Physicians on the Frontlines: Understanding the Lived Experience of Physicians Working in Communities That Experienced a Mass Casualty Shooting

A. Study Design

We performed a qualitative study that sought to understand the lived experiences of physicians who worked at hospitals that had a public mass casualty shooting. Our study protocol was approved by the Yale University Institutional Review Board.

B. Sample Selection and Recruitment

Eligible participants were physicians who worked at hospitals located in communities that experienced a public mass casualty shooting, as defined by the Congressional Research Service. ¹⁰ All major public mass casualty shootings with significant media attention that occurred between 2012 and 2018 were included. Initial study participants were recruited via emails sent to professional listservs as well as advertised on the social media platform, Twitter. Eligible participants contacted the investigators to set up an interview. After the first few interviews, we used the snowball sampling technique to recruit additional participants by referrals from previously enrolled participants. ¹¹ Our team concluded that we had thematic saturation (the point at which no new codes are being generated) after twelve interviews; we then completed five more interviews to confirm saturation.

C. Data Collection

We conducted semi-structured one-on-one interviews of eligible participants by telephone. The interview guide was created by the authors and circulated for edits among content experts. The guide was intentionally broad to accommodate the grounded theory approach. Verbal consent was obtained. Each participant was informed that their interviews would be audio recorded and transcribed with the removal of any identifying information. Participants were not compensated for their involvement.

Four authors conducted interviews from January to May 2020. The interviews consisted of broad, openended questions using a grounded theory approach. See Figure 1, infra. All interviews were audio recorded. Interviews began with basic demographic questions including age, sex, type of training, and years of practice. Questions then addressed the participant's experience during the event. The final set of questions involved the response to the event. These questions addressed the participant's personal reactions and how the reactions of the hospital, community, and nation at large affected their experiences.

OPEN IN VIEWER

Figure 1

Interview Guide

D. Data Analysis

Audio recordings were professionally transcribed. The transcriptions underwent review by the investigators with the audio recording to ensure accuracy of the data. The coding team consisted of an emergency medicine physician and a general surgery resident as the primary coders. Two medical students also participated in the development of the code book and themes as a part of the coding team. We used the constant comparative method of qualitative analysis to develop codes and themes. ¹² The primary coders first read through transcripts and catalogued the transcript data by assigning conceptual codes to different sections. The entire coding committee reviewed these codes at multiple coding meetings, discussing the meaning of the codes and how they relate to each other. We then created a hierarchy of codes, grouping them into themes. These themes, with the subgroups of codes became the codebook for the study. This process was organized on Dedoose Version 8.0.35, a webbased qualitative research software. ¹³

A. Sample

A total of 17 participants with training in emergency medicine, trauma surgery, pediatrics and psychiatry were interviewed. Participant demographics including age, sex, gender, and years of practice are displayed in Table 1, infra. The mean age of participants was 48 years old, and over half were female (53%; 9/17). The mean number of years of practice was 14. Trauma surgery and emergency medicine were the most represented specialties at 35% (6/17) each. The participants came from eight different communities that experienced a mass casualty shooting. Each community was represented by 2-3 participants on average with three incidents being represented by one participant. Some participants had experienced more than one mass casualty shooting. Each incident in these communities involved significant media attention. The deadliest event occurred in Las Vegas, NV on October 1, 2017, which injured 413 people and killed 58 people. In the majority of events, the shooter used a semi-automatic rifle. Additional descriptions and details of all incidents included in this study can be found in Table 2, infra.

OPEN IN VIEWER

Table 1.

Sample Characteristics

Variable	Mean (range) or Frequency
Age, years	48 (29-68)
Number of years in practice	14 (0-34)
Years since mass casualty shooting	3.9 (2-8)
Gender, Female	53% (9/17)
Type of training
Emergency Medicine	35% (6/17)
Pediatrics	6% (1/17)
Psychiatry	24% (4/17)
Trauma Surgery	35% (6/17)

OPEN IN VIEWER

Table 2.

Description of Mass Casualty Shooting Incidents

Location	Date	Number Killed	Number Injured	Type of weapon
Pittsburgh, PA	March 8, 2012	2	7	Semi-automatic handgun
Aurora, CO	July 20, 2012	12	70	Semi-automatic rifle
Sandy Hook, CT	December 14, 2012	26	0	Semi-automatic rifle
Charleston, NC	June 17, 2015	9	0	Handgun
San Bernardino, CA	December 2, 2015	14	22	Semi-automatic rifle
Orlando, FL	June 12, 2016	49	53	Semi-automatic rifle
Las Vegas, NV	October 1st, 2017	58	413	Semi-automatic rifle (modified to shoot like an automatic weapon with a bump fire stock)
Pittsburgh, PA	October 27, 2018	11	6	Semi-automatic rifle and handguns

B. Themes

In discussing the experiences of physicians working in a community that experienced a mass casualty shooting, four major themes emerged: (1) The psychological toll on physicians: “I wonder if I'm broken”; (2) the importance of and need for mass casualty shooting preparedness: “[We need to] recognize this as a public health concern and train physicians to manage it”; (3) massive media attention: “The media onslaught was unbelievable”; and (4) commitment to advocacy for a public health approach to firearm violence: “I want to do whatever I can to prevent some of these terrible events.”

The day of the shooting:

Participants described a variety of strong emotions in both their personal experiences and observations. Reactions to the sheer volume, type of wounds, selfless behavior of the victims, and the pressure on providers to succeed all emerged as themes. One participant described his experience of that day stating:

Many participants described feeling overwhelmed by the sheer volume of patients who were injured by firearms. See Figure 2, infra. One emotion commonly discussed was horror at the event itself, with a distinct feeling that this type of event was different from anything else they had seen in the past:

OPEN IN VIEWER

Figure 2

Theme: Psychological/Emotional Responses (with exemplar quotes)

Participants also noticed a difference in the patients' wounds when compared to their typical practice due to the types of weapons used in the attacks. The nature of the mass casualty shootings and lack of readily available information also led several providers to feel afraid for their own safety. See Figure 2, infra.

The participants also described a heightened sense of pressure to ensure a good outcome for the victims of the shootings. One participant whose hospital only received one patient from a mass shooting described this feeling:

While not every participant described each of these emotions or responses, every single participant described at least one of these reactions to the mass casualty shooting.

Dealing with the aftermath:

Almost all participants described long term psychological consequences in either themselves or colleagues that persisted for many months:

The impact of these events was particularly difficult for participants for a variety of reasons. First, participants described how their roles as perceived leaders within the healthcare system hindered processing their emotions in the immediate aftermath:

This reaction was compounded by the fact that many physicians continued to care for the victims of the mass shooting for days, weeks and months after the event:

This forced re-traumatization and compartmentalization, led several physicians to feel the weight of that day even more forcefully than they otherwise would have. See Figure 2, infra.

Despite these difficulties, participants also described several positive feelings. For many, the outpouring of support locally, as well as from around the country, brought their communities together. See Figure 2, infra. Several participants discussed the satisfaction they gained from taking care of patients that had been so horrifically injured, especially when they were able to achieve a good outcome:

Addressing the aftermath of a mass casualty shooting:

All participants in this study described at least some long-term psychological effects for either themselves or colleagues, including anxiety, flashbacks, and hypervigilance. The majority expressed concern about the development of post-traumatic stress disorder (PTSD) in the weeks to months following the mass casualty shooting — some continuing to describe symptoms years afterwards. Participants identified a wide array of direct support services available from their institution following the mass casualty shooting, including direct provision of therapy and counseling sessions. But participants from three of the eight incidents reported a complete lack of response or recognition. See Figure 3, infra. The importance of having organized, readily-available support systems for providers was repeatedly emphasized:

Individuals who were trainees at the time of the mass casualty shooting were identified as particularly in need of organizational support. In the words of one participant who was a resident at the time of the event:

Community and hospital recognition of providers' work was identified by some as an important source of healing. See Figure 3, infra. However, the unequal distribution of that recognition also caused some stress:

Limitations

There are limitations in this study. As with all qualitative studies, our findings are hypothesis-generating, and may not be generalizable, although we strove to identify outliers and dissenting opinions. Unique to this study, our participants were temporally removed from their experiences, some by many years. Therefore their recollections were subject to recall bias, though our interviewers were trained to prompt participants for concrete details and specific memories. In addition, their accounts are likely influenced by social desirability bias — the inclination to report answers, behaviors, or attitudes that adhere to social norms. ²¹ To mitigate this, all interviewers did not have a personal or professional relationship with the participants and all participants were assured of having anonymity. All interviews were conducted over the phone which increases the anonymity of the interview though may impact the way participants responded or understood questions. Finally, this research was conducted using snowball sampling with referral from one physician to another which may have biased the type of participants that were included within the study, in favor of those who may feel more comfortable speaking about their experiences. To mitigate this, we recruited multiple initial “nodes” of snowball sampling to ensure a variety of participants. Finally, our findings may not be transferable to other healthcare professions, such as nurses and pre-hospital providers, who also care for victims of public mass casualty shootings.