Description of studies
The literature search for clinical guidelines and systematic reviews was conducted in February 2023, and one45 was selected to form the basis of our review. As the literature search of that review was conducted on 1 July 2013, and the search strategy was evaluated to be adequate by means of the AMSTAR 2,39 our search for additional RCTs was set from this date and conducted on 5 May 2023. PRISMA flowcharts for clinical guidelines, systematic reviews and RCTs are provided in online supplemental E, figure S1, S7.1 and S7.2, respectively.
A total of 55 RCTs were included in the Q3 meta-analysis across the primary and secondary outcomes, reporting on 41 different psychotherapies. For meta-analysis, interventions were grouped in the following 11 intervention groups: (1) mindfulness and body-focused psychotherapies, (2) present-centred therapy (PCT), (3) cognitive therapy (CT), (4) interpersonal psychotherapy (IPT), (5) prolonged exposure (PE), (6) CPT, (7) virtual reality exposure, (8) CBT, (9) imagery, (10) NET and (11) eye movement desensitisation and reprocessing (EMDR). For study details, see online supplemental F, table S3.
Risk of bias assessment
Of the 48 RCTs reporting on the primary outcome of PTSD symptoms, 6 were considered to raise some concerns and the remaining to be at a high risk of bias. All 51 RCTs reporting on the primary outcome of depressive symptoms were deemed at a high risk of bias. See online supplemental E, figure S8.1–S8.6 for RoB 2 assessments, and online supplemental figures S9.1–S9.9 for funnel plots.
Primary outcomes
We found a clinically significant difference in treatment effect on PTSD symptoms at the end of treatment, favouring the following psychotherapies over their respective comparators. Mindfulness and body-focused psychotherapies were favoured over mind–body intervention, TAU and waitlist (SMD −0.81, 95% CI −1.28 to −0.34, p=0.0008, I2=41%; 3 trials, n=137, moderate certainty). NET was favoured over trauma-focused psychotherapy, TAU and waitlist (SMD −0.66, 95% CI −1.22 to −0.10, p=0.02, I2=65%; 5 trials, n=175, low certainty) and when removing the active comparators, NET showed an SMD of −1.07 (95% CI −1.22 to −0.10, p<0.00001, I2=0%; 3 trials, n=175). CPT was favoured over trauma-focused psychotherapy, psychotherapy, TAU, sertraline placebo and alternative intervention (SMD −0.76, 95% CI −1.36 to −0.17, p=0.01, I2=95%; 8 trials, n=1564, low certainty) and when removing the outlier sertraline placebo, CPT showed an SMD of −0.32 (95% CI −0.68 to 0.03, p=0.07, I2=86%; 7 trials, n=1475). A statistically significant difference favoured PE over trauma-focused therapy, non-trauma-focused therapy, TAU, waitlist and alternate intervention (SMD −0.35, 95% CI −0.64 to −0.05, p=0.02, I2=78%; 11 trials, n=1567, low certainty) and when removing the outlier waitlist, PE showed an SMD of −0.16 (95% CI −0.32 to 0.01, p=0.06, I2=33%, 10 trials, n=1520, low certainty).
We found a clinically significant difference in treatment effect on depressive symptoms at the end of treatment, favouring the following psychotherapies over their respective comparators. Mindfulness and body-focused psychotherapies were favoured over mind-body intervention, TAU and waitlist (SMD −0.81, 95% CI −1.36 to −0.26, p=0.004, I2=76%; 5 trials, n=247, low certainty). NET was favoured over trauma-focused psychotherapy, TAU and waitlist (SMD −0.77, 95% CI -1.38 to −0.15, p=0.01, I2=69%; 5 trials, n=174, low certainty). When removing the outlier waitlist, NET showed an SMD of −0.45 (95% CI −0.77 to −0.13, p=0.006, I2=0%, 4 trials, n=155). CPT was favoured over trauma-focused psychotherapy, psychotherapy, TAU, sertraline placebo, alternative intervention (SMD −1.21, 95% CI −2.01 to −0.42, p=0.003, I2=97%; 8 trials, n=1663, low certainty). When removing the outlier sertraline placebo, CPT showed an SMD of −0.58 (95% CI −1.18 to 0.01, p=0.05, I2=95%; 7 trials, n=1574). PE was favoured over trauma-focused therapy, non-trauma-focused therapy, TAU, waitlist and alternate intervention (SMD −0.29, 95% CI −0.58 to 0.00, p=0.05, I2=62%; 8 trials, n=1261, very low certainty). When removing the four active comparators, PE showed an SMD of −0.68 (95% CI −1.06 to −0.30, p=0.005, I2=21%; 4 trials, n=165). CBT compared to trauma-focused psychotherapy, psychotherapy, TAU, waitlist, and an online intervention (SMD -0.30, CI 95% -0.58 to -0.02, p = .03, I2=68%, 10 trials, n = 715, very low certainty)
We found a difference in treatment effect on depressive symptoms, favouring trauma-focused therapy and non-trauma-focused therapy over PCT (SMD −1.33, CI 95% –2.61 to −0.04, p=0.04, I2=93%; 5 trials, n=225, low certainty). No significant difference in effect was found on PTSD symptoms at the end of treatment between PCT and trauma-focused psychotherapy.
After removing the outliers described above, there were no significant subgroup differences in any of the therapies for PTSD nor depression outcomes, except for CPT, where significant subgroup differences remained. However, these differences appear to be driven by the active comparators. See online supplemental H for subgroup analyses. All analyses with significant results are reported in online supplemental G, table S3.
