TF-CBT for Long-term PTSD, Major Depressive Disorder and Anxiety Disorders in Victims of Terrorism

This study will be a randomised, controlled, single-blind, single-center, and superiority trial, with two parallel groups (treatment and waiting list) and a primary endpoint based on the percentages of posttraumatic stress disorder, major depressive disorder and/or anxiety disorders, in the post-treatment (primary outcome). Mean levels of posttraumatic, depressive and anxiety symptoms in the post-treatment for both experimental and control groups, and changes in diagnostic percentages and in posttraumatic, depressive, and anxiety symptoms from baseline to 6-month follow-up for the experimental group are considered as secondary outcomes.

Researchers will contact the members of the Spanish Association of Victims of Terrorism and will invite them to an in-person diagnostic interview (pre-treatment evaluation) at a Spanish university. Eligible victims who meet the criteria for long-term posttraumatic stress disorder, major depressive disorder, and/or anxiety disorders will be randomly assigned to either experimental or control conditions. Random numbers generated by the Research Randomizer (Version 4.0) program (Urbaniack \& Plous, 2013) will be used to assign subjects to treatment conditions. Randomisation will be performed as a block with a 1:1 allocation. As part of the informed consent process, the eligible individuals will be told that they will be enrolled in a free and voluntary psychological treatment programme, remaining blind to their allocation to the experimental or control group. Both groups will receive TF-CBT, although the experimental group will do it immediately and the control group, after 4 months of waiting list condition.

The modular TF-CBT therapy programme applied in this study will consist of 16 weekly individual sessions, each lasting 60-90 minutes. It will be organized around a structured protocol and the therapist will have a "checklist" of objectives to be fulfilled within each session. The core treatment component will be exposure, based on the application of Foa's "prolonged exposure" protocol (Foa et al., 2007), but adding specific techniques to control both depressive and anxiety disorders (i.e., progressive muscle relaxation, deep breathing exercises and pleasant activities scheduling), and cognitive restructuring all along the intervention. The TF-CBT intervention will be administered by a team of psychologists with a completed postgraduate training in clinical psychology and/or general health psychology, as well as a specific training in the provision of TF-CBT to victims of terrorism. The therapists will be supervised by senior clinical psychologists and will be monitored weekly.

Calculation of the sample size will be based on diagnostic percentages (dichotomous primary outcome), with a superiority margin of 5% between the two trial arms. In order to achieve an 80% power at the 5% level of significance with equal allocation, the total sample size should be N = 140 (n = 70 in each group). However, taking into account an average drop-out rate of 23.5% in the efficacy studies conducted on victims of terrorism (Garcia-Vera et al., 2015), the aim will be to recruit N = 184.

Contingency tables and chi2 tests will be used to assess post-treatment differences between the experimental vs. control groups in the percentage of individuals meeting diagnostic criteria for posttraumatic stress disorder, major depressive disorder and anxiety disorders, the primary endpoint. The McNemar's test will be carried out for two related samples to compare the percentage of those diagnosed in the pre-treatment and in the follow-up for the experimental group.

To compare the post-treatment symptomatology level of the two groups (treatment and control), analysis of variance will be carried out with repeated measurements on each of the symptomatology measures (posttraumatic, depressive and anxiety), in which the within-subject factor will be the time (pre-treatment and post-treatment), and the between-subject factor will be the experimental condition. As effect sizes, partial eta squared and Hedge's g between two groups at post-treatment will be calculated. Moreover, a pre-follow-up analyses of the symptomatology will be conducted for the experimental group and using repeated measures t-tests, and Hedge's g between pre and follow-up will be also calculated as effect sizes.

Following the approach of Jacobson \& Truax (1991) to examine clinically significant changes, the percentages of patients in each group whose score in the post-treatment (or the follow-up) would be below the cut-off score that marked the step from clinically-significant symptomatology to "sub-syndromal" (the "C-score") will be calculated. This is 29 points for the Posttraumatic Stress Disorder Checklist, specific version (Sanz \& García-Vera, 2015) and 14 for the Beck Depression Inventory-II and the Beck Anxiety Inventory. The Reliable Change Index will be also used (Jacobson \& Truax, 1991; Mcglinchey et al., 2002). According to the Reliable Change Index data calculated by Sanz (2013, 2014; Sanz \& García-Vera, 2015) for the Spanish population, the disorders of the patients whose scores drop by 12 points or more for the Posttraumatic Stress Disorder Checklist, specific version, or 10 points or more in the Beck Anxiety and Depression Inventories, will be designated as "improved". The chi2 tests will be carried out to examine differences between the experimental and control groups for each of these indices of clinical significance.

Per-protocol analyses will be conducted with the completers and Intention-to-treat analyses will be conducted considering all the participants initially enrolled in this study.