Nursing is a practice of caring that involves assessing and addressing human responses to actual or potential health problems [1]. Nurses are required to integrate knowledge, clinical skills, professionalism, and beliefs to provide autonomous and accountable nursing care, and such competencies that are developed through education [2,3]. Although nursing students prepare for their professional roles through theoretical education and clinical practice, the gap between theory and actual clinical settings has been reported to hinder their adaptation to clinical environments and professional performance [4]. Thus, it is essential to develop and implement practical training programs that provide direct experiences in environments similar to real clinical situations.

Simulation−based education is one strategy to address these educational needs, providing nursing students with safe and repetitive practice opportunities [5]. However, due to the high costs of equipment and facilities, limited scenarios, and constraints on time and space, it is difficult to provide sufficient opportunities for all learners [6,7]. Virtual reality (VR)-based simulation has gained attention as an alternative to overcome these limitations. By utilizing computer graphics, sensors, and head-mounted displays (HMDs) [6,8], VR can generate realistic virtual environments, enabling repeated learning and providing individualized feedback without time and place restrictions [9,10].

VR technology can be described as either immersive or non-immersive, depending on how strongly it engages the user. Immersive VR delivers a heightened sense of presence and interaction by using devices such as HMDs [11], whereas non-immersive VR allows users to interactively control virtual environments displayed on a monitor using a mouse or controller [8]. These various forms of VR technology have been applied to simulation-based education to replicate realistic clinical scenarios, facilitating learners’ problem-solving and decision−making skills [12].

VR-based simulations can also be classified by their educational focus into skill-based and situation-based types. Skill-based simulations emphasize the accurate performance of specific nursing procedures, such as nasogastric tube feeding, urinary catheterization, and intravenous injection. In contrast, situation-based simulations focus on non−technical skills, particularly situational awareness, and deal with topics such as sudden changes in a patient’s condition or disaster response [13]. This classification helps to identify how VR contributes to improving either procedural skills or complex situational coping abilities. It also provides an important basis for evaluating the effectiveness of VR in achieving different educational objectives and for developing optimized VR-based educational scenarios.

Recent studies examining the effects of VR-based simulation in nursing education on various outcomes, including nursing skills, knowledge, self−efficacy, and confidence in nursing performance, have reported inconsistent findings [14−16]. The discrepancy in these findings may stem from differences in study design, type of VR, educational content and methods, and the outcome variables measured. Such variations have limited the effective implementation and broader uptake of VR-based approaches in nursing education. In particular, although acquiring nursing skills and knowledge is a fundamental goal of nursing education, the influence exerted by VR-based simulation on these outcomes has been inconsistent, indicating the need for more practical and objective verification.

Therefore, this study aimed to systematically search and analyze studies on VR-based simulation in nursing education published in domestic and international journals, and to conduct a meta-analysis that comprehensively considered the type of VR simulation, educational content, and outcome variables. Through this, the overall effects of VR-based simulation in nursing education are explored, providing evidence to guide the design and development of future VR-based simulation programs.

The methodological approach for this review was informed by widely accepted Cochrane guidance for conducting systematic reviews, and the reporting adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [17,18]. In accordance with recommended practice, this review was also registered in PROSPERO (https://www.crd.york.ac.uk/PROSPERO/view/CRD420251029154).

This study was conducted to examine the effects of VR-based simulation education among nursing students. Based on this, the trends of VR-based simulation studies in nursing education were analyzed. Research on VR-based simulation in nursing education has shown a gradual increase since the first study published in 2016, reaching 10 studies in 2024. This increase can be interpreted as the result of multiple factors acting together, including the advancement and accessibility of VR technology, the growing demand for non-face-to-face education during the coronavirus disease 2019 pandemic, and the need for innovation in nursing education [25].

By study type, non-immersive VR studies (n = 13) outnumbered immersive VR studies (n = 8). Non-immersive studies have steadily increased since 2016, whereas immersive studies have rapidly expanded since 2022. This trend suggests that, while non-immersive VR was initially preferred due to its higher accessibility, the recent development of immersive technologies and the wider distribution of related equipment have led to a sharp increase in immersive VR studies [26]. Analysis by topic type showed that, among skill-based studies, simulator-controlled trials were predominant (n= 9), whereas non-simulator-based studies were limited (n = 3). This reflects a research tendency to compare the effectiveness of VR with conventional simulator-based education. Such control group compositions have an important influence on the interpretation of effect sizes [27]. In contrast, among situation-based studies, non-simulator-based control groups were more common, indicating that VR has often been compared with traditional lecture- or discussion-based education. Therefore, VR has been applied as a complementary tool to simulator-based education in skill-based topics and as a new educational approach to reproduce complex clinical situations in situation-based topics.

According to the meta-analysis results, VR-based simulation showed limited effects on nursing skill performance, whereas it demonstrated beneficial effects on knowledge acquisition. These patterns were further supported by the network meta-analysis, which showed no statistically significant differences among various VR configurations for nursing skill performance, whereas situation-focused VR interventions demonstrated a consistent advantage over simulator-based education for knowledge outcomes.

This suggests that while VR has strengths in promoting cognitive engagement and knowledge internalization [28−30], it may have limitations in transferring tactile sensations, fine motor skills, and muscle memory required for actual skill performance [31].

VR-based simulation offers educational benefits by enabling learners to transform abstract concepts into concrete understanding through realistic visual and auditory experiences [32], and by facilitating problem-solving through repeated decision-making practice in a virtual environment [33,34] However, its current technological constraints limit the accurate reproduction of fine sensory feedback essential for hands-on clinical skills [31].

Among the included studies, some reported significant improvements in both the experimental and control groups after the intervention [35,36], suggesting that it may have been difficult to isolate the independent effect of VR alone [37]. In fact, several meta-analyses that included studies without additional educational interventions in the control group reported that VR was effective in improving skill performance [28,30]. This implies that the learning effect may have resulted more from the structured educational intervention itself than from the inherent effect of VR simulation.

In the subgroup analysis, a statistically significant pooled estimate for nursing skill performance was observed in situation-based VR interventions. However, this finding should be interpreted with caution. The difference between situation-based and skill-based interventions did not reach statistical significance, and when multiple intervention characteristics-including immersion level, educational focus, and comparator type-were simultaneously considered in the network meta-analysis, no VR-based configuration demonstrated a statistically significant advantage over simulator-based education.

These findings suggest that the apparent benefit observed in situation−based topics may not represent a robust or independent effect of VR itself, but rather reflects sensitivity to study design features and comparator characteristics. In particular, the frequent use of non-simulator-based control groups in situation-based studies, in contrast to simulator-based controls commonly used in skill-based studies, may have contributed to the observed variability in effect estimates. Therefore, while situation-based VR interventions may offer contextual advantages for learning, current evidence does not support a definitive conclusion that they are superior for improving nursing skill performance.

Although moderate to substantial heterogeneity was observed for several outcomes, the overall direction of the pooled effects remained consistent after adjustment using the trim-and-fill method. In particular, the knowledge outcome retained its statistical significance following adjustment, while the non-significant findings for nursing skill performance and confidence remained unchanged. These findings suggest that, despite between-study variability, the main conclusions of this meta-analysis are relatively robust to potential publication bias. However, this does not imply that heterogeneity was fully resolved, and the results should still be interpreted with consideration of contextual differences across studies.

This study explored various effects of VR-based simulation on nursing education beyond nursing skill performance and knowledge acquisition. The results showed a positive effect on self-efficacy, whereas no statistically significant difference was observed in nursing performance confidence. These findings contribute to a more comprehensive understanding of the educational utility of VR simulation. First, VR-based simulation education significantly improved self-efficacy, particularly in immersive environments and situation-based topics. This type of simulation enhances learning engagement by stimulating learners’ senses and maximizing a sense of realism as if they were experiencing actual clinical situations [38,39]. The increased sense of realism and immersion strengthens clinical decision-making and coping strategies [40], thereby contributing to the improvement of self-efficacy. In addition, subgroup analyses by control group type showed that VR-based simulation was more effective in enhancing self-efficacy than simulator-based education. This may be because VR can reproduce a wider range of clinical scenarios and enhance interactivity with learners, thereby expanding their overall learning experience [41].

In contrast, VR-based simulation did not show a statistically significant difference in improving nursing performance confidence. A similar trend was observed in the subgroup analysis, which is consistent with the earlier finding that there was no significant difference in nursing skill performance. Some of the studies included in this review modified the performance checklists into self-reported confidence tools for measuring nursing skills [42,43], and such measurement methods are closely related to actual performance ability. Therefore, considering that the improvement in skill performance was limited in this study, the absence of a significant change in performance confidence can be interpreted as a consistent finding. Nevertheless, a significant improvement in nursing performance confidence was observed in the situation-based topics. This suggests that the improvement in confidence through VR simulation is not primarily derived from the enhancement of technical nursing skills, but rather from the direct experience of a learning environment that reproduces complex clinical situations and requires integrative judgment. Nursing students may gain greater educational benefits by practicing independent judgment and decision−making as learners in situations that are difficult to experience directly before entering clinical practice. Therefore, when designing VR simulations to enhance nursing performance confidence, it is important to focus on developing problem-solving and coping abilities applicable to real clinical situations.

This study systematically analyzed the effects of VR-based simulation education for nursing students and provided an integrated perspective on both the educational potential and limitations of VR. Only published RCTs were included in the analysis to ensure objectivity and reliability. In addition, various subgroup analyses-such as those based on the type of immersion and control group were conducted to comprehensively explore the factors influencing the effects. During this process, some findings that contrasted with previous studies were identified, highlighting the limitations of VR and guiding directions for improvement. Furthermore, by providing quantitative evidence for various learning outcomes including knowledge, self-efficacy, and nursing performance confidence, this study provides evidence to inform for future development of tailored VR designs and scenarios.

The limitations of this study are as follows. First, although a comprehensive search strategy was applied, the review process may have been subject to potential selection bias due to the exclusion of non-English publications and unpublished studies. In addition, the number of studies included in some subgroup analyses was limited, which may have introduced uncertainty in estimating the effect sizes.

Second, the included studies showed a high degree of heterogeneity, which may have resulted from differences in intervention types, number of sessions, exposure duration, composition of control groups, and measurement tools. This heterogeneity suggests that contextual factors should be carefully considered when interpreting the pooled results.

Considering these limitations, future research should aim to standardize comparison groups, clarify the dose-response relationship of topics, immersion levels, and exposure intensities, ensure consistent use of objective outcome indicators, and further minimize potential publication bias through strengthened registration and reporting systems.

The meta-analysis of VR-based simulation in nursing education suggested beneficial effects in several domains compared with traditional education. Although no consistent significant differences were found in nursing skill performance or performance confidence, meaningful improvements were observed in knowledge and self-efficacy. In particular, situation-based topics showed positive outcomes in multiple aspects, and immersive VR tended to promote learner engagement and immersion. For educational applications, a blended strategy that appropriately combines VR and simulator-based training according to learning objectives and content is recommended. Given the variability across studies, developing situation-based content using immersive technologies may further enhance clinical decision−making and problem-solving competencies, whereas simulator-based education may remain more effective when the goal is to directly improve psychomotor performance.