Time-Dependent Heterogeneity of Treatment Effects Community-Based Rehabilitation and Hospital-Based Stroke Rehabilitation: A Secondary Meta-Analysis

Purpose: This study aimed to assess the comparative effects of community-based rehabilitation (CBR) and hospital-based rehabilitation (HBR) on activities of daily living (ADL) in stroke survivors, with a specific focus on time-dependent heterogeneity of treatment effects (HTE).Methods: A secondary meta-analysis was conducted using summary data from 16 studies included in a prior systematic review.The main outcome was the Modified Barthel Index (MBI), and the effect size was calculated as the standardized mean difference (SMD).A randomeffects model using restricted maximum likelihood (REML) was employed to account for inter-study variability.Subgroup analysis by intervention duration (2, 3, 6, and 12 months) was performed using meta-ANOVA, and influence and publication bias analyses were conducted to assess the robustness of findings. Results:The overall pooled effect favored CBR over HBR (SMD= 0.60, 95% CI: 0.22-0.97),with substantial heterogeneity (I 2 = 90%).Subgroup analysis revealed that the effect was statistically significant at 3 and 6 months but not at 2 or 12 months.Meta-ANOVA showed no significant differences between subgroup effect sizes.Influence diagnostics indicated robustness of result, and publication bias analysis revealed minimal distortion after trim-and-fill adjustment.Conclusion: CBR demonstrates a significant overall advantage over HBR in improving ADL in stroke patients.While an observed pattern suggests a peak effect within a 3-to 6-month window, the lack of statistical significance between subgroups warrants further investigation of this time-dependent trend.These findings underscore intervention duration as a critical factor in optimizing stroke rehabilitation planning.