Ires Ghielen

111 Meta-analyses on psychotherapies in MS and PD condition (meta-analysis 2). A moderate effect size ( g = 0.51) was found in the first meta-analysis, and a small effect size ( g = 0.36) was found in the second meta- analysis. In both meta-analyses there was considerable heterogeneity, which was probably due to the variability in hours of treatment (range: 5-16 hours), different delivery forms (for example by telephone or face-to-face), differences in comorbidity in the included subjects over all included studies, and specific elements that varied between the interventions (such as psycho education in the study of Okai et al [41]). The heterogeneity decreased when 1) outliers were removed; 2) depression and anxiety outcomes were analyzed separately; 3) only looking at MS patient samples; 4) studies with waitlist control conditions were analyzed; and 5) studies with a low Figure 4. Funnel plot of meta-analysis 1: psychological intervention versus TAU or WL con- ditions. TAU = treatment as usual; WL = waitlist. The imputed studies are shown in black. Figure 5. Funnel plot of meta-analysis 2: psychological intervention versus active control conditions. 6