Table 4.
Grading the evidence of secondary outcomes ranked by class of evidence and year of publication.
| Outcome | First author and year | Intervention and control group | RCTs | Follow-up | Sample size | Hedges' G1 | p-value | I2 (%) | PI 95CI% | SSE/ESB/LS | CE |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Affective symptoms | Jones, Hacker, Meaden et al., 2018 | CBT vs SP-PT (25%), PE (24%), SAT (20%), E-ST (17%), SC (12%), ST (3%) | 6 | <6 months | 400 | −0.24 (−0.67; 0.19) | 0.28 | 81.74 | Null | No/No/Yes | NS |
| CBT vs PE (42%), E-ST (34%), SC (24%) (PANSS) | 3 | 6–12 months | 194 | −0.12 (−0.72; 0.48) | 0.69 | 77.28 | Null | No/No/No | NS | ||
| CBT vs ST (40%), SAT (20%), E-ST (17%), PE (13%), FI (10%) | 7 | >12 months | 379 | −0.14 (−0.36; 0.08) | 0.21 | 14.13 | Null | No/No/No | NS | ||
| Anxiety | Jones, Hacker, Xia et al., 2018 | CBT vs TAU (100%) (BAI) | 3 | >12 months | 335 | 0.12 (−0.10; 0.34) | 0.27 | 0 | Null | No/No/No | NS |
| Depressive symptoms | – | ||||||||||
| Rehospitalisation | Jones, Hacker, Meaden et al., 2018 | CBT vs PE (84%), ST (16%) | 3 | 6–12 months | 78 | −0.08 (−0.31; 0.14) | 0.48 | 0 | Null | Yes/No/No | NS |
| CBT vs PE (34%), ST (21%), CR (18%), BF (14%), E− ST (7%), RT (6%) | 8 | >12 months | 595 | −0.02 (−0.10; 0.06) | 0.64 | 0 | Null | No/No/No | NS | ||
| Jones, Hacker, Xia et al., 2018 | CBT vs TAU (100%) | 6 | >12 months | 648 | −0.12 (−0.28; 0.03) | 0.11 | 0 | Null | No/No/No | NS | |
| Relapse | Jones, Hacker, Xia et al., 2018 | CBT vs TAU (100%) | 5 | 6–12 months | 667 | −0.35 (−0.52; −0.18) | 0.60 × 10−4 | 0 | Not Null | Yes,/No/Yes | IV |
| Jones, Hacker, Meaden et al., 2018 | CBT vs TAU (51%), SC (17%), RT (17%), FI (15%) | 5 | >12 months | 375 | 0.02 (−0.09; 0.13) | 0.72 | 0 | Null | No/No/No | NS | |
| Jones, Hacker, Xia et al., 2018 | CBT vs TAU (100%) | 13 | >12 months | 1538 | −0.14 (−0.29; 0.001) | 0.05 | 51.47 | Null | Yes/No/No | NS | |
| Lynch et al., 2010 | CBT vs TAU (51%), SC (29%), ST (11%), PE (9%) | 8 | 6–36 months | 976 | 0.05 (−0.18; 0.29) | 0.66 | 41.10 | Null | No/No/Yes | NS | |
| Functioning and distress | Bighelli et al., 2023 | CBT vs TAU (64%), inactive control (8%), ST (10%), CR (9%), WL (6%), FI (1%), psychodynamic therapy (1%), PE (1%) | 30 | EoT | 2657 | 0.24 (0.11; 0.38) | 0.50 × 10−3 | 58.18 | Null | No/No.No | III |
| Laws et al., 2018 (distress) | CBT vs TAU (83%), unknown (11%), WL (6%) | 8 | EoT | 573 | 0.37 (0.06; 0.67) | 0.02 | 58.76 | Null | No/No/No | IV | |
| CBT vs TAU (61%), CR (14%), ST (9%), medication (5%), SAT (5%), BF (4%), PE (2%), WL (1%) | 26 | EoT | 1704 | 0.25 (0.10; 0.40) | 0.76 × 10−3 | 54.50 | Null | No/No/No | IV | ||
| Bighelli et al., 2023 | CBT vs TAU (100%) | 19 | EoT | 1682 | 0.34 (0.14. 0.53) | 0.65 × 10−3 | 63.68 | Null | Yes/No/No | IV | |
| Bighelli et al., 2023 | CBT (third-wave, mindfulness (63%), ACT (9%), MCT (28%)) vs mixed control | 20 | EoT | 1391 | 0.58 (0.36; 0.80) | 0.29 × 10−6 | 71.94 | Null | No/No/Yes | IV | |
| Bighelli et al., 2023 | CBT (third-wave, mindfulness (100%) vs mixed control | 12 | EoT | 989 | 0.71 (0.45; 0.96) | 0.57 × 10−7 | 68.70 | Null | No/No/Yes | IV | |
| Bighelli et al., 2023 | CBT (third-wave, MCT (100%)) vs mixed control | 7 | EoT | 155 | 0.45 (0.01; 0.88) | 0.04 | 67.96 | Null | No/No/No | IV | |
| Laws et al., 2018 | CBT vs TAU (61%), CR (14%), ST (9%), medication (5%), SAT (5%), BF (4%), PE (2%), WL (1%) | 16 | 3–18 months | 938 | 0.10 (−0.07; 0.27) | 0.23 | 34.84 | Null | No/No/No | NS | |
| Jones, Hacker, Xia et al., 2018 | CBT vs TAU (100%) | 5 | >12 months | 446 | 0.15 (−0.14; 0.44) | 0.32 | 44.47 | Null | No/No/No | NS | |
| CBT vs TAU (100%) | 5 | 6–12 months | 482 | 0.22 (−0.12; 0.55) | 0.21 | 59.81 | Null | No/Yes/No | NS | ||
| Jones, Hacker, Meaden et al., 2018 | CBT vs NS-C (55%), SAT (45%) | 3 | >12 months | 224 | 0.22 (−0.29; 0.73) | 0.02 | 72.44 | Null | No/No/No | NS | |
| Bighelli et al., 2023 | CBT vs goal-focused supportive contact (56%), SAT (27%), ACT (17%) | 4 | EoT | 229 | 0.01 (−0.40; 0.43) | 0.96 | 58.23 | Null | No/No/No | NS | |
| Quality of life | Laws et al., 2018 | CBT vs TAU (62%), PE (15%), WL (8%), medication (8%), TAU + medication (7%) | 10 | EoT | 626 | 0.04 (−0.12; 0.19) | 0.64 | 0 | Null | No/No/No | NS |
| Social functioning | Barnicot et al., 2020 | CBT vs TAU (65%), ND-SC (35%) | 3 | EoT | 175 | 0.66 (−0.07; 1.39) | 0.08 | 69.71 | Null | No/No/Yes | NS |
Abbreviations Tables 2 and 3: PI: prediction interval, SSE: small study effect, ESB: excess significance bias, LS: largest study, CE: class of evidence, RCT: randomized controlled trial, EoT: end of therapy, ES: effect size, CBT: cognitive behavioral therapy, TAU: treatment as usual, PE: psychoeducation, NS-SC + PE: supportive counseling + psychoeducation, ST: supportive therapy, SC: supportive counseling, SAT: Social Activity Therapy, PANSS: Positive and Negative Syndrome Scale, SST: Social Skills Training, BPRS: Brief Psychiatric Rating Scale, BF: befriending, RT: Recreational Therapy, CR: Cognitive remediation, WL: waiting list, AT: active treatment, BAI: Beck Anxiety Inventory, ND-SC: Non-Directive Supportive Counseling, NS-SC: , FI: , FT: family intervention, GS: group support, ET: exposure therapy, AC: Attention (Placebo) Control, NS-CG: Non-specific counselling group, SP + PT: Standard psychological support + pharmacological therapy, NS-C: Non-specific counselling, E-ST: enhanced supportive therapy, GF-SC: Goal-focused supportive contact, RS: Recreation and Support, GFT: goal focused therapy, Positive and Negative Syndrome Scale, BPRS: Brief Psychiatric Rating Scale, ETAU: enriched treatment as usual.
In some cases the summary estimates from our analysis differed one-hundredth decimal compared to the original findings. We used the data provided by the report of the meta-analysis and expect that variation between estimates adopted from the primary RCTs and estimates provided in the meta-analysis have caused this slight and clinically irrelevant variation.