Redesign of a virtual reality basic life support module for medical training – a feasibility study

Abstract

Background

Healthcare providers, including medical students, should maintain their basic life support (BLS) skills and be able to perform BLS in case of cardiac arrest. Research shows that the use of virtual reality (VR) has advantages such as improved accessibility, practice with lifelike situations, and real-time feedback during individual training sessions. A VR BLS module incorporating these advantages, called Virtual Life Support, has been developed especially for the medical domain. Virtual Life Support was collaboratively developed by software developers and stakeholders within the field of medical education. For this study, we explored whether the first version of this module capitalised on the advantages of VR and aimed to develop an understanding of barriers to feasibility of use.

Methods

This study was conducted to assess the feasibility of employing Virtual Life Support for medical training and pinpoint potential obstacles. Four groups of stakeholders were included through purposive sampling: physicians, BLS instructors, educational experts, and medical students. Participants performed BLS on a BLS mannequin while using Virtual Life Support and were interviewed directly afterwards using semi-structured questions. The data was coded and analysed using thematic analysis.

Results

Thematic saturation was reached after seventeen interviews were conducted. The codes were categorised into four themes: introduction, content, applicability, and acceptability/tolerability. Sixteen barriers for the use of Virtual Life Support were found and subsequently categorised into must-have (restraining function, i.e. necessary to address) and nice to have features (non-essential elements to consider addressing).

Conclusion

The study offers valuable insights into redesigning Virtual Life Support for Basic Life Support training, specifically tailored for medical students and healthcare providers, using a primarily qualitative approach. The findings suggest that the benefits of virtual reality, such as enhanced realism and immersive learning, can be effectively integrated into a single training module. Further development and validation of VR BLS modules, such as the one evaluated in this study, have the potential to revolutionise BLS training. This could significantly improve both the quality of skills and the accessibility of training, ultimately enhancing preparedness for real-life emergency scenarios.

Background

Basic life support (BLS) can be life-saving in the case of cardiac arrest, a major public health issue with approximately 275,000 out-of-hospital cardiac arrests occurring annually in Europe [1]. Outcomes can be significantly improved through proper training in BLS [2]. Studies have demonstrated that frequent repetition, even in short training sessions, improves both the quality of BLS skills and overall attitude toward BLS, such as increased confidence [3, 4]. For non-medical personnel, it is advised to undergo training sessions more than once per year [5]. Increased accessibility to training can facilitate more frequent practice, thus improving proficiency. Making training sessions attractive to attend and increasing accessibility is vital for encouraging attendance. By improving accessibility, more individuals can regularly participate in training which could lead to better preparedness and potentially save lives during cardiac arrests.

Research on the use of Virtual Reality (VR) for BLS training has shown promising results. Using VR-based training instead of traditional classroom-based courses offers at least three major advantages. First, VR improves the accessibility of BLS courses [6], because it allows individuals to practise independently, with or without a mannequin, enabling more frequent practice at convenient times for participants [7]. Second, VR can offer real-time feedback even during individual training, thus enabling individuals to refine their skills in a safe environment without requiring instructors or trainees to be together [7, 8]. Third, VR provides the opportunity to practice lifelike situations. It is possible to create multiple scenarios and add distractions to simulate realistic situations, all within a safe learning environment where no harm occurs even if errors are made [7].

In this study, the term VR is used to refer to fully immersive environments using head-mounted devices (HMDs). We define augmented reality (AR) as systems that integrate 3D virtual objects into 3D real environment (physical world) [9]. Mixed Reality (MR) lacks an explicit definition. However it is often used to describe a mix of real and virtual objects within a single display on a spectrum between a fully real and a fully virtual world (Reality-Virtuality Continuum) [10]. In other words: a view of the real world—physical world—with an overlay of digital elements where physical and digital elements can interact [11]. Extended reality (XR) is an umbrella term that includes AR, MR and VR.

Among the VR modules available, the majority focus on laypeople rather than medical students and healthcare providers e.g. Lifesaver VR. BLS training for laypeople aims to improve the bystander cardiopulmonary resuscitation (CPR) rate by increasing familiarity with the BLS protocol and providing practice opportunities. Medical students are expected to perform BLS and BLS courses are part of their regular medical training [12]. Consequently, a module designed for medical students could focus on the quality of BLS rather than simply motivating them to perform BLS. Previous research has shown that VR can potentially improve educational learning objectives for medical students [7, 13]. Using VR could improve accessibility and encourage students to become more self-motivated learners by creating a sense of urgency [13].

A module tailored for medical students and other healthcare providers should encompass different objectives compared with a module intended for laypersons [6]. BLS modules aimed at medical students are currently in the developmental phase across various medical universities. These VR modules were not ready for comparison at the time of this research and/or did not aim to integrate the earlier stated advantages of VR BLS. Some modules incorporate realistic scenarios but require the use of a computer-connected headset, thereby limiting accessibility [7, 13]. Other modules have used textual feedback during the scenario [14] or Augmented Reality (AR) [15], the latter of which renders the scenario less realistic. AR uses the real world as a framework only, with the possibility for specific virtual elements to be added; for that reason, it is less lifelike. Hence, the development of a new VR-based BLS module with integrated advantages is necessary. Virtual Life Support, a new VR BLS fully immersive module, was developed by VR lab (VR lab BV, Nijmegen, The Netherlands) in consultation with the researchers of this study [16].

Most studies on the development of VR resuscitation modules have used Likert scale surveys, which are quantitative outcome measures [8, 15, 17, 18]. Other studies have explored the effectiveness of VR BLS in randomised controlled trials (RCTs) by comparing VR with other BLS teaching methods, such as classroom-based teaching. The primary outcomes of RCTs on VR BLS are depth and rate of compression, which can be measured using certified mannequins and/or evaluation through a European Resuscitation Council-endorsed checklist [19, 20]. These quantitative methods can help assess the suitability of a module for practical everyday use. Incorporating qualitative methods can provide in-depth insights into user experiences, potential barriers, and facilitators [21].

Prior to validation and introduction in medical training, a module such as the one developed for our study requires testing, optimization, and subsequent comparison to the established gold standard of teaching practices in BLS. A new module is valuable only if it is easily accessible, provides real-time feedback, and offers the opportunity to practice with lifelike situations. Therefore, a feasibility study is needed to explore the use of VR in any practical BLS module [22]. Specifically, this study will investigate the module’s: the accessibility, the effectiveness and accuracy of real-time feedback provided, and the realism of the simulated scenarios.

The aim of this study was to evaluate whether Virtual Life Support is suitable for integrating the stated advantages of VR. As previously noted, most modules that have been developed capture only some of the advantages of VR BLS; creating a VR module that captures all three major advantages while specifically targeting BLS training in the medical domain is challenging. To ensure the major advantages of VR were maximised, Virtual Life Support was collaboratively developed by software developers and stakeholders within the field of medical education. Alongside testing whether the module capitalised on the advantages of VR, the second aim was to develop an understanding of barriers to the applicability (relevance and correctness), acceptability (participant satisfaction and engagement), and tolerability (ease of use without discomfort) of Virtual Life Support as an educational tool for BLS skills training in medical students and ultimately all healthcare providers. To assess this second objective, a primarily qualitative design was chosen.

Methods

Content of virtual life support

Virtual Life Support [16] incorporates several key components to effectively test and refresh students’ skills in individual training. The module includes a calibration process, a lifelike resuscitation scenario, and an overview of performance for each practice. The design minimises the risk of developing simulation sickness by keeping the module under 30 min, ensuring minimal conflict between visual and vestibular sensory systems, and creating a realistic immersive environment where participants walk and move without navigational control [23]. However, simulation sickness remains a concern, potentially reducing accessibility for participants with severe simulation sickness, who might not be able to complete the course.

The calibration process involves adjusting the headband, setting the volume of music and speech, selecting the appropriate mannequin or the correct chest height and aligning the physical setup with the virtual environment. Meta Quest 2 (Meta Platforms, Inc., Menlo Park, California) can be used for XR purposes, the AR/MR function allows participants to see outside environment while wearing the HMD. Version 1.2 of the module uses this MR feature to enable participants to calibrate the mannequin before switching back to full immersive mode.

Once calibration is completed, the scenario begins. The scenario is considered VR since there is no view of the real world, and the mannequin only serves as a haptic tool. Users must perform BLS according to the European Resuscitation Council (ERC) guidelines (see Fig. 1) [20]. The total module takes 10–15 minutes including calibrations. The hands-on resuscitation part, from start of compressions till the participant is relieved by the paramedic, lasts approximately 2 minutes. If a step is not completed within a certain timeframe, a green hologram appears, or verbal feedback is given to help the user perform the next step. The Zoll AED 3 (Zoll International Holding B.V., Elst, The Netherlands), which provides real-time feedback on depth and frequency during compressions [24], is integrated into the module. The performance assessment is displayed on the final screen of the module. Version 1.1 was released at the start of this study, while version 1.2 was released during the interview phase. There are some small differences in content such as ‘’Mixed Reality passthrough view during setup for easier mannequin calibration’’ and ‘’Added period of CPR without an automated external defibrillator (AED)’’ but mainly technical changes, these are described in Appendix 1.

Fig. 1

Overview of the steps in the VR module: Users must ensure the safety of themselves, the victim, and any bystanders by assessing the environment for potential hazards, both in front and behind them. Once the area is deemed safe, they should “check for response” by gently shaking the victim’s shoulders and asking loudly, “Are you alright?” The system tracks whether the user places their hands on the shoulders but does not register the verbal response. The next step is to open the airway and check for breathing. The system verifies hand placement, ensuring that one hand is on the forehead and the fingertips of the other hand are positioned under the chin, after which the user should tilt the head back. The user should look at the chest, listen for breathing, and feel for any breath for at least 10 s. The system ensures that the user maintains this position for the appropriate duration. Once this step is successfully completed, a mobile phone (hands-free mode as shown in the figure) appears on the screen, prompting the user to call emergency services (112). In version 1.1 of the system, an automated external defibrillator (AED) appears immediately after the call. In version 1.2, the user must instruct a bystander to fetch an AED, which becomes available after two cycles of chest compressions and rescue breaths. The user is then prompted to begin chest compressions. The system registers the correct hand placement on the chest and monitors the number, rate (targeting 100–120 compressions per minute), and depth (5–6 cm) of compressions. After 30 compressions, the user is instructed to give two rescue breaths. The system monitors hand placement and the time it takes to switch between compressions and rescue breaths. The cycle of 30 compressions and 2 breaths is repeated. When the AED arrives, users are prompted to turn it on and apply the pads after removing the victim’s clothing. From this point, the user must follow the AED’s verbal instructions. If the user takes too long to complete a step, the system provides visual guidance, such as a green hand hologram, to demonstrate the correct action

Setting

The current first-aid course, which includes BLS, at Radboud University (RU) in Nijmegen consists of an online learning environment with five two-hour live training sessions that involve groups of 15 first-year medical students. Theoretical and practical BLS training begins in the first session and continues during the course following deliberate practice as a learning strategy. Each medical student at Radboud University must pass a theoretical and practical examination to successfully complete the course. BLS training occurs annually during the bachelor’s program [12].

Virtual Life Support was made available for Meta Quest 2 with hand tracking and is widely available at Radboud University for educational purposes. Meaning that the Meta Quest 2 can be used for self-study at the medical faculty, headsets are available in a designated computer lab during opening hours (8:00 AM-10:00 PM). Meta Quest 2 is available at various faculties; technical support is provided to students using these devices. Hand tracking provides information on the actions of the user regarding the steps of the BLS, and the system can then provide feedback that helps the user refine their skills based on this information. The module can be used with or without a physical mannequin by using alternatives such as air- or foam cushions. However, this study used a physical mannequin: the Resusci Anne mannequin (Laerdal Medical, Inc., Stavanger, Norway). The physical mannequin provides sensory feedback without the need for additional haptic technology. Sensory feedback is necessary to acquire the right sense of strength for compressions [25]. Figure 2 shows an example of a participant training with Virtual Life Support while using a Meta Quest 2 and Resusci Anne mannequin.

Fig. 2

Individual using the VR module [16]

Research population

The research population consisted of four groups of stakeholders in medical education: physicians working in acute medical care, educational experts, resuscitation instructors, and medical students (end users). The groups were selected to obtain a mix of feedback on medical content, teaching methods, and end user experiences. The module should be suitable for the medical domain, therefore, feedback from experienced physicians and resuscitation instructors was important in this stage. Both are presumed to possess extensive knowledge of ERC/Dutch Resuscitation Council (NRR) guidelines, including high-quality chest compressions. To be eligible for participation, participants had to be a student or employee at the RU or Radboud University Medical Centre (Radboudumc). Furthermore, participants had to be physically able to perform BLS, 18 years or older, and unfamiliar with this specific module.

Study design

This primarily qualitative study was conducted from September 2022 to January 2023 at Radboudumc (University Hospital, Nijmegen, the Netherlands). This research was deemed exempt from the Medical Research Involving Human Subjects Act, according to the METC Oost-Nederland, case number: 2023–16,599 (Institutional Review Board). Informed consent was obtained from the participants. Participants were informed that they could stop participating at any time. Participants were recruited through purposive sampling to obtain a diverse selection in terms of profession, age, and sex [26]. Participants were invited by email, after which interviews were scheduled. Participants were recruited until thematic saturation was reached, meaning that further data collection would not produce new themes or insights. Participants performed BLS in the VR module under the supervision of the researcher. The test duration ranged from 10 to 15 min.

The use of semi-structured interviews seemed appropriate for exploring expert opinions and end user experiences [27]. Quantitative methods, like Likert scale surveys, were considered but not chosen due to their limited ability to capture detailed, context-specific feedback. They focus on predefined responses, which might not fully address the exploratory nature of our study. Focus groups were considered but not selected to avoid the influence of group dynamics on individual responses. Observational studies were also evaluated but not chosen because they primarily capture behaviours rather than subjective experiences or detailed feedback. Interviews were conducted directly after testing using primarily semi-structured questions, and open-ended questions were subsequently added to clarify the replies (Appendix 2: Interview guide). The interviews had a duration of 10–25 min and were conducted in Dutch. Altered questionnaires designed for examination of VR modules from the Radboud University Health Academy were used as a framework for the interview guide [28]. To obtain specific information on VR BLS, questions were altered based on other qualitative VR BLS studies, including three quantitative questions rated on a 10-point scale where 6 indicated a passing score [8, 17]. For example, “How much did the auditory aspects of the environment involve you?”, transformed to: “How did you experience the spoken instructions?”, and “How much did your experiences in the virtual environment seem consistent with your real-world experiences?” was transformed to an open-ended question focused on BLS: “How realistic was the simulation of BLS if you compare this experience with a real-world experience?” The interview guide was piloted with two experienced researchers and adjusted accordingly. The interviews were recorded in an audio format and transcribed verbatim by the same researcher; quality was randomly checked by another researcher. After the quality check, the audio recordings were deleted. The transcripts were then pseudonymised, labelled, and stored in a secure folder, with access restricted to authorised research team members only.

Data analysis

The interview transcripts were labelled with pseudonyms. Pseudonymization was necessary to analyse differences in the selected groups and demographics, such as previous VR experience and age. Thematic analysis was performed on the interviews. Two researchers used a combined inductive and deductive approach to code the data, whereby the coding scheme was developed as more transcripts were coded [29, 30]. Codes refer to initial tags or labels assigned to segments of the data. Open coding allowed exploration of themes outside the scope of the interviews This method provided insights into both barriers and facilitators. The deductive approach was used to link specific quotes to topics of interest, such as barriers in the introduction and content. The codes were mapped using ATLAS.ti^® (Scientific Software Development GmbH, Berlin, Germany). The researchers assembled the codes independently and compared the results every three or four transcripts, discussing and rearranging discrepant codes until they reached consensus on the codebook. This allowed for contributions of both researchers in developing the code scheme and systematic reflection.

In this study, reflexivity was integral to our approach. It is important to recognise the influence of authors’ contributions on the results [31]. The first author, a 5th year medical student actively involved in teaching BLS and passionate about education, brought a firsthand understanding of the challenges and nuances involved in training BLS. This personal attribute allowed for integration of practical teaching experiences with clinical relevance. Similarly, the second author who served as a sparring partner, shared a comparable background in BLS education, ensuring discussion and critical reflection within the research process. Researchers’ qualifications allowed engaging with participants, promoting trust and transparency throughout the study. The third and fourth authors provided oversight for the analysis and interpretation of the data. Further details about the authors are available in the authors’ information section. By acknowledging the authors’ backgrounds [32], the study aimed to maintain reflexivity, ensuring integrity.

After coding the transcripts, the codes were mapped to capture patterns throughout the data collection, this process facilitated the development of potential themes. The research steering group (consisting of the authors) reviewed the quality of the potential themes as related to the research question [33]. The researchers discussed the relevance of barriers based on their usefulness, the strength of the evidence supporting them, and coherence. After discussing barriers in the research steering group, the topics were subdivided into must-have and nice to have software features. If a barrier significantly hindered or obstructed the effective use of Virtual Life Support in the medical domain, it was categorised as a must-have feature. If the barrier was less essential and did not impede core functionality, it was categorised as a nice to have feature. Facilitators are discussed within the relevant themes, although the primary focus was on barriers. They were included to provide a balanced and representative presentation of the data. Claims were supported using illustrative quotes from the participants, and the included quotes were translated from Dutch to English.

Results

Seventeen interviews were conducted during an eight-week period, involving four physicians, four educational experts, five resuscitation instructors, and four medical students. Thematic saturation was believed to be reached after sixteen interviews, but a seventeenth interview was conducted to confirm that no new themes or insights would emerge. The seventeenth interview did not reveal any additional themes or insights. The researchers carefully selected the quotes that best summarised the barriers. Other results are included in Appendix 3.

Among the participants, four individuals (indicated with asterisks) used version 1.2 of Virtual Life Support, while the remainder used version 1.1. Updates in version 1.2 are described in Appendix 1. Any relevant differences in opinion between participants who used version 1.1 and participants who used version 1.2 (marked with *) are described in the corresponding theme. The same applies to differences between participant groups. None of the participants reported experiencing simulation sickness.

The interviews were transcribed and coded using open coding, as stated in the Methods section. The descriptive codes were categorized into four main themes: introduction, content, applicability, and acceptability/tolerability. These themes were derived from data and represent comprehensive categories that encapsulate the participants’ perspectives and experiences. The quantitative data are placed under the corresponding themes to complement and enhance the understanding of the qualitative data.

Demographics

The participants’ demographic data are displayed in Table 1, age is displayed in age categories to ensure participant anonymity. 41% (n = 7) were female. The average age was 33.1 years, and 70.6% had prior experience with VR. The average years of experience with BLS was 7.8: 21.8 in group A, 3.8 in group B, 3.4 in group C, and 3.3 in group D.

Table 1 Demographics

Theme 1: introduction

In this theme, participants’ opinions and experiences regarding the introduction are explored, focus for this theme is on whether the steps in the introduction adequately prepare participants for the scenario. Authors reflected on the meaning of participants’ feedback while considering their experience in providing instructional guidance and BLS training.

Intuitive introduction

Participants (n = 9) found the introduction of Virtual Life Support to be intuitive and easy to follow, particularly in the end user group. Some participants (n = 7) suggested that verbal instructions, rather than written, would enhance the user experience. B3: “It would be nice if the introduction were audio guided. I would like to hear what I should do, something like: push the blue button. Actual instructions.”

Difficulties with instructions

A few participants (n = 3) encountered difficulties reading the instructions. Participants older than 50 and those without prior experience with VR had difficulties following the instructions during the introduction. The most frequently mentioned instructional issue was that during the calibration process, some participants (n = 4) had difficulty understanding how to confirm the area they were asked to draw during the introduction, others found it not difficult at all. In this step, participants are instructed to define a virtual area within which they can safely move, minimizing the risk of collision with physical objects.

A2: “It worked very intuitively. What I didn’t know was that you can outline the play area and move in and out of it. I find this quite nice.”

Mannequin positioning and MR integration

Proper positioning of the mannequin is essential for accurate hand placement on the mannequin during compressions, according to physicians. Participants suggested that the use of MR could be useful in proper positioning of the mannequin, they suggested that it would be nice if they could see the physical mannequin while placing it within the virtual frame.

Participants who reported that they successfully calibrated the module, mentioned that the mannequin moved during the scenario (n = 9), which makes it difficult to perform compressions correctly.

C4: “I noticed that, especially in the smaller mannequins…this is also the case in real-life…that the mannequin moves during the resuscitation.”

Theme 2: content

Quantitative data

Participants rated the content with a mean score of 7.2 (range 6–8; SD 0.7).

Qualitative data

In this theme, participants’ opinions and experiences regarding the correctness of the content were explored. Authors reflected on participants’ opinions while considering their background and knowledge of BLS. The central question in this theme is: ‘’Is the content of the module accurate and effective in teaching BLS?’’ Participants mentioned that certain steps in the BLS protocol were not sensitive or specific enough. They identified barriers in checking their surroundings, picking up electrode pads, and ensuring correct pad placement. In general, participants (n = 12) mentioned that they were satisfied with the steps included in the protocol, the way the steps were presented, and their ability to adequately perform the steps.

C4: “All the colours were visible and, in my perception, coherent. Yes, you could see things that were stationary, and things in motion, and when you turned your head, everything moved accordingly.”

C1: “It aligns quite well with the actual protocol.”

Regarding electrodes

A3*: “I wanted to pick up the electrode pads, but this did not work the first few times. This was out of sequence.”

D4: “You just moved your hand over the chest, and they automatically stuck in the right spot.”

Missing elements in BLS protocol

Most participants (n = 12) mentioned that Virtual Life Support lacked specific attention to rescuing breaths. Another item that lacked specific attention was performing BLS without the use of an AED. Participants who used version 1.1 reported that they would have benefited from being required to know the appropriate steps even when an AED was not immediately available. This was the case in version 1.2, no one in this version mentioned something about the time until AED’s arrival. Additionally, participants felt there should have been someone in the VR scenario who brought the user the AED. Furthermore, instructions provided by the AED were out of sequence, resulting in an insufficient or excessive amount of time to perform 30 compressions.

C5*: “Sometimes, the AED did not say ‘give two rescue breaths’ even if I was already at 30 compressions.”

Emergency Services Call

There was one barrier that did not bother participants but was noticed. Participants (n = 4) noted that 112, the emergency services number in the Netherlands, was called too late and said they would normally call 112 after checking for a response (during the scenario, a telephone appears which the participant can use to call 112).

Theme 3: Applicability

Quantitative data

Participants rated the usefulness of Virtual Life Support, with a mean score of 7.6 (SD 1.1).

Qualitative data

In this theme, participants’ opinions and experiences regarding the module’s applicability were examined, specifically its usefulness for medical students. Participants (n = 11 including all educationalists) indicated that the module added value to teaching BLS skills to students for multiple reasons, including its realistic feel and the sense of urgency it created.

B2: “Yes, I found it to be highly realistic. I also felt a certain sense of urgency. I experienced a distinct sense of urgency, a feeling that I really needed to take action.”

Self-sufficiency and robot dispatcher

Participants mentioned that working with Virtual Life Support was enjoyable. Some stated that VR has added value because it isolates users from the environment/provides fewer distractions (n = 3) and/or is realistic (n = 9). Some participants would not recommend using this module as a complete substitute for the current repetition module. They said Virtual Life Support could enhance familiarity with the protocol, however, the technical aspects need improvement before implementing this module in the curriculum.

C1: “You learn and store some new information, which you can use for potential real-life experiences.”

Theme 4: Acceptability/tolerability

Quantitative data

Participants rated user satisfaction of Virtual Life Support as 6.6 (SD 1.6).

Qualitative data

In this theme, the acceptability and tolerability of introducing the module as perceived by the participants into the current curriculum were explored. Two barriers were mentioned. Participants (n = 9) found the module to be user-friendly and/or intuitive in general.

A1: “It operates in a very intuitive way, allowing you to see where you need to click. And that works quite well.”

Accuracy of depth measurements

Participants experienced frustrations and raised concerns about the accuracy of the depth measurements taken by the AED. Some users (n = 3) noticed that deliberately moving their heads up and down resulted in better scores. Educationalists considered this to be a significant flaw in the module, as it allowed for potential manipulation of the results. B4*: “There is a problem with the sensitivity for measuring depth, which is disruptive because it doesn’t train the actual skill.”

Module goal

Participants mentioned that the goal of the module was unclear. It was unclear whether they should follow the protocol at their own pace and whether they were tested or trained. A4: “The aim of the module must be clear…if it is testing, you should not provide feedback during testing. Just tell the score at the end.”

Training options

All medical students mentioned that they appreciated practicing the whole BLS protocol. D3:: “I forgot some steps. It was nice to practice with this module because I was forced to do the steps without skipping any.” Participants (n = 7) suggested that introducing variations in the scenarios, such as training without an AED, dealing with difficult bystanders, and incorporating different arrival times for paramedics, would enhance Virtual Life Support’s interest and effectiveness.

Must-have or nice to have

The steering group discussed which barriers need to be fixed to enable the use of Virtual Life Support in the medical domain. The barriers were divided into must-have and nice to have features (Table 2).

Table 2 Overview of barriers in virtual life support, categorised as must-have or nice to have features

Discussion

The objective of this study was to assess the feasibility of adapting an early version of a VR BLS module, called Virtual Life Support, for the medical field, focusing on accessibility, incorporating real-time feedback, and simulating realistic scenarios. This VR BLS module involves cocreation by software developers and stakeholders in medical education. Virtual Life Support was evaluated in a primarily qualitative study with integrated quantitative elements using semi-structured interviews. We developed four themes: introduction, content, applicability, and acceptability/tolerability. Processing the identified barriers labelled as must-have features will be crucial for further validation of the module, while the yield of solutions deemed nice to have features could contribute to optimization for broad use in the medical field.

Facilitators such as the added value of VR to current training methods, and its ability to create a sense of urgency suggest that VR BLS is a potentially powerful educational tool for medical students and other healthcare providers. The findings of this study confirm that improved accessibility is a significant advantage of VR, which is in line with the findings of other research [6]. Various studies have reported that VR modules can be lifelike and create a safe but realistic learning environment [6, 34, 35]. Such realistic simulations result in improved attitudes toward BLS [15, 25]. Participants in this study found the VR module enjoyable and realistic; they felt the urgency of the situation. A VR BLS improves attitudes toward BLS and therefore improves learning efficiency.

The real-time feedback in Virtual Life Support enabled the outlining of barriers to the feasibility of the module, and concrete points of improvement and facilitators were suggested by the participants. The accuracy of real-time feedback from module to user is a challenge that has been both experienced in this research and outlined in other studies [36]. Other studies either used a computer connected to a mannequin and headset, which hinders accessibility, or did not measure the depth [6] or rate of compressions and/or rescue breaths at all, thereby leading to incomplete feedback [7]. Compressions and rescue breaths are essential elements of the BLS protocol, and feedback on these elements is necessary to refine BLS skills. Feedback can be provided during the scenario or afterwards. Providing feedback during the scenario gives users the opportunity to improve skills during the scenario but could be distracting. Providing feedback afterwards enables users to process the feedback, but users must then repeat the scenario to implement the feedback. Virtual Life Support uses a combined method whereby the Zoll AED 3 provides feedback during the scenario, which is in line with reality, and an overall review is given between the scenarios. In this way, the advantages of both methods are optimally utilised.

Strengths and limitations

Stakeholders with various backgrounds participated in this study. Thematic saturation was reached after sixteen interviews, and the seventeenth interview was subsequently added to ensure thematic saturation. Two researchers coded and analysed the data independently, comparing codes every 3–4 interviews to discuss barriers and to minimise inaccuracies by systematic reflection. The researchers used a constant comparative approach to ensure the best possible selection of quotes and codes. Additionally, all themes and barriers were discussed thoroughly within the steering group to reflect on their significance and implications. Therefore, it is unlikely that barriers were missed or categorised incorrectly. As a result, it is expected that all must-have features were identified. We do acknowledge that although barriers are relatively semantic, the must-have and nice to have features are latent and based on interpretation of the research group, this fits within our reflexive approach [32].

By applying qualitative research methods, we developed an understanding of relevant barriers, which led to identification of must-have and nice to have software features to be introduced in the new version of the module. Other studies that explored the use of VR/AR for resuscitation modules primarily relied on Likert scale questionnaires to evaluate their modules [8, 13, 15, 17, 18, 25]. This study could not rely solely on Likert scale questions to evaluate the feasibility of Virtual Life Support, because a deeper understanding of specific barriers was required to adjust such barriers properly. Participants rated the categories as sufficient, even if they experienced barriers. Conducting interviews with multiple stakeholders resulted in broad and interesting points of view, which is a strength of this study.

Virtual Life Support was evaluated in the early stages of development. Stakeholder participation in the development of healthcare products is beneficial for implementation in the real world and for designing and refining interventions [21]. In this context, sixteen barriers were found. The potential to fix barriers was evaluated per barrier by the steering group in consultation with the software developers. The steering group members were excluded from participation as stakeholders to prevent bias. The software developers confirmed that it is realistic to adjust the presented barriers in an improved version of Virtual Life Support. The software developers who contributed their expertise are professionals in the field of VR module development. Therefore, with the confirmation of the software developers, it is expected that it is indeed possible to update Virtual Life Support to create a VR BLS module for medical domains that is easily accessible, provides real-time feedback, and creates lifelike BLS scenarios.

Previous research exploring VR-based BLS without a mannequin indicated skill enhancement was possible, but accurate compression pressure might not be effectively applied [4]. As classroom-based BLS training involves the use of mannequins, we opted against introducing an additional variable, such as training without them. It is unclear whether this module can be used without a mannequin. The effectiveness of Virtual Life Support can be best assessed through comparison with traditional classroom-based BLS lessons, which are the current standard.

Relevance, impact, and recommendations

The results of this study suggest that it is possible to create a VR BLS module for the medical domain that is easily accessible, provides real-time feedback, and creates lifelike BLS scenarios. This study can be used as a framework to evaluate other BLS VR modules. The use of properly tested VR BLS modules as an education tool is encouraged. VR assists students in readiness for real-life BLS scenarios and has the potential to enhance BLS teaching methodologies. Evaluation of an updated version of Virtual Life Support, in which the must-have features have been addressed, is necessary before testing can occur in larger groups of end users. For future research, validating our module’s effectiveness without a mannequin might be interesting. However, it is necessary to prove Virtual Life Support used in combination with a mannequin is effective, before testing with alternatives such as air-cushions or foam-cushions can be conducted.

This study showed that using interviews, instead of quantitative surveys, provides specific data on initial perceptions and feasibility barriers in VR modules. While qualitative methods like focus groups and observation studies were considered for further research, they were not selected due to their higher participant burden and lower scalability. It is therefore advised that primary assessment of VR modules occurs through qualitative methods employing semi-structured interviews, rather than Likert scale inquiries. This approach allows for a deeper exploration of expert and end user experiences, revealing more comprehensive insights into barriers and facilitators. Subsequently, follow-up could utilise open-ended questionnaires, which offer scalability and low participant burden. Large-scale end user testing can be conducted to identify specific errors that might not manifest in small groups. Future studies with VR BLS modules should initially involve testing with stakeholders and end users to identify potential issues. Following stakeholder testing, large-scale end user testing can be conducted to detect any specific errors and thoroughly explore user experiences.

Conclusion

Our research suggests that VR BLS can be used in the medical domain as a valuable educational tool for training medical students on BLS skills. Further development of the VR BLS such as evaluated in this study, is believed to be useful. By interviewing experts and end-users, our study provides in-depth data and insights into topics that might have been missed if a quantitative method was used. Notably, the students in this study were excited to work with a new and innovative learning method that gave them the opportunity to practice their BLS skills. An updated version of the module will be developed and evaluated in a follow-up study.