Robust immediate and limited long-term benefit of prism adaptation on spatial neglect: A Systematic Review and Meta-Analysis of outcomes and predictors

ObjectiveThe first aim was to assess the immediate and long-term effects of prism adaptation (PA) on spatial neglect (SN) via meta-analysis of randomized controlled trials (RCTs). The second aim was to identify potential predictive factors of PA efficacy. MethodsWe searched 4 databases up to March 2025 for RCTs comparing PA or PA+conventional vs. shamPA or shamPA+conventional or only conventional in participants with SN, and assessing behavioral or neuropsychological tests (such as Catherine Bergego Scale (CBS) and Behavioral Inattention Test (BIT)). Cochrane risk-of-bias assessment tool and random-effects meta-analysis were used, and effect size was reported as Weighted Mean Difference (WMD) with 95%CI. ResultsOnly 10 RCTs (356 participants) could be included, 8 of them reporting CBS data and 6 for BIT. Immediate PA effects on SN showed a significant improvement for CBS between the PA group and control group (all studies with prism shift [&ge;]10{degrees}, WMD= -2.13, 95%CI: [-3.93, -0.33], P<0.05), while non-significance for BIT-C. Long-term benefit was not significant for CBS, while significant for BIT-C when the prism shift was [&ge;]10{degrees} (WMD= 12.37, 95%CI: [0.53, 24.21], P<0.05). Linear regressions showed non-significant predictive factors among the participant characteristics or PA intervention parameters. Subgroup analyses for CBS showed a significant immediate improvement in the larger total quantity of prism exposure (number of total trials x prism shift) subgroup (WMD= -2.73, 95%CI: [-5.01, -0.44], P<0.05), whereas subgroup with smaller total exposure showed non-significance. ConclusionEven with stringent inclusion criteria, robust effects of PA on SN were observed in the short-term CBS (all studies with prism shift [&ge;]10{degrees}), mainly derived from studies with total exposure [&ge;]11250{degrees}{middle dot}trials. A significant improvement for long-term BIT-C was observed when the prism shift was [&ge;]10{degrees} (n=2, with 4000{degrees}{middle dot}trials and 21600{degrees}{middle dot}trials). The total quantity of prism exposure ({degrees}{middle dot}trials) may be a useful predictive efficacy factor of efficacy.