With the deepening development of globalization, increased population mobility, accelerated urbanization, and changes in the ecological environment, the frequency, scope, and impact of major emerging infectious diseases (MEIDs) in the world have increased in recent years [1]. MEIDs such as SARS, H1N1 influenza, Middle East respiratory syndrome (MERS), Ebola virus disease (EVD), and COVID-19 have posed significant threats to human health and social stability [2]. Research has indicated that the world will continue to face the growing and increasingly severe challenges posed by MEIDs [3]. Therefore, global preparedness for MEIDs is urgently needed, requiring heightened vigilance, strengthened cooperation, and effective preventive and response measures from every country and region.

Due to their distinctive physiological and psychological characteristics, adolescents exhibit more complex susceptibility to and transmissibility of MEIDs compared with adults, highlighting their vulnerability in public health emergency preparedness [4]. Adolescents spend most of their time in schools characterized by large populations and complex hierarchical structures, alongside frequent collective activities and numerous public spaces on campus. If there is a sudden outbreak of MEIDs, the spread rate is often extremely rapid. It seriously affects the progress of teaching, disrupts the orderly daily management of the campus, and potentially poses serious threats to the lives and safety of individuals within the school [5]. Meanwhile, it may cause widespread psychological panic and social disorder. Currently, the emergency preparedness among adolescents in many countries and regions remains deficient, which is mainly manifested by the lack of disease prevention knowledge, weak response capabilities in emergency situations, and insufficient protective measures and coping resources, which aggravate the threat of MEIDs to adolescents [6]. Therefore, enhancing the emergency preparedness among adolescents for MEIDs is a crucial task in safeguarding the health and social stability of adolescents. It is also an essential condition for achieving the Sustainable Development Goals related to health and well-being [7].

At present, education on emergency preparedness for adolescents is mainly based on traditional classroom teaching, brochures, and lectures [8]. However, surveys indicate that these methods often lack interest, interaction, and practicality, and thus struggle to attract the interest of adolescents [9]. The effectiveness of this education is poor, and the ability of adolescents to respond to MEIDs could not be effectively improved [9,10]. Additionally, the uneven distribution of educational resources means that adolescents in remote and impoverished areas often cannot access high-quality educational services, resulting in problems of educational equity [11].

In recent years, the rapid development of the information technology industry has continuously highlighted the application advantages of the virtual interactive training model, opening new avenues for enhancing emergency preparedness. Virtual interactive technology has been used for mass casualty education and military emergency training as part of urban or national emergency preparedness protocols [12,13] and to promote physical activity to reduce stress [14-16]. Notably, the health care industry has actively carried out virtual interactive training in emergency preparedness for MEIDs [17-19], demonstrating its efficacy in improving the emergency response capabilities of nursing students, medical workers, and emergency management personnel. This is supported by a study within the National Health Service (NHS) in the United Kingdom, which found that virtual reality (VR) training significantly improved the performance of frontline health care workers in infection control measures and personal protective equipment usage during the COVID-19 pandemic compared to traditional training methods [20]. Similarly, a systematic review and meta-analysis confirmed the positive impact of virtual experiments on learning effectiveness across various educational contexts [21]. However, a significant research gap remains. Previous interventions have primarily targeted professional populations within the health care or military sectors. There is a paucity of studies focusing on adolescents in general educational settings. Moreover, existing programs for the public often lack a comprehensive and engaging design that integrates knowledge acquisition, decision-making training, and behavioral simulation into an immersive, storyline-driven experience [22]. This limits their potential to effectively captivate and educate the adolescent demographic.

To fill this gap and better equip adolescents to cope with outbreaks of MEIDs, we designed and developed a virtual interactive system, “People’s War Against Pandemic” [23]. This system allows adolescents to learn essential knowledge and response skills through simulated outbreak scenarios within a narrative framework. This study aimed to test the effectiveness of this virtual interactive system in improving emergency preparedness for MEIDs among adolescents in Luzhou City, Southwest China. Here, we report the results of this study.

Embodied cognition theory posits that cognition arises from the dynamic interaction between the individual, the environment, and tools, and is grounded in sensory and bodily experiences, with perception serving as the basis for action [24]. In this study, the immersive, interactive, and imaginative features of the virtual system foster a deep sense of presence, enhancing emotional engagement. This affective response, mediated by presence, can restructure cognitive frameworks and ultimately promote behavioral change. Furthermore, the system supports the development of new body schemas [25], enabling the projection of real-world motor actions onto sensorimotor experiences within the virtual 3D environment [26]. Consequently, the activation of sensory and motor systems in this embodied context elicits bodily presence, linking physical actions to emotional and cognitive processing, which guides behavior and, in turn, enriches emotional experience, forming a cyclical feedback loop among emotion, cognition, and action (Figure 1).

This study was a single-blind, 2-arm, parallel-group randomized controlled trial (RCT). The CONSORT (Consolidated Standards of Reporting Trials) checklist was used when writing this report (Checklist 1).

The sample size was calculated following a 2-mean test randomized controlled design: n₁ = n₂ = 2[(t_α/2 + t_β)S / δ]². The primary indicator (emergency preparedness for MEIDs) was used to calculate the sample size for this study. A 2-sided test was used to set the test level (α=0.05) and the test efficiency (1 − β = 0.8); then t_α/2=1.96 and t_β=0.84. The SD (S=16.5) and the expected effect size (δ=9.5) were derived from an internal pilot study conducted by our research team prior to the main trial (n=12 adolescents recruited from the same target population). Based on these parameters, the required sample size was calculated as n₁=n₂=48. The total required sample size was 96 participants. Accounting for a 20% dropout rate, the sample size was calculated to be at least 60 participants in each group. It is important to note that this sample size is adequate for testing efficacy under controlled conditions but may not be representative of the entire adolescent population.

This study used a single-center design, with participants recruited from a secondary vocational school in Luzhou, Sichuan. Thereafter, they were randomized into either a control group or an intervention group. This site was selected due to its accessibility and established collaboration, which facilitated the implementation and monitoring of the RCT. Although this approach enhanced the internal validity of the study by controlling for institutional-level confounding factors, it limits the generalizability of our findings to adolescents in other geographic, cultural, or educational settings. Therefore, the results should be interpreted as preliminary evidence of the intervention’s efficacy within this specific context.

The eligibility criteria for participation were as follows: (1) aged 10 to 19 years and (2) informed consent and voluntary participation in this study (participants below 18 y of age required the consent of the guardian). The exclusion criteria were as follows: (1) participants did not cooperate in completing the questionnaire and (2) participants were participating in or had participated in similar studies.

The sociodemographic questionnaire, the EPQ-MEIDs, and the IDSHL Scale were administered at baseline (T1), 2 weeks of intervention (T2), 4 weeks of intervention (T3), and 2 weeks after the intervention (T4). The SUS was administered at 2 weeks of intervention (T3). The baseline data of the participants were collected after the participants or their guardians signed the informed consent. The researchers who collected the data received standardized training.

The data were analyzed using SPSS (version 26.0, IBM Corp). The Shapiro-Wilk test was used to assess the normality of the distribution and the homogeneity of variances. Frequencies, percentages, means, and SDs were used for statistical analysis. The 2-tailed t test, Mann-Whitney U test, or chi-square test, as applicable, was used to compare the demographic characteristics of the participants. The overall comparison of outcome indicators between the 2 groups across 4 time points was conducted using repeated-measures ANOVA, and an independent-samples t test was used to test baseline differences in each index. Significance was defined as P<.05.

A random number table was generated using Excel 2016, and the numbers were divided into the intervention and control groups with an allocation ratio of 1:1. The random numbers were then placed in opaque, sealed envelopes and kept by people who did not participate in the study. Therefore, the allocation sequence was concealed from both participants and researchers at the time of enrollment. For each eligible adolescent, the researcher randomly selected one of the envelopes (for each envelope selected by the researcher, the total number of envelopes would be reduced by 1) and allocated them to the intervention or control group. Participants used a code instead of their real names to conceal allocation information, so that the researchers could not know the grouping of the participants. However, due to the nature of the interventions (virtual interactive system vs traditional lecture), participants became aware of their group assignments upon commencement of the intervention.

Given the overtly different nature of the 2 interventions, blinding of the participants and the trial implementers was not feasible. Therefore, only the data collectors and analysts were blinded; that is, the data collectors were not clear about the grouping, and the data analysts obtained only data marked as A and B groups.

The intervention sessions for the control and intervention groups were scheduled at different times (eg, on different days or in different time slots on the same day) to minimize opportunities for immediate interaction and discussion about the group-specific activities. All participants were explicitly instructed at the beginning of the study not to discuss the specific content of their training sessions (especially the virtual interactive system) with students from other groups until the entire study was completed. The importance of this for the validity of the research was emphasized. The researchers managing the virtual interactive system for the intervention group were different from those primarily involved in data collection, further reducing the risk of inadvertently mentioning the system to control group participants.

The trial was conducted in accordance with the Declaration of Helsinki, approved by the Human Research Ethics Committee of the Affiliated Hospital of Southwest Medical University (KY2022001), and registered with the Chinese Clinical Trials Registry (ChiCTR2200060919). Once selected, the participants signed an informed written consent form, including explanations about the study’s purpose and research participation conditions. All information shared by the participants was kept confidential. A gift was given to participants after completing daily tasks and questionnaires every week, and participants who participated in the whole study were rewarded with 50 RMB (US $7.5).

A total of 120 participants were recruited and randomly allocated to the intervention group (n=60) and the control group (n=60). All 120 participants completed the 6-week assessments. Ultimately, the analysis comprised 120 participants. Figure 3 shows the CONSORT flowchart of the recruitment process.

The demographic information of the participants is presented in Table 1. The mean age was 17.60 (SD 0.56) years for the intervention group and 17.65 (SD 0.52) years for the control group. The following variables did not demonstrate significant differences between the 2 groups: sex, age, ethnicity, family economic income, or place of residence, indicating that the groups were homogeneous.

As shown in Table 2, at baseline, there was no significant difference in the average score of emergency preparedness for MEIDs between the 2 groups (P=.55). However, the average emergency preparedness score for the intervention group at T2, T3, and T4 was higher than that of the control group. The difference was statistically significant (P<.001).

The data were tested for normality and met this assumption. However, Mauchly test of sphericity indicated a violation of the sphericity assumption (W=0.887; P<.05). Therefore, we applied the Greenhouse-Geisser correction to adjust the degrees of freedom and reported the corrected values. The results were as follows—group: F_1,118=109.186 (P<.001; ηp²=0.481), time: F_{2.786,328.790}=129.692 (P<.001; ηp²=0.524), and time × group interaction: F_{2.786,328.790}=23.476 (P<.001; ηp²=0.166), indicating that the interaction effect between time and group on emergency preparedness scores was significant.

The results of the simple effects analysis (Table 3) showed that the emergency preparedness scores of the intervention group differed significantly over time (F_3,177=117.890; P<.001; ηp²=0.666). This demonstrated that the virtual interactive system training had a substantial effect, which improved markedly from baseline to week 4 and was maintained at the 2-week follow-up. Although the emergency preparedness scores of the control group also differed significantly across time (F_{2.508,147.974}=26.136; P<.001; ηp²=0.307), there was no significant difference between T2 and T1 (P=.97, which is >.05), indicating that the emergency preparedness scores increased slowly in the control group, and that the effect was weaker than that in the intervention group.

As shown in Table 4, at baseline, there was no significant difference in the average score of health literacy between the 2 groups (P=.40). However, the average health literacy scores for the intervention group at T2, T3, and T4 were significantly higher than those of the control group; the difference was statistically significant (P=.01, P<.001, and P<.001, respectively).

The 2 sets of data were tested for normality, and the data satisfied Mauchly sphericity test (W=0.943; P>.05). The results were as follows—group: F_1,118=18.480 (P<.001; ηp²=0.135), time: F_3,354=35.047 (P<.001; ηp²=0.229), time × group interaction: F_3,354=5.652 (P<.05; ηp²=0.046), indicating that the interaction effect between time and group on health literacy scores was significant.

The results of the simple effects analysis (Table 5) showed that the health literacy scores of the intervention group differed significantly across time (F_{2.587,152.620}=34.896; P<.001; ηp²=0.372), indicating that the virtual interactive system training had a substantial effect on improving infectious disease-related health literacy. The scores increased markedly from baseline to week 4 and were maintained at the 2-week follow-up. For the control group, health literacy scores also differed significantly across time (F_3,177=6.235; P<.001; ηp²=0.096). Although there was a significant increase from T1 to T2 (P=.04), no significant difference was found between T3 and T2 (P=.16), indicating that the effect of traditional lectures on improving adolescent health literacy was relatively slow and weaker than that of the virtual interactive system.

The total average usability evaluation score of the virtual interaction system by adolescents was 85.75, as shown in Table 6. The item with the highest score on the scale was item 1, “I will be willing to use (login/access) this system frequently,” and the item with the lowest score was item 6, “I thought there was too much inconsistency in this system.”

Historical pandemics have demonstrated that adolescents are not only potential vectors for pathogen transmission but also a critical force in epidemic prevention and control [32]. The participation and cooperation of adolescents in epidemic prevention and control are key factors in curbing the spread of the epidemic [33]. Although health departments and government agencies play a central role in formulating response strategies and providing medical services, these efforts are likely to be in vain without the active response of adolescents. Therefore, effective emergency preparedness for adolescents can significantly reduce the risk of transmission during MEID outbreaks [34].

The results of this study suggest that virtual interactive system training can enhance adolescents’ emergency preparedness more effectively than traditional lectures alone. This finding aligns with previous research conducted among nursing students and health care workers, which also found that virtual simulation training significantly improved emergency response skills and knowledge retention [19,35,36]. While a direct statistical comparison is not feasible due to differing outcome measures, the magnitude of improvement observed in our adolescent population appears substantial. For instance, the intervention group’s emergency preparedness score increased by approximately 20% from baseline to the end of the intervention, a rate of improvement that is comparable to, and in some cases exceeds, the effects reported in studies with professional groups using similar simulation-based training [37]. This suggests that narrative-driven virtual interactive systems could achieve meaningful learning outcomes in a general adolescent population, not just among motivated health care trainees.

Our study extends this evidence by demonstrating that similar benefits can be achieved in a nonprofessional, adolescent population using a narrative-based, publicly-focused platform, a finding corroborated by other research highlighting the efficacy of VR in creating engaging learning environments without physical space restrictions [38]. This can be attributed to the interaction of various factors. On the one hand, “People’s War Against Pandemic” provided rich environmental stimulation through virtual pandemic life scenarios. Moreover, its multidomain task modes offer more targeted stimulation to participants, granting them comprehensive sensory experiences. This helps them familiarize themselves with and master quarantine norms as well as basic self-help and mutual rescue skills. On the other hand, the effectiveness of the system can be explained by the theory of embodied cognition. Unlike passive learning in lectures, our virtual system provides a highly immersive environment that promotes embodied learning. Users learn by doing and making decisions within a simulated context. This approach has been shown to enhance situational awareness, deepen cognitive understanding, and facilitate the transfer of knowledge into behavioral intention and change [39]. Furthermore, the immersive nature of VR and augmented reality systems has been specifically linked to increased user engagement and improved comprehension levels in educational interventions [38], while also being effective in developing empathy through perspective-taking, which could be crucial in fostering compliance with public health measures that protect others.

In this study, after just 2 weeks of training in the intervention group, the emergency preparedness score was close to that of the control group after 4 weeks of education. This may be because the comprehension of scientific knowledge about infectious diseases in lectures was difficult to understand, requiring sustained stimulation over a longer period before effects began to appear [40]. It is evident from this that “People’s War Against Pandemic” could enhance adolescents’ emergency preparedness more rapidly compared to lectures, highlighting its application advantages.

However, it is important to interpret these findings with caution. In this study design, the intervention group received approximately 80 minutes of total educational exposure (a 40-min lecture + 40-min virtual system use), while the control group received only 40 minutes of lecture-based education. Therefore, the observed superior outcomes in the intervention group may be partially or wholly attributable to the increased time-on-task rather than any inherent superiority of the virtual interactive system as a pedagogical modality. This confounding variable precludes a direct comparison of the relative efficacy of the 2 instructional methods under time-equivalent conditions.

Studies have indicated that enhancing IDSHL among adolescents can help alleviate their panic and maintain social stability [41]. Therefore, IDSHL is regarded as one of the critical indicators for evaluating the effectiveness of health education on infectious diseases [42].

The results of this study found that traditional lecture education had a slower effect, while “People’s War Against Pandemic” boasts a more significant advantage in enhancing adolescents’ IDSHL. Jacque et al [43] integrated infectious disease knowledge into a school curriculum and found improvements in health literacy, albeit to a lesser extent than observed in our study. This difference in efficacy may be attributed to the pedagogical approach. While their study used a traditional classroom-based curriculum, our interactive system provided experiential learning through simulation, which is posited to lead to deeper cognitive processing and engagement [21]. The meta-analysis further supports that the impact of virtual experiments on learning effectiveness is significant and can be influenced by factors such as the design and implementation of the virtual experience [21], which aligns with our system’s thoughtful integration of multiple interactive modules. First, most traditional lectures are static, graphic-and-text-based education, lacking interaction [44]. However, “People’s War Against Pandemic” provides opportunities for real-time feedback and dynamic participation, enhancing the learning interest and enthusiasm of adolescents and promoting their active involvement in the practice of infectious disease prevention and control. Second, the storyline and task design of “People’s War Against Pandemic” are continuous and logical, which can progressively enhance adolescents’ health literacy regarding infectious diseases, from basic theoretical knowledge to specific operational practice. Finally, neuroscientists believe that virtual systems involving task failures will lead to emotional arousal, promoting an increase in noradrenaline levels in the amygdala, which helps reinforce the process of memory formation and stabilization. This is beneficial for the retention of knowledge [45].

Furthermore, the observed plateauing of scores at follow-up warrants careful consideration. While common in educational and behavioral interventions, as the initial intensity of the training fades, it highlights the potential need for “booster” sessions or the integration of such virtual training into regular school health curricula to sustain its effects. This pattern underscores that emergency preparedness is not a one-time educational event but a skill that may require periodic reinforcement. Future studies should prioritize long-term follow-up assessments (eg, 6 mo, 1 y) and, where ethically and logistically feasible, use behavioral measures or simulations closer to real-world scenarios to evaluate the true translational impact of the virtual interactive system on actual emergency response behaviors.

The average SUS score in this study was 85.75, indicating that this system had excellent usability [46]. The results showed that item 1 scored the highest, followed by items 7 and 9, suggesting that the system’s design was intuitive, the user interface was friendly, and the operation process was straightforward, fulfilling users’ needs and expectations. Salamah et al [47] pointed out that user satisfaction with a system is a key factor driving their intention to revisit. The innovation of the system, the practicality of the content, and the ease of use are the 3 direct factors affecting user satisfaction. Item 6 (“I thought there was too much inconsistency in this system”) received the lowest transformed contribution score, identifying perceived internal consistency as the relatively weakest aspect of the system and the most useful target for future refinement. Nonetheless, the overall SUS score of 85.75 indicates excellent usability.

First, this study was conducted at a single vocational school in Luzhou, Sichuan, which may limit the generalizability of the findings to adolescents in other regions or educational contexts. However, following the Spatial Statistic Trinity (SST) framework, we have provided comprehensive details to enhance transferability.

Regarding population characteristics, our participants were adolescents aged 16 to 19 years (mean 17.60, SD 0.56) from a lower-middle-income region of China, with demographic profiles typical of vocational school populations in Southwest China (eg, 107/120, 89% female participants; 96/120, 80% rural areas). These characteristics are relevant to similar socioeconomic and educational settings.

The study used a randomized controlled design with concealed allocation, standardized intervention protocols, and complete follow-up (0% attrition), which strengthens internal validity and supports replication in analogous study contexts.

For estimation methods, the study used repeated-measures ANOVA with Greenhouse-Geisser correction where appropriate, reported effect sizes (partial η²), and provided raw data for key outcomes. Transparent reporting of these methods allows researchers to compare our estimates with those from their own settings.

The interaction of these 3 elements suggests that our findings are likely transferable to similar adolescent populations in comparable educational settings (eg, vocational schools in other Chinese cities with similar sociodemographic profiles). However, transferability to dissimilar populations (eg, younger adolescents, different cultural contexts, or nonvocational school settings) requires further empirical testing.

Although the sample size was statistically adequate for this RCT, a larger and more diverse sample would further improve the representativeness of the results. Second, the intervention and follow-up periods were relatively short, which precludes conclusions regarding the long-term efficacy of the virtual interactive system. Future multicenter studies with extended follow-up durations are recommended to validate these preliminary findings and further explore the transferability of the intervention across diverse populations and settings.

The findings of this study demonstrate that, compared with traditional lectures alone, the addition of the “People’s War Against Pandemic” virtual interactive system to lecture-based education led to significantly greater improvements in self-reported emergency preparedness and infectious disease health literacy among adolescents. These results suggest that the system effectively enhanced participants’ perceived readiness and knowledge regarding MEIDs. However, as the outcomes were measured using self-report questionnaires, conclusions regarding actual behavioral changes or objective decision-making skills cannot be drawn from this study. Future research should incorporate objective behavioral assessments or simulated emergency scenarios to evaluate the translation of these self-reported improvements into real-world response capabilities. Due to the unequal total educational time between groups, the observed effects may reflect increased learning time rather than the inherent superiority of the virtual modality. Future research using time-controlled designs is needed to isolate the specific contribution of virtual interactive training. The system has a good user experience and a certain usability, which was worthy of promotion and use.