Immersive virtual reality based on head-mounted display in medical education: a systematic review

Abstract

Introduction

Immersive virtual reality (IVR) based on head-mounted displays (HMD), can construct virtual learning environments for medical education. This study aims to explore the application and effectiveness of IVR in medical education.

Methods

This review followed the PRISMA guidelines and searched the databases of Web of Science, Springer Link, PubMed, IEEE Xplore, and Science Direct for controlled experimental studies published between January 1 2017 and October 1, 2025 on the application of IVR to medical education.

Results

36 studies satisfied the inclusion criteria. Among them 34 reported significantly better objective learning outcomes and 29 showed more positive subjective feedback. The findings suggest that IVR is particularly effective in areas such as clinical surgery training, anatomy education, medical skill development, and nursing education. IVR demonstrates long-term effects in medical knowledge acquirement.

Conclusions

The systematic review reveals that IVR is particularly effective in practice-based medical education. It highlights that VR demonstrates superior long-term knowledge retention and knowledge enhancement. A comprehensive evaluation of IVR use in medical education requires additional perspectives from teachers, and consideration from the dark side of IVR, i.e., technology-induced stress and time. Future IVR implementation in medical education calls for more efforts to address challenges in user adaptability, cost, privacy and regulatory compliance.

Background

In the information age, Virtual Reality (VR) has emerged as a pioneering tool that is transforming various sectors including education. Immersive virtual reality(IVR), an advanced form of VR, allows users situating in interactive three-dimensional environments through head-mounted displays delivering visual and auditory inputs, as well as controllers equipped with haptic feedback [1]. In contrast to earlier VR systems, it features highly realistic environmental rendering, accurate object tracking, multi-sensory real-time feedback (visual, tactile, auditory) [2]. The World Federation for Medical Education updated its global standards in 2023, underscoring the necessity for medical education to integrate resources such as VR, artificial intelligence, and information technology sercices [3]. Medical training faces two key challenges: limited time for medical practice [4]and ethical dilemmas in high-risk/privacy-sensitive procedures [5]. Situated learning theory suggests that immersive, visually realistic scenarios can enhance the integration of knowledge within socially relevant contexts, enabling IVR to support the transfer and application of learning to real-world situations [6].

With ongoing technological progress, IVR has become more operable and cost-effective, alleviating many early limitations [7]. IVR platforms, such as Immersive Virtual Anatomy Laboratory have reduced task completion time by 18% and sustained retention over two months, directly addressing these challenges [8–10]. The COVID-19 pandemic further accelerated the adoption of VR in medical education, as lockdown measures drove a shift toward remote learning and stimulated research and implementation of immersive VR [11]. Comparative studies demonstrate that HMD-based VR classrooms outperform Zoom in presence, collaboration, and knowledge acquisition [11, 12], offering a methodological template for future evaluations. Increasing evidence indicates IVR strong potential to improve educational outcomes [13].

Nonetheless, assessing the effectiveness of IVR in medical education remains a complex task, which is crucial for establishing its superiority over traditional teaching methods. Unfortunately, many existing systematic reviews have focused on the efficacy of traditional VR systems [14]. In recent years, VR has advanced rapidly, with devices becoming lighter and more affordable, promoting its adoption in medical education [15]. However, the effectiveness and sustained impact of these newly developed applications have not been systematically evaluated. This study aims to address this gap by providing an up-to-date review of the latest evidence on IVR in medical education.

Materials and methods

The present study

This review endeavors to scrutinize the impact of IVR in medical education, examining variations in HMD types, the medical disciplines applied, participant demographics, experimental designs, and the effects of IVR interventions. To achieve this, the review poses the following research questions:

RQ1. What types of head-mounted display are adopted in medical education?

RQ2. Which disciplines of medical education does IVR apply to?

RQ3. What are the characteristics of the participants in the IVR experiments?

RQ4. How to design trials of IVR applied to medical education?

RQ5. Will IVR facilitate medical education?

Inclusion and exclusion criteria

Inclusion Criteria:

1. Studies with focused on medical students and medical staff at any stage of education;

2. Studies with IVR as the main intervention in medical education-related purposes, and HDMs usage is must. The details of how IVR is adopted should be explained.

3. Studies with controlled experimental methods and measures at least one effectiveness outcome, such as participants’ skills, knowledge, self-efficacy, satisfaction, or anxiety.

Exclusion Criteria:

1. Studies published in the form of comments, meeting abstracts, bibliography chapters, news, legal texts or letters;

2. Studies published in a language other than English;

3. Studies are not pertinent to medical education, such as training for patients, or teaching medical knowledge to ordinary people;

4. Studies focus on other influencing factors of IVR in medical education, such as the impact of various models and theories.

Search strategy

This review strictly followed Preferred Reporting for Systematic Reviews and Meta-Analysis (PRISMA) statement. The protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) (number: CRD420251025298). It explored 5 databases, including Web of Science, Springer Link, PubMed, IEEE Xplore and Science Direct from January 1, 2017 to October 1, 2025. VR gained significant public attention in 2016, prompting numerous companies to develop related devices. However, as empirical research on VR applications remained scarce that year, the literature search for this study begins in 2017 to better capture subsequent research trends.

For databases, Web of Science, Springer Link and Science direct are comprehensive databases, while PubMed covers biomedical and health research, and IEEE Xplore focuses on computer science. In addition, references to relevant published reviews were checked to include any studies that might be useful and overlooked.

According to the key issue strategy PICOS (research population, intervention measures, control measures, outcome indicators, research design), after a preliminary database search, the following search formulas were composed: (“virtual reality” OR metaverse OR “virtual environment”) AND (train OR teach OR educat* OR learn) AND (medical*) AND (“controlled trials” OR RCT OR experiment).

Critical evaluation

The search was conducted jointly by two graduate students specializing in Medical Information Management. Both researchers had prior experience of English literature retrieval and received systematic training in the screening process. Prior to the search the two researchers developed an initial framework for the screening with the first author. In cases where disagreements arose regarding certain studies, the discrepancies were solved by the third reviewer (the first author). Of the 947 studies identified, 83 duplicate studies were deleted, and 139 studies were ultimately selected for full-text screening. Finally, a total of 36 studies were included for the review. Figure 1 shows the details.

Fig. 1.

Flow Chart of the Screening Process

Synthesis method

The included studies were limited to controlled trials. Substantial heterogeneity was observed across several dimensions, including participants and learning content, composition of control groups, and outcome measures. We did not conduct a meta-analysis for its heterogeneity. Effectiveness was synthesized according to the respective study reports. The data extracted from each study included, the year of publication, the study location, the HDM model, the characteristics of the participants, the intervention adopted, the evaluation of the effectiveness, main results, etc. The details are presented in Appendixes 1–2.

Quality assessment

Two reviewers independently assessed the risk of bias using the Cochrane risk of bias tool [16]. The assessment included the following domains: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessors, completeness of outcome data, and selective outcome reporting.

Studies were rated for overall risk of bias as follows:

1. High risk of bias: if at least one key bias domain was rated as high.

2. Unclear risk of bias: if two or more key bias domains were rated as unclear.

3. Low risk of bias: if no high-risk biases were present and less than two unclear risks remained.

Results

Risk of bias

In total, our assessment identified 11 studies as being at high risk of bias, while 25 were deemed to have unclear or low risk. Many studies had the absence of detailed information regarding the blinding of participants and the blinding of outcome assessors. For participant blinding, there were exchanges between some of the experimental and control groups that prevented confidentiality of the intervention conditions in the different groups. Furthermore, the absence of blinding between instructors and participants may have introduced assessment bias, particularly in outcomes involving subjective evaluation. Additionally, some trials exhibited a risk of bias in random sequence generation, particularly those that allocated participants based on pre-existing characteristics, such as grouping by professional seniority, rather than through genuine randomization [17].

Most of the experiments provided descriptions of random sequence generation, allocation concealment, handling of incomplete outcome data, and selective reporting, which were assessed to be at low risk of bias. The specific results of these assessments are shown in Fig. 2.

Fig. 2.

Risk of Bias Graph and Summary

General results

This review encompassed a total of 36 studies. Regarding the year of publication, there has been fluctuation in the number of studies since 2017. However, an overall upward trend has been observed in recent years, as illustrated in Fig. 3. Notably, the publication reached the top in 2022 with ten studies. For further details, refer to Appendix 1.

Fig. 3.

Publication Year of the Included Studies

In terms of research sites, 26 of 36 studies (72%) were conducted in developed countries, with only ten studies taking place in developing nations. Specifically, China occupied a leading position with six studies, and followed closely by the United States with five studies. There were four studies from Denmark and South Korea, while three studies from Japan. Two studies from Australia and Turkey. In addition, Ireland, Austria, Germany, Canada, Portugal, Switzerland, Singapore, Spain, Brazil and Iran each contributed one study.

RQ1: IVR and Head-mounted display

All included studies utilized HMDs to create immersive virtual environments. Specifically, 30 studies employed high-end HMDs, three utilized lower-end models, and another three did not specify the exact display models. A detailed list of the HMDs is provided in Appendix 1.

Among the high-end HMDs, 15 studies utilized the Oculus series, eight employed the HTC VIVE series, two used Meta Quest series, and one study used the Pico Neo series. These premium HMDs, which offer the capability to connect with additional controllers, touchpads, or models for haptic feedback and are generally priced over $300, provide a relatively high level of immersion. Neves’ study used the M300XL Vuzix series HMD [18], which not only offers enhanced network performance and immersion but also accommodates users wearing vision correction glasses. Furthermore, three studies incorporated IVR surgery simulators. Zhang Z used the EYEsi VR magic ophthalmic surgery simulator for teaching ophthalmic microsurgery [19], while Martin Frendø employed the visible ear simulator version 3.0 [20]. McKinney B constructed a virtual operating room using the Osso VR series HMD, known for its professional environmental effects and guidance feedback [21].

Two studies used the Oculus Gear VR headset with relatively low immersion, but its advantage is that the cost is low and the price is only $99 [22, 23]. In addition, another three studies explicitly stated the use of VR HDM, but did not provide a specific model of HDM [24–26].

RQ2: medical disciplines

For the 36 studies, 24 were in the clinical medicine, five focused on nursing education and another seven on medical skills. The pedagogical content and education disciplines of each study are shown in Appendix 1.

In 24 clinical surgical studies, 14 are teaching medical surgery. Two studies concentrated on unicompartmental knee arthroplasty, utilizing the SawBone model to assess surgical procedures [21, 26]. Knotting and suture techniques in clinical surgery were the focus of two additional studies, including the reef knot [23]and single-line interrupted suture [18]. A study on teaching how to manipulate surgical instruments and set up an instrument Table [27]. The remaining nine studies covered a range of surgical topics, including total hip arthroplasty [28], cochlear implant surgery [20], aneurysm clipping [17], lumbar transforaminal epidural block [29], ophthalmic microsurgery [19], laparoscopic salpingectomy [30], cesarean Sect [31]., acute intrapartum shoulder dystocia [32]and orthognathic surgery [33].

Six studies addressed the teaching of anatomy education [34–39]. Two study focused on cardiopulmonary resuscitation [25, 40]. The remaining two studies explored distinct areas. Sapkaroski’s study examined VR communication scenarios with claustrophobic patients [41], and Mansoory’s study delved into oral medicine, employing Unity 3D to create models for dental surgery instruction [22].

In the realm of nursing education, three of the five studies targeted specialized populations for distinct nursing services, including memory rehabilitation nursing for elderly patients with dementia [24], care for patients with arrhythmias [42], and neonatal resuscitation care [43]. Lo’s study developed varied scenarios for nasogastric tube feeding without limiting the care recipients [44]. While, Zabaleta’s study focuses on developing nursing competencies for operating room nurses [45].

Seven studies focused on medical skills with medical examinations and videography, such as bronchoscopy training [46], basic clinical ultrasound skills training [47], radiography [48], X-ray localization [49], interventional radiology [50], Radiation safety [51] and cardiac point-of-care ultrasound [52].

RQ3: characteristics of experimental participants

The total number of the participants in the 36 studies was 1,982, ranging from 14 medical students [28] to 144 nursing students [40]. The participants adhered strictly to the experimental protocols, completing both the experiments and subsequent evaluation analyses. Further details are presented in Appendix 1.

22 of the studies involved medical students, and nine studies engaged doctors or nurses as the participants. Predominantly, the participants were first- and second-year college students or novice doctors, as the recruitment criteria stipulated that participants should not have prior exposure to the relevant knowledge or skills. Additionally, five studies included both medical students and doctors, facilitating comparative analyses among participants with varying levels of experience. For instance, Greuter observed no significant effect of 3D VR models on aneurysm detection among doctors, whereas students experienced a significant reduction in measurement time (26.95 ± 35.39 vs. 59.16 ± 44.60 s, p = 0.015) [17].

In addition, seven studies have some drop-outs either for time conflicts or physical discomfort. Most studies have a dropout rate of less than 10%, but in Moll’s study [37], 23 of the 120 participants (19%) dropped out of the experiment because of conflicting course schedules, and eight students dropped out because of illness [25].

RQ4: IVR experiment design

All experimental designs included in this review are controlled trials. Among them, 34 are RCTs, and one study is a controlled experiment [42], and another one was a non-randomized controlled experiment [43]. Detailed experimental information is provided in Appendix 2.

A total of 32 studies compared the efficacy of IVR teaching with traditional teaching methods, including 2D images, PowerPoint presentations, electronic resources, lectures, and traditional model materials. Two studies compared IVR with ordinary VR [36, 46]. Additionally, one study established three groups to compare IVR, ordinary VR simulation, and online lectures [43]. This review also considered studies where immersive VR served as a control group compared to other learning methods, such as an ophthalmic microsurgery study that assessed whether grape simulation could achieve the effects of immersive VR course training [19].

Regarding intervention duration, 13 studies did not specify the duration of the IVR intervention, with the shortest intervention time of 20 min [38], and the longest intervention time of 120 min [20]. The vast majority of the experimental designs focused solely on the short-term learning outcomes among medical students, with assessments conducted immediately after a single teaching session. Only three studies included a follow-up knowledge retention test administered two months after the intervention, all of which indicated better long-term knowledge retention with VR. For example, Stepan’s study conducted a retention test two months post-intervention [38]. Moreover, most IVR interventions were just one-time learning, only two study performed ten exercises [33, 39].

RQ5: the effectiveness of VR in medical education

This review evaluates whether IVR can facilitate the development of medical education based on two key aspects: objective effectiveness and subjective effectiveness. Objective measures include metrics such as task completion time, operational accuracy, and test scores, whereas subjective measures encompass self-reported perceptions, satisfaction levels, and learning confidence. IVR is considered beneficial for advancing medical education if it demonstrates non-inferiority or superiority compared to traditional teaching methods.

Based on the key evaluation indicators, IVR in medical education was effective in most studies, or at least served as a viable alternative to traditional teaching methods. 36 studies examined objective effects, with only two reporting that VR effects were inferior. In contrast, 30 studies focused on subjective effects, with only one indicating worse outcomes. The distribution of results is illustrated in Fig. 4, and the specific results of each study are detailed in Appendix 2.

Fig. 4.

Distribution of Subjective and Objective Results

In terms of objective effects, 34 studies demonstrated either an enhancement or no significant difference between the experimental and control groups’ results. Notably, no adverse outcomes were observed in nursing education, anatomy education, or communication and collaboration education. Among the two studies that reported worse results, Frederiksen compared IVR with traditional VR and found that the high cognitive load associated with IVR significantly decreased performance for most simulator metrics [30]. Similarly, Kato’s study on radiographic technology, found that skill proficiency related to palpation and patient interaction was significantly lower in the VR group compared to the conventional X-ray group [48].

In terms of subjective effects, the IVR group performed better, with most of the studies mentioning that using IVR improves interest in learning, self-confidence. In addition, two studies focused on cognitive load, but one study showed that although the IVR group scored significantly higher than the control group in terms of external goals, task value, and satisfaction, they also experienced higher cognitive load [44]. Frederiksen reported that cognitive load during IVR and traditional VR simulation increased by 66% and 58%, respectively (the difference was significant, p < 0.001). In the IVR group, mild stressors led to a further increase in cognitive load by 15.2%, severe stressors by 43.1%, while the traditional VR group just experienced an increase of 23% [30].

Finally, seven studies reported adverse effects of VR on medical education, such as discomfort and motion sickness. In a study on aneurysm clipping, one participant (5%) described IVR-related side effects such as dizziness or nausea [17]. In Mitani’s study, one student (1%) had VR sickness [50]. In a study using VR to teach cardiac anatomy, four participants (17%) reported physical symptoms including nausea, headache, and dizziness, with one participant discontinuing the VR experience after 10 min due to dizziness [37]. One study found that although there was no vertigo, a small number of users experienced other adverse effects, including user fatigue and neck discomfort caused by using HMDs [35]. Besides, in Zabaleta’s study, 30 participants (54%) reported one or more mild symptoms, in descending order: blurred vision (25 participants), difficulty focusing vision (14 participants), and sweating (10 participants) [45].

Discussion and conclusions

The effectiveness of IVR in medical education

Consistent with previous studies, this review concluded that IVR is predominantly utilized for clinical surgery education, particularly in teaching scenarios such as surgical manipulation, nursing, anatomy, clinical diagnosis, and doctor-patient communication [41]. This preference is largely attributed to the technical attributes of VR, with its interactivity allowing learners to rehearse in a simulated setting, thereby enhancing their procedural skills and decision-making capabilities.

Furthermore, this review revealed that VR can enhance subjective learning experiences, such as interest and confidence. The increase in confidence is likely due to the accumulation of practice sessions [53]. Subjective feelings, including satisfaction and motivation, can not only directly influence learning outcomes [54], but also serve as mediating or moderating variables on learner competence, which subsequently impacts academic achievement [55].

Long-term knowledge retention and knowledge enhancement

Unlike previous reviews, this review specifically investigates the effectiveness of VR in knowledge enhancement among populations with varying levels of prior experience, as well as its long-term knowledge retention outcomes. Extant studies mainly investigate the effectiveness of new medical student’ learning new medical knowledge using VR. Only six studies have compared the effectiveness of VR for students with varying levels of experience [17]. However, the essence of education extends beyond the mere transmission of new knowledge. In line with the theory of lifelong learning, education also underscores the reinforcement of knowledge, its long-term retention, and the deepening and enhancement of that knowledge [56].Although VR has shown potential in facilitating the acquisition of new knowledge in medical education, no in-depth research has yet explored its effectiveness in advancing medical students’ knowledge and skills from novice to expert levels. Consequently, there is an imperative for future research to address this gap, thereby enabling a more holistic assessment of the value and potential of VR technology in medical education.

This review believes that the superior retention of knowledge is a distinct advantage of VR technology over other teaching methods, an area that has received scant attention among current research. Only Koucheki’s and Xie’s conducted knowledge retention tests with two-month post-experiment, revealing that the VR group performed comparatively better [24, 35]. This outcome may stem from the immersive VR experience stimulates students’ interest and enthusiasm, enabling them to engage more deeply with the learning content, and thereby enhancing the learning effect [57, 58]. Furthermore, this emotional engagement is more likely to trigger intrinsic learning motivation [58], foster enthusiastic participation, diminish the perceived difficulty of learning materials, and bolster knowledge retention and learning transfer [57].

Evaluating the effectiveness from the teachers’ perspective

Existing studies predominantly evaluate teaching effectiveness from the students’ perspective. While they often overlook the role of another key participant in the educational process—the teacher. Studies indicate that the introduction of new teaching methodologies and the associated stress of integrating new technologies can lead to physical, social, and psychological challenges for teachers [59]. One such challenge is burnout syndrome, characterized by exhaustion and burnout resulting from increased demands [60]. This syndrome can significantly impact teachers’ performance levels, thereby affecting the overall effectiveness of medical education [61].

In light of these findings, this review advocates for the inclusion of teachers’ perspectives within existing IVR assessment frameworks to evaluate the true effectiveness of VR in medical education comprehensively [62]. Factors such as teachers’ acceptance of IVR technology, the pressure they experience in utilizing it, and their efficiency in preparing for lectures are crucial in assessing the impact of IVR on medical education [63]. An evaluation strategy that incorporates these factors can not only uncover the direct influence of IVR technology on student learning outcomes but also delve into its potential effects on teaching activities and overall teaching efficiency. This approach offers a more holistic and scientifically grounded basis for the continued refinement and expansion of IVR technology within the medical education sector.

Future of IVR in medical education

The widespread application of IVR technology in medical education must address ethical, legal, and social concerns. The Ethical Guidelines for Virtual Reality Technology Research and Development issued by China’s Ministry of Science and Technology in April 2025 provide crucial regulatory frameworks for the advancement of VR-based medical education. These guidelines also highlight the necessity of balancing technological innovation with ethical compliance in future IVR applications for medical training. Furthermore, it is recommended to promote the formulation of industry standards to facilitate deeper integration between IVR technology and traditional medical education. Issues such as protecting the privacy of feedback data generated by students using IVR devices [64], and the potential for IVR education to exacerbate educational inequities must be considered [65]. Currently, no studies indicate adherence to national legal standards for the application of IVR in medical education. Given the critical nature of medical education to public health, the introduction of policies and standards is essential when implementing new technologies like IVR to ensure their effectiveness and safety.

Limitations

This review had limitations. Solely English publications were included, which could have led to the exclusion of relevant studies in other languages. Furthermore, the focus of this review was on compared trials in the field. Not all gray literature was screened, which could have also led to the exclusion of suitable studies. Future research could incorporate comparative experimental studies between VR and other technologies, such as Augmented Reality (AR) and Mixed Reality (MR), to further analyze the effectiveness of VR in medical education.

Abbreviations

VR

Virtual Reality

IVR

Immersive virtual reality

HMD

Head-mounted display

PRISMA

Preferred Reporting for Systematic Reviews and Meta-Analysis

PROSPERO

Prospective Register of Systematic Reviews

Appendix: Studies included in the review

The 36 publications included in the systematic review are listed in Appendix 1-2. Prisma Checklist is Appendix 3.

Appendix 1

Head-mounted displays, educational content and participants

Serial Number	Author+year	Title	Area	Head-mounted display	Educational content	Experimental participants	Num	Drop out
1	Xie Z,2023 [24]	Using Virtual Reality in the Care of Older Adults With Dementia: A Randomized Controlled Trial	China	Using VR HDM (brand and model not specified)	Memory rehabilitation for elderly dementia patients	sophomores majoring in Intelligent Health and Elderly Care Service Management	38	2 did not complete the experimental study due to illness.
2	Lo YT,2022 [44]	Effectiveness of immersive virtual reality training in nasogastric tube feeding education: A randomized controlled trial	Taiwan, China	HTC VIVE Focus Plus	Nasogastric tube feeding	>= 20-year-old nursing students	107	None
3	Frend M, 2022 [20]	Cochlear Implant Surgery: Virtual Reality Simulation Training and Transfer of Skills to Cadaver Dissection-A Randomized, Controlled Trial	Denmark	Visible Ear Simulator (VES) 3.0 version	Cochlear implant (CI) surgery	ENT resident physicias	18	None
4	Kim SK,2023 [42]	Constructing a Mixed Simulation With 360° Virtual Reality and a High-Fidelity Simulator: Usability and Feasibility Assessment	the republic of Korea	Oculus Quest 2	Providing care for patients with arrhythmia	Nursing students	48	One participant answered insufficiently
5	Greuter L, 2021 [17]	Randomized study comparing 3D virtual reality and conventional 2D on-screen teaching of cerebrovascular anatomy	Switzerland	HTC Vive, Valve HTC Earphones	Aneurysm clipping surgery	Neurosurgery resident doctors and medical students in grades 4-6	20	None
6	McKinney B,2022 [21]	Virtual Reality Training in Unicompartmental Knee Arthroplasty: A Randomized, Blinded Trial	U.S.A	Osso VR	Surgery for unicompartmental knee arthroplasty (UKA)	Orthopedic resident physicians with different training years (PGY 1 to 5)	22	None
7	Neves Lopes V,2022 [18]	Telestration in the Teaching of Basic Surgical Skills: A Randomized Trial	Portugal	M300XL Vuzix，Corp，Rocheste	Basic surgical skills include five suture techniques: single line intermittent suture, cross pad suture, horizontal pad suture, vertical pad suture, and simple continuous suture.	No professional grade specified	20	None
8	Koucheki R,2023 [35]	Immersive Virtual Reality and Cadaveric Bone are Equally Effective in Skeletal Anatomy Education: A Randomized Crossover Noninferiority Trial	Canada	Oculus Quest 2	Skeletal Anatomy Education	First year medical students	50	None
9	Kim JY,2023 [29]	Virtual reality simulator's effectiveness on the spine procedure education for trainee: a randomized controlled trial	the republic of Korea	Oculus Quest 2	Lumbar Transforaminal Epidural Block (LTFEB)	First or second year anesthesiology and pain medicine resident physicians	20	None
10	Andersen AG, 2023 [46]	Preparing for Reality: A Randomized Trial on Immersive Virtual Reality for Bronchoscopy Training	Denmark	HTC VIVE Pro Eye	Bronchoscopy training	Junior doctors who received training in the first year	34	one did not complete the intervention
11	Zhang Z,2023 [19]	Skills assessment after a grape-based microsurgical course for ophthalmology residents: randomised controlled trial	China	EYEsi VR magic	Ophthalmic microsurgery	First year ophthalmic resident physicians	83	None
12	Mansoory MS,2022 [22]	A study to investigate the effectiveness of the application of virtual reality technology in dental education	Iran	Oculus Gear VR	Dental education	Six-year dental students	50	none
13	Frederiksen JG,2020 [30]	Cognitive load and performance in immersive virtual reality versus conventional virtual reality simulation training of laparoscopic surgery: a randomized trial	Denmark	Oculus Rift VR	Laparoscopic salpingectomy	First year resident physicians	31	None
14	Stepan K, 2017 [38]	Immersive virtual reality as a teaching tool for neuroanatomy	U.S.A	Oculus Rift VR	Clinical Anatomy Teaching	First and second year medical students	66	None
15	Patel N,2021 [37]	Stereoscopic virtual reality does not improve knowledge acquisition of congenital heart disease	U.S.A	HTC Vive Pro	Anatomy of congenital heart disease (CHD) in the atrioventricular canal	Having a background in anatomy or medicine, without specifying any other information	51	None
16	Moll-Khosrawi P,2022 [25]	Virtual reality as a teaching method for resuscitation training in undergraduate first year medical students during COVID-19 pandemic: a randomised controlled trial	Germany	Using VR HDM (brand and model not specified)	Basic Life Support (BLS) Training	First year undergraduate medical students	88	31 participants were excluded due to course or physical reasons
17	Rosenfeldt Nielsen M, 2021 [47]	Clinical Ultrasound Education for Medical Students: Virtual Reality Versus e-Learning, a Randomized Controlled Pilot Trial	Denmark	Pico Neo Interactive Inc	Basic Clinical Ultrasound Skills Training	Medical students without specified major grade	20	None
18	Yang SY,2022 [43]	The effects of neonatal resuscitation gamification program using immersive virtual reality: A quasi-experimental study	the republic of Korea	Oculus Rift VR	Neonatal resuscitation nursing	Nursing students before obtaining their license	83	Two VR volunteers did not participate in program,
19	Kim HJ,2024 [31]	Immersive virtual reality simulation training for cesarean section: a randomized controlled trial	the republic of Korea	Oculus Quest 2	Cesarean section surgery (CS)	Medical students in their third year of medical school (18/105), interns (1/105), resident physicians (57/105), and faculties and staffs	105	None
20	Kurul R,2020 [36]	An Alternative Method for Anatomy Training: Immersive Virtual Reality	Türkiye	Oculus Rift VR	Anatomy skills	Students majoring in Physical Therapy and Rehabilitation in their first year of undergraduate studies	72	None
21	Zaid MB,2022 [26]	Virtual Reality as a Learning Tool for Trainees in Unicompartmental Knee Arthroplasty: A Randomized Controlled Trial	U.S.A	Using VR HDM (brand and model not specified)	Surgery for unicompartmental knee arthroplasty (UKA)	Resident orthopedic physician or fourth year medical students	22	None
22	Sapkaroski D,2022 [41]	Immersive virtual reality simulated learning environment versus role-play for empathic clinical communication training	Australia	Oculus Rift VR	Clinical communication	First year medical students majoring in radiation technician	79	None
23	Falcone V,2024 [32]	Impact of a virtual reality-based simulation training for shoulder dystocia on human and technical skills among caregivers: a randomized-controlled trial	Austria	Oculus Quest 2	Acute shoulder dystocia (SD) during labor	Resident physicians, attending physicians, midwives, and medical students in their final year at the medical school	61	Three people did not complete it
24	Hu KC,2020 [34]	Impact of virtual reality anatomy training on ultrasound competency development: A randomized controlled trial	Taiwan, China	HTC VIVE	Anatomy	Third year medical students in medical school	101	None
25	Wan T, 2024 [33]	Effectiveness of immersive virtual reality in orthognathic surgical education: A randomized controlled trial	China	HTC VIVE Pro	Orthognathic Surgical Procedures	Fifth grade medical students	20	None
26	Kato K,2022 [48]	Radiography education with VR using head mounted display: proficiency evaluation by rubric method	Japan	HTC Vive Pro	Radiographic technology	First grade students at the Radiological Technician Training School	30	None
27	Yoganathan S,2018 [23]	360° virtual reality video for the acquisition of knot tying skills: A randomised controlled trial	Ireland	Oculus Gear VR	Knotting skills, especially one handed reef knots	First year graduate medical students	40	None
28	Hooper J,2019 [28]	Virtual Reality Simulation Facilitates Resident Training in Total Hip Arthroplasty: A Randomized Controlled Trial	U.S.A	Oculus Rift VR	Total hip arthroplasty (THA)	First year orthopedic resident physicians in graduate school	14	None
29	Sapkaroski D,2019 [49]	Quantification of Student Radiographic Patient Positioning Using an Immersive Virtual Reality Simulation	Australia	Oculus Rift VR	Hand front back and oblique hand X-ray positioning	First grade radiology students	76	None
30	He Y,2024 [39]	Enhancing medical education for undergraduates: integrating virtual reality and case-based learning for shoulder joint	China	HTC VIVE-Pro	Anatomy	Third-year undergraduates	82	None
31	Mitani H,2025 [50]	Effectiveness of a virtual reality-based interventional radiology simulator for medical student education	Japan	Using VR HDM (brand and model not specified)	Interventional radiology (IR)	Groups of four to five students	97	Two students in the VR–IR simulator group were excluded due to VR sickness and simulator malfunction.
32	Bodur G,2025 [40]	Evaluating the effectiveness of virtual reality simulation in cpr training for nursing students: A randomized controlled trial	Turkey	Oculus Quest 2 headsets	Cardiopulmonary resuscitation（CPR）	Nursing students	144	None
33	Mwangi W,2025 [51]	Comparative Effectiveness of Immersive Virtual Reality and Traditional Didactic Training on Radiation Safety in Medical Professionals: A Crossover Study	Japan	Meta Quest 2 HMD	Radiation safety	Medical professionals from cardiac catheterization laboratories and orthopaedic theatres	39	None
34	Khoo C,2025 [52]	Self-Directed Virtual Reality-Based Training versus Traditional Physician-Led Teaching for Point-of-Care Cardiac Ultrasound: A Randomized Controlled Study	Singapore	Oculus Quest 2 headsets	Cardiac point-of-care ultrasound (POCUS)	Medical students with no prior formal ultrasound training	43	None
35	Zabaleta J,2024 [45]	Clinical trial on nurse training through virtual reality simulation of an operating room: assessing satisfaction and outcomes	Spain	Meta Quest 2 HMD	Nursing	Nurses without prior thoracic surgery experience	56	None
36	Camargo,CP 2025 [27]	Virtual reality and traditional training in surgical instrumentation: A non-inferiority comparative study	Brazil	Oculus PiCO2	Manipulate the surgical instruments and set an instrument table	First-year surgery programs of both genders	24	None

Appendix 2

Experimental process and results

Author+ year	Experimental group	Control group	Duration	Indicators	Objective	Subjective	Adverse effects
Xie Z,2023 [24]	VR simulation environment role-playing	using conventional prop methods	Not specified	Assessment of PANAS Scale Scores and Final Exam Scores at Each Stage	The PA score showed an increase over time, with the intervention group showing a more significant increase. The NA score decreases over time, with little difference between groups.	The average total SUS score was 83.68 (SD=9.44), exceeding the threshold of 60, indicating that the system provided a user-friendly experience.	Not available
Lo YT,2022 [44]	IVR simulation	2D video	35 mins	Knowledge test results (pre-test and post test), learning motivation, self-efficacy, cognitive load, and satisfaction	The difference in scores between the two groups of knowledge tests was not significant, but both groups showed a significant improvement in knowledge scores after intervention: the IVR group scored 7.75 to 8.85 (t=-6.48, p<0.001), while the control group scored 7.35-8.72 (t=-5.45, p<0.001), but the difference between the groups did not reach statistical significance (t=-0.54, p>0.05).	The cognitive load and satisfaction of the IVR group were significantly higher than those of the control group (t=2.335 and t=2.297, p<0.05), and the effect size was moderate (d=0.456 and 0.458)	Not available
Frend M, 2022 [20]	CI VR simulator	Laboratory teaching	120 mins	Anatomy performance score, self orientation, and VR performance	The average score of the intervention group was 22.9 points, with the highest score of 44 points, which was 5.4% higher than the control group's 21.8 points (P=. 51). On average, the intervention group required 1.3 sessions of assistance during the corpse drilling process; In the control group that received help 1.9 times, the incidence of this condition was 41% higher (P=. 21).	The addition of cochlear implantation virtual reality training to basic mastoidectomy virtual reality simulation training did not lead to a significant improvement of performance or self-directedness in this study.	Not available
Kim SK,2023 [42]	360 ° VR simulation	Teacher lecture discussion	120 mins	Assessment of mastery of arrhythmia nursing knowledge, decision-making ability, anxiety level, and emotional response	The mixed simulation group showed greater progress in knowledge。	With higher decision-making abilities in "knowing and acting" (P=. 025) and "seeking information from teachers" (P=. 049), and lower anxiety in "utilizing resources to collect information" (P=. 031). The study participants achieved good empathy (3.28 ± 0.72) and enjoyed the program (4.56 ± 0.60). They are satisfied with the plan (4.48 ± 0.65).	Not available
Greuter L, 2021 [17]	3D VR	2D images	Not specified	Accuracy of aneurysm detection time and location description, subjective impression	Overall, compared with 2D images, the detection time of aneurysms in 3D VR models is shorter, with a statistically significant trend (25.77 ± 37.26 vs 45.70 ± 51.94 seconds, p=0.052). No significant difference was observed among residents (3D VR 24.47 ± 40.16 vs 2D 33.52 ± 56.06 seconds, p=0.564), while among students, the use of 3D VR models significantly shortened the detection time of aneurysms (26.95 ± 35.39 vs 59.16 ± 44.60 seconds, p=0.015).	Most participants (90%) preferred the 3D VR models for aneurysm detection and description.	One participant (5%) described VR related side effects such as dizziness or nausea.
McKinney B,2022 [21]	Immersive VR for UKA surgical training	Traditional technique guidelines for UKA surgical training	60 mins	The completion quality, time, accuracy, and subjective motivation of surgical steps	Doctors trained using immersive VR correctly performed more steps (33 vs. 27, step<0.01) and completed their procedures in significantly faster time (26.7 vs. 35.4 minutes, page<0.01). They also scored higher in all global assessment categories, reaching significance in 4 out of 5 categories.	Subjective questionnaire responses demonstrated positive feedback within both groups with a trend toward virtual reality.	Not available
Neves Lopes V,2022 [18]	Remote VR teaching	Traditional on-site guidance teaching	Not specified	Time, quality standards, self-assessment, and confidence	Most sutures do not show statistically significant differences. However, this situation exists in both cross mattress sutures (p=0.05) and simple continuous sutures (p=0.01), which affects the total time spent performing all exercises (p=0.04). In these projects, students in the telemetry group have faster speeds.	On a global scale, participants in remote teaching groups found the device to be very comfortable to use (4.30 [SD 0.68]). Three out of ten students have no difficulty using smart glasses, with an average classification of 3.70 (SD 1.34).	Not available
Koucheki R,2023 [35]	Immersive VR learning of skeletal anatomy	Real bone model learning of skeletal anatomy	30 mins	Knowledge test scores, subjective feelings	The percentage increase in scores between pre intervention and post intervention knowledge tests was 15.0% in the upper limb IVR group and 16.7% in the upper limb skeletal group (p=0.286). For the lower limbs, the IVR score increased by 22.6%, and the score for the cadaver group increased by 22.5% (p=0.936). 79% of participants found that IVR is most valuable for teaching 3D orientation, anatomical relationships, and key landmarks.	Most participants support the combination of traditional methods and IVR technology for learning skeletal anatomy (LSM>3).	No dizziness, but a small number of users have noticed other adverse effects, including user fatigue and neck discomfort caused by using HDM
Kim JY,2023 [29]	VR simulation LTFEB training	plastic spine covered with foam education	60 mins	Checklist score, overall score, duration of surgery (in seconds), number of C-arms taken, and satisfaction score	The group using the simulator showed higher overall rating scores (P=0.014), less program time (P=0.025), reduced C-arm usage (P=0.001). Compared with the pre training scores, both groups showed an increase in their checklist scores (V group, P=0.004; C group, P=0.041). There was no statistically significant difference in the degree of change in checklist scores between the two groups before and after training.	Higher overall satisfaction scores (P=0.007)	Not available
Andersen AG, 2023 [46]	Trained bronchoscopy simulator using HMD in iVR environment	Trained on the same bronchoscopy simulator but without using HMD	30 mins	Diagnostic integrity (%), structured progression score, program time (seconds), hand movement measurement, heart rate variability, Surg TLX total score.	The pre group scored significantly higher in terms of diagnostic integrity (100 i.qr. 100-100 vs. 94 i.q.r. 89-100, p-value=0.03) and structural progression (16 i.qr. 15-18 vs. 12 i.qr. 11-15, p-value=0.03), but in terms of surgical time (367 seconds standard deviation [SD] 149 vs. 445 seconds SD 219, p-value=0.06) or hand movement (-1.02 IQR -1.03- [-1.02] vs. -0.98 IQR -1.02- [-0.98], p-value=0.27). The heart rate variability of the control group decreased (5.76 i.q.r. 3.77-9.06 vs. 4.12 i.q.r. 2.68-6.27, p=0.25). There was no statistically significant difference in the total score of Surg TLX between the two groups.	There was a slight tendency for the HRV to be lower in the control group (median 5.76, i.q.r. 3.77–9.06) than in the iVR group (median 4.12, i.q.r. 2.68–6.27). This was however not significant (p = 0.25). In addition, there was no significant difference in total Surg-TLX points between the two groups (p = 0.77).	Not available
Zhang Z,2023 [19]	Training on grape based microsurgery course	VR based course training	Not specified	Suture performance score, changes in confidence after training, economic investment, and usability analysis	Group A participants were superior to Group B participants in terms of suture span (p<0.05) and suture thickness (p<0.05). Meanwhile, participants in Group A exhibited better tissue protection during the incision suture task (p<0.05). There was no statistically significant difference in the total score of corneal incision suturing and circular joint capsule incision between group A and group B (6.50 ± 0.1 vs 6.29 ± 0.1, 6.19 ± 0.2 vs 6.15 ± 0.2; p=0.26 and 0.87, respectively).	Group A showed a more positive attitude to withstand the training for more than 4 hours (p<0.001), as well as a higher willingness to receive more times of the training in the future (p<0.001).	Not available
Mansoory MS,2022 [22]	Teaching using VR technology	Traditional face-to-face and demonstration techniques were used for teaching	Not specified	Skill test scores, subjective satisfaction	The average score of the experimental group students (16.92 ± 1.12) was significantly higher than that of the control group (16.14 ± 1.18).	Most students (76%) are very satisfied with using this technology in their learning process.	Not available
Frederiksen JG,2020 [30]	Immersive VR training	Traditional VR teaching	Not specified	Surgical completion time, damage to surrounding tissues (thermal therapy injury and blood loss), efficiency of instrument movement, motion sickness	Immersive VR also leads to significant performance degradation of most simulator metrics.	Compared to the baseline, the cognitive load during immersive VR and traditional VR simulations increased by 66% and 58%, respectively (p<0.001). In the immersive VR group, mild stressors further increased cognitive load by 15.2%, severe stressors increased by 43.1%, while the traditional VR group (severe stressors) increased by 23%.	Not available
Stepan K, 2017 [38]	Immersive VR teaching	Learning using text and 2D images	20 mins	Pre test, post test and retention test scores, learners' subjective feelings	There was no significant difference in anatomical knowledge between the two groups before intervention, after intervention, or retention testing.	The VR group found that the learning experience was significantly more attractive, enjoyable, and useful (all<0.01), and scored significantly higher in motivation assessment (<0.01).	Not available
Patel N,2021 [37]	Use 3D VR headset to view heart model Control group	Use desktop computer interface to view the same heart model	30 mins	Differences in post test scores. Participants' ratings of usability and comprehension level	The median score difference among VR participants was 12 (IQR 9-13.3), while the DT group was 10 (IQR 7.5-12). No difference was found in score improvement (p=0.11).	VR participants had a higher impression of its interface usability than DT participants (median 8 to 7, p=0.01). VR participants had a higher impression of their understanding of the topic than desktop participants (median 7 to 5, p=0.01).	4 subjects (17%) reported physical symptoms, including nausea, headache, and dizziness. One subject stopped the VR experience after 10 minutes due to dizziness.
Moll-Khosrawi P,2022 [25]	After the web-based BLS training, an additional 35 minutes of VR BLS training were conducted	Only received web-based BLS training	55 mins	No blood flow time (an indicator of BLS quality), overall quality of BLS, and learning gain for undergraduate students	The duration of no blood flow in the intervention group (p=0.009) was significantly lower, with a difference of 28% (95% - CI [8%; 43%]) between the two groups. The overall BLS performance of the intervention group was also significantly better than that of the control group, with an average difference of 15.44 points (95% - CI [21.049.83]), p<0.001. In CSA, undergraduate students in the intervention group reported significantly higher learning gains.	Interestingly, the VR group reported the highest learning gain for item seven (“I feel competent with the use of the AED”) and the learning gain was even three-fold higher than in the control group.	Not available
Rosenfeldt Nielsen M, 2021 [47]	Immersive VR ultrasound skill training	Ultrasound skill training using electronic learning platform	60 mins	Objective Structured Assessment on Ultrasound Skills score. Score for hand eye coordination ability	Compared with the online learning group (n=9; 126 [95% CI, 113 to 138]; mean difference, 17 points [95% CI, 4 to 30]), the virtual reality group (n=91; 126 [95% CI, 113-138]; mean difference, 17 points [95% CI, 4 to 30]; P<0.01) did not show a significant impact on hand eye scores (mean difference, 3 points [95% CI, -3 to 9]; P=0.32).	91% of the virtual reality group hope to have more opportunities for virtual reality learning.	Not available
Yang SY,2022 [43]	Received immersive VR neonatal resuscitation gamification program based on Keller's ARCS model	Control group 1: Received high fidelity neonatal resuscitation simulation and online neonatal resuscitation program lecture Control group 2: Only received online neonatal resuscitation program lecture	100 mins	Knowledge of neonatal resuscitation nursing, problem-solving and clinical reasoning abilities, confidence in neonatal resuscitation nursing performance, anxiety level and learning motivation	After intervention, the neonatal resuscitation knowledge [F (2)=3.83, p=. 004] and learning motivation [F (2)=1.79, p=. 025] in the virtual reality group and simulation group were significantly higher than those in the control group, while the problem-solving ability [F (2)=2.07, p=. 038] and self-confidence [F (2)=6.53, p<. 001] in the virtual reality group were significantly higher than those in the simulation group and control group.	The anxiety of the simulation group [F (2)=16.14, p<. 001] was significantly lower than that of the virtual reality group and the control group	Not available
Kim HJ,2024 [31]	Receive VR simulation training, focusing on PROM management and CS practice	Watch a video demonstration that includes clinical scenario descriptions and CS surgical records	Not specified	Confidence level score test score	The VR group also achieved significantly higher scores in the mini test [median (interquartile range), with the VR group at 42 (37-48); The control group was 36 (32-40), P<0.001.	After intervention, compared with the control group, the VR group had higher confidence scores in all four aspects, including managing PROM patients, performing CS as an operator, and understanding the indications and complications of CS.	Not available
Kurul R,2020 [36]	3D virtual reality device dissection training	VR image demonstration	30 mins	Knowledge test scores and subjective feelings	Compared with the VR group (P<0.001) and the control group (P<0.001), the scores were significantly higher after the test. The study found that the difference between pre-test and post test results was significantly higher in the VR group (P<0.001).	88.8% of students answered “I agree” or “I strongly agree” to the “I enjoyed studying anatomy with virtual reality” sentence with a mean score of 1.69 ± 0.92.	Not available
Zaid MB,2022 [26]	Immersive commercial VR learning module	Traditional technical guidelines and surgical videos	45 mins	Surgical time, surgical score, confidence level	There was no difference in the average surgical time between the guideline group and the VR group (guideline=42.4 minutes, while VR=43.0 minutes; P=0.9) or the average total OSATS (guideline=15.7 vs VR=14.2; P=0.59).	Most trainees believe that virtual reality will be a useful tool for resident physician education (77%), and if available, they will use virtual reality for case preparation (86.4%).	Not available
Sapkaroski D,2022 [41]	VR simulation and communication with claustrophobia patients	Role playing and communication with patients	Not specified	Self efficacy	The average performance of the VR training group (TVR and CVR) was better than that of the role-playing group (5% and 11%);	The results may demonstrate the capacity for immersion into an emotional narrative in a VR environment to increase the user's susceptibility for recalling and selecting empathic terminology.	Not available
Falcone V,2024 [32]	VR based shoulder dystocia simulation training	Learning theory through PowerPoint presentation	Not specified	Surgical time, quality rating, and subjective perception	The HELP-RER (2) scores of the control group subjects were significantly higher [7 (95% CI, 6-7) compared to 6.5 (95% CI, 6-7); p=0.01)], the time from diagnosis to delivery was shorter [85.5 seconds (95% CI, 66.5-98.25) vs. 99 seconds (95% CI, 75-120); p=0.02)],	The subjective workload of reporting was lower [57 (95% CI, 43.75-68.75) compared to 68 (95% CI, 56-79); p=0.04],	Not available
Hu KC,2020 [34]	VR anatomy training	Learning using traditional teaching materials	60 mins	Accuracy, speed, and confidence of ultrasound scanning	Participants in the intervention group (median=16; quartiles 13 to 19) scored significantly higher in ultrasound task performance tests than the control group (median=10; quartiles 7 to 14; Mann Whitney U=595; P<0.01). In subgroup analysis, the intervention group performed significantly better in six out of ten ultrasound tasks. Compared with the control group, participants in the intervention group also showed greater improvement in ultrasound image recognition MCQ testing (Mann Whitney U=914; P<0.05).	The use of VR for learning focused regional anatomy may help gain a better awareness of the disposition and spatial relationship of anatomical structures, and an enhanced understanding of ultrasonographic visualization windows.	Not available
Wan T, 2024 [33]	IVR simulation	Traditional teaching methods	10 classes, each session lasting up to 40 mins.	Score, program duration, timeout count, instrument selection error count, position and angle error count, and the number of prompts required for the next step	The VR group scored higher than the traditional group (94.67 points vs. 87.65 points). Compared with the control group, the VR group completed the program faster, with fewer instrument selection and angle errors. No difference was observed in the number of prompts provided by the system between the two groups.	Unmeasured	Not available
Kato K,2022 [48]	Teaching using HMD-VRC	using conventional X-ray physics equipment	Not specified	Skill proficiency score, subjective questionnaire results	Compared with the HMD-VRC group, the RP group showed significantly higher scores in elbow joint lateral # 3 and 7, as well as PA chest X-ray photography # 6, 8, 11, and 12.	The use of HMD-VRC for learning received positive feedback.	Not available
Yoganathan S,2018 [23]	360 degree VR video	2D video	20 mins	Knot rating	Compared with the traditional teaching group, the VR video teaching group had significantly better knot tying scores (median knot tying score: 5.0 vs 4.0, p=0.04). When combined with face-to-face teaching, this difference still existed (median knot tying score: 9.5 vs 9.0, p=0.01). More people in the VR group than in the 2D group constructed a complete reef knot (17/20 vs 12/20).	Unmeasured	Not available
Hooper J,2019 [28]	VR-THA	Traditional standard learning materials	45 mins	Written test results, THA surgical technique performance score for cadavers	There was no significant difference in the improvement of test scores between the VR group and the control group (P=. 078). In multivariate regression analysis, the VR cohort showed a significant improvement in overall cadaver THA scores (P=. 048). Compared with the control group, the VR queue showed greater improvement in each specific score category, but this trend was only statistically significant for technical performance (P=. 009).	Unmeasured	Not available
Sapkaroski D,2019 [49]	Immersive VR simulation	Traditional clinical role-playing scenarios	Not specified	Comparison of Hand Positioning in PA Hand Projection	The first group of students showed an average improvement of 36% in finger separation (P<0.001), an average improvement of 11% in palm flatness (P ≤ 0.001. The comparison of position projection results showed that there was no statistically significant difference in localization between the two queues (P=0.171)	Unmeasured	Not available
He Y,2024 [39]	virtual reality learning	Case-based learning（CBL）	120 mins，10 times	Test scores、knowledge retention、 interaction, learning resources, skill improvement, course recommendation, and overall satisfaction	there were no significant differences in the scores of students between groups (P > 0.05)	Analysis using the Kruskal–Wallis H test revealed significant differences (P < 0.05) among groups in six key areas	Not available
Mitani H,2025 [50]	VR–IR simulator training	Conventional verbal explanations and educator demonstrations 、 technical achievement scores	120 mins	Procedure time、fluoroscopic time、 patient peak skin dose、the amount of contrast media	There were no significant differences between the VR–IR simulator group and the conventional group regarding total procedure time (median [25–75% interquartile range]: 13.5 [11.8–14.5] vs. 14.3 [12.3–16.8] minutes, p = 0.11), fluoroscopic time (10.1 [8.5–13.0] vs. 11.0 [8.6–13.7] minutes, p = 0.31), and patient peak skin dose (276 [243–373] vs. 303 [239–395] mGy, p = 0.57)	Unmeasured	One student have VR sickness
Bodur G,2025 [40]	VR group	Traditional simulation group	Not mentioned	Knowledge, psychomotor skills, attitudes, learning styles and self-directed learning skills	In the control group, the mean pre-test score was 56.67 and the post-test score was 81.54. In the experimental group, the pre-test score was 52.27 and the post-test score was 79.39. In the post-test, the scores for both groups increased significantly. However, the differences between the groups were not statistically significant according to t-test (p > 0.05)	Students in the VR group outperformed the control group on several critical psychomotor CPR skills measured by OSCE.	Not available
Mwangi W,2025 [51]	VR training	Didactic training	60 mins, 4 times	Satisfaction, engagement and confidence	VR training was significantly higher (4.8 ± 0.4) than for didactic training (3.6 ± 0.7)	Participant satisfaction and engagement were higher with VR training (p < 0.001), and confidence in applying safety practices increased significantly following VR training (p < 0.001).	Not available
Khoo C,2025 [52]	Virtual-reality (VR) simulator	Physician-led (PL) teaching	30 mins	MCQ scores	At 1 month post-training, the PL group similarly had a greater mean change from baseline MCQ scores, but this difference was not statistically significant (p = 0.12). For practical scores, the VR group scored higher than the PL group, although this difference was not statistically significant (p = 0.06).	Unmeasured	Not available
Zabaleta J,2024 [45]	VR basic formation	Basic formation	Not specified	Number of Completed Tasks、Individual Task Scores 、 Time Spent in the Simulator 、Overall Score、Simulator Sickness Questionnaire (SSQ) and satisfaction test	Participants achieved a median evaluation mode score of 480 points (IQR = 32 points). The experimental group (520 points) achieved an overall higher score than the control group (440 points; P = .04).	Overall satisfaction rating for the experience was 8.72 (Range 6–10). The likelihood of recommending this experience to a colleague received a score of 8.96 (Range 6–10).	30 participants reported one or more symptoms, all classified as mild (Supplementary Table S1). The most prevalent undesired effects were, in descending order: blurred vision (25 participants), difficulty focusing vision (14 participants), and sweating (10 participants).
Camargo,CP 2025 [27]	VR	Text and practice	30 mins	Test scores、recommendation、 satisfaction	Regarding the difference between pre-test and post-test, there was no difference in all groups: Text, Practice, and Virtual reality	Regarding recommendation, the VR group showed higher scores compared to the Text group	Only one participant (1/8) showed nausea in the virtual groups

Appendix 3

PRISMA Checklist

Section and Topic	Item #	Checklist item	Location where item is reported
TITLE
Title	1	Identify the report as a systematic review.	1
ABSTRACT
Abstract	2	See the PRISMA 2020 for Abstracts checklist.	2
INTRODUCTION
Rationale	3	Describe the rationale for the review in the context of existing knowledge.	4
Objectives	4	Provide an explicit statement of the objective(s) or question(s) the review addresses.	4
METHODS
Eligibility criteria	5	Specify the inclusion and exclusion criteria for the review and how studies were grouped for the syntheses.	5
Information sources	6	Specify all databases, registers, websites, organisations, reference lists and other sources searched or consulted to identify studies. Specify the date when each source was last searched or consulted.	6
Search strategy	7	Present the full search strategies for all databases, registers and websites, including any filters and limits used.	6
Selection process	8	Specify the methods used to decide whether a study met the inclusion criteria of the review, including how many reviewers screened each record and each report retrieved, whether they worked independently, and if applicable, details of automation tools used in the process.	7
Data collection process	9	Specify the methods used to collect data from reports, including how many reviewers collected data from each report, whether they worked independently, any processes for obtaining or confirming data from study investigators, and if applicable, details of automation tools used in the process.	8
Data items	10a	List and define all outcomes for which data were sought. Specify whether all results that were compatible with each outcome domain in each study were sought (e.g. for all measures, time points, analyses), and if not, the methods used to decide which results to collect.	8
	10b	List and define all other variables for which data were sought (e.g. participant and intervention characteristics, funding sources). Describe any assumptions made about any missing or unclear information.	NA
Study risk of bias assessment	11	Specify the methods used to assess risk of bias in the included studies, including details of the tool(s) used, how many reviewers assessed each study and whether they worked independently, and if applicable, details of automation tools used in the process.	8
Effect measures	12	Specify for each outcome the effect measure(s) (e.g. risk ratio, mean difference) used in the synthesis or presentation of results.	NA
Synthesis methods	13a	Describe the processes used to decide which studies were eligible for each synthesis (e.g. tabulating the study intervention characteristics and comparing against the planned groups for each synthesis (item #5)).	NA
	13b	Describe any methods required to prepare the data for presentation or synthesis, such as handling of missing summary statistics, or data conversions.	NA
	13c	Describe any methods used to tabulate or visually display results of individual studies and syntheses.	8
	13d	Describe any methods used to synthesize results and provide a rationale for the choice(s). If meta-analysis was performed, describe the model(s), method(s) to identify the presence and extent of statistical heterogeneity, and software package(s) used.	NA
	13e	Describe any methods used to explore possible causes of heterogeneity among study results (e.g. subgroup analysis, meta-regression).	NA
	13f	Describe any sensitivity analyses conducted to assess robustness of the synthesized results.	NA
Reporting bias assessment	14	Describe any methods used to assess risk of bias due to missing results in a synthesis (arising from reporting biases).	9
Certainty assessment	15	Describe any methods used to assess certainty (or confidence) in the body of evidence for an outcome.	9
RESULTS
Study selection	16a	Describe the results of the search and selection process, from the number of records identified in the search to the number of studies included in the review, ideally using a flow diagram.	7
	16b	Cite studies that might appear to meet the inclusion criteria, but which were excluded, and explain why they were excluded.	11
Study characteristics	17	Cite each included study and present its characteristics.	10
Risk of bias in studies	18	Present assessments of risk of bias for each included study.	NA
Results of individual studies	19	For all outcomes, present, for each study: (a) summary statistics for each group (where appropriate) and (b) an effect estimate and its precision (e.g. confidence/credible interval), ideally using structured tables or plots.	10
Results of syntheses	20a	For each synthesis, briefly summarise the characteristics and risk of bias among contributing studies.	10
	20b	Present results of all statistical syntheses conducted. If meta-analysis was done, present for each the summary estimate and its precision (e.g. confidence/credible interval) and measures of statistical heterogeneity. If comparing groups, describe the direction of the effect.	NA
	20c	Present results of all investigations of possible causes of heterogeneity among study results.	NA
	20d	Present results of all sensitivity analyses conducted to assess the robustness of the synthesized results.	NA
Reporting biases	21	Present assessments of risk of bias due to missing results (arising from reporting biases) for each synthesis assessed.	10
Certainty of evidence	22	Present assessments of certainty (or confidence) in the body of evidence for each outcome assessed.	NA
DISCUSSION
Discussion	23a	Provide a general interpretation of the results in the context of other evidence.	18
	23b	Discuss any limitations of the evidence included in the review.	22
	23c	Discuss any limitations of the review processes used.	22
	23d	Discuss implications of the results for practice, policy, and future research.	21
OTHER INFORMATION
Registration and protocol	24a	Provide registration information for the review, including register name and registration number, or state that the review was not registered.	NA
	24b	Indicate where the review protocol can be accessed, or state that a protocol was not prepared.	NA
	24c	Describe and explain any amendments to information provided at registration or in the protocol.	NA
Support	25	Describe sources of financial or non-financial support for the review, and the role of the funders or sponsors in the review.	23
Competing interests	26	Declare any competing interests of review authors.	23
Availability of data, code and other materials	27	Report which of the following are publicly available and where they can be found: template data collection forms; data extracted from included studies; data used for all analyses; analytic code; any other materials used in the review.	NA

From: [16].

Authors’ contributions

Wei Zhang conceptualized the study, designed the methodology, performed literature screening/data extraction, conducted formal analysis, and drafted the original manuscript. Ziyun Ding developed search strategies, validated data extraction, performed quality assessment of included studies, and contributed to manuscript revision. Maxim Bakaev spearheaded technical analysis of VR hardware/software specifications in medical applications. Olga Razumnikova led bias assessment for technologically-focused studies and analyzed data in VR training systems. Magdalena Kludacz-Alessandri contributed to European literature inclusion, validated methodological rigor, and edited the discussion on international implementation challenges. Jianjie Wu supervised the entire project, resolved academic disputes during screening, and finalized the manuscript for submission. All authors collaborated on the selection of the final paper collection and contributed to crafting the conclusion. The final version of the paper received approval from all authors.

Funding

This paper is supported by National Natural Science Foundation of China (Project No. 72104087), Chunhui Project from Ministry of Education in China (Project No. HZKY20220326) and Key Project of Higher Education Scientific Research Planning Project of Chinese Association of Higher Education(Project No. 24YJ0301).

Data availability

Not applicable.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.