Background Virtual reality (VR) technology has emerged as a transformative therapeutic innovation in healthcare, providing immersive three-dimensional, multisensory environments that facilitate experiential learning, rehabilitation, and engagement. While numerous studies highlight its benefits, empirical research on VR-based interventions for autism spectrum disorder (ASD) and physical therapy (PT) remains limited by methodological inconsistencies, small sample sizes, and variable intervention designs. These limitations constrain the generalizability of findings and delay VR integration into standard therapeutic practice. Objective This study conducted a systematic review and meta-analysis to assess the therapeutic effectiveness of VR interventions within ASD and PT contexts. It aimed to determine the extent of improvement achieved through VR-based therapies, identify methodological and practical challenges affecting clinical adoption, and classify the dominant types and features of VR interventions applied across both domains. Methods Following PRISMA 2020 guidelines, a comprehensive search was performed across PubMed, Scopus, and Web of Science for studies published between 2020 and 2025. Eligible peer-reviewed articles provided empirical evidence on VR interventions for autistic individuals or patients undergoing physical rehabilitation. Twenty-three studies meeting inclusion criteria underwent methodological quality assessment and systematic data extraction. Quantitative synthesis used standardized effect size calculations (Hedges' g) and subgroup analyses to evaluate intervention outcomes and contextual moderators. Results The meta-analysis showed a significant positive impact of VR interventions across both domains (Hedges' g = 0.66, p 0.001), representing a moderate-to-large therapeutic effect. Among autistic participants, major improvements appeared in social communication, emotion recognition, and adaptive behaviors. In PT populations, VR interventions enhanced balance, gait, and upper-limb motor recovery. Fully immersive systems and programs exceeding eight weeks produced the most substantial clinical benefits. Quality analysis showed that 56.5% of studies met the highest standards, with others slightly lower due to missing control groups or incomplete ethical reporting. Conclusion VR is an effective adjunctive therapy that enhances motivation, engagement, and adherence. Despite barriers of cost, training, and protocol variability, its scalability and adaptability support broad clinical use. AI-driven personalization, mobile VR, and tele-rehabilitation offer promising future directions. However, these findings should be interpreted with caution given the median sample size of 45, moderate heterogeneity (I 2 = 44.3%), 43.5% of studies without a control group, and statistically significant funnel plot asymmetry indicative of potential publication bias (Egger's test: p = 0.0031).
Tawil et al. (Thu,) studied this question.