Ires Ghielen

108 Chapter 6 In addition, subgroup analyses were conducted (table 2) to investigate differences in effect size for disease type, control condition, and risk of bias (high vs low). Significantly larger effect sizes were found in MS patient populations, TAU control condition, and studies with a high risk of bias. Meta-regression analyses on risk of bias (β = 0.08, 95% CI = -0.18 – 0.33, p >0.05) did not show a significant relationship with effect size. Meta-analysis 2: psychological therapy versus active control condition Figure 3 shows the forest plot of the standardized effect sizes of psychological therapies on psychological distress in MS patients, compared with an active control condition. The mean effect size ( g ) was 0.36 (95% CI = 0.13 — 0.58) with a heterogeneity estimate (I²) of 40 (95% CI = 0 — 74). To investigate the treatment effects for the types of investigated intervention, two separate meta-analyses were conducted on CBT-based treatments and MBTs. The treatment effect for CBT-based treatments was highest ( g = 0.45, 95% CI = 0.26 — 0.64 with I² = 0, 95% CI = 0 — 73), followed by a small effect for MBTs ( g = 0.06, 95% CI = -0.56 — 0.68 with I² = 68, 95% CI = 0-89) (see table 2). Publication bias No evidence for publication bias was found in both meta-analyses. Inspection of the funnel plots did not indicate significant publication bias (figures 4 & 5). Duval & Tweedie’s trim-and-fill procedure resulted in the imputation of four studies in the first meta-analysis, and no imputations in the second meta-analysis, according to a random model. Egger’s regression intercept indicated no significant publication bias (p > 0.05 in both analyses). Discussion In this study, we investigated the effectiveness of CBT and MBT on psychological distress in patients with PD and MS by conducting two main meta-analyses of randomized controlled trials. There were no RCTs found studying these therapies in HD populations. Nineteen studies were included in the analyses, of which twelve compared the treatment of interest with a TAU or waitlist condition (meta-analysis 1), and seven studies compared the treatment of interest with an active control