Implementation of a workshop for mass casualty incident triage training using an immersive virtual reality simulation

Abstract

Objective

We offered a workshop at the 2023 annual meeting of the Society for Academic Emergency Medicine to teach the Sort–Assess–Lifesaving Interventions–Treatment/Transport (SALT) triage protocol for responding to mass casualty incidents (MCIs) using an immersive virtual reality (VR) simulator. Here, we report workshop outcomes.

Methods

After a 1‐h didactic on the basics of triage protocols, workshop participants rotated through three skill stations at which learners learned how to use the VR headset and controllers, practiced applying SALT triage skills through a tabletop exercise, and then finally used our VR simulator for training responses to MCIs. During their encounter with VR, participants applied their new knowledge to triaging and treating the victims of an explosion in a virtual subway station. After a brief orientation, participants entered the scene to treat and triage virtual patients who had various life‐threatening (e.g., acute arterial bleed, penetrating injury, pneumothorax, amputations) and non–life‐threatening injuries (lacerations, sprains, hysteria, confusion). The simulator generated a performance report for each workshop attendee to be used for debriefing by a skilled facilitator.

Results

Participants were mostly trainees (residents), all of whom properly initiated their encounter with global sort commands (walk and wave) to identify the most critically injured. On average, participants correctly treated 92% of 18 injuries, with all bleeding injuries being properly controlled (tourniquets or wound packing). On average, participants correctly tagged 87.7% of 11 patients, but only took the pulse of 67% of the 11 patients. Learners had difficulty with cases involving embedded shrapnel and properly tagging patients who were stable after treatments.

Conclusions

Our VR simulator provided a practical, portable, reproducible training and assessment system for preparing future emergency medical systems (EMS) medical directors to teach their EMS professionals the triage and lifesaving intervention treatment skills needed to save lives.

INTRODUCTION

As mass casualty incidents (MCIs) proliferate in the United States, we increasingly rely on skilled first responders, who deliver accurate and timely triage and to reduce overall morbidity and mortality of victims. ¹ , ² , ³ , ⁴ Recent MCIs have shown that limited lifesaving interventions, primarily hemorrhage control, swift extrication, and immediate transport, should be priorities for care. ⁵ , ⁶ , ⁷ , ⁸ Training first responders to integrate response skills into a dynamic, unstable, and active disaster is challenging and difficult to simulate. Further complicating first responder training is the lack of a standard triage protocol or use of protocols that are outdated. ⁹ , ¹⁰ Transitioning first responders to best evidence triage protocols will require effective, best evidence‐based practices for training that are being actively researched and reported in the literature. ¹¹ , ¹²

The Sort–Assess–Lifesaving Interventions–Treatment/Transport (SALT) triage protocol is evidence‐based and endorsed by the Centers for Disease Control and Prevention. It is becoming widely adopted due to its simplicity and expedited hemorrhage control. ¹³ , ¹⁴ SALT is also endorsed by major organizations like the American College of Emergency Physicians. ¹³ , ¹⁴ Leaders in emergency medicine should consider promoting SALT triage as a uniform standard and support efforts to train the first responder workforce to mastery levels of performance through experiential learning. ¹⁵ , ¹⁶ While MCIs are currently low frequency, effective preparedness of those on the frontlines will result in high yields in the form of maximizing lives and limbs saved.

To fulfill this objective, we developed an interactive, virtual reality (VR) simulation that features a realistic environment for training MCI triage skills using universal avatars, programmed to simulate life‐threatening and non–life‐threatening injuries. ¹² Performance reports are provided to facilitate deliberate formative feedback. ¹² This paper describes our delivery of a workshop that offered participants the opportunity to experience the VR simulator for SALT triage training during the Society for Academic Emergency Medicine Annual Meeting, 2023.

METHODS

Workshop learning objectives

Learning objectives for the workshop were as follows: (1) compare and contrast the two predominant triage protocols to highlight the specific advantages of SALT triage, (2) recognize and apply lifesaving interventions (LSIs) needed during MCIs, (3) prioritize patient assessment by likelihood of life threat by applying a global sort, and (4) synthesize SALT triage protocol into action by triaging and treating an MCI using VR.

Workshop teaching methods

We introduced participants to the principles of disaster response, triage protocols, LSIs, and the first responder's role. We presented the evolution of triage protocols and identified strengths and weaknesses of each. ⁹ , ¹³ , ¹⁴ , ¹⁷ The SALT protocol was emphasized given expert endorsement and evidence base. ¹⁰ Participants were assigned to groups and rotated through stations to prepare for the MCI VR simulator: (1) using the VR equipment, MetaQuest 2 headset, ¹⁸ and practicing how to use the hand controllers to navigate the virtual space with a popular, mainstream VR game; and (2) applying the SALT triage protocol on paper cases. ¹⁹

We introduced participants to the simulator through a tutorial to practice first responder tasks in a nonchaotic setting. The tutorial covered the use of navigation features (teleporting and gliding), pulse taking, and applying tools or tags to patients. Participants also practiced interacting with the virtual patients.

Once prepared, participants were immersed in the full‐scale simulation, a subway platform status post–bomb blast. As first responders, participants had to triage and treat the scene using global sort commands, then assessing, treating, and triaging 11 patients. We used patient cases that were specifically designed for beginner's SALT triage training. ¹² , ¹⁹ The virtual scenario can be programmed for various levels of trainee expertise: from novice (low levels of chaos, e.g., noise, debris, and darkness; fewer patients; lower acuity injuries) to expert (high chaos levels, more patients, higher acuity of patient injuries). For this event, we set the chaos level and difficulty level of the virtual patients to moderate. ¹² , ²⁰ A successful first responder took a pulse on and tagged all 11 patients and delivered critical LSIs.

After the simulation, participants were debriefed by an expert in disaster response. The performance report presented patients in the order assessed. Debrief sessions are formative in nature and give the participant the opportunity to explain the clinical reasoning behind their triage and treatment decisions (Appendix S1).

Data analysis

We analyzed participant's VR performance to assess workshop outcomes. Participants were scored on the use of sort commands, pulse taking, correct application of LSIs, and application of correct triage label. We evaluated workshop outcomes with descriptive statistics from IBM‐SPSS Statistics for Windows, version 28.0. ²¹ Research on the use of our VR simulator for training and assessment of SALT triage skills was approved by our institutional review board (study number: 2020B0128).

RESULTS

Most (82%) of the 13 participants in the VR simulation workshop were emergency medicine residents. Among the other participants, there was one each: a medical student, an attending, a fellow, and one who did not answer. Their performances are summarized in Table 1. All participants correctly executed global sort commands and applied triage tags to all patients in the scene.

TABLE 1.

Descriptive performance statistics and item statistics from the MCI VR workshop: performance of participants by patient case.

Case	Item or task	Correct answer	Percentage correct	Wrong answers, error analysis
Sort the Scene—Step 1.	Issue a command that says: “If you are able to walk, please walk to the safe zone at the bottom of the stairs.”	Issue walk command	100	0
Sort the Scene—Step 2.	Issue a command that says: “If you can hear my voice, please wave your hand in the air.”	Issue wave command	100	0
Patient 1 (Bob_0) is still with no signs of life (no pulse, no breathing, no movement).	Take pulse	Yes	92.3	1, no
	Injury 1: face shrapnel	No treatment	100	0
	Injury 2: L wrist amputation	No treatment	100	0
	Injury 3: L chest collapse	No treatment	100	0
	Injury 4: stomach puncture	No treatment	100	0
	Injury 5: R shin amputation	No treatment	100	0
	Injury 6: L thigh laceration	No treatment	100	0
	Tag	Black	100	0
Patient 2 (Gary_1) is still. He suffered from embedded shrapnel in the forehead, above the eye, and is moribund with apneic respirations.	Take pulse	Yes	100	0
	Injury: face shrapnel	No treatment	100	0
	Tag	Gray	84.6	2, black
Patient 3 (Lily_2) is a waver. She has suffered a shin amputation with significant blood loss and is nearing hemorrhagic shock.	Take pulse	Yes	61.5	5, no
	Injury: R shin amputation	Tourniquet	100	0
	Tag	Red	69.2	4, yellow
Patient 4 (Gary_3) is still and suffers from a right chest collapse.	Take pulse	Yes	92.3	1, no
	Injury: R chest collapse	Needle decompression	100	0
	Tag	Red	92.3	1, yellow
Patient 5 (Lily_4) is a waver who experienced pulsatile bleeding from a puncture would to the left flank.	Take Pulse	Yes	76.9	3, no
	Injury: L side puncture	Wound pack	100	0
	Tag	Red	92.3	1, yellow
Patient 6 (Mike_5) can wave but had a puncture wound to the right flank and thigh laceration with pulsatile bleeding.	Take Pulse	Yes	46.2	7, no
	Injury: L side puncture	Wound pack	100	0
	Injury: R thigh laceration	Tourniquet	100	0
	Tag	Red	100	0
Patient 7 (Gloria_6) is a waver who suffered from shrapnel embedded in her right calf. The wound is not bleeding but would if removed.	Take pulse	Yes	69.2	4, no
	Injury: R calf shrapnel	Gauze wrap	38.5	1, tourniquet; 7, none
	Tag	Yellow	76.9	2, red
Patient 8 (Mike_7) is a waver who has a puncture wound in the left thigh that has resulted in pulsatile bleeding and a right calf laceration.	Take pulse	Yes	38.5	8, no
	Injury: L thigh puncture	Tourniquet	100	0
	Injury: R calf laceration	Gauze wrap	53.8	6, none
	Tag	Yellow	92.3	1, red
Patient 9 (Gloria_8) is a walker who suffered from pulsatile bleeding from a laceration on her forearm.	Take pulse	Yes	38.5	8, no
	Injury: L forearm laceration	Gauze wrap	46.2	4, tourniquet; 3, none
	Tag	Green	46.2	6, yellow
Patient 10 (Gary_9) has hearing loss from bilateral tympanic membrane ruptures.	Take pulse	Yes	61.5	5, no
	Injury: ear bleed	No treatment	100	0
	Tag	Green	92.3	1, yellow
Patient 11 (Helga_10) has a small abrasion to the forehead but is otherwise unharmed.	Take pulse	Yes	38.5	8, no
	Injury: forehead scrape	No treatment	100	0
	Tag	Green	100	0
Sort Information 1.	Used sort information to treat still patients first	1–2–4	76.9	4–2–3; 7–4–3
Sort Information 2.	Used sort information to treat walkers last	9–10–11	76.9	6–11–9; 6–8–5; 8–9–11
Treatment performance (18 injuries).		Mean (±SD) correct: 91.7% (±6.8%)
Triage tag performance (11 patients).		Mean (±SD) correct: 87.7% (±9.2%)
Pulse‐taking performance (11 patients).		Mean (±SD) correct: 67.5% (±24.4%)

Abbreviations: MCI, mass casualty incident; VR, virtual reality.

More than three‐fourths (76.9%) correctly prioritized “still” unresponsive patients first. More than three‐fourths (76.9%) also properly treated the “walking” patients last. Most participants (11 of 13, 84.6%) did not take the pulse of every patient. The majority (11 of 13, 84.6%) took the pulse of the patients who were too injured to move (static patients) but skipped those who could walk or wave (10 of 13, 77%). Participants more accurately treated patients with more serious injuries, like major hemorrhage or chest collapse, compared to those with less severe injuries like minor lacerations. Only 38.5% properly stabilized the embedded calf shrapnel wound that was not actively bleeding and only about half (54%) properly treated a calf laceration.

As we might expect, some participants were especially challenged in tagging patients who did not precisely conform to the SALT triage algorithm. For example, the woman with the shin amputation (Patient 3) presented with significant blood loss, tachypnea, and near hemorrhagic shock. She should have been tagged red, for immediate, but nearly a third tagged her yellow once they controlled bleeding with a tourniquet. Another challenging patient was the woman with active bleeding from a forearm laceration. Once her bleeding was controlled, this woman should have been tagged “green” for minimal. However, nearly half tagged her “yellow” for delayed.

DISCUSSION

Performance reports suggested that presimulation training was effective at teaching and reinforcing SALT to workshop participants. All participants executed global sort commands, a key concept in SALT triage. Participants effectively used sort information to prioritize patient assessment, and most correctly applied treatments or avoided LSI for all injuries.

While many participants failed to take pulses on every patient, this omission rarely changed their ability to treat or triage. Examples can be seen with Patients 6 and 11, who had low rates of measured pulses but a 100% rate of correct triage tag. This is a novel finding since SALT triage requires a pulse check but, in this realistic interactive environment, many participants skipped this step. This is likely explained by the participant's ability to assess patients globally through inspection and interaction. They felt rushed and made rapid assessments based on patient appearance, including consciousness, and implicitly decided that pulse taking was not necessary for triage.

Most triage errors were related to incorrect triage labeling caused by nuanced or dynamic patients or incorrect application of SALT rules. Pulsatile bleeding of an extremity that was controlled with a tourniquet reduced that patient's immediacy. Alternatively, errors also involved a patient whose hemorrhage had been controlled but only after severe blood loss causing hemorrhagic shock. This patient should have remained “immediate.” This finding suggests that novice learners may be prone to “tagging errors” and would benefit from additional reinforcement of the importance of withholding labeling until after treatment.

This workshop confirmed the simulator as a feasible tool for SALT training. The system is portable and can identify nuances in triage skills, critical for skill development. Future efforts should be made to research the efficacy of the simulator by conducting randomized controlled group experiments comparing the simulation training to current methods, such as in‐person (analog) large‐scale simulation events.

LIMITATIONS

This evaluation covered a single conference workshop with only 13 self‐selected emergency medicine trainees, who may have been particularly motivated to learn SALT. Deployment at other venues with other participant types may lead to different results. Also, this evaluation focused on analysis of trainee behaviors rather than intentionality.

CONCLUSIONS

The ability of first responders to prioritize critical patients and apply lifesaving interventions saves lives. A workshop using an immersive virtual reality simulation was effective in teaching and reinforcing these mass casualty triage concepts.

AUTHOR CONTRIBUTIONS

Nicholas E. Kman, Jillian McGrath, Ashish R. Panchal, and David P. Way contributed to the development of the virtual reality program used in the workshop. Nicholas E. Kman and Jillian McGrath contributed to the development of the other teaching materials and the design of the workshop. Nicholas E. Kman, Jillian McGrath, and David P. Way contributed to drafting the original manuscript. Matthew Malone and Travis Sharkey‐Toppen contributed to the refinement and delivery of the workshop. All authors contributed to writing and revising the final version of the manuscript.

FUNDING INFORMATION

First Responder, the virtual reality program used for this workshop, was developed with the support of the U.S. Agency for Healthcare Research and Quality grant R18HS025915.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Supporting information

Appendix S1.

Kman NE, McGrath J, Panchal AR, Malone M, Sharkey‐Toppen T, Way DP. Implementation of a workshop for mass casualty incident triage training using an immersive virtual reality simulation. AEM Educ Train. 2024;8(Suppl. 1):S70–S75. doi: 10.1002/aet2.10939