Navigating Virtual Reality in Stroke Rehabilitation: Scoping Review of Diverse Intervention Effects.

Background: Virtual reality (VR) technology has been increasingly explored in stroke rehabilitation due to its immersive and interactive features. However, considerable heterogeneity exists in intervention designs, study populations, and outcome measures, limiting the feasibility of conducting a systematic review. Objective: This scoping review aims to comprehensively map randomized controlled trials (RCTs) investigating the use of VR interventions in stroke rehabilitation, with particular focus on upper limb function, gait and balance, cognitive function, and quality of life. Methods: Following the Arksey and O'Malley framework and PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines, we conducted a scoping review of RCTs published in databases from their inception to January 4, 2025. Seven databases were searched, including PubMed, Embase, Web of Science, CNKI, Wanfang, VIP, and CBM. Studies were included if they met predefined eligibility criteria, including adult patients with stroke receiving VR-based rehabilitation and a randomized controlled trial design. Exclusion criteria included non-Chinese or non-English literature, literature with unavailable full text, studies with duplicate publication or data, and studies that were irrelevant to the research topic or did not incorporate VR technology in their intervention measures. Data extracted included intervention type, sample size, training frequency and duration, outcomes, and study setting. Due to significant heterogeneity across studies, a narrative synthesis approach was used. No formal risk of bias or quality appraisal was conducted. Results: Fifteen RCTs involving 804 patients with stroke were included. Intervention modalities varied significantly in terms of type, content, frequency, and duration. Nonimmersive VR (NIVR) interventions were more frequently applied in studies targeting upper limb function and gait training, while fully immersive VR (FIVR) was assessed in 2 head-mounted display (HMD)-based trials, whereas most studies used NIVR for upper-limb and gait-related outcomes. Many studies reported positive trends in motor function, cognitive performance, gait balance, and quality of life. However, findings were inconsistent, and not all outcomes reached statistical significance. Mild adverse events, such as fatigue or dizziness, were occasionally reported; however, no serious events occurred. Conclusions: This scoping review outlines the research status of VR in stroke rehabilitation. VR may offer potential benefits; however, existing studies have limitations, including substantial heterogeneity in intervention protocols, limited long-term follow-up, and baseline imbalances in some trials. In addition, because this review did not include a formal quality or risk-of-bias assessment, the observed effects should be interpreted as preliminary signals rather than definitive evidence of efficacy, and the certainty of the evidence cannot be determined. Future research should standardize outcome measures, improve methodological rigor, and incorporate quality and risk-of-bias evaluation to strengthen the evidence base and support clinical implementation.