from a tourniquet,^{Reference Hsu, Chang and Perez5} so caution seems warranted before increasing federal spending on these first aid efforts instead of on primary prevention measures that might stop such shootings from occurring in the first place.

Reading the work of Giwa and colleagues is uncomfortable, but it is meant to be. Practicing medicine in a country that accepts nearly 40,000 deaths from firearm-related violence each year means that many of us have considered what we might do, how we might act, where we might hide if a shooter showed up in our hospital. Yet, our response cannot be to act as if such events are foregone conclusions. Let the lessons learned from this (and all) epidemics include a commitment to focusing on prevention rather than a demand that our health care workers become martyrs because we failed to act.