An overview of systematic reviews on mental health promotion, prevention, and treatment of common mental disorders for refugees, asylum seekers, and internally displaced persons

Abstract Background Migrants who have been forced to leave their home, such as refugees, asylum seekers, and internally displaced persons (IDP), are likely to experience stressors which may lead to mental health problems. The efficacy of interventions for mental health promotion, prevention, and treatment may differ in this population. Objectives With this overview of systematic reviews, we will map the characteristics and methodological quality of existing systematic reviews and registered systematic review protocols on the promotion of mental health and prevention and treatment of common mental disorders among refugees, asylum seekers, and IDPs. The findings from this overview will be used to prioritise and inform future Cochrane reviews on the mental health of involuntary migrants. Methods We searched Ovid MEDLINE (1945 onwards), Ovid Embase (1974 onwards), Ovid PsycINFO, ProQuest PTSDpubs, Web of Science Core Collection, Cochrane Database of Systematic Reviews, NIHR Journals Library, CRD databases (archived), DoPHER, Epistemonikos, Health Evidence, 3ie International Initiative for Impact Evaluation, and PROSPERO, to identify systematic reviews of mental health interventions for involuntary migrants. We did not apply any restrictions on date, language, or publication status to the searches. We included systematic reviews or protocols for systematic reviews of interventions aimed at refugees, asylum seekers, and internally displaced persons. Interventions must have been aimed at mental health promotion (for example, classroom‐based well‐being interventions for children), prevention of mental health problems (for example, trauma‐focussed Cognitive Behavioural Therapy to prevent post‐traumatic stress disorder), or treatment of common mental disorders and symptoms (for example, narrative exposure therapy to treat symptoms of trauma). After screening abstracts and full‐text manuscripts in duplicate, we extracted data on the characteristics of the reviews, the interventions examined in reviews, and the number of primary studies included in each review. Methodological quality of the included systematic reviews was assessed using AMSTAR 2. Main results The overview includes 23 systematic reviews and 15 registered systematic review protocols. Of the 23 published systematic reviews, meta‐analyses were conducted in eight reviews. It was more common for the search strategy or inclusion criteria of the reviews to state that studies involving refugees were eligible for inclusion (23/23), than for asylum seekers (14/23) or IDPs (7/23) to be explicitly mentioned. In most reviews, study eligiblity was either not restricted by participant age (9/23), or restricted to adults (10/23). Reviews commonly reported on studies of diagnosis or symptoms of post‐traumatic stress disorder or trauma (11/23) and were less likely to report on depression or anxiety (6/23). In 15 reviews the intervention of interest was focused on/ specific to psychological therapy. Across all 23 reviews, the interventions most commonly identified from primary studies were general Cognitive Behavioural Therapy, Narrative Exposure Therapy, and a range of different integrative and interpersonal therapies. Even though many reviews included studies of participants without a diagnosis of a mental health problem, they often assessed mental health treatments and did not usually distinguish between promotion, prevention, and treatment in the review aims. Together the 23 systematic reviews included 336 references, of which 175 were unique primary studies. Limitations to the methodological quality of reviews most commonly related to reporting of selection criteria (21/23), absence of a protocol (19/23), reporting of study design (20/23), search strategy (22/23), and funding sources of primary studies (19/23). Authors' conclusions Gaps exist in the evidence on mental health interventions for refugees, asylum seekers, and internally displaced persons. Most reviews do not specify that internally displaced persons are included in the selection criteria, even though they make up the majority of involuntary migrants worldwide. Reviews specific to mental health promotion and prevention of common mental disorders are missing, and there is more evidence available for adults or mixed populations than for children. The literature is focused on post‐traumatic stress disorder and trauma‐related symptoms, with less attention for depression and anxiety disorders. Better quality systematic reviews and better report of review design and methods would help those who may use these reviews to inform implementation of mental health interventions.

Together the 23 systematic reviews included 336 references, of which 175 were unique primary studies. Limitations to the methodological quality of reviews most commonly related to reporting of selection criteria (21/23), absence of a protocol (19/23), reporting of study design (20/23), search strategy (22/23), and funding sources of primary studies (19/23).

Authors' conclusions
Gaps exist in the evidence on mental health interventions for refugees, asylum seekers, and internally displaced persons. Most reviews do not specify that internally displaced persons are included in the selection criteria, even though they make up the majority of involuntary migrants worldwide. Reviews specific to mental health promotion and prevention of common mental disorders are missing, and there is more evidence available for adults or mixed populations than for children. The literature is focused on post-traumatic stress disorder and trauma-related symptoms, with less attention for depression and anxiety disorders. Better quality systematic reviews and better report of review design and methods would help those who may use these reviews to inform implementation of mental health interventions.

P L A I N L A N G U A G E S U M M A R Y An overview of systematic reviews on mental health interventions for involuntary migrants
Refugees, asylum seekers, and internally displaced persons are involuntary migrants, who have o en experienced distress when forced to leave their home, on the journey, and in the process of settling in a host country or new environment. Mental health promotion, prevention, and treatments for mental health problems such as depression, anxiety, and post-traumatic stress disorder may work di erently in these groups of people than for the general population. This overview of systematic reviews summarises the characteristics of reviews available on this topic, to help us determine which research questions are the most important to address in future Cochrane reviews.
We searched for systematic reviews and protocols of systematic reviews on mental health promotion, prevention, and treatment of mental health problems for refugees, asylum seekers, and internally displaced persons. Mental health promotion may, for example, involve a classroom-based well-being intervention for children. An example of prevention is trauma-focused therapy to prevent post-traumatic stress disorder. Treatment may, for example, include psychological therapy for depression. We found 23 systematic reviews and 15 protocols of reviews in progress. Together the 23 published systematic reviews included 336 references, 175 of which were unique studies. Reviews more commonly included refugees and asylum seekers than internally displaced persons, and were more frequently focused on adults than children. There was more attention on the treatment of post-traumatic stress disorder than there was for mental health promotion or prevention, or for the treatment of depression or anxiety. Studies of Cognitive Behavioural Therapy, Narrative Exposure Therapy, and integrative and interpersonal therapies were most likely to be included in reviews.
The quality of reviews was limited by a range of issues, many of which related to poor reporting of the review methodology.
The evidence available from systematic reviews may not match the need for evidence-based interventions for the mental health of involuntary migrants. Review authors should consider relevant groups such as internally displaced persons, children, and people with depression or anxiety, and relevant interventions such as those for mental health promotion or prevention, and treatments other than psychological therapy.

B A C K G R O U N D
The United Nations estimates there are around 40 million internally displaced persons (IDPs), 25 million refugees, and three million asylum seekers worldwide, and their numbers are growing (UNHCR 2019). While most research on involuntary migrants takes place in high-income countries, most live in low-and middle-income countries (Wainberg 2017). In addition to experiences in the country of settlement, the circumstances in which people are forced to leave their homes are likely to be extremely stressful and o en unsafe. A large priority setting exercise by the World Health Organization (WHO) Global Forum for Health Research identified people exposed to violence or trauma as a top priority for intervention in global mental health (Sharan 2009).
Compared to the general population, migrants who were forced to leave their home are more likely to experience common mental disorders. The e icacy of psychological therapies (talking therapies) may be di erent in this population. Apart from language and cultural barriers, the availability of treatment and access to treatment may be additionally restricted, depending on the host country. Even in the UK, a high-income country with a National Health Service, refugees and asylum seekers may, for example, face organisational and logistic as well as cultural, or language barriers to care. For example, the lack of a permanent home address might make it di icult for migrants to register with a general practitioner (GP) and receive notification letters of medical appointments and subsequent results (Fassil 2015).
The Cochrane Global Mental Health Satellite aims to support the production, dissemination, and implementation of systematic reviews relevant to mental health in low-and middle-income countries (Barbui 2017). This includes reviews on the e ectiveness of mental health promotion and the prevention and treatment of common mental disorders for refugees, asylum seekers, and internally displaced persons. To ensure that Cochrane Reviews on this topic fill important gaps in the literature, we undertook an overview of systematic reviews, sometimes called a scoping review or review of reviews, to produce a map of the evidence that is currently available. Rather than synthesising data on the e ectiveness of interventions from individual studies, in this overview of systematic reviews we describe the characteristics of systematic reviews, published or ongoing (including registered protocols), on mental health promotion, prevention, and treatment of common mental disorders for refugees, asylum seekers, and internally displaced persons. The resulting evidence map highlights the breadth and depth of the evidence and helps to identify priority research questions and inform the development of Cochrane Reviews on this topic.

Description of the condition
Common mental disorders considered in this review include all depressive and anxiety disorders, including post-traumatic stress disorder (PTSD). We are furthermore interested in mental health promotion and prevention of these conditions, as well as symptoms of mental health problems without a formal diagnosis.
Major depressive disorder is characterised by a period of at least two weeks of depressed mood, and is nearly always accompanied by a persistent loss of interest or pleasure in activities which were previously considered enjoyable (APA 2013). A range of symptoms may accompany these key features of depression, including weight loss or weight gain, insomnia or hypersomnia, psychomotor agitation or retardation, fatigue, loss of energy, feelings of excessive guilt and worthlessness, diminished concentration, and recurrent thoughts of death (APA 2013). Other depressive disorders listed in the Diagnostic and Statistical Manual of Mental Disorders, Fi h Edition  include those which occur in specific situations (for example, premenstrual dysphoric disorder), disruptive mood dysregulation disorder in children, and persistent depressive disorder (previously also called dysthymia; symptoms last at least two years). Bipolar disorder is not categorised as one of the depressive disorders, although depressive episodes occur as part of bipolar disorder.
Symptoms of depression and anxiety may be present simultaneously (APA 2013). Anxiety disorders, such as generalised anxiety disorder, and trauma-related disorders, such as PTSD, are treated as separate types of disorders in DSM-5 (APA 2013). For this review, we considered anxiety disorders (including phobias and panic disorder) and trauma-and stressor-related disorders (reactive attachment disorder, disinhibited social engagement disorder, PTSD, acute stress disorder, and adjustment disorder).
Anxiety disorders share symptoms of excessive fear, worry, and anxiety, and related behavioural changes. Fear is the emotional response to a perceived imminent threat, which may be real or not, whereas anxiety is the anticipation of a threat in the future. Fear is o en associated with immediate and quick responses and behaviours, including panic attacks, whereas anxiety is associated with tension, stress, and behaviours of caution and avoidance. Depending on the type of anxiety disorder and varying between patients, other symptoms may include fatigue, restlessness, irritability, di iculty sleeping, and impaired concentration. Generalised anxiety disorder and PTSD may cooccur (APA 2013).
PTSD can develop a er experiencing a traumatic event, or recurring or chronic traumatic experiences. These include experiences or events witnessed first-hand, as well as contact with others exposed to trauma. PTSD may develop immediately, shortly a er the trauma occurs, or more than six months a er the traumatic event (delayedonset) (APA 2013). PTSD symptoms include: re-experiencing traumatic events or moments (nightmares, memories, feelings, reactions); avoidance (of people, places, conversations, feelings); hyperarousal (insomnia, irritability, poor concentration); and negative thoughts and feelings (less positive feelings, loss of interest in pleasurable activities, feeling distant from others) (APA 2013).
Even though their circumstances and experiences will di er, all refugees and asylum seekers have le their country of origin because of a well-founded fear of persecution, conflict, violence, or other dangerous circumstances (UNHCR 2019). All are likely to have experienced adverse circumstances and insecurity in their home country, challenges associated with the migration journey, and challenges upon arrival and through resettlement processes in a new country. A review of Afghan refugees resettled in industrialised countries identified a range of common adverse experiences that impact on mental health, such as witnessing atrocities, losing family members, stressful escape and transit experiences, living in refugee camps, cultural and language barriers, mental health stigma, unemployment, financial hardship, and loss of status, culture, and identity (Alemi 2014).
Studies on prevalence rates of mental illness among migrants, including refugees and asylum seekers, report widely varying estimates. A review of refugees and labour migrants reported pooled prevalence estimates of 44% depression and 40% anxiety among refugees, compared to 20% and 21% respectively among labour migrants (Lindert 2009). A review including 17 studies of adult refugees resettled in Western countries found a prevalence rate of PTSD of around 9% (Fazel 2005). A recent study of 1000 Syrian refugee children and adolescents living in Lebanon and Jordan found that 46% had developed PTSD (Khamis 2019).
People who are internally displaced have been forced to leave their homes over serious safety concerns, and are staying elsewhere but within their country of origin. Prevalence rates of PTSD reported for this group include 54% of adult internally displaced persons in northern Uganda (Roberts 2008), and 56% for people who fled from the a ermath of a tsunami in Sri Lanka (Ranasinghe 2007).

Description of the interventions
This overview of systematic reviews includes interventions relating to mental health promotion, prevention of common mental disorders, and treatment of common mental disorders.

Mental health promotion
Mental health promotion usually targets the entire population (universal), but may target high-risk populations such as refugees, asylum seekers, and internally displaced persons (selected health promotion). It considers outcomes related to positive aspects of functioning and well-being rather than ill health, and in this way it is assumed to lower the risk of developing mental disorders (Tol 2015). Mental health promotion interventions include those delivered at an individual level or in a group-based format. For example, activities to encourage good mental health and development for children may take place in the classroom or in refugee camps. Programmes might be delivered in villages or neighbourhoods, for example in low-and middle-income countries a ected by humanitarian crises.

Prevention of common mental disorders
Prevention may be universal, selective (focused on vulnerable individuals or groups), or indicated prevention (for those with symptoms but no diagnosis of mental health problems) (Tol 2015). Whereas mental health promotion interventions are likely to encourage good general mental health, prevention can either be focused on general mental health or on specific common mental disorders. Children may receive trauma-focused cognitive behavioural therapy (CBT) for the prevention of PTSD, which can be delivered in groups in the case of a large-scale shared trauma (NICE 2018). An o en used prevention intervention is singlesession psychologically-focused debriefing; however, this is not recommended for the prevention of PTSD in adults or children as it may increase rather than decrease the risk of PTSD and depression (Rose 2002).

Treatment of common mental disorders
Many interventions aimed at improving symptoms of common mental disorders are available. This overview may identify many di erent interventions for the treatment of depression, anxiety, and PTSD. We therefore briefly summarise the most commonly used interventions, and courses of treatment recommended by the UK National Institute for Health and Care Excellence (NICE).

Cognitive Behavioural Therapy (CBT)
Certain types of CBT may apply specifically to this population, such as Narrative Exposure Therapy (NET), trauma-focused CBT, stress inoculation therapy or training, and culturally sensitive CBT. Trauma-focused CBT can be used for those diagnosed with PTSD, or those with PTSD symptoms, while NET is most o en used for those with complex or multiple traumas.

Other psychotherapy
Therapies for common mental disorders, depending on the severity of symptoms and specific diagnosis, range in intensity from active monitoring, psychoeducation and low-intensity psychological interventions (relaxation exercises, counselling/nondirective supportive therapy, self-help, behavioural activation) to high-intensity psychological interventions (interpersonal therapy, psychodynamic therapy) (Kendrick 2012). Patients with PTSD may be o ered Eye Movement Desensitization and Reprocessing therapy (EMDR). Some argue that arts-based programmes and expressive and creative therapies (music, drawing, play) may increase accessibility and reduce social stigma among refugee children (McDonald 2017). Creative writing and 'writing for recovery' approaches are used for adults and children in the treatment of PTSD (Baker 2018).
Treatment may be delivered to individuals, couples, or groups. In low-and middle-income countries or settings with limited resources, task-shi ing and multi-agency collaborative treatments may be more appropriate than one-to-one therapy led by highly trained mental health professionals (Silove 2017). Lay counsellors or health workers who have undertaken a short training programme may deliver counselling, behavioural therapy, or social community interventions. Task-shi ing of the delivery of interventions to less specialised workers makes it more feasible to deliver mental health treatments in low-resource settings, and may increase sustainability of implemented programmes in such settings over time.

Transdiagnostic approaches
Over the last few years, experts in global mental health have called for a move away from the traditional system of categorising patients and treatments according to diagnosis, to a more integrated 'transdiagnostic approach' of treatment according to similarity in symptoms. In low-and middle-income countries in particular; this approach may allow for a better use of

Medication
Antidepressants might be used for depression and anxiety for children and adolescents when first-line talking therapies have not worked or in the case of severe symptoms or where talking therapies are not available (NICE 2019). For adults, medication may be indicated, particularly for more severe forms of PTSD, anxiety and depression, and if someone has a preference for drug treatment. For adults with PTSD, antipsychotics may be prescribed to treat disabling psychotic symptoms or psychotic symptoms unresponsive to other treatments in PTSD (NICE 2018).

Library
Trusted evidence. Informed decisions. Better health.

How the intervention might work
As the types of interventions identified may vary widely, we describe below the hypothesised working mechanisms of the psychological and pharmacological interventions most commonly used to treat anxiety, depression, and PTSD.

Cognitive Behavioural Therapy (CBT)
CBT for depression, anxiety, and PTSD addresses patterns of thought, particularly negative thoughts and beliefs, and aims to change this way of thinking as well as changing behaviours that may accompany negative patterns of thought (Beck 1979).
In NET, a type of CBT, the patient is guided through the construction of an autobiographical narrative, with a focus on traumatic experiences . The creation of a coherent, chronological timeline of personal events is thought to help process the traumatic event (Robjant 2010;. This is a form of exposure therapy, in which a therapist exposes a patient to a traumatic situation, event, or memory. Exposure may be gradual or all at once, and may be aided by images or virtual reality. When a fear is activated by facing it, this fear can then be reprocessed as the patient becomes used to the exposure (habituation), and symptoms are reduced (Foa 2016). Stress inoculation therapy is an example of non-trauma focused therapy, derived from CBT, and was designed to help people cope with stress.
Trauma-focused CBT was originally developed for children and adolescents who su ered from sexual abuse and is now used for children, adolescents, and adults (Cohen 2012). It di ers from generic CBT in that it is recognises the influence of the child's family, it addresses problems (cognitive, behavioural, somatic, relational) relating to the trauma, and it is adaptable and mindful of family and community values and culture. Components of the intervention, such as engagement with the narrative of the exposure and education on trauma, can be adapted to the age of the patient.
Transdiagnostic CBT is designed for multiple mental disorders, and for people with multiple mental disorders, to target the common elements of multiple and co-occurring illnesses. It is based on the idea that certain cognitive and behavioural elements are shared across a range of mental health problems and diagnoses (Mansell 2009). Transdiagnostic CBT is both a type of CBT and a 'transdiagnostic approach', but is considered part of the 'CBT' category in this overview.

Third-wave CBT and behavioural approaches
Third-wave CBT approaches di er from the original, traditional model of CBT. They target the individual's relationship with cognitions and emotions, and focus on the function of cognition such as thought suppression or experiential avoidance (an attempt or desire to suppress unwanted internal experiences, such as emotions, thoughts and bodily sensations) (Hofmann 2008). Strategies used to change thinking processes include acceptance and commitment therapy, compassionate mind training, mindfulness-based therapy, and dialectic behaviour therapy. Behavioural therapies, for example behavioural activation, seek to achieve change in behavioural patterns and activities rather than cognitive patterns (Kanter 2012).

EMDR
EMDR involves treatment in which the therapist instructs the patient to focus on associations with trauma through images, memories, emotions, and thoughts, while simultaneously using visual (rapid eye movements), auditory, or tactile stimuli. This bilateral stimulation is hypothesised to facilitate reprocessing of the disturbing information associated with traumatic memories a er which symptoms reduce (Shapiro 2017). There is an ongoing discussion as to whether the bilateral stimulation is an active ingredient of the therapy and a variety of working mechanisms have been proposed. Some argue that relaxation in response to a stimulus in the absence of danger leads to positive mental and physiological changes, while others argue that traumatic images are made less vivid and emotional as the working memory is used for tasks performed simultaneously during EMDR (Landin-Romero 2018).

Social skills and assertiveness
The social interactions in di erent contexts are the focus of social skills training and assertiveness training for anxiety and depression (Jackson 1985).

Psychodynamic therapies
Grounded in psychoanalytic theory (Freud 1949), psychodynamic therapy uses the therapeutic relationship to explore and resolve unconscious conflict through the redirection of emotions to the therapist (transference) and interpretation, with relief of symptoms as an indirect outcome.

Creative therapies
Creative therapies may use writing, music, arts, dance/ movement, or drama to recall traumatic memories and process trauma associated with PTSD in a non-verbal way. Mechanisms of action are thought to include relaxation, activation and expression of memories and emotions, facilitating a sense of control and empowerment resulting from creating art, exposure through symbolic art, and rebuilding of self-esteem (Baker 2018).

Interpersonal, cognitive analytic, humanistic, and other integrative therapies
Humanistic therapies focus on the therapeutic relationship, and therapist values of empathy, genuiness, and unconditional positive regard are hypothesised to facilitate patient insight and change in symptoms (Rogers 1951). Integrative therapies, including counselling, interpersonal therapy, and cognitive analytic therapy, form a group of therapies that combine components of di erent psychological therapy models, for example from CBT, psychodynamic therapy, and person-centered approaches (Stiles 2008).

Transdiagnostic approaches
Transdiagnostic approaches vary in terms of their key mechanisms of action and may borrow from and combine di erent treatment approaches. Examples of transdiagnostic approaches implemented in low-and middle-income countries in the last Library Trusted evidence. Informed decisions. Better health.
Cochrane Database of Systematic Reviews few years include Problem Management Plus and CETA (Common Elements Treatment Approach).
Problem Management Plus combines psychoeducation, motivational interviewing, problem-solving therapy and behavioural techniques. Problem-solving therapy and behavioural therapy helps people to manage the day-to-day practical problems (such as work, relationships) associated with mental illness. Psychoeducation educates patients both on the e ects of adversities on mental health and the rationale of the treatment, while motivational interviewing is used to promote engagement with the treatment (Dawson 2015).
CETA was developed to be delivered by people who are not mental health specialists, in settings with limited resources. Elements of CETA include encouraging engagement with the intervention, psychoeducation on symptoms and the intervention, relaxation strategies, behavioural activation to encourage participation in rewarding activities, coping with emotions, and exposure therapy (Murray 2014). These elements can be delivered in di erent combinations to address various symptoms.

Medication
Antidepressants a ect the activity of neurotransmitters such as serotonin and noradrenaline, which in turn is hypothesised to a ect the regulation of mood and emotions. Selective Serotonin Reuptake Inhibitors (SSRIs) reduce the reabsorption of serotonin by the brain, which can increase positive feelings. Serotonin and Noradrenaline Reuptake Inhibitors (SNRIs) block the reabsorption of serotonin and noradrenaline. Tricyclic antidepressants (TCAs) are an older class of antidepressants which are no longer commonly used. Most TCAs work by preventing the reuptake of serotonin or noradrenaline, or both (Feighner 1999).

Why it is important to do this overview
Refugees, asylum seekers, and internally displaced persons are a large and vulnerable group of people, who are more likely than the general population to su er from a common mental disorder. At present, no Cochrane Review exists on interventions for the promotion of mental health, or the prevention or treatment of common mental disorders in this population. Future Cochrane Reviews may focus on mental health promotion, prevention or treatment, across several common mental disorders, for a wide range of interventions, in di erent age groups and populations, across di erent settings. This overview provides an evidence map of systematic reviews conducted on this topic, to identify priority research questions and inform the development of Cochrane Reviews.

O B J E C T I V E S
To map the characteristics and methodological quality of existing systematic reviews and registered review protocols on the promotion of mental health and prevention and treatment of common mental disorders among refugees, asylum seekers, and internally displaced persons.
Whereas a systematic review would normally seek to answer questions related to the e ectiveness or e icacy results of included studies, this overview of systematic reviews provides a description of the depth and breadth of the literature available and does not answer questions of e ectiveness. Data on study characteristics were extracted to give an overview of systematic reviews, ongoing or published, on this topic.
This overview is part of a Cochrane Global Mental Health satellite project to identify priorities for Cochrane Reviews in global mental health. We will produce an evidence map and a lay summary of literature identified in the overview, to provide a basis to engage with stakeholders within and outside of academia to prioritise Cochrane Reviews of mental health of refugees, asylum seekers and internally displaced persons. This will ensure that the Cochrane Global Mental Health Satellite takes forward research questions seen as a priority by stakeholders to promote a strong evidence base in global mental health.

M E T H O D S
Our overview summarises systematic reviews that include a wide range of participants, interventions, comparators, and outcomes. We followed general principles for conducting an overview of reviews, for example in the search strategy, screening of reviews, and appraisal of the methodological quality of included reviews. Other methods, however, such as the appraisal of primary studies and synthesis of results, are not relevant to the objectives of this overview. The methodology used for this overview therefore also draws on guidance from the Campbell Collaboration on evidence and gap maps (

Types of studies
Systematic reviews and protocols of systematic reviews registered in the PROSPERO online database were eligible for inclusion. Reviews had to be clearly identified by the authors as a 'systematic review' or 'meta-analysis' in either the title or abstract of the review; and the authors had to present evidence of a systematic search including a search strategy. Cochrane reviews and systematic reviews with and without a meta-analysis were eligible for inclusion. We included systematic reviews regardless of the study Library Trusted evidence. Informed decisions. Better health.
Cochrane Database of Systematic Reviews design and methodology of the primary studies. To be included, reviews had to address the evaluation of one or multiple relevant interventions. Systematic reviews were included regardless of the number or breadth of databases searched.

Types of participants
We included reviews of studies involving refugees, asylum seekers, and internally displaced persons of all ages. We adopted the following definitions of the UN Refugee Agency (UNHCR), which are derived from the 1951 Convention on the Status of Refugees (UNHCR 2019).
• Refugee: a person who, owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group, or political opinion, is outside the country of his nationality, and is unable to or, owing to such fear, is unwilling to avail himself of the protection of that country. • Asylum seeker: an individual who is seeking asylum, but whose claim has not yet been finally decided on. • Internally displaced persons: persons or groups of persons who have been forced or obliged to flee or to leave their homes or places of habitual residence, in particular as a result of, or in order to avoid the e ects of, armed conflict, situations of generalised violence, violations of human rights or natural or human-made disasters, and who have not crossed an internationally recognised border.
Depending on the type of intervention (promotion of mental health, prevention, treatment), participants may have been either diagnosed with depression, anxiety, or PTSD, or experience symptoms associated with one or more of these disorders, or not have any reported symptoms. Although treatment would be expected to be primarily given to participants with a diagnosed common mental disorder, we accepted reviews of treatment interventions with participants without a diagnosis or with elevated symptoms only.
Only systematic reviews including studies with the above population groups were eligible. If samples were mixed, for example including studies conducted in disaster zones and including internally displaced persons, the review was not eligible for inclusion in our overview.

Types of interventions
All interventions to promote mental health, or to prevent or treat common mental disorders, were eligible for inclusion. We considered common mental disorders to include anxiety disorders, including PTSD, and depressive disorders, as described in the Description of the condition section above. Eligible interventions included psychotherapies and medication, individual or group treatments, as well as interventions delivered by professionals and lay health workers. We categorised interventions as follows, according to the classification presented in the Description of the interventions section.
• Mental health promotion • Prevention of common mental disorders • Treatment of common mental disorders: CBT, other psychotherapy, transdiagnostic approaches, medication.
We planned to adapt this classification if interventions were identified that would not fit any category.
We included only interventions aimed at the promotion of mental health, or the prevention or treatment of common mental disorders, or a combination of these approaches. For example, we excluded reviews of studies evaluating the e ects of nutrition or physical activity on mental health outcomes unless the main aim of the intervention was to promote or improve mental health.
Interventions were included whether they were targeted at specific groups within the population of refugees, asylum seekers, and internally displaced persons or not, but interventions not aimed at this population were excluded.
We included systematic reviews focussing on general mental health or wellbeing without specifying particular mental health conditions eligible for inclusion.

Comparator
All types of comparators were eligible for inclusion. This included any other type of intervention including those part of 'treatmentas-usual', no intervention (including waiting list), and any type of placebo.

Types of outcome measures
Reviews that reported any mental health-related outcomes were eligible, irrespective of the measure used or length of followup. This included outcomes relating to symptoms (e.g. severity of anxiety symptoms), diagnosis (e.g. recurrence of depression), functioning, disability, quality of life, and adverse events (for example, hospitalisation or suicidal attempts). Reviews reporting on outcomes relating to constructs of positive psychological constructs, such as wellbeing, were also eligible for inclusion.

Search methods for identification of reviews Information sources
We searched the following bibliographic databases using key terms relating to the population (refugees, asylum seekers or internally displaced persons; and mental health, including depression, anxiety, PTSD), together with a filter for systematic reviews (Appendix 1). We checked the reference lists of included systematic reviews to identify additional evidence which may have been missed by the searches.

Data collection and analysis
We de-duplicated, uploaded and screened records in Covidence so ware (Covidence 2020).

Selection of reviews
Two review authors (EU, BC) independently screened titles and abstracts against inclusion criteria. We obtained full-text manuscripts for all titles that were selected during this process, contacting study authors if necessary. Full-text articles were screened by two review authors (EU, BC) independently, and we resolved disagreements through discussion, with a third review author (RC) to arbitrate if necessary. We recorded reasons for excluding full-text articles, and present a ' Table 1: Characteristics of excluded studies'. We collated multiple reports of the same systematic review.
We based selection of reviews on the inclusion and exclusion criteria relating to types of studies, participants, and interventions. We included systematic reviews regardless of reported outcomes, date and language of publication, and study quality.
For the NIHR Journals Library, PROSPERO, and Epistemonikos, abstracts could not be downloaded and imported in to Covidence. Results from these searches were screened by one reviewer (EU) on the website and, if relevant, records were added to Covidence for full-text screening in duplicate (see Di erences between protocol and review).

Data extraction and management
We created a data extraction sheet in Microso Excel to collect data from included systematic reviews, and two authors piloted the data extraction sheet by entering data from the first three included systematic reviews, making adjustments if necessary.
We recorded the following information: • Publication information: first author, year of publication, research group; • Type of review: Cochrane or non-Cochrane, published protocol (yes/no), meta-analyses (yes/no); • Population of interest at review level: involuntary migrant population (refugees, asylum seekers, internally displaced person), age (adult/child/mix), mental health diagnosis (PTSD, anxiety, depression, mix, other); • Countries of primary studies eligible for inclusion in the review: included countries/regions of origin, included study settings; • Intervention type (psychological, pharmaceutical, other) eligible for inclusion in the review, specific interventions identified in the review (see Description of the interventions); • Comparators eligible for inclusion at review level (no treatment, placebo, waiting list, treatment-as-usual, other treatment); • Intervention provider at review level (professional, lay worker, mix, other including non-specialist from a non-governmental organisation); • Types of primary studies (RCTs, other designs); • References of included primary studies.
A separate spreadsheet was used to record the methodological quality of the included systematic reviews. For protocols of systematic reviews not yet completed, we collected as much of this information as possible.
We developed a guidance document with an explanation of each of the data extraction items and their categories for all authors taking part in data extraction, to ensure that authors extracted the same data using the same categories for the di erent variables of interest.
We did not extract data on the e ects of interventions from the included systematic reviews or their included primary studies, because the aim of the overview was to map review characteristics.
We planned to extract references of the included primary studies from each systematic review so that we could create a matrix of primary studies in included systematic reviews or order to assess overlap in the primary studies reported by various reviews. We did not plan to extract information from the primary studies, because for this overview we were interested in characteristics of the reviews rather than the primary studies included in the reviews.
Data extraction was performed in duplicate by two reviewers (EU, BC, LR, and FVJ), and we resolved any disagreements through discussion, with the involvement of a third author if required (MP, CB, RC).
Since this overview does not report outcome data, we did not include a 'Summary of findings' table.

Assessment of methodological quality of included reviews
We used AMSTAR 2 to critically appraise included systematic reviews (Shea 2017). This tool is suitable for reviews including randomised and non-randomised studies. It includes 16 domains relating to the research question, review design, search strategy, study selection, data extraction, justification for excluded studies, description of included studies, risk of bias, sources of funding, meta-analysis, heterogeneity, publication bias, and conflicts of interest.
AMSTAR guidance for the following domains is particularly relevant to the interpretation of our assessment findings. Cochrane Database of Systematic Reviews • Data extraction: Data extraction should be performed in duplicate by two reviewers, or with a sample performed in duplicate and good agreement between reviewers. • Meta-analysis: If no meta-analyses are performed, authors are still expected to discuss any potential heterogeneity in the results and how this may a ect the conclusions of the review.
For many of the AMSTAR 2 domains, a positive response is only possible if the required information is reported in the review paper or protocol. The quality rating of the design and conduct of a review therefore depends heavily on the quality of the reporting of a review. We contacted authors in case of missing information and used this information to inform the quality assessment. For protocols of reviews not yet completed, we did not perform a quality assessment.
We used findings from the AMSTAR 2 critical appraisal to understand the certainty of the evidence base of systematic reviews, which in turn informs what future systematic reviews and primary research is needed.
Our approach deviates from the AMSTAR 2 guidance as we only used the individual domains and we did not produce ratings of overall confidence in the findings of each review. Since our overview did not extract data on the findings of reviews, we did not deem ratings in the confidence of review findings appropriate.
In the discussion section of this review, we specifically consider the characteristics of two high-quality systematic reviews included in our overview. We did not prespecify criteria for high quality reviews. The two highlighted reviews received the highest quality assessment rating for the majority of AMSTAR 2 domains (see Di erences between protocol and review).

Data synthesis
We reported results as a narrative synthesis of the characteristics of included systematic reviews.
We included the following.
• A table of all characteristics of included systematic reviews specified in the 'Data extraction and management' section. • A description of ongoing reviews with study characteristics based on registered or published review protocols. • An inventory of all interventions and comparators included in the identified systematic reviews. • An assessment of overlap in primary studies included in the selected reviews. • A figure of the evidence and gaps in the evidence.

R E S U L T S
Searches were conducted on the 4th of September 2019. Two authors (EU and BC) screened titles and abstracts of 4613 records and 63 were included for full-text screening (EU, BC) ( Figure 1). The most common reasons for exclusion were that the review included the wrong population (N = 8), for example, migrants instead of refugees, asylum seekers, and internally displaced persons only, and that the review was not a systematic review or systematic review protocol (N = 8). All excluded studies with reasons for exclusion are listed in the excluded studies table (Table 1). For three studies, we could not include findings because no full text could be obtained (Anders 2016; Khan 2018; Piegenschke 2019). Reviews and protocols were published in English, except for two reviews in German (Anders 2016; Piegenschke 2019), and one review in Chinese (Liu 2009a).

Library
Trusted evidence. Informed decisions. Better health.

Description of included reviews
We included 23 published systematic reviews including one review of systematic reviews, and 15 protocols of additional ongoing or planned systematic reviews registered in PROSPERO (https:// www.crd.york.ac.uk/prospero/). In this section we first describe the registered protocols and then the published reviews.

Registered protocols of ongoing systematic reviews
We found fi een protocols of systematic reviews registered in PROSPERO, which were planned or still ongoing. One of these is an unpublished Cochrane review protocol in progress (Soltan 2018). Twelve protocols specified plans to conduct meta-analyses; seven of these included only RCTs ( In this section, we summarise the review questions and eligibility criteria of registered review protocols. If protocols did not specify certain aspects of inclusion and exclusion criteria, we assumed these criteria were not used to select studies. For example, if trial setting was not specified, we assumed trials conducted in any setting were included.
An overview of all included protocols can be found in Table 2.

Scope and selection criteria
Most of the protocols stated that primary studies in any setting would be eligible for inclusion (10/15). One review planned to include studies from Europe (Danmole 2017), two planned to include studies conducted in high-income countries (Soltan 2018; Wright 2019), one to include studies from low-and middle-income countries (Jaroudy 2018), and according to one protocol studies from low-income countries in the Middle East were eligible for inclusion (Alzaghoul 2019).

Study participants
All fi een review protocols specified that primary studies of refugees were eligible for inclusion. Twelve specified that asylum seekers were included, and four reviews listed internally displaced persons as participants eligible for inclusion ( Some of the review protocols focussed on participants with a specific diagnosis, such as PTSD or trauma-related illness (5/15), or a range of diagnoses, such as depression, anxiety, and PTSD (Jaroudy 2018; Lawton 2019). Other review protocols included any mental health problem or diagnosis (4/15) or measured mental health outcomes in participants who did not necessarily have a mental health condition (4/15).

Interventions and comparators
While in some review protocols any interventions were eligible for inclusion (4/15), others focused on specific or several types of therapy (NET, CBT, arts-based, low-intensity) (6/15), or a broad range of interventions (3/15). One review protocol focused on preventative interventions, and included only community-based interventions (Soltan 2018). Another review protocol was of community-based interventions delivered by lay workers (Aslam 2018). All other review protocols did not specify the intervention provider eligible for inclusion.
For most review protocols, any comparator was eligible for inclusion (11/15), while for four review protocols eligible comparators were specified. One review protocol included any comparator except pharmacological treatment (Alzaghoul 2019); the other three included a range of comparators such as no intervention, treatment-as-usual, waiting list, or other therapy (Aslam 2018; Hameed 2017; Miyazaki 2018).

Review status
In November 2019, we contacted authors of all review protocols registered in PROSPERO to enquire about the status of the review. Three authors did not respond (Danmole 2017; Jaroudy 2018; Miyazaki 2018). The author of one protocol indicated that their review would not be completed or published (Hameed 2017). Five reviews were ongoing at the time of our enquiry and five reviews were either nearing submission for publication or under review with a journal.

Completed (published) systematic reviews
None of the published systematic reviews were Cochrane reviews. All were published in the last ten years, between 2009 and 2019. Meta-analyses were conducted in eight reviews, four of which included only RCTs (Crumlish 2010; Lambert 2015; Morina 2019; Turrini 2017). Published reviews included in this overview are summarised in Table 3.

Scope and selection criteria
Reviews mostly included studies from any setting. Two reviews included only studies from high-income countries (Nosè 2017) or developed countries (Nakeyar 2016).

Study participants
Refugees were explicitly included in the search strategy or selection criteria of all reviews, asylum seekers in 14 out of 23 reviews, and internally displaced persons in seven out of 23 reviews. Reviews included study participants of any age (9/23), children or children and adolescents only (4/23), or adults only (10/23). Reviews most commonly focused on a diagnosis or symptoms of PTSD or trauma (11/23). Others included various disorders such as PTSD, anxiety, and depression (5/23), or measured mental health in general or included any mental health problem (7/23). Figure 2 shows the number of systematic reviews identified by type of mental health condition, participant age group, and category of refugee.

Library
Trusted evidence. Informed decisions. Better health.

Figure 2. Evidence and gap map Interventions and comparators
Most reviews were focussed on psychological interventions, sometimes called psychosocial interventions by review authors, or psychotherapy (15/23), and two of these reviews also included pharmacological treatments (Crumlish 2010; Turrini 2019). One review included only studies of pharmacological interventions (Sonne 2017), two focused on community-based interventions (Tyrer 2014; Williams 2011), and five included any intervention (Eberle-Sejari 2015; Liu 2009a; Nocon 2017; Sims 2017; Van Wyk 2014). Although many reviews included studies of participants who had not been diagnosed with a mental health condition, none of these focused on mental health promotion or prevention interventions. Given the nature of the interventions, it appeared these reviews focussed on treatment of mental health conditions or symptoms rather than prevention.
Most reviews included any type of comparator (21/23). One considered active comparators or placebo for pharmacological interventions, and any comparator for psychological interventions (Crumlish 2010). One review did not include any comparators as it was a review of studies without a control group (Van Wyk 2014).

Number of included studies and unique studies
In total, the 23 systematic reviews included in this overview comprised 336 references to primary studies, of which 175 were unique primary studies (see Appendix 2). A large number of studies was included in only one of the systematic reviews (N = 113), while the two most frequently included RCTs were included in nine systematic reviews (Neuner 2008;).

Methodological quality of included reviews
We assessed the methodological quality of all included systematic reviews in duplicate using AMSTAR 2. Assessments of all reviews per AMSTAR 2 domain are shown in Figure 3.

Study selection
Study screening and selection was performed in duplicate, or partly in duplicate with good agreement between reviewers, for five reviews (Liu 2009a; Mitra 2019; Naseh 2019; Turrini 2017; Turrini 2019). Four reviews did not comprehensively describe all included studies (Alfadhli 2016; Mitra 2019; Nakeyar 2016; Williams 2011) and, for 17 reviews, no list of excluded studies with reasons for exclusion was provided.

Data extraction
Data extraction was performed in duplicate for five reviews (Liu 2009a; Naseh 2019; Nosè 2017; Turrini 2017; Turrini 2019). In the other 18 reviews, data extraction was not done in duplicate.

Risk of bias
Risk of bias was not assessed in all reviews. Eleven reviews reported on the key criteria for RCTs (generation of the allocation sequence, selection bias) or non-RCTs (methods, selection bias). Four reviews also reported on funding sources of included studies (Nosè 2017; Thompson 2018; Turrini 2017; Turrini 2019).

Meta-analysis
Fi een reviews did not perform meta-analyses, and various domains of AMSTAR 2, therefore, do not apply to these reviews. Three reviews performed meta-analyses but did not use appropriate methods, for example without a justification for combining studies (Gwozdziewycz 2013; Morina 2019; Naseh 2019).
Of the eight reviews in which meta-analyses were reported, in three reviews the impact of potential biases on the results was discussed (Nocon 2017; Nosè 2017; Turrini 2019). Two reviews with metaanalyses assessed the risk of publication bias (Nosè 2017; Turrini 2019).
Out of the 15 reviews for which no meta-analyses were carried out, ten included a discussion of the potential impacts of heterogeneity of the evidence.

Conflicts of interest
Thirteen reviews either reported any conflicts of interest or reported that the authors had no conflicts of interest.

Characteristics of high-quality systematic reviews
Two systematic reviews with meta-analyses received the highest quality assessment rating for 14 out of 16 domains (Nosè 2017) and 13 out of 16 domains (Turrini 2019). These reviews were conducted by authors from the same institution, with overlap in the author team. Their inclusion criteria specified either RCTs only (Turrini 2019) or RCTs and non-randomised trials with a control group (Nosè 2017). Both specified in their eligibility criteria that refugees and asylum seekers were included, either adults only (Nosè 2017) or of all ages (Turrini 2019). Together these reviews covered primary studies of psychological, social, or rehabilitation interventions for PTSD (Nosè 2017) and diagnoses or symptoms of PTSD, depression, and anxiety (Turrini 2019). Any comparators were eligible for inclusion. Out of the 40 references included in these two reviews, there were 28 unique primary studies.

E ect of interventions
In this overview of systematic reviews, we did not report on the e ectiveness of interventions in the included reviews.

Summary of main results
This overview of systematic reviews included 23 completed reviews and 15 protocols of reviews registered in PROSPERO, which were planned and/or ongoing.
The 23 included systematic reviews comprised 336 references, of which 175 were unique primary studies. All were published between 2009 and 2019. Four systematic reviews included RCTs only. Reviews included studies from any setting, except for one review of studies from high-income countries and one review of studies from developed countries. As for the population, inclusion criteria were more o en specific in their stated inclusion of refugees than asylum seekers or internally displaced persons ( Figure 2). Most reviews focused on psychological therapies, and there was more evidence on interventions for the treatment of PTSD or trauma-related symptoms from reviews of any age or adults only than there was for children, people with depression or anxiety, or prevention or mental health promotion (Figure 2). Pharmacological treatments were considered in only two reviews. Interventions most frequently reported in reviews included Library Trusted evidence. Informed decisions. Better health.

Cochrane Database of Systematic Reviews
CBT-based approaches, integrative and interpersonal therapies, trauma therapies (including testimony therapy), and creative therapies. Less evidence was available for transdiagnostic therapy, psychodynamic therapy, education, and medication. Review elements compromising the methodological quality of the reviews included the absence of reference to a review protocol, the lack of a comprehensive search strategy, and single screening and/or data extraction of studies included in the review.
Most of the 15 review protocols of ongoing systematic reviews included studies from any setting on interventions for the treatment of common mental disorders. Specific settings included Europe, high-income countries, low-and middle-income countries and low-income countries in the Middle East. Study participant eligibility criteria specified refugees, or refugees and asylum seekers, but only four out of 15 protocols listed internally displaced persons as participants eligible for inclusion. Interventions were most o en aimed at treating (symptoms of) PTSD or trauma, or the reviews focussed on interventions for mental health in general. Where interventions eligible for inclusion in the reviews were specified, these included various types of psychological therapy, preventative interventions, and community-based interventions. For the majority of protocols, any comparator was eligible for inclusion.

Overall completeness and applicability of evidence
A wide range of research questions could be asked in relation to mental health promotion, prevention, and treatment of common mental disorders for refugees, asylum seekers, and internally displaced persons. This overview of systematic reviews shows that some of these questions have received limited attention in reviews published to date. We did not identify many reviews seeking to assess the e icacy of treatments for children or for depression and anxiety ( Figure 2). Inclusion criteria of reviews o en did not include internally displaced persons, and few of the included primary studies evaluated pharmacological treatments. The di erence between mental health promotion, prevention, and treatment of common mental disorders was usually not explicitly made by review authors, but most reviews included interventions typically delivered as treatment, although some can and are used as preventive interventions.
These gaps in the evidence partly reflect the availability of primary data and partly reflect decisions made in the design of the reviews. For example, the greater availability of data on therapies for the treatment of PTSD rather than anxiety or depression appears to be influenced by a greater evidence base for PTSD in this population. The absence of internally displaced persons from many reviews however is likely to be due to the restricted selection criteria and search terms applied by review authors, which o en did not include all three groups of involuntary migrants and instead focussed on refugees.
Most systematic reviews included primary research not limited to RCTs. Although the literature appears to include many nonrandomised and uncontrolled study designs, it is unlikely that robust conclusions on the e ectiveness of treatments can be drawn from these studies.
We assessed the quality of the reviews using AMSTAR 2. For many of the AMSTAR 2 domains, reviews can only achieve a high rating if review design and methods are clearly reported. A review scoring lower on this quality assessment may therefore either be of poorer methodological quality, su er from substandard reporting, or both. Although we did not report on the findings from the reviews, it is clear that the majority of included reviews were lacking both in methodological quality and the transparancy of reporting of the review findings. This would hinder any application of the evidence in practice. For example, without evidence of a comprehensive search strategy, it is unclear whether all relevant primary studies have been included. The lack of reporting on conflicts of interest of review authors makes it di icult to assess the risk of bias a ecting review findings.

Quality of the evidence
Most reviews did not register or publish a protocol. This made it di icult to appraise the methodology of the reviews. For most reviews, participants, interventions, comparators, and outcomes (PICO) were not made explicit, the study design was not fully explained, and the search strategy was not comprehensive ( Figure  3). Most author teams did not perform study selection and screening in duplicate. Excluded studies were not usually reported, and included studies were not always described in su icient detail. For reviews in which a meta-analysis was performed, there was evidence that methods applied were not appropriate in three out of eight reviews. In most reviews, potential limitations of primary studies were discussed, but formal 'Risk of bias' assessments, including reporting of primary study funding sources, were not carried out. For ten out of 23 reviews, no conflict of interest statement was included.
Two systematic reviews were ranked as high-quality across the majority of the assessed domains and may therefore be more informative to future research on this topic (Nosè 2017;Turrini 2019). In line with other reviews included in this overview, these two reviews focussed on refugees and asylum seekers, and included either adults or participants of all ages. One review covered interventions for PTSD only, while the other included diagnoses or symptoms of PTSD, depression, and anxiety.

Potential biases in the overview process
It is possible that we did not identify all relevant reviews and review protocols. Protocols were only searched through the PROSPERO website, while authors may have registered review protocols on other online platforms. Reviews were only included if the title or abstract specified that a 'systematic review' was conducted and if there was evidence of a systematic search including a search strategy. Reviews on a slightly di erent population, for example, of studies conducted with people living in settings of humanitarian crises, could be informative for the population of involuntary migrants. We did not include such reviews unless it was made explicit that participants were refugees, asylum seekers, or internally displaced persons.
The list of primary studies from the included reviews is not a comprehensive overview of the literature. Other relevant studies will have been published that were not included in reviews. This overview of systematic reviews will guide the development of new Cochrane reviews, to identify these primary studies.
A limitation of our overview is that we could not obtain a full-text manuscript for three reviews which were eligible for inclusion in the overview (Anders 2016; Khan 2018; Piegenschke 2019).

Library
Trusted evidence. Informed decisions. Better health.

Cochrane Database of Systematic Reviews
Some of the primary studies in the systematic reviews that were included may not be relevant to the topic of this overview. We selected reviews based on the eligibility of the review and review methods, rather than assessing the eligibility of primary studies. For example, some of the studies may discuss mental health of our target population without evaluating an intervention. Also, there may be multiple reports of the same primary study using di erent manuscript titles among the 175 unique references we identified.
Two of our authors (MP, CB) are also co-authors for some of the included reviews and review protocols registered in PROSPERO (Nosè 2016; Nosè 2017; Turrini 2017; Turrini 2019; Turrini 2019a). In line with Cochrane guidance, these authors were not involved in any of the data extraction or quality assessments of reviews in this overview. We therefore do not expect the involvement of these authors in the overview to bias our findings.

Agreements and disagreements with other studies or reviews
In 2017, researchers including two of our co-authors (MP, CB) published a review of systematic reviews on the prevalence and treatment of common mental disorders in asylum seekers and refugees (Turrini 2017). This umbrella review is included in our overview and comprises 14 systematic reviews on the e icacy of mental health interventions. We identified the same reviews, but excluded six reviews because they were not specifically of involuntary migrant populations. We included eight of the same reviews and 15 additional reviews, including those published a er April 2017.
Turrini 2017 reported predominantly on reviews of NET and di erent types of CBT, and included fewer studies of other interventions such as EMDR, trauma-focused therapy, testimony therapy, and antidepressants. Primary studies on PTSD were more frequently included in their reviews than studies on depression and anxiety, despite their finding that depression and anxiety were at least as frequent as PTSD among refugees and asylum seekers. These gaps in the evidence are lin line with our findings.

Implications for practice
This overview did not focus on the e icacy of interventions and therefore cannot inform practice directly. It does indicate, however, that the evidence available to decision-makers in clinical practice and policy is predominantly of limited quality. Only two systematic reviews, reporting on interventions for common mental disorders in refugees and asylum seekers (Turrini 2017) and PTSD for refugees and asylum seekers in high-income countries (Nosè 2017), achieved a high-quality rating across most domains.
Most systematic reviews included primary studies other than RCTs, which would make it di icult to draw conclusions on the e ectiveness of interventions. For meta-analyses conducted in reviews with severe methodological limitations, the produced e ect estimates may give the impression of evidence readily applicable to practice, while these findings may be biased.
For evidence to inform practice, the evidence base should provide information for populations, settings, and interventions relevant to practice. Our overview shows that important groups such as internally displaced persons, children, and people with depression and anxiety are less likely to be considered in systematic reviews. Evidence on mental health promotion and the prevention of mental health problems, which could be an important avenue for early intervention a er resettlement, is largely absent.

Implications for research
Many of the 23 included systematic reviews, and the registered review protocols, focused on similar interventions for similar populations. This is illustrated by our finding of 175 unique primary studies among 336 references identified in reviews. Meanwhile, several relevant groups were underrepresented in reviews, leading to gaps in the evidence. Based on the evidence identified in this overview, we see potential for future reviews to address the following research questions for the population of involuntary migrants.
1. What is the e icacy of prevention and treatment of common mental disorders other than PTSD?
Treatment for (symptoms of) PTSD was included in two highquality reviews and may therefore not require further evidence synthesis at this time. However, since PTSD is by no means the only mental health problem facing involuntary migrants, reviews of interventions for anxiety and depressive disorders as well as transdiagnostic approaches are required. For established therapies such as CBT and NET, several reviews including a range of primary study designs were identified. To answers questions on the e icacy of interventions, systematic reviews of RCTs would be most useful.

What is the e icacy of mental health promotion and prevention and treatment of common mental disorders for children?
Most reviews we identified did not focus on children (Figure 2), even though di erent interventions are available and appropriate for children as well as adults and the e icacy of interventions may di er between children and adults.

What is the acceptability of interventions for involuntary migrants?
Systematic review authors may wish to consider outcomes such as dropout rates, cultural appropriateness, and cost-e ectiveness. Many of the interventions identified in this overview were not developed for involuntary migrants, which raises unanswered questions about the appropriateness of interventions for this population. Particularly where there are indications of limited e icacy of interventions, measures of acceptability may indicate whether adaptations of interventions to the population or setting are required. Most involuntary migrants live in low-and middleincome countries, where resources are limited and transdiagnostic or task-shi ing approaches may be more appropriate than traditional, resource-heavy psychological therapies.
To answer any of these research questions, we suggest two ways to strengthen the existing evidence base. Firstly, systematic review authors should consider the explicit inclusion of refugees, asylum seekers, and internally displaced persons in their objectives, selection criteria, and search terms. Internally displaced persons form the largest group of involuntary migrants globally, yet were o en not explicitly included in systematic reviews. Secondly, high quality reviews with transparant and complete reporting of review design and methods would aid anyone using these reviews to inform decisions on the implementation of new and existing Library Trusted evidence. Informed decisions. Better health.
Cochrane Database of Systematic Reviews mental health interventions in practice. For example, the online registration or publication of a review protocol, a description of selection criteria, and an assessment of the quality or risk of bias of included studies are key review elements which should not be missing from any systematic review.

A C K N O W L E D G E M E N T S
We thank the editorial team of the Cochrane Common Mental Disorders Group (CCMD) Group for their guidance during protocol development. We developed the search strategies with Sarah Dawson, the CCMD Information Specialist.

Alzaghoul 2019
Alzaghoul A, McKinlay A. E ectiveness of psychological interventions for treating post-traumatic stress disorder in children and adolescents a ected by war, conflicts and manmade disasters in low/middle income countries in the Middle East. A systematic review of randomized controlled trials. crd.york.ac.uk/prospero/display_record.php?RecordID=140370 (accessed 09 September 2019).

Aslam 2018
Aslam RW, Dowrick C, Hagan R, Waqas A, Winrow EW, Edwards R. Barriers and facilitators to uptake of psychosocial interventions delivered by lay therapists to improve mental health and wellbeing of asylum seekers and migrants: a systematic review. crd.york.ac.uk/prospero/ display_record.php?RecordID=104453 (accessed 10 September 2019).

Jaroudy 2018
Jaroudy S, Thornicro G, Molyneaux E. Are psychosocial interventions e ective for reducing symptoms of depression, anxiety or PTSD in displaced or refugee adults in low or middle income countries who have been a ected by conflict? A systematic review protocol and planned meta-analysis. crd.york.ac.uk/prospero/display_record.php?RecordID=102081 (accessed 09 September 2019).

Soltan 2018
Soltan F, Vanderbloemen L. Systematic review of communitybased interventions for improving health and well-being in refugee children and adolescents a er resettlement in developed countries. crd.york.ac.uk/prospero/ display_record.php?RecordID=99102 (accessed 09 September 2019).

Thompson 2018
Thompson CT, Vidgen A, Roberts NP. Psychological interventions for post-traumatic stress disorder in refugees and asylum seekers: a systematic review and meta-analysis. Clinical Psychology Review 2018;63:66-79.

Tribe 2019
Tribe RH, Sendt KV, Tracy DK. A systematic review of psychosocial interventions for adult refugees and asylum seekers. Journal of Mental Health 2019;28 (6)

Williams 2011
Williams ME, Thompson SC. The use of community-based interventions in reducing morbidity from the psychological impact of conflict-related trauma among refugee populations: a systematic review of the literature. Journal of Immigrant and Minority Health 2011;13(4):780-94.

Library
Trusted evidence. Informed decisions. Better health.

Albane 2019
Albane B, Lisé C. Education in emergencies in humanitarian responses, results of creative or artistic educational practices in conflict areas and in camps promoting the well-being of children: protocol for a systematic review (meta-synthesis). crd.york.ac.uk/prospero/display_record.php?RecordID=124913 (accessed 09 September 2019).

Ho 2018
Ho R, Clark A, Saunders S. What are the facilitators and barriers to access mental health services for refugees and asylum seekers: a systematic review. crd.york.ac.uk/prospero/ display_record.php?RecordID=88629 (accessed 10 September 2019).

Murray 2010
Murray KE, Davidson GR, Schweitzer RD. Review of refugee mental health interventions following resettlement: best practices and recommendations. American Journal of Orthopsychiatry 2010;80 (4)     An overview of systematic reviews on mental health promotion, prevention, and treatment of common mental disorders for refugees, asylum seekers, and internally displaced persons (Review)  Table 3. Included systematic reviews (Continued) Cochrane Library Trusted evidence. Informed decisions. Better health.

A D D I T I O N A L T A B L E S
An overview of systematic reviews on mental health promotion, prevention, and treatment of common mental disorders for refugees, asylum seekers, and internally displaced persons (Review)  Table 3. Included systematic reviews (Continued)

D I F F E R E N C E S B E T W E E N P R O T O C O L A N D R E V I E W
We did not search Ovid Global Health and BIREME/PAHO LAtin America and the Caribbean DAtabase (LILACS) because we did not have the access and resources required to do so.
For the NIHR Journals Library, PROSPERO, and Epistemonikos abstracts could not be downloaded and imported in to Covidence. Results from these searches were screened by one reviewer (EU) on the website and, if relevant, records were added to Covidence for full-text screening in duplicate.
We had planned to create a matrix of primary studies from all included reviews, to show overlap between reviews and identify unique primary studies included in the reviews. However, because of the large number of studies identified in the included reviews, a matrix type figure would have been too large and not informative for this overview. Instead, we reported the total number of records identified and the number of unique records.
In the results and discussion sections of this review, we added a paragraph highlighting the characteristics of two included reviews rated 'high quality' in response to a peer-reviewer suggestion. Although we had not planned to highlight high-quality reviews, including this information meant we could point out which reviews might be considered more informative to future research on this topic.