Over the past two decades, mass casualty shooting incidents, defined as an incident where existing hospital and emergency medical services are overwhelmed by the number and severity of casualties, have increased dramatically in both number of events and victims.1 The genesis of this continued assault on our society is clearly multifactorial, but may be facilitated by the widespread availability of high-velocity military-type assault weapons and large capacity magazines.2 Some events are acts of terrorism, some acts of revenge, others acts of racial hatred. Many are subsequently found to involve some degree of mental illness.3 The true motive is rarely determined as the assailant either commits suicide or is killed during the law enforcement response. Due to this complexity, a solution for preventing these events, although widely sought after, remains elusive. Regardless, mass casualty shootings continue to plague our 24-hour news cycle with recurring scenes of death, injury, horror, and loss.
The immediate consequences of these events thrust local hospitals and trauma centers into the national spotlight. Effective, planned, and practiced mass casualty response involving all hospitals, government agencies, and other organizations who contribute to the regional trauma system is imperative to achieving organizational readiness and maximizing victim survival. From large metropolitan cities to small rural towns, no hospital or municipality is immune to the possibility that it may need to respond to a disaster event with multiple casualties. In recent years, the nature of such events has shifted from civilian bus, train, or aircraft accidents to military-style active shooter events and bombings. Regardless, all hospitals must have a comprehensive plan for responding to a mass casualty incident.
On June 12, 2016, at approximately 2:00 AM, an armed assailant entered the Pulse Nightclub in Orlando, FL, and initiated an assault that took the lives of 49 people and injured 53.4 The majority of victims were able to flee the scene or were rescued by law enforcement. During the two and a half hour standoff that ensued, the gunman prevented his hostages and other critically injured victims from receiving timely trauma care until a SWAT team entered the building, rescued the remaining hostages, and killed the assailant.
Orlando Regional Medical Center, the regional Level I trauma center in Central Florida and located a mere three blocks from the Pulse Nightclub, was the primary hospital responding to this mass casualty event. Less critically injured patients, and those who escaped and later sought medical attention, were taken to one of two local community hospitals (AdventHealthOrlando and Dr. P. Philip's Hospital, an affiliate of Orlando Health). The first wave of 36 victims was transported to the Level I trauma center within the first hour of injury. The second wave of 26 victims was transported to one of the three hospitals following resolution of the hostage situation just after 5:00 AM or self-transported after fleeing the scene that morning. A third wave of 40 victims trapped within the nightclub, widely anticipated by law enforcement to arrive for care by the three hospitals, never materialized as all of these victims had succumbed to their injuries.
We have performed a detailed analysis of victim presentation patterns, surgical resources required, operative volumes and timing, and patient severity of injury to report the magnitude of this mass casualty event and the patient care resources required to respond. While each mass casualty event is unique, we believe this information will be of benefit to hospitals who strive to prepare and train for future similar events. A subsequent article will address the hospital disaster planning lessons learned including incident command, triage, personnel mobilization, communication, surge capacity, blood banking, security, media relations, and family and staff counseling.
A retrospective review was performed of all known victims injured in the Pulse Nightclub shooting who either presented to one of three Orlando hospitals for treatment or died following the event. Institutional review board approval for this analysis with a waiver of informed consent was received from both Orlando Health and AdventHealth. Autopsy reports for all victims who succumbed to their injuries were obtained from the District Nine Medical Examiner's Office. The list of known survivors was compiled from records of the mass casualty response maintained by each hospital. Each patient chart or autopsy report was carefully reviewed by one of six authors (C.P.S., M.L.C., K.S., H.E., J.A.I., M.G.). Where questions remained regarding operative care or injuries sustained, the chart was reviewed by the authors and a consensus decision was reached.
The victims were stratified into three groups: survivors who received care at the regional Level I trauma center, survivors who received care at one of two local community hospitals, and decedents. Victim demographics, mechanism of injury, injuries sustained, time to presentation, operative procedures and surgical specialty required, blood utilization, bed utilization, intensive care unit and hospital length of stay, and survival were collected. As these were high-velocity gunshot wounds, the incidence of “through-and-through” bullet wounds was common. To avoid the potential bias and confusion associated with counting total bullet wounds, the number of bullets striking each victim, defined as “bullet impacts,” was carefully determined from reviewing the patient charts, the medical examiner's forensic analyses, personal experience as the patient's surgeon, or speaking with the operating surgeon. Victim severity of injury was assessed using Abbreviated Injury Score (AIS), Injury Severity Score (ISS), New Injury Severity Score (NISS), and American Association for the Surgery of Trauma (AAST) Organ Injury Grade (for organs where injury scoring systems are defined). Organ injury grades were determined by viewing the patient's radiographic images, personal experience, speaking with the operating surgeons, or reviewing the medical examiner's wound descriptions.
Statistical analyses were performed using Minitab 18 Statistical Software (Minitab Inc., State College, PA). Categorical data were analyzed using Fisher's exact test and continuous data were analyzed using Mann-Whitney U test or Kruskal-Wallis test. Data are reported as mean ± standard deviation, median (interquartile range) or percentage as appropriate. A p value less than 0.05 was considered statistically significant.
There are 102 known victims of this mass casualty incident (Fig. 1). Thirty-nine victims succumbed to their gunshot wounds and were found deceased within the nightclub while one victim fled and died in an adjacent parking lot. Nine victims arrived at the Level I trauma center in extremis of which five were pronounced dead upon arrival and four underwent brief trauma resuscitations before being pronounced dead. Fifty-three victims survived to receive definitive medical care, 40 at the Level I trauma center, and 13 at one of two local community hospitals. Twenty-nine victims required admission to the Level I trauma center, and five victims were admitted to a community hospital. The remaining 19 victims were treated and discharged from an emergency department that morning. It is likely that there were additional victims with minor lacerations, abrasions, or contusions sustained in the frantic exodus from the nightclub that did not seek medical care. Hundreds of victims, first-responders, and health care workers required psychological counseling in the days and weeks that followed.5
There were no age or sex differences between survivors and decedents. The victims of this mass casualty incident were primarily young men (Table 1). The exact number of rounds fired by the gunman in the Pulse Nightclub that morning remains unclear but has been estimated by law enforcement to be between 180 and 250. Many of these high-velocity projectiles passed through multiple victims resulting in a greater number of wounds than rounds fired. As a result, 296 individual bullet impacts can be attributed to these rounds. The decedents sustained significantly more bullet impacts (4 ± 3; range, 1–13) than survivors (2 ± 1; range, 1–5) (p = 0.008). Survivors who required hospital admission were typically admitted for management of gunshot wounds while those who were discharged from the emergency departments were commonly treated for tangential gunshot wounds, falls, lacerations, exposure to other victim's blood, or anxiety.
Due to the proximity of the Level I trauma center to the Pulse Nightclub, the first victim arrived 15 minutes after the shooting began at 2:00 AM. The last victim arrived at 5:40 AM following rescue of the hostages. The first wave of 36 victims arrived at the Level I trauma center over a span of 38 minutes. Twenty-five percent of the victims transported to the Level I trauma center arrived by police vehicle with an unknown number of victims carried the three blocks to the emergency department by bystanders. The remaining victims were transported to the Level I trauma center from the on-scene triage area by ambulance once the area was deemed safe. Thirteen victims were transported by ambulance or drove themselves to one of the two community hospitals arriving between 4:25 AM and 8:12 AM.
At the Level I trauma center, due to the volume of presenting patients and reports from law enforcement that there were numerous additional victims to follow, initial triage and trauma evaluations were performed using physical examination, limited plain radiographs, and bedside focused abdominal sonography for trauma examinations. Computed tomography was not utilized in the initial trauma assessments, but was used later in the morning once patient acuity had decreased and the true number of presenting patients was known. At the community hospitals, two patients received computed tomography imaging during their initial evaluation.
The majority of victims sustained a gunshot to multiple body regions (Fig. 2). All 49 decedents sustained gunshot wounds while only 48 of the 53 survivors who required medical care sustained gunshot wounds. Injuries to the head, face, chest, and abdominal AIS body regions were significantly more common among decedents (p < 0.0001). Decedents sustained a total of 205 individual bullet impacts while survivors sustained 91. Total bullet impacts to the head (11% vs. 2%; p < 0.01) and chest (19% vs. 11%; p = 0.09) predominated among decedents while those to the abdomen (20% vs. 14%; p = 0.23) and extremities (56% vs. 67%; p = 0.1) were more common in survivors. Mean AIS scores, denoting the severity of the ballistic injuries, were significantly higher for the head and chest body regions among decedents (Table 2). Both ISS and NISS were significantly higher among decedents when compared with survivors presenting to either the Level I trauma center or community hospitals (p < 0.0001).
The AAST Organ Injury Grading Scales do not exist for head, spinal cord, and orthopedic injuries. As a result, the AAST Organ Injury Grading Scale system could be applied to only 39 decedents with 133 organ injuries and 16 survivors with 42 organ injuries (Table 3). Decedents were significantly more likely to have chest wall and thoracic vascular AAST Organ Injury Grades while survivors were significantly more likely to have colon injuries.
Of patients admitted to the Level I trauma center, 83% underwent operation in the first 24 hours (Table 4). Five of these patients required more than one operation in the first 24 hours due to the complexity of their injuries with one patient requiring four separate procedures. The majority of operations in the first 24 hours were either exploratory laparotomies or orthopedic procedures for ballistic fractures. Two operating rooms were open and staffed at the time of the first victim's arrival to the Level I trauma center with two additional operating rooms available by 3:00 AM and two more rooms by 4:00 AM. No victim who survived to receive operative care succumbed to their injuries. Fifty-two percent of the patients were transferred directly from the emergency department to the operating room. Patients requiring exploratory laparotomy for abdominal penetrating injuries arrived in the operating room within 130 ± 92 minutes of presentation (range, 29–374 minutes). The duration of the initial emergent laparotomies was 72 ± 61 minutes (range, 18–250 minutes). Of the 13 emergent laparotomies performed, no intra-abdominal injuries were found in two (15%) patients. Seven (54%) of these patients were managed with an open abdomen and temporary abdominal closure. These abdomens remained open for 3.4 ± 2.5 days (range, 2–9 days) before definitive closure. Patients requiring orthopedic repair of non–life-threatening injuries arrived in the operating room within 764 ± 248 minutes of presentation (range, 418–1080 minutes). The duration of initial orthopedic operation was 155 ± 100 minutes. In the first 24 hours of hospital admission, 171 units of packed red blood cells, 116 units of fresh frozen plasma, 25 units of platelets, and 90 units of cryoprecipitate were transfused. Specialists utilized in the first 24 hours included trauma surgeons, emergency room physicians, orthopedic surgeons, anesthesiologists, hand surgeons, vascular surgeons, interventional radiologists, intensivists, and hospitalists. Specialties required at the Level I trauma center in subsequent days included plastic surgeons, urologists, neurosurgeons, neurologists, infectious disease consultants, gastroenterologists, physiatrists, and psychiatrists.
Of patients admitted to the community hospitals, none had injuries to the head, chest, or abdomen. Eighty percent underwent an orthopedic operation in the first 24 hours. These patients arrived in the operating room within 637 ± 171 minutes of presentation (range, 418–821 minutes). The duration of initial operation was 54 ± 33 minutes (range, 22–84 minutes). In the first 24 hours of hospital admission, 1 unit of packed red blood cells was transfused. Specialties utilized at the community hospitals in the first 24 hours included general surgeons, emergency room physicians, anesthesiologists, orthopedic surgeons, and plastic surgeons.
The primary intent of this analysis is to evaluate both the injuries sustained and resources necessary to respond to a mass casualty event such as the Pulse Nightclub shooting. Given the increasing frequency of these events, we hope these data are useful to both trauma centers and community hospitals in identifying the crucial resources necessary to respond to such events and improve victim survival.6
The various phases of this mass casualty response required different surgical and clinical services. Within the first 24 hours, trauma, orthopedic, and hand surgeons were the most needed surgical specialties. Emergency room physicians were crucial to the initial triage and resuscitation of victims. Anesthesiologists, intensivists, and interventional radiologists were similarly essential. Intensivists and hospitalists rapidly transferred or discharged appropriate inpatients to increase hospital bed capacity given the initial law enforcement reports that the total number of victims would exceed 150. A vascular surgeon was required for one patient with an axillary artery injury that required endoluminal stenting. No neurosurgical interventions were required in the first 24 hours as gunshot wounds to the head from these high-velocity rounds were uniformly fatal, preempting any opportunity for neurosurgical intervention. Similarly, there was no requirement for acute neurosurgical intervention for spinal injuries within the first 24 hours. Had this been a mass casualty event involving blunt force trauma, such as an earthquake, tornado, hurricane, and so on, the need for acute neurosurgical coverage in the first 24 hours for blunt head or spinal cord injury may have been very different. In the ensuing days following the incident, the availability of additional surgical subspecialties became vital including plastic surgeons and urologists. These phased differences in surgical staffing should be factored into the development of any hospital disaster response plan.
Initial triage at the Level I trauma center was performed by a senior emergency medicine physician as victims arrived outside the emergency department. Operative triage of patients was performed by the trauma surgeon who served as the surgical triage officer in the emergency department. This individual remained in the trauma resuscitation bay and emergency department throughout the event. Patients were triaged for surgical intervention based on wounding pattern, vital signs, and mental status. Those deemed most acute were sent to the operating theater for abdominal exploration by the trauma surgeons called in for the response. This served two purposes. First, efficiency was maximized by delivering appropriate patients to the waiting trauma surgeons and operating room staff who were able to focus on preparing for rather than transporting the next patient. Second, risk to the responding trauma surgeons and hospital staff was minimized as there was briefly concern for an active shooter within the emergency department at the Level I trauma center.
Two of the emergent laparotomies performed were nontherapeutic. Although this poses a potential waste of resources and unnecessary morbidity on victims, all patients taken for emergent laparotomy presented with what clinically appeared to be an acute abdomen and chest or abdominal wounds consistent with a high likelihood of ballistic injury. Morrison et al.7 reported a nontherapeutic laparotomy rate of 21% among British combat causalities. Velmahos et al.8 reported a nontherapeutic laparotomy rate of 14% in a series of 1,856 civilian gunshot wound patients. Our nontherapeutic laparotomy rate (15%), therefore, appears to be consistent with previously published data.
There are clear variations in injury patterns between survivors and decedents that may enable hospitals to better prepare for the staffing and equipment needs required to respond to a similar mass casualty incident. TariVerdi et al.9 demonstrated through simulation that a hospital's ability to respond to mass casualty incidents depends not only on the number of critical patients but also on their arrival pattern and injury type. Classification of our patient population by injury type and AAST Injury Grade may help predict the types of injuries that may be encountered following a mass shooting. The AAST Injury Grade Scales do not address brain, spinal, or orthopedic injuries. These injuries were common in this incident and played a significant role in patient outcome. Extremity injuries were especially common among both survivors and decedents. We suggest that Injury Grading Scales for these three injury types should be developed in conjunction with our neurosurgical and orthopedic surgery colleagues.
Smith et al.10 recently analyzed the publicly available autopsy reports of the deceased Pulse Nightclub victim population, but did not have the benefit of the decedent's medical records, the survivor's medical records, or the personal experience of surgeons who treated these victims. They also analyzed the decedent's bullet wounds differently, counting both entry and exit wounds as opposed to only bullet impacts as we did. We believe counting both entry and exit wounds potentially exaggerates the severity of a patient's ballistic injuries. As a result, they report slightly different data on the same patient population. They also focused on pre-hospital care while we focused on hospital-based care of mass casualty victims. They calculated a median NISS score for decedents of 43 (interquartile range, 34–57). Our median decedent NISS score was similar at 41 (interquartile range, 28–50), and our median survivor NISS score was 20 (interquartile range, 10–27). We identified differences in ISS and NISS among our surviving patients, but essentially identical ISS and NISS for the decedents. The complexities of these differences are not fully understood, but a previous study has suggested that NISS outperforms ISS as a predictor of mortality and complications in penetrating trauma given that injuries may cluster in a similar body region due to wounding pattern.11
In a subsequent, retrospective review of 213 autopsy reports from 19 mass shooting events, Smith et al.12 identified that “… the vast majority of persons involved in civilian public mass shooting events will die at the scene.” Based on autopsy reports of decedents, their study does not include victims who are able to flee the scene or those rescued by law enforcement. An estimated 200 potential victims of the Pulse Nightclub shooting were able to escape by themselves and law enforcement and emergency medical services were able to rescue approximately another 40 survivors and transport them to local hospitals. Smith et al. identified that deaths were most commonly due to gunshot wounds to the head and chest. These findings are consistent with our results and likely represent an important consideration in the initial triage and care of mass shooting victims where survival may be best optimized by focusing on patients with gunshot wounds to the abdomen and extremities rather than head and chest body regions.
The response to this catastrophic mass casualty incident required extensive collaboration between the Level I trauma center, local community hospitals, local/state/federal law enforcement, fire rescue/emergency medical services, and local/state/federal government.4 Our community has a long history of performing full-scale regional exercises as a tool for disaster response preparation. These exercises have fostered relationships between hospitals and governmental entities that ease the flow of information and enhance real-time decision making during crises. The annual full-scale regional exercises 3 months prior to the Pulse Nightclub incident was an active shooter scenario involving numerous hospitals/agencies and over 600 volunteer victims. The lessons learned during that exercise were of vital importance during the response on the morning of June 12. This reinforces the importance of preparation and training in optimizing victim outcome from a mass casualty incident.
The proximity of the Pulse Nightclub to the Level I trauma center, as well as the location of the two community hospitals, positively impacted patient survival. Patients in hemorrhagic shock from their ballistic injuries were able to receive rapid hemorrhage control due to earlier presentation to definitive care. Such outcomes may not have been possible had the distance to definitive care been greater. While law enforcement attempted to rescue all victims with signs of life during the hostage standoff phase of the attack, there were victims whose presentation to definitive trauma care was delayed due to their entrapment in the club with the shooter. Whether earlier transport and trauma care for these patients would have resulted in improved survival is speculative and relies on a supposition that was not valid in this situation. An analysis of the prehospital care provided to these victims is not the focus of this article. We agree with previous authors that early extrication of mass casualty shooting victims and rapid transport to definitive trauma care should always be the goal.
As the number of victims from this mass casualty incident became clear, the surge capabilities of each hospital were implemented. At the Level I trauma center, operating room staff from the adjacent pediatric Level I trauma center (Arnold Palmer Hospital) and women's hospital (Winnie Palmer Hospital for Women and Babies) were brought to Orlando Regional Medical Center to rapidly staff additional operating rooms. Our incident command leadership, in conjunction with the trauma medical director, discussed whether to transfer patients to one of two Level II trauma centers (each approximately 25–30 minutes away). As the capabilities of the Level I trauma center had not been exceeded, this did not become necessary, but may have had the third wave of 40 victims survived to arrive at a hospital.
As with any retrospective review, our analysis does have potential limitations. Documentation during a mass casualty incident is inherently limited by the need to focus on the large number of presenting casualties. This may be mitigated somewhat by the fact that the authors were intimately involved in the care of these victims and personally familiar with their injuries. Due to the frenetic nature of the morning, we do not have definitive details on the transport of each victim. As we reviewed the autopsy reports for the decedents, we found that some medical examiners included more detail than did others and this could potentially have impacted the accuracy of the organ injury scoring.
Our hope is that the experience and analysis of our response to the Pulse Nightclub shooting may be of value to those who aim to prepare for similar mass casualty incidents. Given the increasing propensity for mass shootings, as well as other environmental catastrophes, all hospitals should plan and practice how they would respond to a mass casualty or disaster event.
C.S., M.C., K.S., J.I. participated in study design. C.S., M.C., K.S., J.I., S.E., H.E., M.G. participated in data collection. C.S., M.C., K.S., J.I., S.E., W.H., M.L., M.L. participated in article preparation and editing.
Paula Velduis, Advent Health.
The authors declare no funding or conflicts of interest.
1. World Health Organization. Mass casualty management systems, 2007 pg 30.
2. Callcut RA, Robles AM, Kornblith LZ, Plevin RE, Mell MW. Effect of mass shootings on gun sales—a 20-year perspective. J Trauma Acute Care Surg
3. Philpott-Jones S. Mass shootings, mental illness, and gun control. Hastings Cent Rep
4. Cheatham ML, Smith CP, Ibrahim JA, Havron WS, Lube MW, Levy MS, Ono SK. Orlando Regional Medical Center responds to pulse nightclub shooting. Bull Am Coll Surg
5. Havron WS 3rd, Safcsak K, Corsa J, Loudon A, Cheatham ML. Psychological effect of a mass casualty event on general surgery residents. J Surg Educ
6. Russo RM, Galante JM, Holcomb JB, Dorlac W, Brocker J, King DR, Knudson MM, Scalea TM, Cheatham ML, Fang R. Mass casualty events: what to do as the dust settles? Trauma Surg Acute Care Open
7. Morrison JJ, Poon H, Garner J, Midwinter MJ, Jansen JO. Nontherapeutic laparotomy in combat casualties. J Trauma Acute Care Surg
8. Velmahos GC, Demetriades D, Toutouzas KG, et al. Selective nonoperative management in 1,856 patients with abdominal gunshot wounds: should routine laparotomy still be the standard of care? Ann Surg
9. TariVerdi M, Miller-Hooks E, Kirsch T. Disaster Med Public Health Prep
10. Smith ER, Shapiro G, Sarani B. Fatal wounding pattern and causes of potentially preventable death following the pulse night Club shooting event. Prehosp Emerg Care
11. Smith BP, Goldberg AJ, Gaughan JP, Seamon MJ. A comparison of Injury Severity Score and New Injury Severity Score after penetrating trauma: a prospective analysis. J Trauma Acute Care Surg
12. Smith ER, Sarani B, Shapiro G, et al. Incidence and cause of potentially preventable death after civilian public mass shooting
in the US. J Am Coll Surg
. 2019;229:244–251. (In press).