Effectiveness of virtual reality in training of in-service nurses : a systematic review and a meta-analysis

With the advent of the information technology era, innovation in nursing education and training strategies must keep pace with the times to support the professional development of nurses. This review aimed to assess the effects of virtual reality (VR) training on nurses’ knowledge levels, practical skills, self‑efficacy, and critical thinking abilities. A comprehensive literature search was conducted across international databases (PubMed, Web of Science, Embase, and the Cochrane Library) and Chinese databases (SinoMed, CNKI, Wanfang, and VIP), covering all relevant studies in English and Chinese published up to April 2026. RoB 2.0 and ROBINS-I were used to assess risk of bias for randomized controlled trials (RCTs) and quasi-experiments respectively. RCTs were analyzed as the primary evidence base with complete meta-analytic procedures, including subgroup analysis, sensitivity analysis, and publication bias assessment. Quasi-experimental studies were synthesized separately as supplementary evidence to support the overall findings. Standardized mean differences (SMD) was used for effect size pooling, and the GRADE system was adopted to evaluate evidence certainty stratified by study design. 30 studies met the inclusion criteria and were included in the review, including 16 randomized controlled trials and 14 quasi‑experimental studies. The meta-analysis indicated that, relative to conventional training, VR training may improve nurses’ knowledge levels (SMD = 1.74, 95% CI: 1.10–2.38, p < 0.05), practical skills (SMD = 2.19, 95% CI: 1.35–3.02, p < 0.05), self-efficacy (SMD = 0.29, 95% CI: 0.03–0.54, p = 0.57), and critical thinking ability (SMD = 1.85, 95% CI: 0.93–2.77, p < 0.05). High heterogeneity and publication bias were observed for knowledge and practical skills outcomes. The Duval and Tweedie trim-and-fill method revealed no evidence of publication bias-related missing studies, with adjusted pooled effect sizes remaining consistent with the original estimates. Evidence certainty was evaluated using the GRADE system. The certainty of evidence was rated as very low for knowledge level and practical skills, high for self‑efficacy, and moderate for critical thinking ability. VR training may produce beneficial effects on nurses’ knowledge levels, practical skills, self-efficacy and critical thinking ability. Nevertheless, high heterogeneity, publication bias and small sample existed for outcomes, and the GRADE assessment shows very-low-to-high evidence certainty across different indicators. These findings should be interpreted with caution.