Effects of virtual reality interventions on anxiety symptoms in women undergoing gynecological examinations and surgeries: a multi-level dose–response meta-analysis

Objective Using a systematic review and multi-level meta-analysis, this study quantified the effects of virtual reality (VR) interventions on anxiety symptoms among women undergoing gynecological examinations or surgeries. Furthermore, it examined potential sources of effect heterogeneity and the relationship between intervention parameters and therapeutic efficacy, thereby providing an evidence-based reference for the rational application of VR interventions in clinical settings. Methods This study followed the PRISMA guidelines. We conducted a systematic search of the PubMed, Web of Science, PsycINFO, and Cochrane Library databases up to January 5, 2026, and included randomized controlled trials (RCTs) that evaluated the effects of VR interventions on anxiety symptoms among women undergoing gynecological examinations or surgeries. A multi-level random-effects model was used to pool effect sizes, specifically the SMD. We also conducted heterogeneity assessments, publication bias tests, subgroup analyses, meta-regressions, and Bayesian dose–response analyses. Results Twelve randomized controlled trials published between 2019 and 2025 were included. The meta-analysis showed that the VR intervention significantly reduced anxiety symptoms in women during the intervention period, with a small-to-moderate effect (SMD = −0.47, 95% CI − 0.73 to −0.21, p 0.001). The quality of evidence was rated as moderate according to the GRADE criteria. Subgroup analyses suggested more consistent effects in non-invasive procedures, with VR alone, and when anxiety was assessed using the Numerical Rating Scale (NRS). Subgroup analyses suggested more consistent effects for interventions lasting 5–10 min, whereas no statistically significant effect was observed in highly invasive procedures. Linear meta-regression identified country/region, anxiety assessment tool, and methodological quality as significant moderators. Non-linear analyses suggested potential relationships with age, sample size, and intervention duration. A potential duration range of 9–13 min was identified, but this finding remained exploratory. Conclusion Virtual reality interventions were associated with reduced anxiety symptoms, although substantial heterogeneity may limit the generalisability of the pooled effect. As a non-pharmacological digital intervention, VR may serve as a promising adjunct to routine care, particularly in less invasive gynecological settings. Further multicenter studies with larger samples and standardized designs are needed to verify long-term effects and clarify optimal intervention parameters. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/ .