Mass casualty triage: using virtual reality in hazardous area response teams training

Governance and legal

NHS England funds the National Ambulance Resilience Unit (NARU) to manage the NHS major incident response by ensuring the ambulance service meets the core standards of the Emergency Preparedness, Resilience and Response (EPRR) framework (Makin and Groves, 2020; NHS England, 2023)' NARU trains and manages the HART teams, which are located within the 10 ambulance trusts in England (NARU, 2022).

Ambulance services have legal obligations under the Civil Contingencies Act 2004 and the NHS Act 2006, which was further revised with the Health and Social Care Act 2012, to adequately prepare for these types of incidents and this involves training the first responders in triage (Bennett, 2018).

Hazardous area response teams

Hazardous area response teams (HARTs) were created following the London terror attacks in 2005. HART paramedics are all registered with the Health and Care Professions Council (HCPC) and undergo additional specialist training to work in complex and high-risk environments. HART paramedics ensure that NHS ambulance services can operate within the hot zone/inner cordon and work closely with partner agencies. HART's primary role is to save lives by rapidly locating, triaging and treating casualties. HART operatives are expected to make decisions in complex, high-stress situations, and to rapidly and accurately triage casualties at an MCI (NARU, 2022).

Currently, triage training for HART teams consists of presentations, paper-based exercises, practical skills stations and exercises using low-fidelity sand manikins with observations written on casualty cards (Figure 1). Following this training, each team takes part in at least one large-scale exercise each year involving live actors as simulated patients.

Mass casualty incidents and triage

An MCI can be defined as an incident with a greater number of casualties than the responding resources can handle (Ryan et al, 2018; Khorram-Manesh et al, 2021). Triage is implemented to identify casualties requiring immediate, life-saving interventions and categorise casualties as priority 1 (P1), priority 2 (P2) or priority 3 (P3), and identify deceased casualties (NARU, 2014; Khorram-Manesh et al, 2021).

Ambulance services in the UK use the National Ambulance Medical Directions (NASMeD) triage sieve (NARU, 2014) (Figure 2). All casualties have triage cards attached to them, identifying which triage category they are in and limited interventions to address catastrophic haemorrhage using tourniquets, haemostatic dressings and pressure bandages (Craigie et al, 2020) (Appendix 1). Deceased casualties' diagnosis of death is documented on black triage cards with the date, time, clinician's initials and PIN, and that the casualty's airway has been opened to confirm that they are not breathing (Appendix 1). These actions are crucial to document to mitigate against future potential litigation in public inquiries, where both commander and clinician actions would be subject to scrutiny (Craigie et al, 2020).

Manikins

The manikins (Figure 1) are extremely cumbersome and offer no real-life feedback to the clinician during assessment for catastrophic haemorrhage, respiratory rate, pulse rate or capillary refill time. All these physiological assessments are key components of the NASMeD triage sieve (Figure 2) (NARU, 2014).

Unlike with VR, when clinicians are using manikins for triage training, they refer to a casualty card which offers no real-time insight into patient emotion or physiological status (Ferrandini Price et al, 2018). This lack of feedback and real-life experience is likely to inhibit the HART paramedic's preparedness for an actual MCI or large-scale exercise, as environmental conditions between theory-based training and real-life experience differ widely. (Ferrandini Price et al, 2018). High-fidelity manikins that offer real-time feedback to clinicians are available but the cost could be prohibitive when several are required for simulation (Massoth et al, 2019).

Virtual reality

Over the past decade, various studies have shown that virtual reality (VR) has increasingly been used in medical training for doctors, nurses and paramedics (Bilek et al, 2021; Cicek et al, 2021). VR is a pedagogic resource that complements the holistic training health professionals undergo for MCI triage (Rees et al, 2020; Bilek et al, 2021). Evidence suggests that VR is a cost-effective method of training and is comparable to a live simulation (Luigi Ingrassia et al, 2015; Mills et al, 2018).

VR enables participants to learn in an experiential way by being immersed into an environment in which incidents in the real world can be replicated (Bilek et al, 2021; Behmadi et al, 2022; Boros et al, 2022). Objective data can also be provided to the user, facilitating individual and organisational learning by allowing structured debriefing (Luigi Ingrassia et al, 2015; Ferrandini Price et al, 2018; Bilek et al, 2021).

The global use of VR is increasing across the gaming industry and educational institutions (Bilek et al, 2021; Cicek et al, 2021). VR equipment is becoming more affordable, with the industry estimated to be worth $196 trillion (Cicek et al, 2021). It has been estimated that by 2026 the healthcare VR market could be worth $2.4 trillion alone (Boros et al, 2022). With increased adoption of VR, it could become more financially viable for ambulance services to incorporate VR into training (Boros et al, 2022).

The author was motivated to develop marauding terrorist attack (MTA) incident training because such attacks are current threats to the UK (Makin and Groves, 2020). This MSc project aimed to narrow the evidence gap in the use of VR in mass casualty triage training, and understand if VR could improve training, in turn improving HART preparedness for an MCI.

Aim

The aim of this study was to investigate whether VR could improve the overall effectiveness of HART triage training by allowing clinicians to apply the triage sieve process to casualties as they would at a real incident by using a VR head-mounted display (HMD) and haptics in the controllers to feel respiratory and pulse rates (Figure 3). Clinicians can interact with casualties and select interventions such as tourniquet application, wound packing, applying pressure bandages, opening airways and placing casualties in the recovery position.

It is hoped that VR will not only improve the effectiveness of triage training but lead to increased confidence at applying the triage sieve process, therefore increasing preparedness for an MCI.

Outcomes

The primary outcome was to establish whether VR would be an effective supplementary training resource to facilitate preparedness for response to an MCI.

The secondary outcome was to assess the percentage increase in confidence HART paramedics gained at responding to an MCI and triaging casualties following the VR simulation.

VR software development

The author could not locate VR software that focused on MCI triage and incorporated the NASMeD triage sieve (Figure 2) that is used by UK NHS ambulance services (NARU, 2014). He partnered with a Manchester-based VR company. Subsequently, an MTA MCI triage scenario was specified and co-developed for use in this MSc project (Appendix 2).

An HMD was required for the study. The Oculus Quest 2 HMD (Figure 3) was selected as it is an accessible, portable and standalone tether-less device that does not require a computer connection (Rees et al, 2020). Furthermore, the Oculus HMD's hand controllers use haptics to provide feedback to clinicians when assessing casualties for respirations and pulses, in addition to interacting with equipment within the VR scenario (Rees et al, 2020) (Appendix 2).

An MTA scenario was developed (Appendix 2) involving a firearms attack at a football stadium. To ensure the participants felt fully immersed, casualties (unless deceased or unconscious) could talk to participants and inform them of what injuries they had. Distractions were added with background noises from sirens and radio chatter.

The software was developed over a 5-month period through various stages and was tested by the author and paramedics from outside the study group. The final version was completed on 1 May 2022.

Ethics

Review against the NHS Health Research Authority (2023) website showed that the project was not research and should be classified as a service improvement. Queen Mary University London's (QMUL) research and ethics department were informed of the study and provided a letter for the North West Ambulance Service (NWAS) research and development team who maintained oversight of the study. NWAS also gave approval through a letter to the author (Appendix 3).

Site and participants

HART paramedics from Manchester were the target participants of this study. However, it is hoped that the findings of the study can be generalisable to NHS ambulance services that may choose to incorporate VR into triage training for frontline staff to facilitate national preparedness for an MCI. The Journal of Paramedic Practice, a peer-reviewed UK journal, was the target journal.

Sampling and inclusion criteria

The author had a potential sample size of 42 operational HART paramedic participants. To allow for maternity leave and sickness absence, the author required a minimum sample size of 36 HART paramedic participants.

Inclusion criteria allowed only operational HART paramedics to take part in the study. Participants were HART paramedic team leaders, specialist paramedics and operatives.

Data definitions

Data was collected via an online, simplified 3-point Likert scale which enable the author to easily and accurately measure participant responses to answer the primary and secondary outcomes.

A table was populated with the data. As shown in Table 1, the data demonstrates the number and percentage of participant responses for agree, disagree and neutral responses to the Likert survey.

Power and sample size

Recruiting 36 participants with the projected increments in the Likert scale would mean the study approached 100% power. Chi-square testing was used to access proportions. Significance was set at alpha 0.05.

An incentives, participants were given a reusable coffee cup and entered a raffle to win an Oculus Quest 2 which was drawn once they had completed the study. This ensured 100% participation and near 100% power.

Descriptive methodology

Artwork was commissioned and a poster for the study was developed with a link to the study website which contained the information pertaining to the study. The paramedics could then sign the participation information sheet and volunteer to take part in the study (Appendix 4).

The design was a before and after educational study; participants completed pre and post surveys consisting of questions and statements with which they agreed, disagreed or remained neutral. Figure 4 shows each stage of the study and data collection points from the pre and post study surveys. All survey responses were anonymous. (Appendix 5).

The study started with a 16-question survey incorporating simplified 3-point Likert scale questions; it identified demographics, experience and confidence of participants with regards to MCIs, triage and VR (Appendix 6).

Following this, VR was demonstrated and participants were made familiar with the VR equipment. This included a tutorial on a virtual patient where participants practised interventions and interactions in the virtual environment before entering the MTA scenario (Figure 5).

Once the participants were confident with all interventions and interactions, they could decide to enter the triage scenario. The scenario consisted of 15 casualties. There were three P3 walking wounded casualties and two survivor reception people (who are automatically guided past the participant upon entering the scenario). The remaining 10 were divided into P1, P2 and dead casualties, who were to be triaged by the participants. Once they had completed the exercise, the participants would see a visual pop-up displaying accuracy and time taken (Figure 6).

On completion of the tutorial and scenario, the participants completed a fnal online survey consisting of 11 questions using a simplified 3-point Likert scale. To allow participants to give qualitative feedback and their opinions about the VR study, the author included space for a free text response at the end of the survey (Appendix 6).

The data collected from the pre and post study online surveys were stored in accordance with QMUL and NWAS research and development team policies. The study lead ensured that all data protection procedures were followed throughout.

Analytic methodology

The author liaised with his year 3 supervisor and statisticians to analyse the study data. Statistical analysis and plotting were performed using Stata 17MP.

Descriptive statistics for categorical data were performed using proportions. Key proportions were reported with the two-sided binomial exact of 95% confdence interval (CI). For questions where no participants responded with disagree, only agree or neutral, a one-sided CI of 97.5% was used. Univariable analysis was performed using Fisher's exact test (for categorical variables) (Appendix 6).

Bias and mitigation

Data collection bias was mitigated by using the Microsoft Forms online survey tool to collate responses. By using this online technology, neither the study lead nor anyone else could tamper with data. A detailed breakdown of bias and mitigation can be found in the project plan (which can be made available from the author on reasonable request).

Participant characteristics

Figure 7 shows participant demographics including gender, age, years of experience as a paramedic/HART paramedic role within HART and level of familiarity with VR.

Main results

The results from the anonymous simplified 3-point Likert scale are shown in Table 1.

Virtual reality technology

Over the past few years, VR technology has developed quickly, becoming a proven technology in military training, healthcare and many commercial industries (Luigi Ingrassia et al, 2015; Mills et al, 2018; Rees et al, 2020; Bilek et al, 2021; Cicek et al, 2021; Behmadi et al, 2022; Boros et al, 2022).

Increasingly, paramedics and other clinical professions use VR technologies in training (Mills et al, 2018; Rees et al, 2020; Behmadi et al, 2022; Boros et al, 2022).

The author wanted to understand and bridge the current evidence gap in HART paramedic MCI triage training and VR technology by understanding its effectiveness at improving the training process and increasing confidence. To the author's knowledge, this is the first study investigating VR MCI triage and UK HART paramedics.

Primary findings

The primary aim of the study was to understand whether VR could improve HART MCI triage training. To do this, the author identified primary and secondary outcomes.

To answer the primary outcome, the author compiled responses to three statements in the post-survey questionnaire (Table 1).

The findings of this study clearly address the primary outcome and support the findings of Birtill et al (2021), who state technologies such as VR when incorporated in training will improve not only paramedic training but also patient care (Birtill et al, 2021).

Secondary findings

The secondary outcome of the study focused on confidence. HART paramedics are trained to work in hazardous situations and expected to deliver high-quality prehospital clinical care (NARU, 2022). To be able to operate to the required standard HART paramedics need to be confident in themselves, their teammates and their training (NARU, 2022).

In the pre survey, participants were asked how confident they felt about responding to an MCI and triaging casualties. Thirty-one (86%) (95% CI) said they were confident and five (14%) (95% CI) said they were not (Appendix 6).

The post VR scenario survey found that 31 (86%) (95% CI) said their confidence had increased when attending an MCI and applying the triage process. This shows that VR increased paramedic self-confidence, which underlines the importance of incorporating VR into training.

This theme supports the findings of Mills et al (2018) and Behmadi et al (2022), who stated that incorporating VR into MCI training increases paramedics' confidence regarding attending an MCI (Mills et al, 2018; Behmadi et al, 2022).

Immersion in VR

All participants reported feeling fully immersed in the MTA scenario (Appendix 2). Previous studies suggest that transferring the learning from VR to the real world requires participants to feel fully immersed, with scenarios that are close to reality (Rees et al, 2020; Bilek et al, 2021; Cicek et al, 2021; Behmadi et al, 2022; Boros et al, 2022).

Furthermore, if participants feel fully immersed in a VR environment, this will enable them to respond in a realistic way suggesting experiential learning could be transferred to the real world (Rees et al, 2020; Behmadi et al, 2022; Boros et al, 2022).

Statistical significance and links to previous studies

This study demonstrated statistical signifcance that VR can increase the effectiveness of HART MCI triage training. Taking part in the VR scenario appeared to increase participants' confdence; however, this was not statistically signifcant.

Overall, the results are encouraging, supporting previous studies of paramedics highlighting VR as an effective training tool for MCI triage (Luigi Ingrassia et al, 2015; Mills et al, 2018; Ferrandini Price et al, 2018; Bilek et al, 2021; Behmadi et al, 2022).

Adaptability and portability

VR is a portable training resource and new software can be easily developed and updated as required by organisations (Luigi Ingrassia et al, 2015; Mills et al, 2018; Rees et al, 2020; Bilek et al, 2021; Birtill et al, 2021; Pedram et al, 2021; Behmadi et al, 2022; Boros et al, 2022).

VR would be useful to help implement the new NHS ten-second triage tool (TST), which is due to be operational in April 2024. If VR were adopted by NARU and NHS ambulance services, staff training would be enhanced and individual trusts and NARU would be able to collect the data.

VR can also be used by an infnite number of clinicians across multiple locations, enabling organisations to ensure staff are adequately trained (Mills et al, 2018; Bilek et al, 2021; Behmadi et al, 2022).

Furthermore, capturing data from simulations and identifying trends and training needs will constantly improve training and ensure the requirements in the EPRR framework are met (Makin and Groves, 2020).

Cost-beneft analysis

Replicating the scenario developed for this study in a real-life training exercise would be costly, requiring signifcant fnancial and time resources (Lowe et al, 2020). Previous cost-beneft analyses have shown that, over the course of a year, VR training could be 90% cheaper than replicating the scenario in a real-world training environment (Mills et al, 2018).

This fndings of the study supplement existing evidence and validate the current bank of evidence supporting the use of VR in paramedic MCI training. The author believes that VR is a key training tool, enabling organisations to train staff in a cost-effective way while immersing them in a stressful environment without causing risk to themselves or patients.

Next steps

The author would like to continue to contribute to the development of further triage modules, covering each of the triage phases and represent the different MCIs that could potentially happen (Ryan et al, 2018; Craigie et al, 2020). The author would also like to focus on other areas of paramedic practice such as HART competencies, commander training, clinical skills and chemical, biological, radiological, nuclear and explosive (CBRNe). VR technology and these simulations should be used to train police, and fire and rescue service colleagues in the new TST. As of April 2024, all emergency service staff will be using TST to assess casualties at MCIs. The author believes that this is a positive step and will improve patient outcomes following an MCI.

To further progress the adoption of VR, a relationship with an academic institution would be crucial to enable further academic research to be conducted into the uses of VR in paramedic practice and the wider healthcare system.