摘要
This letter comments on the randomized controlled trial conducted by Kolit et al. (2025), in which they examined the effectiveness of a virtual reality (VR) program combined with conventional occupational therapy (COT) for children with developmental coordination disorder (DCD). I suggest a more cautious interpretation of the findings.First, only gross motor skills and praxis showed between-group advantages for the VR group, whereas cognitive, visual–perceptual, and activity measures did not differ between groups. Therefore, conclusions regarding the superiority of VR across all domains may not be fully supported.Second, the inference of treatment effectiveness appears to rely heavily on within-group pre–post changes and associated effect sizes. In a randomized trial, such within-group improvements, which may occur in both groups because of maturation or nonspecific factors, cannot replace between-group comparisons as the primary basis for causal inference.Third, the authors assessed numerous outcomes without explicitly addressing multiplicity or adjusting for multiple testing, which increases the risk of Type I error and overinterpretation of marginally significant results. I recommend that future reports clearly specify primary and secondary outcomes, apply appropriate multiplicity control, and present confidence intervals (CIs) alongside adjusted p values. A more conservative interpretation would strengthen the contribution of this trial and provide a more precise estimate of the potential role of VR as an adjunct to pediatric occupational therapy for children with DCD. Such rigor will facilitate synthesis and guide clinical decision-making in occupational therapy practice.I read Kolit et al.’s (2025) report of their randomized trial with interest and appreciate its International Classification of Functioning, Disability and Health–aligned outcome mapping, pragmatic Wii Fit Plus protocol with clear task progressions, and single-blind design—these are valuable contributions to pediatric rehabilitation.While acknowledging the study’s strengths, I offer the following observations to help clarify and contextualize the interpretations:Clarify between-group inferences. In their Discussion section, Kolit et al. (2025) conclude that adding VR to COT was “more effective in improving motor, cognitive, sensory (praxis), and perceptual (visual) skills, as well as activity levels.” However, the posttest between-group comparisons (Kolit et al., 2025, Table 4) showed statistically significant advantages only for the Test of Gross Motor Development (2nd ed.; TGMD–2) and the Sensory Integration and Praxis Test (SIPT); results for the Childhood Executive Functioning Inventory (CHEXI [working memory, p = .43; inhibitory control, p = .88), Motor-Free Visual Perception Test (3rd ed.; MVPT–3; p = .07), and Functional Independence Measure for Children (WeeFIM; p = .74) were not significant. A more precise summary would state that VR + COT demonstrated clear superiority for gross motor skills and praxis, with no confirmed between-group advantage for executive function, visual perception, or activity independence.Distinguish within-group change and effect sizes versus superiority. Table 3 of Kolit et al.’s (2025) article presents large within-group effect sizes in the study arm for several outcomes (e.g., TGMD–2, d = 2.7; CHEXI, d = 0.8; WeeFIM, d = 0.8) and a moderate effect for the MVPT–3 (d = 0.7). Because their analyses relied on Wilcoxon (within-group) and Mann–Whitney (between-group) tests with p < .05, without an explicit Group × Time interaction or baseline-adjusted models, within-group gains alone are insufficient to infer cross-group superiority. I suggest reporting baseline-adjusted between-group estimates (e.g., analysis of covariance [ANCOVA] or rank-based ANCOVA/mixed models) or directly comparing change scores, accompanied by CIs, to buttress causal interpretations.Address multiplicity in the analysis of multi outcomes. Numerous outcomes were evaluated and SIPT subtests administered, yet no multiplicity adjustment is described. Several posttest p values are borderline (e.g., SIPT Verbal Command, p = .040; SIPT Postural Praxis, p = .020; SIPT Sequencing, p = .020) and could attenuate under Holm–Bonferroni or false discovery rate control. Prespecifying primary endpoints (e.g., for the TGMD–2) and providing corrected p values and CIs would align the statistical evidence with the clinically meaningful narrative already suggested by the motor and praxis findings.In sum, Kolit et al.’s (2025) trial offers promising, clinically relevant evidence that VR augments COT for gross motor skills and praxis in children with DCD while signaling encouraging but as-yet-unconfirmed advantages in executive function, visual perception, and activity outcomes. Clarifying the claims along these lines would further strengthen an already-valuable contribution.