Psychological therapies for post‐traumatic stress disorder in children and adolescents

Objectives

This is a protocol for a Cochrane Review (intervention). The objectives are as follows:

To assess the effects (i.e. benefits and harms) of psychological therapies for post‐traumatic stress disorder and complex post‐traumatic stress disorder in children and adolescents.

Background

Description of the condition

Post‐traumatic stress disorder (PTSD) is a severe and often chronic mental disorder (Galatzer‐Levy 2018; Kessler 2017), characterised by persistent stress reactions following exposure to (i.e. experiencing or witnessing) one or more traumatic events (APA 2022; WHO 2022a). The Diagnostic and Statistical Manual of Mental Disorders (DSM‐5‐TR) defines a traumatic event as “exposure to actual or threatened death, serious injury, or sexual violence” (APA 2022), while the International Classification of Diseases (ICD‐11) emphasises the subjective perception of having been exposed to “an extremely threatening or horrific event or series of events” (WHO 2022a). Across both diagnostic manuals, PTSD is characterised by the following three symptom clusters: (1) a sense of current threat (e.g. hypervigilance and increased reactivity), (2) re‐experiencing the trauma through flashbacks or nightmares, and (3) avoidance of internal and external reminders of the traumatic event (APA 2022; WHO 2022a). In addition, the DSM‐5‐TR lists negative changes in cognitions and mood as a fourth symptom cluster, while in the ICD‐11, a similar symptom cluster is represented in the distinct diagnosis of complex PTSD (CPTSD). Specifically, CPTSD comprises ICD‐11’s three symptom clusters of PTSD, plus three symptom clusters labelled 'disturbances in self‐organisation' (DSO): (1) difficulties in regulating emotions, (2) a negative sense of self, and (3) interpersonal difficulties (Cloitre 2020; WHO 2022a).

Epidemiological studies indicate that 31% to 62% of people have been exposed to at least one traumatic event before the age of 18 years (Landolt 2013; Lewis 2019; McLaughlin 2013). However, while most report acute stress reactions in the first weeks following trauma exposure, these reactions are usually temporary (Galatzer‐Levy 2018; Hiller 2016). Nevertheless, studies estimate that 7% to 25% of trauma‐exposed children and adolescents develop PTSD (Alisic 2014; Landolt 2013; Lewis 2019; McLaughlin 2013). The broad range of prevalence estimates may be the result of methodological variations (e.g. different assessment strategies and sample sizes), cultural differences between study populations, and variations in the prevalence of different types of trauma exposure (i.e. different sources, severities, or frequencies of exposure). Few studies have examined the prevalence of CPTSD, but current evidence from both paediatric (Redican 2022), and adult populations (Cloitre 2018a; Cloitre 2019; Hyland 2021), suggests that rates of CPTSD might be slightly higher than rates of PTSD; bearing in mind that some of those currently diagnosed with overlapping conditions – such as PTSD with affective comorbidity, or borderline personality disorder – may be encompassed in this prevalence estimate (Atkinson 2024; Cloitre 2014). Evidence suggests that the higher rates of CPTSD may be attributed to CPTSD being more strongly associated with the types of trauma that pose the highest risk for psychopathology (i.e. multiple, prolonged, interpersonal, or early‐life traumas), while the reverse seems to be true for PTSD and lower‐risk traumas (i.e. those that occur in isolation, later in development, or are not interpersonal) (Cloitre 2019; Hyland 2021; Hyland 2023; Kessler 2017; Koenen 2017; Liu 2017).

A possible explanation for this association was recently proposed by the Memory and Identity Theory of CPTSD, which suggests that the stronger association between high‐risk traumas and CPTSD may be explained by high‐risk traumas being more likely to disrupt normal identity development and result in negative views of oneself (i.e. a DSO symptom) (Hyland 2023). According to this theory, differences in vulnerability (e.g. social support, adaptive coping strategies, previous psychopathology, etc.) can moderate the effects of trauma exposure on memory and identity (Hyland 2023). For example, resilient individuals may not develop traumatic memories despite trauma exposure (i.e. develop neither PTSD nor CPTSD), or they may develop a traumatic memory, but be able to maintain a positive view of themselves despite exposure to high‐risk trauma (i.e. develop PTSD rather than CPTSD) (Hyland 2023). In contrast, vulnerable individuals may develop CPTSD, even from exposure to low‐risk trauma (Hyland 2023).

The high prevalence of PTSD and CPTSD in children and adolescents has significant societal and personal costs, and without adequate treatment, the prognosis for recovery is poor (Hiller 2016; Scheeringa 2011; WHO 2022b). Despite this, research suggests that PTSD is underdiagnosed in children and adolescents, who may have limited ability to articulate their thoughts and feelings, often present with different symptoms than adults, and for whom few validated assessment tools are available (Løkkegaard 2023; Mikolajewski 2017; Scheeringa 2012). For example, children and adolescents may present with more somatic complaints (e.g. stomach aches or headaches), generalised fears, behavioural changes (e.g. aggression or repetitive play), or developmental regression (e.g. loss of verbal skills) (APA 2022; Kaminer 2005; WHO 2022a). Furthermore, PTSD and CPTSD often co‐occur with other psychiatric disorders, such as depression, anxiety, or substance‐use disorders, which can complicate treatment and worsen outcomes (APA 2022; WHO 2022b). Finally, while PTSD and CPTSD can present at any stage of life, the onset of psychopathology during childhood or adolescence may be particularly detrimental, and can have severe consequences for the acquisition of the cognitive, social, and emotional skills needed to thrive in life (WHO 2022b). Therefore, it is essential to identify effective and developmentally sensitive psychological interventions for the treatment of PTSD and CPTSD in children and adolescents.

Description of the intervention

The effectiveness and acceptability of psychological therapies for the treatment of PTSD in adult populations is well established, with all major guidelines providing strong recommendations for manualised, trauma‐focused psychotherapies as first‐line treatment for PTSD in adults (specifically, different trauma‐focused variants of cognitive and behavioural therapies, or eye movement desensitisation and reprocessing (EMDR)) (APA 2017; ISTSS 2020; NICE 2018; Phoenix Australia 2020; VA/DoD 2023). However, significantly less attention has been given to child and adolescent populations, and while many of the same treatments used for adult populations have been applied to the treatment of children and adolescents, the unique needs, abilities, and symptoms of these younger populations should be considered. Importantly, youth is a developmentally sensitive period in life, characterised not only by an increased vulnerability to stressors, but also by heightened plasticity, creating a window of opportunity for intervention efforts (Uhlhaas 2023).

Both trauma‐focused and generic psychological therapies have been used in the treatment of children and adolescents with PTSD, but guidelines currently recommend trauma‐focused cognitive behavioural therapies as first‐line treatment for this population (ISTSS 2020; NICE 2018; Phoenix Australia 2020). In these therapies, the therapist and client collaborate to identify, cope with, and challenge distorted and maladaptive cognitive and behavioural patterns associated with the trauma (Beck 2021; Ehlers 2000; Ehlers 2005). To this end, therapists use components such as psychoeducation, cognitive restructuring, relaxation techniques, and imaginal or in vivo exposure exercises, to help the client form more adaptive cognitions and behaviours (Bryant 2021; Robertson 2004; Sones 2011).

Of note, adaptations of psychological therapies for children and adolescents often include a stronger emphasis on therapeutic alliance, conjoint child‐parent sessions, teaching stabilisation skills to regulate affect and behaviour, and incorporating visual materials or other forms of expression to support communication (Ahmad 2008; Cohen 2018; de Roos 2017; Ruf 2010).

Definition of psychological therapies

We will include studies of both trauma‐focused and non‐trauma‐focused psychological therapies in this review. We will define psychological therapies in accordance with the recently proposed definition of psychological treatments (herein, used synonymously with therapies) by Division 12 of the American Psychological Association, Society of Clinical Psychology (Tolin 2024). According to this definition, a psychological therapy is:

“...an intervention consisting of specific actions between a person or persons and a mental health professional or designee, with the intent of engaging cognitive, emotional, behavioral, or interpersonal processes, in the service of modifying health or functional outcomes, and whose core assumptions about its procedures and mechanisms of change are founded in psychological science and consistent with scientific understanding” (Tolin 2024).

Importantly, this definition specifies that psychological therapies happen between people, i.e. are facilitated by bidirectional interaction between client and provider. In other words, psychological therapies are relational, which excludes therapies that are self‐administered, or based on one‐sided communication (e.g. purely instructional or educative communication). Furthermore, as this definition does not include interventions engaging biological or physiological mechanisms, psychological therapies do not include interventions such as psychopharmacology, brain stimulation, or interventions based purely on physical activity. We provide an overview of common categories of psychological therapies below, recognising that there are other ways of categorising these therapies, as several categories overlap, and as some therapies do not fit neatly within a single category, but draw upon several (Zarbo 2016).

Cognitive and behavioural therapies

These therapies focus on identifying and changing maladaptive patterns of cognition or behaviour through methods such as cognitive restructuring and exposure techniques (Beck 2021). Common examples include cognitive behavioural therapy, cognitive processing therapy, prolonged exposure, narrative exposure therapy, imagery‐based therapies, and skills training. In some of the more recent adaptations of cognitive behavioural therapies, known as third‐wave therapies, focus has shifted towards the client’s relationship to their thoughts, feelings, and behaviours, focusing on awareness and acceptance as a moderator of meaningful change (Hayes 2021). In addition to traditional cognitive and behavioural techniques, third‐wave therapies incorporate strategies such as acceptance and mindfulness, often targeted experiential avoidance, psychological flexibility, and emotional regulation (Hayes 2021). Common examples of third‐wave therapies include acceptance and commitment therapy, mindfulness‐based therapies, dialectical behaviour therapy, and metacognitive therapy.

Psychodynamic therapies

These interventions aim to understand current behaviour, emotions, cognitions, and relationships in the context of subconscious processes, early experiences, attachment patterns, and internalised conflicts (Summers 2024). Common examples include classic psychoanalysis, modern psychodynamic therapies, mentalisation‐based therapy, and attachment‐based therapies (Summers 2024).

Humanistic therapies

These interventions emphasise self‐discovery, personal growth, and making meaning of life experiences (Cain 2016; Cooper 2016). They are typically focused on building a strong therapeutic relationship, based on empathy, acceptance, and genuineness, and they tend to have a positive and holistic focus (Cain 2016). Common examples include existential therapy, person‐centred therapy, gestalt therapy, narrative therapy, emotion‐focused therapy, and non‐directive/supportive therapy.

Systemic therapies

These therapies conceive psychopathology and its treatment in the context of the social systems in which the client lives, their relationships, and their interactions with others (von Sydow 2010). These therapies typically take the form of family therapy, but may also focus on and include other relevant people or social systems (von Sydow 2010).

Body‐oriented and somatic psychotherapies

These therapies emphasise the connection between mind and body (Röhricht 2009). They typically involve exploring and increasing awareness and tolerance of bodily sensations and expressions related to one’s cognitions and emotions (Röhricht 2009). Of note, body‐oriented and somatic psychotherapies may be particularly relevant for the treatment of PTSD and CPTSD, as these conditions are associated with a high prevalence of somatic symptoms (Afari 2014; Møller 2021). Common examples include sensorimotor therapy and somatic experiencing.

Expressive therapies

In these therapies, client‐provider interactions are facilitated through creative or activity‐based forms of expression, such as playing with toys, or engaging in roleplay, art, music, or dance (Johnson 2016; Seymour 2016). These therapies may be particularly useful when working with young children who have limited ability to communicate verbally, or may find it challenging to do so (Seymour 2016).

Integrative therapies

Many established therapies integrate techniques from multiple psychological approaches and may be referred to as integrative (Zarbo 2016). This is true, for example, of several trauma‐focused therapies, such as EMDR or accelerated resolution therapy, both of which incorporate aspects of cognitive and behavioural therapies, somatic therapies, psychodynamic therapies, and more (Shapiro 2018; Waits 2017).

How the intervention might work

Despite empirical evidence supporting the effectiveness of various psychological therapies, we still have limited knowledge about how even our most studied interventions produce change (Cuijpers 2019; Kazdin 2007). A comprehensive account of possible mechanisms of change is beyond the scope of this review, but here we will briefly highlight some of the more widespread theories.

Perhaps most pervasive is the ongoing debate about whether psychological therapies work through the specific techniques of a given intervention (e.g. fear extinction via exposure therapy), or through common factors underlying all psychological therapies (e.g. a strong therapeutic alliance or the client’s expectation that a therapy will be effective) (Cuijpers 2019; Tschacher 2014; Wampold 2015). However, it has been suggested that the antagonism between the common and specific factors view is likely misguided, and that both contribute to therapeutic effectiveness (Cuijpers 2019; Tschacher 2014). In this view, common and specific factors are intrinsically correlated, with specific techniques being the very instruments by which therapists facilitate the expression of common factors (Tschacher 2014). As such, some therapeutic approaches may be more effective than others, whether by their specific techniques alone, or by their ability to facilitate a given common factor (Tschacher 2014).

In the case of PTSD, cognitive and behavioural therapies are often found to have the strongest effect (APA 2017; ISTSS 2020; NICE 2018; Phoenix Australia 2020; VA/DoD 2023). These therapies are typically based on the cognitive model by Ehlers and Clark, according to which, the effectiveness of these therapies, in adults with PTSD, is facilitated by modifying the client’s negative misappraisals associated with their trauma (Ehlers 2000). In addition, Meiser‐Stedman and colleagues argue that reducing safety‐seeking behaviours (i.e. avoidance symptoms) is also an essential factor (Meiser‐Stedman 2017). Some studies have also offered tentative support for the cognitive model of PTSD in children and adolescents (e.g. Meiser‐Stedman 2002). Many of the psychological therapies used in treating PTSD include specific techniques, such as psychoeducation to normalise PTSD reactions; emotional processing of the traumatic experience (meaning‐making); exposure elements to desensitise the client to trauma reminders; cognitive restructuring to help the client rethink assumptions and reactions to the traumatic experience; and coping skills training (e.g. problem‐solving, safety planning, and anxiety management) (AACAP 2010; Carr 2004; Robertson 2004; Sones 2011). Furthermore, many PTSD treatments modified for use with children and adolescents either include caregivers in the therapy or offer adjunct therapy to caregivers (Gutermann 2016; Gutermann 2017). This may enhance the effectiveness of therapies by improving caregivers’ ability to recognise trauma symptoms in their child, and their ability to help their child deal with emotional distress and related behavioural problems (AACAP 2010).

Research on the role of memory, particularly in the context of PTSD and CPTSD, has also provided promising insights into psychological therapies’ potential mechanisms of change (Brewin 2006; Brewin 2015; Hyland 2023; Lane 2015). One view suggests that memories are malleable, providing an opportunity for therapy to update maladaptive memory representations with more adaptive information; a process called memory reconsolidation (Lane 2015). An alternative view argues that memory representations cannot be permanently altered, but that alternative and more adaptive memory representations can be created, and whether the old or new memory representation then determines behaviour is decided through a process of retrieval competition, with the most accessible memory having the greatest influence (Brewin 2006; Brewin 2015; Lane 2015). By practising recalling these more adaptive representations, especially in the presence of cues that would normally trigger the activation of old traumatic memories, their retrieval is thought to be strengthened (Hyland 2023). The previously mentioned Memory & Identity Theory of CPTSD aligns well with the theory of retrieval competition (Hyland 2023). In CPTSD, the sense of self is often experienced as fragmented and unstable, and past adversity may not be fully understood and integrated within a comprehensive self‐understanding (Hyland 2023). According to the Memory & Identity Theory, neither identities nor memories can be permanently altered, but the way they are experienced and evaluated can be changed to make them more or less dominant in the retrieval competition (Hyland 2023). A more nuanced view suggests that neither reconsolidation nor retrieval competition alone can explain the full range of therapeutic changes that clients with PTSD or CPTSD undergo through psychological therapy (Brewin 2015).

Why it is important to do this review

This review will build upon the previous systematic review conducted by Gillies and colleagues (Gillies 2012). While Gillies and colleagues concluded that psychological therapies can help children and adolescents with PTSD, it was not possible to say whether one type of psychotherapy was superior. An updated review is needed to identify and incorporate the substantial number of new, eligible studies published since the aforementioned review. Furthermore, we will use a new and expanded search strategy, along with updated analytical tools.

Treatment of PTSD in children and adolescents has also been the topic of a number of more recent systematic reviews and meta‐analyses. However, as with the review by Gillies and colleagues, the literature searches are now outdated for several of these reviews (e.g. Gutermann 2016; Gutermann 2017; Morina 2016; Morina 2017). Amongst more recently published reviews, several limited their focus to specific populations, such as children and adolescents from low‐ and middle‐income countries (Alozkan‐Sever 2023; Uppendahl 2020), forced migrants or refugees (Molendijk 2024), maltreated youths (Bennett 2021), or children and adolescents exposed to natural and man‐made disasters (Xie 2024). Similarly, other reviews focussed on specific psychological therapies, such as trauma‐focused cognitive behavioural therapy (Thielemann 2022; Thielemann 2024; Xian‐Yu 2022), EMDR (Chen 2018), school‐based interventions (Tareke 2023), or specific therapy modalities, such as group therapy (Davis 2023). While some reviews used broader inclusion criteria, they still had significant limitations, such as including only a few studies (Arora 2024), or potentially overlooking relevant studies by searching only a few databases or using narrow search terms (Bastien 2020; Hoppen 2023; Hoppen 2024; Mavranezouli 2020).

Finally, we did not identify any reviews that assessed the certainty of evidence using the GRADE approach, or used Trial Sequential Analysis, and many reviews used either no form of risk of bias assessment (Gutermann 2016; Gutermann 2017; Molendijk 2024; Morina 2016), or did not use Cochrane's updated risk of bias tool, RoB 2 (Bastien 2020; Chen 2018; Hoppen 2023; Hoppen 2024; Mavranezouli 2020; Morina 2017; Xian‐Yu 2022). The lack of these assessments neglects the effect of systematic errors and the quality of evidence. As such, there is a need for an up‐to‐date, more comprehensive and methodologically rigorous review that incorporates GRADE assessments and controls for random errors with RoB 2 (Wetterslev 2017).

Objectives

To assess the effects (i.e. benefits and harms) of psychological therapies for post‐traumatic stress disorder and complex post‐traumatic stress disorder in children and adolescents.

Methods

Criteria for considering studies for this review

Types of studies

We will include randomised clinical trials (RCTs), including parallel‐group, cross‐over, and cluster‐randomised designs. We will not include quasi‐randomised trials (i.e. trials using methods of randomisation that are not truly random, such as allocation based on date of birth, medical record number, or order of recruitment). We will not restrict our searches by language, date, or publication status. We will seek translation of all relevant sections of studies that are not in English or a Scandinavian language.

Types of participants

We will include studies of children and adolescents in any setting, of any sex, gender, or race, with a mean age of up to 18 years, and a diagnosis of post‐traumatic stress disorder (PTSD) or complex post‐traumatic stress disorder (CPTSD). PTSD must be diagnosed in accordance with the diagnostic criteria specified in the International Classification of Diseases (ICD‐10, ICD‐11) or the Diagnostic and Statistical Manual of Mental Disorders (DSM‐III, DSM‐III‐R, DSM‐IV, DSM‐IV‐TR, DSM‐5, DSM‐5‐TR). CPTSD must be diagnosed in accordance with the diagnostic criteria specified in the ICD‐11. The diagnosis of PTSD or CPTSD must be determined through a structured or clinician interview, or by a baseline score above the clinical cutoff on a validated scale (provided that trial authors used the scale as a diagnostic tool to determine eligibility). We will consider CPTSD to be present whether it is reported as CPTSD or as PTSD plus all three disturbances in self‐organisation (DSO) symptoms (officially diagnosed, or as measures above a clinical cutoff on a validated scale). We will include participants both with and without comorbid psychiatric disorders. We will exclude studies in which participants are reported to have an intellectual disability (i.e. IQ below 70).

To be eligible for inclusion, at least 70% of study participants must be diagnosed with either PTSD or CPTSD, while the remaining participants must have subsyndromal symptoms of PTSD or CPTSD. This cutoff was based on the cutoffs used in recognised clinical guidelines for the treatment of PTSD (e.g. ISTSS 2020; Phoenix Australia 2020). Studies in which fewer than 70% of participants were diagnosed with PTSD or CPTSD, or studies in which the remaining participants did not have subsyndromal symptoms of PTSD or CPTSD, will be included if the data for diagnosed participants are reported as a separate subset of the data. Similarly, studies including participants outside our specified age range will be included if data for participants with a mean age of up to 18 years are reported as a separate subset of the data.

Types of interventions

Experimental interventions

We will include any psychological therapy, regardless of theoretical orientation, provided at least two of the four criteria defined by Karatzias and colleagues are met (Karatzias 2019).

1. An established school/approach to psychotherapy is cited.

2. A description of the therapy, referencing a psychological process, is provided.

3. A treatment manual, used to guide treatment delivery, is cited.

4. Active components of the treatment are identified.

We define psychological therapies in accordance with the recently proposed definition by Division 12 of the American Psychological Association, Society of Clinical Psychology (Tolin 2024). While it is not possible to make an exhaustive list of all psychological therapies encompassed by this definition, below is a list of some common categories of psychological therapies. These categories overlap to some degree, and some therapies do not fit neatly within a single category but draw upon several (see Description of the intervention for Division 12's definition of psychological therapies, as well as details on the intervention categories listed below and the therapies within them).

1. Cognitive and behavioural therapies

2. Psychodynamic therapies

3. Humanistic therapies

4. Systemic therapies

5. Body‐oriented and somatic psychotherapies

6. Expressive therapies

7. Integrative therapies

We will include eclectic psychotherapies (i.e. combinations of multiple psychological therapies), provided all intervention components are clearly described (see Karatzias 2019). Psychological interventions combined with non‐psychological interventions (e.g. pharmacological interventions, physical activity interventions, or self‐help), will only be eligible if the non‐psychological intervention is applied equally (or is equally accessible) to both the experimental group and the control group, to cancel out its effects.

The interventions may be delivered by anyone professionally trained in delivering psychological therapies or the specific therapy used (e.g. a psychologist, licenced therapist, school counsellor, or nurse). Interventions may also be delivered or mediated by a non‐professional designee (e.g. a parent or teacher), provided the designee receives training and supervision from a professional, to ensure competent delivery of the therapy.

The interventions may be delivered in any setting (e.g. in/outpatient healthcare settings, community settings, school settings), by any mode of delivery (e.g. in‐person or remotely), and in any format (e.g. individually, group‐based, dyadically). Interventions of any duration and intensity are eligible, provided they include a minimum of two sessions in total. Interventions may focus exclusively on the child or adolescent, or include participation in or parallel sessions for relevant others (e.g. parents or teachers).

Comparator interventions

The experimental intervention must be compared to one of the following:

1. An inactive control (i.e. waiting list, no treatment, sham interventions);

2. Treatment as usual (i.e. any form of standard treatment, whether pharmacological, psychological, any other type of intervention, or a combination of interventions);

3. Another psychological intervention (see definition and restrictions specified for the experimental intervention);

4. Pharmacological interventions;

5. Other interventions (i.e. any non‐pharmacological and non‐psychological interventions, e.g. physical activity, acupuncture, etc.);

6. Any type of add‐on therapy, provided it is equally applied or accessible to both groups.

Types of outcome measures

Outcomes may be either self‐rated by the participants, or observer‐rated by clinicians, caregivers, or teachers. Data from different assessors will be synthesised separately. We will only use measures reported to be valid and reliable in a peer‐reviewed journal, except in the case of spontaneous or systematic reporting of attrition, and serious and nonserious adverse events and effects.

We will make all reasonable attempts to obtain raw data on non‐reported outcomes by contacting study authors, but study eligibility will not be dependent on the reporting of outcomes. To provide an overview of the full scope of RCTs conducted on this topic, studies reporting none of the outcomes of interest to this review will be included as empty studies.

Timing of outcome assessment

We will extract outcomes measured at pre‐treatment, post‐treatment, short‐term follow‐up (up to three months), medium‐term follow‐up (longer than three months, up to six months), and long‐term follow‐up (longer than six months). If a trial reports follow‐up outcomes at more than one time point within a time frame, we will use the outcome reported at the latest time point within the time frame.

Hierarchy of outcome measures

If studies report multiple validated measures of the same outcome, we will prioritise measures in the following order for populations with a mean age below 13 years: clinician‐rated scales, teacher‐rated scales, caregiver‐rated scales, and self‐rated scales. For populations with a mean age above 13 years, we will prioritise measures in the following order: clinician‐rated scales, self‐rated scales, teacher‐rated scales, and caregiver‐rated scales. If several validated measures are collected in the same way (e.g. two self‐rated measures), we will prioritise the outcome measure most frequently used across studies. If availability across studies is equivalent, we will prioritise the outcome measure with the best psychometric properties.

Primary outcomes

1. Severity of PTSD symptoms: measured with a validated instrument based on diagnostic criteria for PTSD, as defined by the DSM or ICD manuals, e.g. the Child and Adolescent Trauma Screen (CATS; CATS‐2) (Sachser 2017; Sachser 2022); the Child PTSD symptom Scale for DSM‐5 (CPSS‐5) (Foa 2018); Clinician‐Administered PTSD Scale for DSM‐5 – Child/Adolescent Version (CAPS‐CA‐5) (Nader 1996; Pynoos 2015); the Child Trauma Screen (CTS) (Lang 2017); UCLA Child/Adolescent PTSD Reaction Index for DSM‐5 (Kaplow 2020); or the International Trauma Questionnaire – Child and Adolescent Version (ITQ‐CA) (Cloitre 2018b; Haselgruber 2020)

2. Proportion of participants with serious adverse events/effects: reported spontaneously or systematically, or using a validated instrument. According to the International Council for Harmonisation's Guideline for Good Clinical Practice (ICH 2016), a serious adverse event is any untoward medical occurrence which, at any dose, results in death, is life‐threatening, requires inpatient hospitalisation or prolongation of existing hospitalisation, results in persistent or significant disability or incapacity, or a congenital anomaly or birth defect. We will consider all other adverse outcomes to be nonserious. In line with chapter 19 of the Cochrane Handbook for Systematic Reviews of Interventions, we use the term adverse events for unfavourable outcomes occurring during or after psychological therapy, for which a causal relationship between the therapy and the event cannot be ascertained, and adverse effects for unfavourable outcomes for which there is a clear indication of a causal relationship with the therapy (Peryer 2024).

Secondary outcomes

1. Depressive symptoms: measured with a validated instrument, e.g. the Beck Depression Inventory (BDI; BDI‐II) (Beck 1961; Beck 1996); the Hospital Anxiety and Depression Scale – Depression Subscale (HADS‐D) (Zigmond 1983); Hamilton Rating Scale for Depression (HDRS/HAM‐D) (Hamilton 1960); Center for Epidemiologic Studies – Depression Scale (CES‐D) (Radloff 1977); Depression Anxiety Stress Scales – Depression Subscale (DASS‐21‐D; DASS‐42‐D) (Lovibond 1995); Patient Health Questionnaire‐9 (PHQ‐9) (Kroenke 2001); the Montgomery Åsberg Depression Rating Scale (MADRS) (Montgomery 1979); the Children’s Depression Rating Scale (CDRS) (Mayes 2010); the Children's Depression Inventory (CDI) (Kovacs 1985); Depression Self‐Rating Scale for Children (DSRS) (Birleson 1987); Short Mood and Feelings Questionnaire (SMFQ) (Angold 1995)

2. Anxiety symptoms: measured with a validated instrument, e.g. the Revised Children's Anxiety and Depression Scale (RCADS) (Chorpita 2000); the Spence Children’s Anxiety Scale (SCAS) (Spence 1998); Depression Anxiety Stress Scales – Anxiety Subscale (DASS‐21‐A; DASS‐42‐A) (Lovibond 1995); State‐Trait Anxiety Inventory for Children (STAI‐C) (Spielberger 1973); the Hospital Anxiety and Depression Scale – Anxiety Subscale (HADS‐A) (Zigmond 1983); Revised Children’s Manifest Anxiety Scale (RCMAS) (Reynolds 2008)

3. General functioning: measured with a validated instrument, e.g. the Children's Global Assessment Scale (CGAS) (Shaffer 1983); Global Assessment of Functioning (GAF) (APA 2000); the World Health Organization Disability Assessment Schedule (WHODAS) (Üstün 2010)

4. Quality of life: measured with a validated instrument, e.g. the Child Health Questionnaire (CHQ) (Landgraf 2002); Pediatric Quality of Life Inventory 4.0 (PedsQL) (Varni 2001); Kidscreen‐27 (Ravens‐Sieberer 2007)

5. Attrition: number of participants dropping out following randomisation

6. Proportion of participants with nonserious adverse events/effects: measured with either standardised psychometric rating scales, such as the Systematic Assessment for Treatment Emergent Events (SAFTEE) (Levine 1986), or through spontaneous reporting. For details on the distinction between adverse events and effects, see our description of the primary outcome serious adverse events or effects.

Exploratory outcomes

The following outcomes will be purely exploratory.

1. Severity of CPTSD symptoms: measured with a validated instrument based on the ICD‐11 diagnostic criteria for CPTSD, e.g. the International Trauma Questionnaire – Child and Adolescent Version (ITQ‐CA) (Cloitre 2018b; Haselgruber 2020); the Child and Adolescent Trauma Screen 2 (CATS‐2) (Sachser 2022)

2. Severity of DSO symptoms: measured with a validated instrument, e.g. the ITQ‐CA

3. Dissociative symptoms: measured with a validated instrument, e.g. the Child Dissociative Checklist (CDC) (Putnam 1993); the Adolescent Dissociative Experiences Scale (A‐DES) (Armstrong 1997)

Search methods for identification of studies

A new literature search will be conducted, based on a revised search strategy, and including databases not searched in the review by Gillies and colleagues (Gillies 2012). Therefore, we will not restrict our searches by date of publication. Nor will we restrict searches by language or publication status.

Electronic searches

The following electronic databases and registers will be searched:

1. Cochrane Central Register of Controlled Trials (CENTRAL; latest issue);

2. MEDLINE Ovid (1946 onwards);

3. Embase Ovid (1974 onwards);

4. PsycINFO Ovid (1806 onwards);

5. Cumulative Index to Nursing and Allied Health Literature EBSCOhost (CINAHL; 1937 onwards);

6. Latin American and Caribbean Health Science Information database (LILACS; 1982 onwards);

7. ERIC EBSCOhost (Education Resources Information Center; 1966 onwards);

8. Web of Science Core Collection Clarivate Analytics (1900 onwards);

9. Sociological Abstracts ProQuest (1952 onwards);

10. ProQuest Dissertations and Theses A&I (1973 onwards);

11. Library Hub Discover (discover.libraryhub.jisc.ac.uk/);

12. Networked Digital Library of Theses and Dissertations (NDLTD; ndltd.org);

13. National Institute for Health and Care Research (NIHR; bepartofresearch.nihr.ac.uk/);

14. US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (clinicaltrials.gov);

15. World Health Organization’s International Clinical Trials Registry Platform (ICTRP; trialsearch.who.int/Default.aspx);

16. ISRCTN Registry (isrctn.com);

17. Australian New Zealand Clinical Trials Registry (ANZCTR; anzctr.org.au/BasicSearch.aspx);

18. EU Clinical Trials Register (clinicaltrialsregister.eu/).

Our MEDLINE Ovid search strategy is outlined in Appendix 1. The search strategy was developed by DBI in collaboration with a research librarian, and subsequently reviewed by co‐authors MQ, MSJ, and OJS. We will modify this search strategy for each database to ensure the use of appropriate syntax and controlled terms.

Searching other resources

We will handsearch relevant journals, including: Journal of Traumatic Stress; Journal of Trauma & Dissociation; Psychological Trauma: Therapy, Research, Practice, and Policy; the American Journal of Psychiatry; JAMA Psychiatry; British Journal of Psychiatry; ACTA Psychiatrica Scandinavica; Journal of the American Academy of Child and Adolescent Psychiatry; the Journal of Clinical Psychiatry; Journal of School Psychology; School Mental Health; and JAMA Pediatrics. We will restrict our handsearches of journals to publications from the past five years (from the search date).

We will also handsearch abstracts of key conferences for PTSD and trauma (e.g. the International Society for Traumatic Stress Studies, ISTSS; the International Society for the Study of Trauma and Dissociation, ISSTD), and make all reasonable attempts to obtain any relevant unpublished data. We will restrict our handsearches of conference abstracts to publications from the past five years (from the search date).

To identify additional relevant studies, we will also handsearch the reference lists of: (1) included studies; (2) relevant systematic reviews (i.e. those cited in the background section Why it is important to do this review and any relevant reviews identified through our searches for trials); and (3) the most recent version of relevant clinical guidelines published by: ISTSS; the National Institute for Health and Care Excellence (NICE); Phoenix Australia; the U.S. Department of Veterans Affairs (VA/DoD); the American Psychological Association (APA); and Scandinavian health organisations.

Finally, we will also contact experts working in this field of research and enquire about ongoing or unpublished trials.

Data collection and analysis

The review will be conducted in accordance with the guidelines provided by the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2024a), and analyses will be performed using Cochrane's statistical software, Review Manager (RevMan) (RevMan 2025). We will refer to the thresholds for statistical and clinical significance recommended by Jakobsen and colleagues (Jakobsen 2014).

Selection of studies

To reduce title and abstract screening activities by the review authors, we will first use Cochrane's Screen4Me service. The service assists in selecting relevant studies and reduces the workload by an estimated 50% to 70%. The Screen4Me process comprises three components: (1) known assessments (matches records in the search results to records that have already been screened and labelled as either RCTs or non‐RCTs by Cochrane Crowd, Cochrane's crowdsourcing platform where contributors from around the world help to identify RCTs and other types of healthcare‐related research); (2) the RCT classifier (a machine‐learning model that distinguishes RCTs from non‐RCTs); and, if appropriate, (3) further use of Cochrane Crowd to help identify RCTs amongst remaining records. Use of the third component will depend on the number of results remaining after the first two steps. For more information about Screen4Me and the evaluations that have been done, please go to the Screen4Me webpage on the Cochrane Information Specialist’s portal. For more detailed information regarding evaluations of the Screen4Me components, see: Marshall 2018; Noel‐Storr 2020; Noel‐Storr 2021; Thomas 2021.

Two review authors (DBI, MQ, SKN, MTK, MSJ, JPS, or LM) will independently screen the titles and abstracts of the remaining records. We will resolve any uncertainties or disagreements during title/abstract screening through consensus, or by consulting a third review author (OJS). Where necessary, we will seek translation of any records that are not in English or a Scandinavian language.

Two review authors (DBI, MQ, SKN, MTK, MSJ, JPS, or LM) will independently screen the full‐texts of all records judged to be potentially eligible during title/abstract screening. If there are uncertainties about a study's eligibility, which cannot be resolved through consensus or by consulting a third review author (OJS), we will make all reasonable attempts to contact the study authors for clarification. Studies excluded during full‐text screening will be listed in a characteristics of excluded studies table, along with their reasons for exclusion. We will also complete a PRISMA flowchart, summarising the status of all records identified (Page 2021). Records excluded by Screen4Me will be included in the PRISMA flowchart.

We will use Covidence software to facilitate screening, and to keep track of included and excluded records (Covidence 2024).

Data extraction and management

We will develop data extraction forms in Covidence to facilitate standardisation of data extraction (Covidence 2024). Data extraction forms will be pilot tested on five studies, and revised if necessary (e.g. if anything is unclear, if data are missing from the form, or if data are superfluous). All review authors who will extract data will independently pilot test the form to ensure consistency between extractors.

Two review authors (DBI, MQ, SKN, MTK, MSJ, JPS, or LM) will independently extract data into the forms in Covidence. Any uncertainties or disagreements will be resolved through discussion, or by consulting a third review author (OJS). The review authors DBI and OJS will transfer the extracted data into RevMan (RevMan 2025). If data are unclear or missing in the published trial reports, we will make all reasonable attempts to contact the study authors for clarification, or to obtain the missing data. If there are multiple reports of the same trial, we will collate all reports of the trial when extracting data, and count it as a single trial while listing all publications in the references.

We will extract the following data from the included studies:

1. Practical details: corresponding author (including contact information), sources of funding, potential conflicts of interest, and any comments (i.e. our own or from study authors);

2. Study details: year published, country, setting, and study design;

3. Population details: inclusion and exclusion criteria, group differences at baseline, age, sex, PTSD versus CPTSD diagnosis (including how it was established), type(s) of trauma exposure, comorbidity, and sample size;

4. Intervention and comparator: components, mode of delivery/format (e.g. remote or face‐to‐face delivery; group or individual therapy), duration (in weeks), number of sessions, frequency of sessions, family involvement, professional background of therapist, and whether concurrent psychotropic medication was used or allowed; and

5. Outcomes: measures of treatment effect (see details in Measures of treatment effect), instruments used (including definitions of upper and lower limits, direction, thresholds, and who the rater was), and time points measured.

Assessment of risk of bias in included studies

Two review authors (DBI, MQ, SKN, MTK, MSJ, JPS, or LM) will independently assess risk of bias in the included studies. Discrepancies between judgements will be resolved through discussion. If review authors are unable to reach a consensus, a third review author will be consulted (OJS). We will make all reasonable attempts to contact trial authors regarding missing information required to make risk of bias judgements.

We will use Cochrane’s RoB 2 tool to assess risk of bias (i.e. potential systematic errors) in the included studies (Higgins 2024b; Sterne 2019). Of interest to this review is the effect of assignment to intervention (the intention‐to‐treat effect).

RoB 2 assesses risk of bias for individual outcomes of interest, rather than the overall risk of bias for the entire study. We will complete risk of bias assessments for the post‐treatment measures of the six outcomes most central to this review: (1) PTSD symptoms, (2) serious adverse events/effects, (3) depressive symptoms, (4) general functioning, (5) quality of life, and (6) nonserious adverse events/effects. These outcomes will be presented in a summary of findings table.

RoB 2 is structured into the following five domains, through which bias may occur:

1. Risk of bias arising from the randomisation process;

2. Risk of bias due to deviations from the intended interventions;

3. Risk of bias due to missing outcome data;

4. Risk of bias in the measurement of the outcome;

5. Risk of bias in the selection of the reported results.

To assess each domain, we will answer the series of signalling questions recommended for RoB 2, using the response options: yes, probably yes, probably no, no, or no information. Based on these answers, the RoB 2 algorithm will suggest one of three risk of bias judgements for each domain: low risk of bias, some concerns, or high risk of bias. We will either verify these judgements or provide clear justifications for any changes made to them.

Finally, we will make an overall risk of bias judgement for each outcome, based on the five domain‐level judgements. For an overall judgement of low risk of bias, all five domains must be judged to be at low risk of bias. For an overall judgement of some concerns, at least one domain must be judged to raise some concerns, and none may be judged to be at high risk of bias. For an overall judgement of high risk of bias, at least one domain must be judged to be at high risk of bias, or multiple domains must be judged to raise some concerns in a way that substantially lowers confidence in the results.

We will use the RoB 2 Excel tool for randomised trials to record and manage our risk of bias assessments (available from www.riskofbias.info/welcome/rob-2-0-tool/current-version-of-rob-2). We will summarise all risk of bias judgements, and any relevant supporting information, in risk of bias tables. If we include cluster‐randomised trials or cross‐over trials, we will use the test versions of the RoB 2 Excel tools for cluster‐randomised trials (available from www.riskofbias.info/welcome/rob-2-0-tool/rob-2-for-cluster-randomized-trials) and cross‐over trials (available from www.riskofbias.info/welcome/rob-2-0-tool/rob-2-for-crossover-trials).

Measures of treatment effect

Continuous data

We will use the mean scores and standard deviations (SD) of each intervention group and calculate the mean difference (MD) with 95% confidence intervals (CI). We will calculate the standardised mean difference (SMD) with 95% CI, as well as 95% prediction intervals (PI) using a random‐effects model, should different outcome measures be used to measure the same construct in the trials. We will calculate MDs or SMDs based on post‐treatment data and follow‐up data (in separate analyses). Where the direction of a scale is opposite to most of the other scales, we will multiply the corresponding mean values by −1 to ensure adjusted values. If the trials do not report means and SDs, but report other values, such as t‐tests, P values, or 95% CI, we will try to transform these into SDs.

Dichotomous data

We will summarise dichotomous data as risk ratios (RR) with 95% CI and 95% PI using a random‐effects model.

Skewed data

As a meta‐analysis is based on assumptions of normality, we will check all continuous data for skewness before including them in a meta‐analysis. For a scale that starts from zero, an SD that is more than half of the mean value suggests skewness, while an SD that is more than the full mean value is considered strong evidence of a skewed distribution (Altman 1996). Studies with strong evidence of skewness will not be included in the meta‐analyses, but will instead be reported narratively (see Data synthesis).

Unit of analysis issues

Repeated observations

We will calculate study estimates based on post‐treatment data. We will conduct separate analyses of data from different time point measurements (i.e. pre‐treatment, post‐treatment, short‐term follow‐up, medium‐term follow‐up, and long‐term follow‐up).

Cross‐over trials

We intend to incorporate data from randomised cross‐over trials until the occurrence of the first cross over (Curtin 2002). We will exclude outcomes from subsequent periods due to the potential carryover effects from the preceding treatment(s), and we will not consolidate repeated participant observations in a single meta‐analysis.

Cluster‐randomised trials

In cluster‐randomised trials, we expect investigators to report their findings with due consideration of clustering effects, using measures such as robust standard errors or hierarchical linear models. If there is ambiguity about whether a cluster‐randomised trial is adequately controlled for clustering, we will contact the trial investigators for clarification. If we find that appropriate controls for clustering are lacking, we will request and re‐analyse individual participant data, using multilevel models designed to account for clustering effects. Subsequently, we will analyse effect sizes and standard errors in RevMan. If there is insufficient information to control for clustering, we will enter outcome data into RevMan using individuals as the units of analysis, and then conduct a sensitivity analysis to assess the potentially biasing effects of inadequately controlled cluster‐randomised trials (Donner 2002); for more details, see Sensitivity analysis. If individual participant data are not available, we will look for information on intraclass correlation coefficients to adjust for the potential clustering effects.

Multiple‐arm studies

To avoid unit of analysis issues in trials with multiple arms, we will combine all relevant experimental intervention groups into a single experimental group and all relevant comparator intervention groups into a single comparator group.

Adjustment for multiplicity

We will adjust the P values and CIs for multiplicity due to the many secondary outcome comparisons, following the method described by Jakobsen and colleagues (Jakobsen 2014).

Dealing with missing data

We will make all reasonable attempts to obtain any missing data by contacting trial authors. We will report when we obtain this information from the trial authors in the risk of bias tables.

We will evaluate the methods used for handling the missing data, and we will evaluate whether it is likely that the missing data influenced the results of outcomes of interest. When possible, we will calculate effect sizes based on intention‐to‐treat (ITT) data. If only completer analysis data are reported, we will calculate effect sizes using these. We will discuss each trial’s methodology for dealing with missing continuous data (e.g. last observation carried forward, best‐worst case scenario, or multiple imputation). If dichotomous data are not based on ITT data, we will add the number of participants lost in each group to the participants with unfavourable results, acting on the assumption that most participants are not lost at random. If we cannot obtain missing raw data from the trial authors after two attempts at contacting them, but the data are reported in graphical form, we will extract approximate numerical data points from the graphs using the online open‐source software Plot Digitizer (plotdigitizer.sourceforge.net).

If data are not reported in a usable way, and require other forms of processing before being analysed, we will consult a statistician.

We will assess the robustness of results derived from statistically processed data through sensitivity analyses (see Sensitivity analysis).

Assessment of heterogeneity

We will assess clinical heterogeneity by using the Clinical Diversity in Meta‐analyses (CDIM) tool (Barbateskovic 2021). This tool uses a scale with four domains and a total of 11 items. The four domains are: (1) setting (country, time of conduct, status, location); (2) population (age, comorbidities, inclusion/exclusion criteria, sex); (3) intervention (duration, intensity, type of intervention, co‐interventions, type of control interventions); and (4) outcomes (type of outcome, timing, raters) (Barbateskovic 2021). We will evaluate the methodological heterogeneity by comparing the design of trials. We will investigate statistical heterogeneity by both visual inspection of the forest plots and the I² statistic (Higgins 2003). We will judge I² values between 0% and 40% to indicate little heterogeneity, between 30% and 60% to indicate moderate heterogeneity, between 50% and 90% to indicate substantial heterogeneity, and between 75% and 100% to indicate considerable heterogeneity (Deeks 2024). We will also assess statistical heterogeneity using the Chi² test (P < 0.10) and report Tau² (an estimate of between‐study variance).

We intend to undertake meta‐analysis even if there is substantial statistical heterogeneity, but we will interpret the results with caution, discuss possible risks or concerns, and possible impact on results, and investigate them by conducting sensitivity and subgroup analyses (see Subgroup analysis and investigation of heterogeneity). Where there is substantial clinical heterogeneity, we will make trial‐by‐trial decisions on whether the heterogeneity makes the data inappropriate for meta‐analysis. Studies judged to be inappropriate for meta‐analysis will instead be included in a narrative synthesis (see Data synthesis).

Assessment of reporting biases

Reporting biases arise when the dissemination of research findings is influenced by the nature and direction of results. To explore the possibility of publication bias, we will provide funnel plots for meta‐analyses with 10 or more included trials, and perform Egger’s statistical test for small‐study effects (Egger 1997).

We will also use the Risk of Bias – Missing Evidence (RoB‐ME) tool, to assess for risk of bias due to missing evidence (Page 2023). RoB‐ME focusses on both selective non‐reporting and selective non‐publication, and their impact on meta‐analyses (Page 2023). This will help us explore how our selected outcomes will be impacted by the P‐value, magnitude, or direction of the study results, and consider the potential for missing studies across the review.

Data synthesis

We will include data in our analyses regardless of risk of bias judgements. We intend to perform meta‐analysis even if there is substantial statistical heterogeneity, and we will make trial‐by‐trial decisions on whether clinical heterogeneity makes a given study inappropriate for meta‐analysis. Where there is substantial statistical or clinical heterogeneity, we will interpret the results with caution, discuss possible risks or concerns and possible impact on results, and investigate the effect of heterogeneity through sensitivity and subgroup analyses.

We will use the statistical analyses recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2024). To give more weight to studies with less variance (mostly larger studies), we will use the inverse‐variance method, which increases the precision of the pooled effect estimate (Deeks 2024). We will pool interventions and comparators in groups that are as clinically homogenous as possible. We will use the random‐effects model for our meta‐analyses, as we expect a high degree of clinical heterogeneity to be present in most cases, but not so substantial as to prevent pooling in principle. Where the amount of clinical heterogeneity makes it inappropriate to pool trials in meta‐analysis, we will provide a narrative description of the trial results. If data pooling seems feasible, we will pool the primary studies' effects and calculate their 95% CIs. If a trial provides more than one measure of the same outcome, we will prioritise outcomes as per the hierarchy described in the Types of outcome measures section. We will group interventions into the comparisons listed below.

Main comparisons

1. Psychological therapies versus an inactive control (i.e. waiting list, no treatment, sham psychotherapy)

2. Psychological therapies versus other psychological therapies

Secondary comparisons

1. Psychological therapies versus treatment as usual

2. Psychological therapies versus pharmacological interventions

3. Psychological therapies versus other interventions

4. Psychological therapies versus an add‐on therapy that is equally applied or accessible to both groups

When data cannot be synthesised in a meta‐analysis due to incomplete data or clinical heterogeneity, we will provide a narrative synthesis following the methodological guidance outlined in Chapter 12 of the Cochrane Handbook for Systematic Reviews of Interventions (McKenzie 2024), and the reporting guidance specified in the Synthesis Without Meta‐analysis (SWiM) reporting guideline (Campbell 2020).

For a visual presentation of our findings, we will summarise effect estimates using forest plots, effect direction plots, Trial Sequential Analysis plots, and tabular presentation of study characteristics.

Subgroup analysis and investigation of heterogeneity

We will conduct subgroup analyses on the primary outcome measures to investigate heterogeneity in the subgroups listed below (see also Assessment of heterogeneity). We will compare subgroups using the formal test for subgroup differences in RevMan.

1. Type of intervention: we will first pool data on a general level (e.g. by treatment modality and trauma‐focused versus non‐trauma‐focused), and then group data into more specific groups of interventions (e.g. prolonged exposure, EMDR, narrative exposure therapy), based on the available evidence.

2. Type of trauma: e.g. sexual abuse, physical trauma, natural disasters

3. Age: younger than 7 years of age, 7 to 12 years of age, older than 12 years of age

4. Psychiatric comorbidity: with comorbidity compared to without comorbidity

5. Duration of treatment: less than 3 months, 3 to 6 months, longer than 6 months

6. Diagnosis: PTSD compared to CPTSD

7. Percentage diagnosed: 100% diagnosed with PTSD/CPTSD compared to < 100% diagnosed with PTSD/CPTSD

Sensitivity analysis

We will conduct sensitivity analyses on the primary outcome measures to assess the impact of heterogeneity on the overall pooled effect estimate by systematically removing studies (outliers) that are contributing to the heterogeneity.

We will conduct sensitivity analyses (by excluding/including trials one by one) to determine whether our findings are sensitive to the following:

1. Decisions made during the review process: e.g. based on available data, we may choose to divide treatment‐as‐usual into smaller groups, organised by content or format

2. Impact of bias: trials at low risk of bias compared to trials at uncertain or high risk of bias

3. Type of model used for analysis: repeating the analysis using the fixed‐effect model to test the robustness of the results

4. Imputed data: comparing analyses with available outcome data with those using an ITT approach

5. Skewed data: whether including data that may have been skewed affected the analysis

Diversity‐adjusted required information size and Trial Sequential Analysis

Trial Sequential Analysis (TSA) is a methodology that combines a diversity‐adjusted required information size (DARIS) calculation for a meta‐analysis with the threshold for statistical significance (Brok 2008; Brok 2009; Thorlund 2009; Wetterslev 2008). TSA serves as a mechanism to assess the statistical reliability of data in the cumulative meta‐analysis by adjusting P values for sparse data and repetitive testing on accumulating data (Brok 2008; Brok 2009; Thorlund 2009; Wetterslev 2008). We will only use TSA as a sensitivity analysis for the GRADE assessment of imprecision (see Summary of findings and assessment of the certainty of the evidence for details on the GRADE approach). The DARIS will be calculated a priori to determine the number of participants needed in order not to downgrade for imprecision in the meta‐analysis. We will undertake TSAs for the primary outcomes based on the a priori assumptions listed below (Brok 2008; Brok 2009; Thorlund 2009; Wetterslev 2008; Wetterslev 2009).

For the continuous primary outcome (severity of PTSD symptoms):

1. The standard deviation of the primary outcome;

2. An intervention effect is equivalent to Hedge’s g of 0.5;

3. A type I error of 2.5% (alpha) for the outcome because of two primary outcomes (Jakobsen 2014);

4. A type II error of 10% (beta), ensuring a minimum of 90% power;

5. The diversity of the meta‐analysis.

For the dichotomous primary outcome (proportion of participants with serious adverse events):

1. Proportion of participants with the outcome in the control group;

2. A relative risk reduction of 20%;

3. A type I error of 2.5% (alpha) for the outcome because of two primary outcomes (Jakobsen 2014);

4. A type II error of 10% (beta), ensuring a minimum of 90% power;

5. The diversity of the meta‐analysis.

For the GRADE assessment of imprecision, we will downgrade three levels for imprecision if the accrued sample size is below 33% of DARIS; two levels if the accrued sample size is between 33% and 65% of DARIS; and one level if the accrued sample size is between 66% and 100% of DARIS. We will not downgrade if the cumulative Z curve crosses boundaries for benefit, harm, futility, or DARIS.

For the primary continuous outcome, severity of PTSD symptoms, we will transform the MD and SD from the different rating scales used in the analysis to the MD and SD of a commonly used scale, using the following formula: MD = SMD*SD (Thorlund 2011).

Summary of findings and assessment of the certainty of the evidence

Using GRADEpro GDT, we will create a summary of findings table for each of the two main comparisons (GRADEpro 2024). In these tables, we will present the most clinically meaningful outcomes of this review (listed below), measured at post‐treatment.

Primary outcomes

1. PTSD symptoms

2. Serious adverse events/effects

Secondary outcomes

1. Quality of life

2. General functioning

3. Depressive symptoms

4. Non‐serious adverse events/effects

For each of these outcomes, the summary of findings tables will provide information on the certainty of evidence, the magnitude of the effect of the interventions examined, and the sum of available data. We will use risk ratios (RR) and either standard mean differences (SMD) or mean differences (MD) as measures of effect, and we will use TSA to assess imprecision (Jakobsen 2014; Korang 2020).

We will use the GRADE approach to assess the certainty of the body of evidence for each of the outcomes listed above (Hultcrantz 2017; Neumann 2024; Schünemann 2013; Schünemann 2024). GRADE judgements are based on an assessment of five domains: (1) risk of bias, (2) consistency of effect, (3) imprecision, (4) indirectness, and (5) publication bias. We will use the overall risk of bias for each of the outcomes to inform the GRADE assessments (see Assessment of risk of bias in included studies for details on RoB 2).

With GRADE, randomised trials are initially judged to provide high‐certainty evidence, but we will downgrade by one level in the case of serious limitations in one of the five domains listed above, and by two grades in the case of very serious limitations. We will interpret the grades of evidence in accordance with GRADE guidance (Hultcrantz 2017; Neumann 2024; Schünemann 2013; Schünemann 2024):

1. High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.

2. Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.

3. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect.

4. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

Two review authors (DBI, MQ, SKN, MTK, MSJ, JPS, or LM) will independently undertake GRADE assessments. Discrepancies will be resolved through discussion, or by consulting a third review author (OJS) when necessary. We will justify and document all decisions to downgrade or upgrade the certainty of the evidence in footnotes, and provide comments to aid the reader’s understanding when necessary.

We will ensure that our interpretation and reporting of results are informed by the GRADE assessments, by creating the summary of findings tables before writing the results, discussion, and conclusion of the review.

Appendices

Appendix 1. Preliminary search strategy (MEDLINE Ovid)

The following MEDLINE Ovid search strategy was developed by the first author (DBI) in collaboration with a research librarian. The search strategy was subsequently modified for each database, and a research librarian was consulted to ensure the use of appropriate syntax and controlled terms.

MEDLINE Ovid search strategy

The definition of trauma used in this review is not limited to the types of trauma included in our search‐string. For example, life threatening illness is listed by the DSM‐5‐TR as a potentially traumatising experience but is not included in this search‐string due to the excessive number of irrelevant references retrieved by search terms related to somatic illness. However, this does not mean that we will not include such studies, should they be retrieved by our other search terms.

Descriptor	#	Search terms
Core PTSD and trauma terminology	1	exp "stress disorders, traumatic"/ or "trauma and stressor related disorders"/
	2	(trauma* or posttrauma* or psychotrauma* or PTSD or PTSS or CPTSD).ti,ab,kf.
	3	(stress adj (disorder? or syndrome? or reaction?)).ti,ab,kf.
	4	((disturb* adj2 self‐organi#ation) or DSO).ti,ab,kf.
	5	(((extreme or complex or chronic or acute) adj stress) or DESNOS).ti,ab,kf.
	6	((("enduring personality change") adj3 catastroph*) or EPCACE).ti,ab,kf.
	7	(reexperienc* or re‐experienc* or flashback? or hypervigil*).ti,ab,kf.
Historical PTSD‐terminology	8	(shellshock* or shell‐shock*).ti,ab,kf.
	9	((war or combat or battle or torture) adj (neuros#s or syndrome?)).ti,ab,kf.
Adverse childhood experiences	10	"adverse childhood experiences"/
	11	(adverse adj (childhood or experience? or "life event?")).ti,ab,kf.
	12	(adversit* or ACEs).ti,ab,kf.
Victimisation	13	"disaster victims"/ or "crime victims"/
	14	(victim* or polyvictim*).ti,ab,kf.
Abuse and neglect	15	exp "child abuse"/ or "physical abuse"/ or "emotional abuse"/
	16	((abuse? or abusive) adj2 (survivor? or child* or p?ediatric or physical* or emotional* or sexual* or psychologic* or domestic* or verbal* or famil* or parent* or dating or partner* or relation* or interpersonal* or household? or home? or environment?)).ti,ab,kf.
	17	(neglect* adj2 (survivor? or child* or p?ediatric or physical* or emotional* or psychologic* or domestic* or famil* or parent* or abuse? or abusive or household? or home? or environment?)).ti,ab,kf.
	18	(maltreat* or mistreat*).ti,ab,kf.
Sexual/non‐sexual violence and harassment	19	violence/ or exp "sex offenses"/ or incest/ or "sexual harassment"/ or exp "harassment, non‐sexual"/ or "exposure to violence"/ or "domestic violence"/ or exp "gender‐based violence"/ or "intimate partner violence"/ or exp "gun violence"/ or exp "ethnic violence"/
	20	(rape? or molest* or incest* or assault* or traffick* or harass* or bully* or bullied or cyberbull* or hazing or stalk* or kidnap*).ti,ab,kf.
	21	(sex* adj2 (forced or unwanted or coerc*)).ti,ab,kf.
	22	(violen* adj2 (survivor? or child* or physical* or emotional* or sexual* or psychologic* or domestic* or verbal* or famil* or parent* or dating or partner* or relation* or interpersonal* or gender‐based or household? or home? or environment?)).ti,ab,kf.
War and terror exposure	23	exp terrorism/ or refugees/ or "war exposure"/ or exp "armed conflicts"/ or exp "war crimes"/ or torture/
	24	(war or wars or refugee? or "forc* displace*" or "child soldier?" or "moral injur*" or torture? or hostage? or terror* or shooting? or bomb* or explosion?).ti,ab,kf.
Disaster exposure	25	disasters/ or exp "natural disasters"/
	26	(disaster? or catastrophe? or hurricane? or tornado? or cyclon* or earthquake? or tsunami? or flood* or fire? or wildfire? or volcan* or avalanche?).ti,ab,kf.
Accidents	27	"accidents, traffic"/ or "accidents, aviation"/
	28	(accident? or crash*2 or collision?).ti,ab,kf.
Grief and bereavement	29	exp bereavement/
	30	(grief or griev* or bereave*).ti,ab,kf.
Condition	31	or/1‐30
Psychological therapies (unspecified)	32	exp psychotherapy/ or counseling/
	33	(psychotherap* or psycho‐therap* or talk‐therap* or counsel?ing or coaching).ti,ab,kf.
	34	((psychosocial* or psycho‐social* or psychologic* or mental* or talk‐based) adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique?)).ti,ab,kf.
Cognitive and behavioural therapies	35	((cognitive adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique? or restructuring or reappraisal? or remediation or processing or behavio?r*)) or CBT).ti,ab,kf.
	36	((behavio?r* adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique? or activation? or modification? or dialectic*)) or DBT).ti,ab,kf.
	37	((exposure adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique? or prolonged or narrative or written or virtual or VR or augmented or imaginal or graded or gradual or interoceptive)) or KIDNET).ti,ab,kf.
	38	("implosive therapy" or "flooding therapy" or "imaginal flooding" or "stress inoculation" or "virtual reality" or "augmented reality").ti,ab,kf.
	39	((metacognitive or meta‐cognitive) adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique? or remediation or processing)).ti,ab,kf.
	40	((imagery adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique? or rescripting or rehearsal or guided)) or ImRs).ti,ab,kf.
	41	(third‐wave or mindfulness or meditat* or "acceptance and commitment" or "compassion‐focus?ed" or MBCT or MBSR).ti,ab,kf.
	42	(schema? adj3 (therap* or maladaptive or mode?)).ti,ab,kf.
Psychodynamic therapies	43	psychoanalysis/
	44	(psychoanaly* or psycho‐analy* or psychodynamic* or psycho‐dynamic*).ti,ab,kf.
	45	(((interpersonal or attachment or relational) adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique?)) or IPT).ti,ab,kf.
Systemic therapies	46	((system? or systemic or multisystem* or milieu) adj (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique?)).ti,ab,kf.
	47	((system?‐level or system?‐based) adj (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique?)).ti,ab,kf.
	48	(((family or parent or caregiver) adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique?)) or CFTSI).ti,ab,kf.
Humanistic therapies	49	((humanistic or existential or gestalt or person‐cent?red or client‐cent?red or patient‐cent?red or non‐directive or nondirective or narrative or emotion* or solution‐focus?ed) adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique?)).ti,ab,kf.
Expressive therapies	50	((expressive or experiential or creative or play or doll? or toy? or puppet* or art or paint* or drawing or music* or dance or roleplay* or role‐play*) adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique?)).ti,ab,kf.
	51	(sandplay or sand‐play or sandtray or sand‐tray or theraplay or thera‐play or psychodrama or psycho‐drama).ti,ab,kf.
	52	(writing adj1 (therap* or intervention? or treatment? or approach or approaches or program? or programmes? or technique? or trauma or expressive)).ti,ab,kf.
Body‐oriented and somatic psychotherapies	53	"mind‐body therapies"/
	54	((body‐oriented or mind‐body or somatic or psychosomatic or psycho‐somatic or sensorimotor) adj3 (therap* or interventions? or treatment? or approach or approaches or program? or programme? or technique?)).ti,ab,kf.
	55	"somatic experiencing".ti,ab,kf.
Integrative therapies and psychotherapeutic components	56	((integrative or eclectic) adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique?)).ti,ab,kf.
	57	(psychoeducat* or psycho‐educat*).ti,ab,kf.
	58	((skill? or stabili#ing or stabili#ation or supportive or problem‐solving) adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique? or training)).ti,ab,kf.
	59	(trauma adj (focus* or informed or based or specific or sensitive)).ti,ab,kf.
	60	((PTSD or trauma) adj (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique?)).ti,ab,kf.
	61	(memory adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique? or training or reconsolidation)).ti,ab,kf.
	62	((desensiti#ation adj2 (reprocessing or eye‐movement)) or EMDR).ti,ab,kf.
	63	("accelerated resolution" or "emotional freedom" or "thought field").ti,ab,kf.
	64	(crisis adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programmes? or technique?)).ti,ab,kf.
	65	(debrief* or "critical incident stress" or "traumatic incident reduction").ti,ab,kf.
Setting and delivery	66	"community mental health services"/ or "school mental health services"/ or "mental health teletherapy"/ or "distance counseling"/
	67	((community‐based or school‐based) adj3 (therap* or intervention? or treatment? or approach or approaches or program? or programme? or technique?)).ti,ab,kf.
	68	((remote* or online or ehealth or e‐health or mhealth or m‐health or e‐mental or tele* or web or app or virtual* or mobile or internet or phone or smartphone or video‐based or digital*) adj3 ( therap* or intervention? or treatment? or approach or approaches or program? or programme? or techniques?)).ti,ab,kf.
	69	(teletherap* or telepsycho*).ti,ab,kf.
Intervention	70	or/32‐69
Children and adolescents	71	exp child/ or "child, abandoned"/ or "child, adopted"/ or "child, foster"/ or "child, orphaned"/ or "child, unwanted"/ or "child, hospitalized"/ or "child, institutionalized"/ or infant/ or adolescent/ or "adolescent, hospitalized"/ or "adolescent, institutionalized"/
	72	(child? or children? or childhood? or kid? or boy? or girl? or teen* or preteen* or youth? or youngster? or adolescen* or preadolescen* or pube* or prepube* or juvenile? or minors or p?ediatric or toddler* or infant* or underage? or under‐age? or schoolchild* or school‐age? or grader? or schooler? or preschooler? or highschooler? or kindergartener? or orphan*).ti,ab,kf.
	73	((student? or pupil? or age?) adj2 (school? or preschool? or highschool? or kindergarten? or pre‐k or grade)).ti,ab,kf.
Age group	74	or/71‐73
Adapted from Cochrane's highly sensitive search strategies for identifying randomised trials in MEDLINE Ovid and Embase Ovid; 2023 revisions (Lefebvre 2024). We adapted these two search strategies into one that better aligned with psychological therapies, could be used more consistently across our included databases, and maintained high sensitivity.	75	exp “controlled clinical trial”/
	76	"random allocation"/
	77	"comparative study"/
	78	(trial or RCT).ti,ab,kf.
	79	(random* or quasirandom*).ti,ab,kf.
	80	"cross‐over studies"/
	81	(crossover or cross‐over).ti,ab,kf.
	82	placebo.ti,ab,kf.
	83	"double‐blind method"/ or "single‐blind method"/
	84	((singl# or doubl#) adj blind*).ti,ab,kf.
	85	((parallel or intervention or control) adj group?).ti,ab,kf.
	86	((alternate or group? or intervention? or patient? or subject? or participant? or client?) adj5 (match or matched or assign* or allocation)).ti,ab,kf.
	87	"sample size"/ and control.ab.
	88	(assigned or allocated).ti,ab.
	89	(compare or compared or comparison or comparing).ti,ab,kf.
	90	"control groups"/
	91	(control* adj7 (study or design or trial?)).ti,ab,kf.
	92	"human experimentation"/ or "therapeutic human experimentation"/
	93	or/75‐92
	94	(exp animals/ or exp "animal experimentation"/) not (human?.ti,ab,kf. or humans/ or exp "human experimentation"/)
	95	(nonrandom* not random*).ti,ab,kf.
	96	(("systematic review" or "scoping review" or "narrative review" or "integrative review" or "umbrella review" or "rapid review" or "literature review") not (trial or study)).ti.
	97	or/94‐96
Study design	98	93 not 97
	99	31 and 70 and 74 and 98

Contributions of authors

All authors contributed to the development of this protocol. SCBD, DBI, MTK, and OJS drafted the initial version of the protocol. DBI and OJS drafted the revised version of this protocol, with written contributions from MSJ and LM (Why it is important to do this review), CG (Sensitivity analysis), and SKN (Selection of studies). MQ provided thorough methodological guidance throughout the writing of this protocol. DBI developed the search strategy in collaboration with a research librarian (see Acknowledgements), and MQ and MSJ reviewed the search strategy. All authors read and edited the protocol and provided expert guidance on the content and methodology. All authors approved the final draft of this protocol. Ole Jakob Storebø (OJS) is the guarantor for the review.

Sources of support

Internal sources

• Internal support, Denmark

  Many of the authors worked on this review during office time.

• Psykiatriledelsens Forskningspulje, Denmark

  DBI's and MTK's salary was partially covered by Region Zealand's Psychiatric Management's Research Fund (in Danish: Psykiatriledelsens Forskningspulje).

External sources

• University of Southern Denmark, Denmark

  DBI's and MTK's salary was partially covered by the University of Southern Denmark, through a scholastic stipend.

Declarations of interest

All authors declare no conflicts of interest. Christian Gluud (CG) is a Cochrane editor, but was not involved in the editorial process of this review. Should an author of this review turn out to be an author of a study that is potentially eligible for inclusion in this review, this will be noted in the final review, and said author will not take part in eligibility decisions, data extraction, or risk of bias and GRADE assessments of that study.

New