https://orcid.org/0000-0003-3012-4584
Figures
Abstract
Group psychosocial interventions can be a scalable treatment for depression, anxiety, and post-traumatic stress disorder (PTSD) in low- and middle-income countries (LMICs) but effects are heterogeneous. Little is known about how intervention mechanisms and context interact to generate different outcomes in different populations. We conducted a realist systematic review, combining traditional systematic review methods with analysis of intervention mechanisms and contextual factors to further understanding of complex interventions. This involved: (i) a scoping review to build initial theory and inform data extraction and analysis; (ii) systematically searching six databases for randomised controlled trials (RCTs) of group psychosocial interventions for participants aged 0–19 years in LMICs (21 November 2022, updated 26 April 2024); (iii) extracting data on outcomes, contextual factors, and intervention mechanisms to build hypotheses about how context interacts with mechanisms to generate outcomes (context-mechanism-outcome configurations; CMOCs); (iv) narratively synthesising CMOCs into wider programme theories about how and why psychosocial interventions work; (v) meta-analyses and meta-regressions to assess trends and test CMOCs. We included 38 RCTs with data for 6,086 participants (52% female, mean age 13). These data informed 14 programme theories including theories that interventions work best when adapted for participant cognitive level, incorporating age-appropriate activities, accounting for local gender-specific issues, and being culturally adapted. Pooled post-intervention effect sizes were -0.72 (95% CI -1.01 to -0.42, p < 0.001) for depression, -0.90 (95% CI -1.57 to -0.23, p = 0.014) for anxiety, and -0.71 (95% CI -1.05 to -0.38, p < 0.001) for PTSD. The only significant results in meta-regressions were larger effect sizes for older children and in populations without exposure to conflict for depression symptoms only. Socio-demographic and contextual factors may influence how interventions work and help to explain the heterogeneity of effects. More high-quality RCTs with moderation and mediation analyses are needed to explore the transferability of these interventions.PROSPERO registration: CRD42022364043.
Citation: Jokinen T, Abassi M, Hodsoll J, Atmore KH, Bonell C, Rose-Clarke K (2026) Group psychosocial interventions for anxiety, depression, and post-traumatic stress disorder in children and adolescents in low- and middle-income countries: A realist systematic review and meta-analysis of randomised controlled trials. PLOS Ment Health 3(1): e0000533. https://doi.org/10.1371/journal.pmen.0000533
Editor: Olumide Thomas Adeleke, Bowen University, NIGERIA
Received: March 12, 2025; Accepted: December 16, 2025; Published: January 29, 2026
Copyright: © 2026 Jokinen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The datasets used and/or analysed during the current study are available in Supporting Information.
Funding: TJ is funded by an NIHR Academic Clinical Fellowship (ACF-2021-17-009). KR-C is funded by a UKRI Future Leaders Fellowship (MR/W00285X/1). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Child and adolescent mental health is a growing global health priority. The World Health Organisation (WHO) estimates that one in seven adolescents experiences a mental disorder at any given time, accounting for 13% of the global burden of disease [1], and there is likely to be a greater burden of subsyndromal distress that never presents to services. Mental health problems in childhood and adolescence have been associated with issues including poor school achievement and substance misuse [2]. Many mental health problems persist or worsen in later life, leading to significant morbidity and mortality with associated economic burdens [3].
Globally, depression is the fourth highest cause of disability-adjusted life years lost among 10- to 24-year olds. Anxiety disorders are the most prevalent mental disorders among adolescents, affecting 3.6% of 10- to 14-year-olds and 4.6% of 15- to 19-year-olds [4]. Many children and adolescents are exposed to trauma and go on to develop symptoms of post-traumatic stress disorder (PTSD). A systematic review estimated rates of up to 87% for symptoms of PTSD among children and adolescents in low- and middle-income countries (LMICs) exposed to traumatic experiences [5].
Most of the world’s children and adolescents live in LMICs. In these settings, the WHO recommends psychosocial interventions as first-line treatment for common mental disorders due to their relative safety compared to pharmacological treatments, accessibility and efficacy [6]. Group interventions are useful because multiple adolescents can be treated at the same time by one facilitator with knock-on cost savings, and may offer additional advantages over individual therapy in terms of the group climate, cohesion and alliance. Group interventions may reduce stigma, potentially allowing participants to speak more openly about problems and receive support from others [7]. Despite WHO recommendations, a recent umbrella review of psychosocial interventions for mental health conditions in LMICs found evidence of effectiveness for children and adolescents to be lacking and inconsistent [8]. A meta-analysis of psychotherapy interventions for youth in LMICs highlighted potential benefits, particularly relative to interventions in high-income countries, but inconsistent effects and significant risk of bias [9]. Developing and testing more effective and scalable interventions is therefore a priority, but a major challenge is the lack of understanding about how interventions work, and the contextual and individual factors that affect outcomes [10]. Knowing what works, for whom, and in what settings is crucial for policymakers to make informed decisions for their local context.
We aimed to address this research gap using a ‘realist systematic review’ approach combining realist synthesis with systematic searching and theory-led statistical testing [11]. This modification of the traditional realist review approach [12,13] adds rigour by including all pertinent studies rather than stopping at theoretical saturation, considering study designs in inclusion criteria, and synthesising outcome, moderation and mediation data to determine what light these shed on theorised mechanisms [11]. Key realist terminology is summarised in Table 1.
Our main question was: what are the participant characteristics, contextual factors and intervention mechanisms reported by studies of psychosocial interventions for depression, anxiety and PTSD among children and adolescents in LMICs; and how do they interact to explain differences in effects of such interventions?
Subsidiary questions were:
- How are interventions theorised to work in terms of context-mechanism-outcome configurations (CMOCs)?
- What does qualitative research nested within intervention studies suggest about possible refinements to these CMOCs?
- What do statistical analyses reported by these studies suggest about these CMOCs?
Methods
We used a modified realist systematic review approach which involved: (i) a scoping review to identify initial theories of interest to focus data extraction and build a theoretical framework to inform subsequent formulation of CMOCs; (ii) a systematic search to include all pertinent studies; (iii) data extraction focused on identifying theory, moderation and mediation analyses, and quantitative outcome data; (iv) a risk of bias assessment; (v) formulating CMOCs and programme theories using reciprocal and refutational synthesis; and (vi) a meta-analysis and meta-regressions of quantitative outcome data to assess overall patterns and test CMOCs and programme theories. Methods are described in more detail below.
We adhered to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 2020 statement and Realist And MEta-narrative Evidence Syntheses: Evolving Standards (RAMESES) publication standards [15,16]. The review protocol was registered on PROSPERO in October 2022 (CRD42022364043).
Scoping review
We conducted the scoping review in September 2022 to identify initial theories of interest to focus data extraction and build a theoretical framework to inform subsequent formulation of CMOCs. We included the five systematic reviews in the umbrella review by Barbui et al [8]. We also included 17 reviews that were excluded from this umbrella review on the basis that they did not conduct a meta-analysis. We included any psychosocial intervention for children and adolescents in LMICs. In the scoping review we included individual or group interventions for any mental disorder. As most studies focused on outcomes for depression, anxiety, and PTSD, we selected these as the outcomes for the systematic search in order to ensure a degree of consistency and generalisability within and across findings. Key findings are summarised here in the Methods section because they informed our approach to data extraction and quantitative analysis in the main review.
We synthesised prospective theories about intervention mechanisms and contextual factors suggested by authors of included studies, and explored to what extent they were supported by findings from related study process evaluations and subgroup/moderation/mediation analyses. Theories could be summarised under four overarching themes requiring further exploration: gender and age [17,18]; social protective factors [19,20]; cultural adaptation [21–23]; and location of intervention delivery [24,25].
Systematic review
We conducted a systematic review to identify all RCTs of group psychosocial interventions for depression, anxiety, and PTSD in children and adolescents in LMICs in non-clinical settings. Although a pure realist approach would usually involve less stringent restrictions on included study types, we chose to include only RCTs in order to maximise rigour in the quantitative analyses and subsequent conclusions.
Search strategy
Our inclusion criteria were:
- Population: children and adolescents with a mean age between 0 and 19 years [1].
- Intervention: group psychosocial interventions (‘interpersonal or informational activities, techniques, or strategies that target biological, behavioral, cognitive, emotional, interpersonal, social or environmental factors with the aim of reducing symptoms…and improving functioning or well-being’) [26].
- Comparator: wait-list control, treatment as usual (TAU), or enhanced treatment as usual (ETAU).
- Outcomes: depression, anxiety, and/or PTSD measured using validated rating scales or locally developed and commonly used alternatives. We did not restrict this to formal diagnostic criteria, but included trials of participants with clinically elevated symptom scores, consistent with other recent meta-analyses [9,27].
- Design: randomised controlled trials (RCTs) conducted in non-clinical settings in LMICs [28].
We did not apply language or date restrictions. We included treatment interventions only and excluded trials of mental health promotion or prevention interventions.
TJ searched MEDLINE, EMBASE, PsycInfo, Global Health (Ovid), Social Policy and Practice (Ovid), and the Cochrane Central Register of Controlled Trials from inception date up to 21 November 2022 (updated 26 April 2024). We combined search terms in title and abstract fields related to children and adolescents, psychosocial interventions, and mental health conditions, and used the Cochrane LMIC filter [29]. Full search strategies are provided in S1 Appendix.
After removal of duplicates, TJ and MA independently screened study titles and abstracts and subsequently full texts, resolving disagreements through discussion. We sought not-yet published results from eligible trials by searching for trial protocols and then contacting authors.
Data extraction
Using Covidence software TJ extracted information on: study characteristics (country, setting); study design (individual or cluster RCT, number of arms); participants (study population, number per arm, age, gender); intervention (type, theory, location, facilitators, training, exposure, duration, attendance); control condition; outcomes (type, rating scales, mean and standard deviation at baseline/endpoint/follow-up or mean change, rates of follow-up); and theory (insights relating to mechanisms and contextual factors identified by study authors). Data were checked by MA. We contacted study authors to obtain missing data.
Risk of bias assessment
TJ and MA used the Cochrane Risk of Bias 2 Tool for randomised trials to independently assess risk of bias [30], with disagreements resolved through discussion. No studies were excluded based on risk of bias, but the assessment did inform the synthesis.
Synthesis of theory
We narratively synthesised theory by building on the themes from the scoping review. We summarised theories of change from included studies, formulating these into CMOCs. We extracted qualitative data including from participant and facilitator interviews and focus groups to identify key concepts, which we used to expand and develop the CMOCs. Where available, we narratively synthesised within-study moderation and mediation analyses and used these to test CMOCs. To raise the level of abstraction, we compared and contrasted CMOCs, noting commonalities shared across similar CMOCs to identify salient features, while highlighting essential differences unique to specific interventions and contexts. We iteratively synthesised these into broader categories of programme theories, organised under the themes identified in the scoping search. We used a reciprocal synthesis method, similar to constant comparison, to examine concepts across studies, CMOCs, and programme theories, and identify where one set of concepts could incorporate another. We used refutational synthesis to explore inconsistencies and contradictions [31].
Statistical methods
We conducted a meta-analysis of post-intervention standardised mean differences, adjusted for sample size (Hedges’ g) to synthesise results by intervention type. We used R version 4.3.2 and the metafor, meta, metasens, and dmetar packages. Given the wide range of settings and treatments, we expected some degree of heterogeneity and therefore used a random effects model. As the summary measure was a standardised mean difference, we used a normal-normal model fit using Restricted Maximum Likelihood (REML). Heterogeneity was quantified with the tau-squared and I-squared statistics. Prediction intervals estimated the expected range of effect sizes (ES) for future studies. For the two studies with two treatment arms we combined the trial of TRT + parenting with trials of TRT alone; [32] and for the other trial we used a multivariate model allowing within-study clustering (assuming correlation to be 0.5).[33] We grouped treatments under five categories: CBT (including CBT, TF-CBT, TRT); IPT; WfR and m-WET; eclectic (EASE, YRI, Shamiri, CBI, BAND, REBT); coping-based interventions (guided self-help, mindfulness, mind-body skills, bibliotherapy, didactic therapy). We conducted several sensitivity analyses; we excluded outliers (trials reporting outcomes with 95% confidence interval (CI) that did not overlap with 95% CI of the pooled effect) and used a Baujat plot to identify studies with high influence on pooled ES according to their contribution to heterogeneity. To investigate aspects of programme theories we tested 6 select study level covariates which could potentially explain the differences in ES across studies: type of intervention (CBT-based, IPT-based, etc.); age of participants (early, middle, and late adolescence); proportion of females in the study; whether study took place in a conflict zone; intervention setting (school, community centre, NGO, or refugee camp); and risk of bias. We used Funnel plots and Egger’s regression test as potential indicators of publication or small study bias. Where feasible we looked at these within type of intervention since heterogeneity in addition to publication bias can lead to funnel plot asymmetries.[34] We used limit meta-analysis to adjust for potential publication bias.[35]
Results
Study characteristics
Fig 1 presents the PRISMA diagram. After full-text review, we included 39 reports related to 38 RCTs across 24 LMICs, with data for 6,086 participants. Fifty-two percent of participants were female, with a mean age of 13. Two RCTs targeted children under 10; 27 targeted adolescents aged 10–19; and nine spanned both age ranges. Eight RCTs were feasibility or pilot trials. Table 2 presents study characteristics. The list of excluded studies is available in S1 Text.
Eight RCTs tested group cognitive behavioural therapy (CBT) including its adaptations ‘Coping Cat’ (n = 2) and Rational Emotive Behaviour Therapy (REBT; n = 1); eleven tested trauma-focused CBT (TF-CBT) including its adaptation ‘Teaching Recovery Techniques’ (TRT; n = 5). One tested interpersonal therapy (IPT). Six combined CBT and IPT frameworks: ‘Early Adolescent Skills for Emotions (EASE; n=3); ‘Youth Readiness Intervention’ (YRI; n = 1); ‘Belonging against Negative Thinking and Depression’ (BAND; n = 1); and Shamiri (n = 1). Five tested the eclectic ‘Classroom Based Intervention’ (CBI). Three tested narrative/written exposure therapy (NET/WET; WfR/m-WET); two mindfulness-based interventions; one didactic therapy; one guided self-help; and one bibliotherapy. Comparators were waitlist (n = 25); TAU (n = 7); and ETAU (n = 7); one RCT had both ETAU and TAU arms [36].
28 RCTs were in schools; four in community centres; five in offices of non-governmental organisations; and one in a refugee camp. Intervention duration ranged from three days to 20 weeks (median seven weeks). Most involved weekly sessions. Intervention facilitators included lay facilitators (n = 15); teachers (n = 3); school counsellors (n = 2); psychologists (n = 18); and psychiatrists (n = 1). Follow-up periods ranged from one week to twelve months.
The most frequent outcome was depression (n = 28); then PTSD (n = 24); and anxiety (n = 13). There was a large amount of diagnostic overlap and comorbidity, with most RCTs addressing two (n = 13) or all three (n = 7) of these outcomes.
Synthesis of theory: Programme theories
We expanded the initial four themes from the scoping review into five (age, gender, social protective factors and stability, cultural adaptation, and location of intervention delivery), under which we organised programme theories built from CMOCs, summarised in Table 3. We separated the combined theme of age and gender from the scoping review into two separate themes, because these contained very different programme theories.
Theme 1 - Age
Interventions worked best when content was matched to participants’ level of cognitive development (programme theory 1.1), and when they consisted of age-appropriate activities (programme theory 1.2). We reformulated theories of change suggested by authors of included reports as six CMOCs. After analysing evidence from focus groups and interviews, we added an emphasis on participant motivation [33] and adaptation of language [56]. Using a reciprocal synthesis method across CMOCs, we identified from participant, facilitator, and expert feedback that the cognitive developmental level of participants was central to determining the appropriate level of complexity that they could access. This defined programme theory 1.1, which was supported by moderation analyses from two RCTs of an active, more behavioural CBI intervention which found that younger children benefitted more [66,67]. We separated out motivation and engagement, related to fun and interesting age-appropriate activities and less to cognitive complexity, to form programme theory 1.2 [40].
Theme 2 - Gender
Gender-specific issues were important contextual factors triggering particular intervention mechanisms (programme theory 2.1). We formulated three CMOCs. Qualitative data from facilitator interviews highlighted the role of aggression and a need for more validation among boys in Brazilian favelas [40]. To refine CMOCs, we distinguished between tendencies to externalising expressions of distress among boys compared to girls being more comfortable talking about emotional problems in groups. We synthesised the CMOCs into programme theory 2.1, recognising that gender-specific issues vary across contexts and therefore cannot be generalised, but in all cases require careful matching of intervention content. Moderation analyses identified gender as significant in five RCTs [17,44].
Theme 3 - Social protective factors and stability
We identified important roles for environmental stability (programme theory 3.1), chronic stressors and psychological reserves (programme theory 3.2), and levels of social protective factors (programme theory 3.3). We formulated nine CMOCs. Longer-term follow-up data emphasised the importance of participants consolidating and generalising learned skills, which is easier in stable settings compared to situations with ongoing conflict or instability in the home environment; programme theory 3.1 [63]. Process evaluation data highlighted the role of ongoing stressors, such as trauma exposure, in limiting capacity to experience Reductions in distress [37]; programme theory 3.2, supported by significant results in a moderation analysis for a different RCT [66]. The role of parents and parenting practices was a key interacting mechanism in two mediation analyses [37,45], forming the basis of programme theory 3.3. This was refuted by another CMOC supported by moderation analyses of an RCT in Uganda, which hypothesised that formerly abducted boys who experienced more stigma and social isolation found more meaningful opportunities and made proactive efforts to discuss their experiences and reintegrate into social life, leading to greater improvements in depression [42]. Through refutational synthesis, we further qualified programme theory 3.3, suggesting that extreme levels of instability might create a particularly urgent impetus.
Theme 4 - Cultural adaptation
Programme theories centred on matching local conceptualisations of distress (programme theory 4.1), reducing stigma (programme theory 4.2), and engaging local stakeholders (programme theory 4.3). We formulated five CMOCs. Qualitative data from focus groups and expert panels in the included reports, [41] consistent with existing literature, led us to differentiate ‘superficial’ from ‘deep’ cultural adaptation (going beyond simply translating language to considering metaphors, content, concepts, goals, and methods, and their applicability to a unique cultural context) [70]. We compared the effectiveness of interventions with deep adaptation to the lack of effect of an intervention that failed to match local, predominantly somatic, conceptualisations of distress; programme theory 4.1 [23]. We developed programme theories 4.2 and 4.3 by considering the applicability of the remaining CMOCs on a more abstract level. There were no relevant moderation or mediation analyses.
Theme 5 - Location of intervention delivery
As the majority of RCTs were conducted in schools, our programme theories predominantly relate to mechanisms particularly pertinent to this setting. Nonetheless, other non-clinical and established community settings can also be conducive for the activation of mechanisms. Key mechanisms included normalising difficulties and reducing stigma (programme theory 5.1), building on existing networks of group and peer support (programme theory 5.2), identification and recruitment of participants (programme theory 5.3), empowerment of stakeholders (programme theory 5.4), and dissemination of learning (programme theory 5.5). We formulated eight CMOCs building on participant and facilitator feedback from focus groups. We expected that public and private schools might have different levels of scope for activating protective factors; however moderation analyses of school type in two RCTs failed to find significant results [38,39]. Consequently, when synthesising CMOCs into programme theories we focused on broader networks of support rather than individual school-based protective factors (programme theory 5.2).
One RCT suggested a theme related to lay-facilitator engagement, noting key factors promoting facilitator engagement: working with a reputable organisation, co-working with another facilitator to create a collegial atmosphere, and being given an opportunity to develop skills. These were hypothesised to promote fidelity to the intervention model, leading to improved outcomes [38]. However, data were insufficient to develop this theme.
Risk of bias
We rated nineteen RCTs ‘low risk’; 15 ‘some concerns’; and 5 ‘high risk’ of bias (S1 Table). The most common source of risk of bias was missing outcome data. Risk of bias had no significant effect on results in the meta-analysis.
Effect of psychosocial interventions
Depression symptoms: the pooled effect size was -0.72 (95% CI -1.01 to -0.42, p < 0.001), with high heterogeneity (I-squared = 0.89, 95% CI 0.85 to 0.92); see Forest plot (Fig 2). In univariate regressions, only age and exposure to conflict showed significant associations with ES. ES were larger for older participants versus younger (difference = 0.6, 95% CI -1.14 to -0.057, p = 0.032) and in non-conflict-affected populations versus conflict-affected populations (difference = 0.6, 95% CI 0.05 to 1.15, p = 0.033). The Funnel plot showed clear asymmetry (S1 File), and Egger’s test was significant at p < 0.001 confirming small study effects. Following adjustment using limit meta-analysis the pooled effect was no longer significant (-0.31, 95% CI -0.71 to 0.10, p = 0.137).
Anxiety symptoms: the pooled effect size was -0.90 (95% CI -1.57 to -0.23, p = 0.014), with high heterogeneity (I-squared = 0.93, 95% CI 0.89 to 0.95); see Forest plot (Fig 3). In univariate regressions no study characteristics were associated with ES. The Funnel plot showed some asymmetry (S1 File), and Egger’s test was significant at p = 0.022 confirming small study effects. Using limit meta-analysis to adjust for small study effects reduced the overall ES to -0.3.
PTSD symptoms: the pooled effect size was -0.71 (95% CI -1.05 to -0.38, p < 0.001), with high heterogeneity (I-squared = 0.91, 95% CI 0.88 to 0.93); see Forest plot (Fig 4). In univariate regressions only intervention type was significantly associated with effect size (p = 0.006). NET/WET-based interventions had the greatest effect sizes, with eclectic the smallest. The Funnel plot showed clear asymmetry (S1 File), and Egger’s test was significant at p = 0.002 confirming small study effects. Adjusting ES for small study effects using limit meta-analysis meant the overall pooled effect was no longer significant (0.11, 95% CI -0.58 to 0.36, p = 0.653).
In sensitivity analyses to remove outliers, heterogeneity (I-squared) and effect sizes decreased but remained significant (p < 0.05) across depression, anxiety, and PTSD outcomes (S2 File).
Among our programme theories, only themes one (age) and three (social protective factors and stability) yielded significant results in univariate regressions. For depression symptoms, ES were larger for older children and in populations without exposure to conflict. Full meta-regression results are available in S3 File.
Discussion
This is, to our knowledge, the first realist systematic review of group psychosocial interventions for depression, anxiety, and PTSD among children and adolescents in LMICs. Previous reviews have found heterogeneous results; it is clear that psychosocial interventions can be highly effective, but can also have very little impact [20,71,72]. Insufficient attention has been paid to the interaction of contextual factors and mechanisms that explain some aspects of this heterogeneity [73]; this review has sought to make initial suggestions and inform further research.
Overall results revealed a large effect of psychosocial interventions, with substantial heterogeneity. Our meta-analysis found slightly larger pooled effect sizes and similar heterogeneity compared to a recent meta-analysis exploring the same outcomes [74], and slightly smaller pooled effects compared to one with a narrower focus on psychotherapy interventions [9]. Treatment type was associated with the size of effects on PTSD symptoms, with narrative/written-exposure-therapy-based interventions having the greatest effect and CBI the least. The overlap between outcomes reflects both rates of comorbidity and non-specificity of diagnostic criteria, rating scales, and screening instruments. Previous reviews have argued for transdiagnostic interventions having wider applicability and feasibility in addressing a range of common problems rather than disorder-specific criteria [25,75].
We identified programme theories and constituent CMOCs that hypothesise key aspects of the success or failure of interventions, summarised under five themes of interest. Children and adolescents of different ages respond differently to intervention content. Interventions should be designed to match the developmental level of the target population. Children and younger adolescents benefit most from behavioural interventions that are practical and focus on active skills, whereas older adolescents may be better able to understand and apply more cognitive interventions. This is consistent with trajectories of development moving towards a greater capacity for metacognitive engagement towards later adolescence. In certain contexts where the cultural milieu may place less emphasis on introspective analysis of cognitive patterns, however, such a linear trajectory may not apply [76,77]. These differences may help to explain some of the heterogeneity of findings across the included studies. If interventions with a heavy cognitive focus are to be used, one approach could be to combine behavioural and cognitive elements into a single intervention, beginning with behavioural emphases to promote engagement and initial gains, which then allows effective socialisation to the cognitive model [53]. Interventions should involve age-appropriate, enjoyable activities, such that participants will attend, engage, and work to apply learned skills in real-life contexts.
Gender can affect intervention relevance. Different activities for boys and girls might promote better engagement. Different skills might need to be emphasised to target particular behavioural or cognitive patterns that feature in local gender norms, and gender-specific problems might need to be considered [40,44]. Gender may also affect participation in group therapy [78]. Effective interventions might therefore look different for gender-segregated groups, or incorporate gender-specific components in mixed groups [77]. Little is known about how this relates to gender diverse and non-binary gender categorisations, especially in LMICs. None of the included studies in our review incorporated data for gender diverse participants, and more research is needed.
Interventions might work best when children and adolescents have a higher level of social protective factors (including financial security and social status) and environmental stability that allows them to engage with therapy, implement learned skills outside of sessions, and receive support to continue making changes in the post-intervention period [19,23,79]. Special attention may be needed to enable children and adolescents with ongoing stress and trauma to derive benefit. Notably, there was a discrepancy in programme theory 3.3 (children and adolescents with recourse to a higher level of social protective factors and support will be better able to engage with interventions and make use of learned skills); a refuting CMOC suggested that it was those boys that experienced the highest levels of stigma and isolation that benefited most. Possibly this could be related to the unique context of the study in a displaced-persons camp with formerly-abducted child soldiers, or because these boys had higher baseline levels of symptoms and hence greater room for improvement [42]. Perhaps extreme levels of instability and trauma history, alongside stigma associated with abduction, compelled even the most traumatised adolescents to engage and benefit. Meta-regressions showed that for depression, interventions appeared to be less effective for conflict-affected populations. The number of studies was too small to provide definitive conclusions, and more research is needed to refine hypotheses about the relevance of ongoing conflict and trauma exposure. More work is needed to better understand the potential relevance of factors such as uneven economic development, rural-urban differences, adversities such as bereavement, and ongoing trauma exposure, in order to allow for more precise insights.
Culturally adapted interventions should go beyond simple translation of language to include culturally relevant metaphors, content, concepts, goals, and methods [70]. Doing so promotes engagement, reduces stigma, and considers local conceptualisations of distress and appropriate responses to these. Otherwise, content may be less relevant to participants’ problems. However, our CMOCs related to this theme were formulated primarily from qualitative data and we were unable to test programme theories statistically; our findings contrast with a meta-analysis that found locally adapted interventions to be less effective than direct imports [9]. One potential explanation for this inconsistency is that adaptations to interventions are rarely described in detail by authors and may differ greatly in their nature and effectiveness. It might be that deep adaptations are more effective than superficial adaptations but have been less used in past studies. Better reporting is needed to facilitate evaluation of different kinds of adaptation and to identify appropriate methodology to adapt interventions, ensuring adaptations optimise rather than worsen outcomes.
The school setting may provide unique advantages. This is consistent with findings from a systematic review of school-based interventions in LMICs showing overall effectiveness for mental health conditions; and a realist review of universal school-based interventions in humanitarian contexts for non-clinical wellbeing [80,81]. In schools, participants can be easily identified. Difficulties can be normalised, potentially reducing stigma, and teachers and principals can be empowered as stakeholders. Existing peer support networks and supportive group dynamics can be built upon, creating an environment conducive to therapeutic work that is rich in curative factors [82]. Networks of information sharing can allow those that are not actively participating in an intervention to benefit indirectly. However, this depends on school climate; schools can also be hostile environments with children subject to bullying and corporal punishment, as well as worsening of stigma. Implementing lay-facilitator delivered interventions in under-resourced school settings can be challenging however, and issues including facilitator training, supervision, and sustainability warrant further attention.
Limitations of the review relate to the included RCTs and the review methods. Only 19 of the 38 included RCTs were deemed low risk of bias. Although this is a common issue in RCTs testing complex psychosocial interventions, the risk of bias potentially undermines the robustness of pooled effect sizes. Effect sizes were not significantly affected after adjusting for overall risk of bias, but did become non-significant after accounting for small study samples in the limit meta-analysis. There is therefore a need for cautious interpretation of our findings as well as a need for more high-quality RCTs to be conducted. Few RCTs explored differential outcomes for subgroups, possibly because they were underpowered. Consequently, we were unable to quantitatively test all our CMOCs through meta-regressions. Other potentially important mediators, such as facilitator training, session dose, frequency, and intervention duration were not able to be tested and could be further investigated. The meta-regressions done at the study level, without individual participant data, are prone to the ecological fallacy and ignore randomisation, and are thus limited in how severely they can test CMOCs. Individual participant data meta-analyses might be helpful, although data on important contextual factors may not have been collected. Therapy manuals might set out theorised mechanisms, but this may not match what happens in reality, and few RCTs were set up to explicitly test mechanisms. The large amount of comorbidity and diagnostic overlap in the included RCTs made it difficult to separate out what mental health conditions were being treated beyond non-specific psychological distress. Screening tools and rating scales tend to prioritise sensitivity over specificity and therefore trials tend to include a heterogenous mixture of psychopathology. However, the universality of psychiatric diagnoses across cultures and contexts is debatable, and perhaps too zealous an emphasis on strict diagnostic criteria risks a mismatch with local conceptualisations of distress.
Key strengths of this review include a systematic search strategy and the integration of realist analysis with quantitative methods. We took a different approach to that proposed by Bonell et al.; [11] we used an initial scoping search to structure our data extraction and formulation of CMOCs and we did not synthesise process evaluation data, because this was lacking in the included reports.
Our findings call for further research into psychosocial interventions for children and adolescents in LMICs. There is an urgent need to rigorously and systematically test theory by setting out targeted CMOCs explicitly. More attention to mediators, moderators, and subgroup analyses can help to delineate mechanisms and identify important contextual factors, which might inform new theory. Each of the themes identified in this review merits further exploration and could serve as a starting point for realist evaluations. Although we know that schools can be a good location for intervention delivery, variation between schools needs exploration, with attention to the factors, including school climate, that make interventions particularly successful in one school over another. This might help better match interventions, as well as suggest changes that could make the school environment more conducive. Inconsistencies within the theme ‘social protective factors and stability’ also require further exploration. Qualitative research is called for, to better understand the mechanisms and processes through which different psychosocial interventions bring about change, and how these mechanisms might interact on a deeper level in particular intervention contexts. This includes attention to ‘deep’ cultural adaptation and exploring what this really means for a mental health intervention.
Our findings, in conjunction with what is already known, can be of help to policymakers and researchers alike. The paucity of resources in many LMIC settings, as well as the urgent unmet clinical need with its potential for wide ranging and lasting implications, means that any intervention chosen must be a good fit for the context and target population, and we hope that our suggestions can contribute towards this.
Supporting information
S1 Appendix. Systematic review search strategy.
https://doi.org/10.1371/journal.pmen.0000533.s001
(PDF)
S1 Text. List of studies excluded at full-text screening, with reasons.
https://doi.org/10.1371/journal.pmen.0000533.s002
(PDF)
S1 Table. Risk of bias ratings: Cochrane risk of bias 2 tool.
https://doi.org/10.1371/journal.pmen.0000533.s003
(PDF)
S2 Table. Full data extracted from studies used for meta-analysis.
https://doi.org/10.1371/journal.pmen.0000533.s004
(XLSX)
S4 File. PRISMA checklist.
From: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. https://doi.org/10.1136/bmj.n71
https://doi.org/10.1371/journal.pmen.0000533.s008
(PDF)
References
- 1. World Health Organisation. Mental health of adolescents. 2021. Accessed 2024 April 9. https://www.who.int/news-room/fact-sheets/detail/adolescent-mental-health
- 2. Veldman K, Bültmann U, Stewart RE, Ormel J, Verhulst FC, Reijneveld SA. Mental health problems and educational attainment in adolescence: 9-year follow-up of the TRAILS study. PLoS One. 2014;9(7):e101751. pmid:25047692
- 3. Blakemore SJ. Adolescence and mental health. The Lancet. 2019;393(10185):2030–1.
- 4. Global Burden of Disease Collaborative Network. Global burden of disease study 2019 (GBD 2019) Disease and Injury Burden 1990-2019. 2020. http://ghdx.healthdata.org/record/ihme-data/gbd-2019-disease-and-injury-burden-1990-2019
- 5. Yatham S, Sivathasan S, Yoon R, da Silva TL, Ravindran AV. Depression, anxiety, and post-traumatic stress disorder among youth in low and middle income countries: A review of prevalence and treatment interventions. Asian J Psychiatry. 2018;38:78–91.
- 6. Dua T, Barbui C, Clark N, Fleischmann A, Poznyak V, van Ommeren M. Evidence-Based Guidelines for Mental, Neurological, and Substance Use Disorders in Low- and Middle-Income Countries: Summary of WHO Recommendations. PLOS Med. 2011;8(11):e1001122.
- 7. Marmarosh CL. Ruptures and repairs in group psychotherapy: from theory to practice. Int J Group Psychother. 2021;71(2):205–23.
- 8. Barbui C, Purgato M, Abdulmalik J, Acarturk C, Eaton J, Gastaldon C, et al. Efficacy of psychosocial interventions for mental health outcomes in low-income and middle-income countries: an umbrella review. Lancet Psychiatry. 2020;7(2):162–72. pmid:31948935
- 9. Venturo-Conerly KE, Eisenman D, Wasil AR, Singla DR, Weisz JR. Meta-analysis: The Effectiveness of Youth Psychotherapy Interventions in Low- and Middle-Income Countries. J Am Acad Child Adolesc Psychiatry. 2023;62(8):859–73.
- 10. Weisz JR, Venturo-Conerly KE, Fitzpatrick OM, Frederick JA, Ng MY. What four decades of meta-analysis have taught us about youth psychotherapy and the science of research synthesis. Annu Rev Clin Psychol. 2023;19:79–105.
- 11. Bonell C, Taylor B, Berry V, Filho SRP, Rizzo A, Farmer C, et al. Re-orientating systematic reviews to rigorously examine what works, for whom and how: Example of a realist systematic review of school-based prevention of dating and gender violence. Res Synth Methods. 2023;14(4):582–95. pmid:37287195
- 12. Hunter R, Gorely T, Beattie M, Harris K. Realist review. International Review of Sport and Exercise Psychology. 2022;15(1):242–65.
- 13. Pawson R, Greenhalgh T, Harvey G, Walshe K. Realist review--a new method of systematic review designed for complex policy interventions. J Health Serv Res Policy. 2005;10 Suppl 1:21–34. pmid:16053581
- 14. Wong G. Getting to grips with context and complexity - the case for realist approaches. Gac Sanit. 2018;32(2):109–10. pmid:28735621
- 15. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.
- 16. Wong G, Greenhalgh T, Westhorp G, Buckingham J, Pawson R. RAMESES publication standards: realist syntheses. BMC Med. 2013;11:21. pmid:23360677
- 17. Jordans MJD, Komproe IH, Tol WA, Kohrt BA, Luitel NP, Macy RD, et al. Evaluation of a classroom-based psychosocial intervention in conflict-affected Nepal: a cluster randomized controlled trial. J Child Psychol Psychiatry. 2010;51(7):818–26. pmid:20102428
- 18. Wergeland GJH, Riise EN, Öst LG. Cognitive behavior therapy for internalizing disorders in children and adolescents in routine clinical care: A systematic review and meta-analysis. Clin Psychol Rev. 2021;83:101918.
- 19. Purgato M, Gastaldon C, Papola D, van Ommeren M, Barbui C, Tol WA. Psychological therapies for the treatment of mental disorders in low- and middle-income countries affected by humanitarian crises. Cochrane Database Syst Rev. 2018;7(7):CD011849. pmid:29975811
- 20. Purgato M, Gross AL, Betancourt T, Bolton P, Bonetto C, Gastaldon C, et al. Focused psychosocial interventions for children in low-resource humanitarian settings: a systematic review and individual participant data meta-analysis. Lancet Glob Health. 2018;6(4):e390-400.
- 21. Jordans MJD, Pigott H, Tol WA. Interventions for Children Affected by Armed Conflict: a Systematic Review of Mental Health and Psychosocial Support in Low- and Middle-Income Countries. Curr Psychiatry Rep. 2016;18(1):9. pmid:26769198
- 22. McMullen J, O’Callaghan P, Shannon C, Black A, Eakin J. Group trauma-focused cognitive-behavioural therapy with former child soldiers and other war-affected boys in the DR Congo: a randomised controlled trial. J Child Psychol Psychiatry. 2013;54(11):1231–41. pmid:23738530
- 23. Tol WA, Komproe IH, Susanty D, Jordans MJD, Macy RD, De Jong JTVM. School-based mental health intervention for children affected by political violence in Indonesia: a cluster randomized trial. JAMA. 2008;300(6):655–62. pmid:18698064
- 24. Berger R, Gelkopf M. School-based intervention for the treatment of tsunami-related distress in children: a quasi-randomized controlled trial. Psychother Psychosom. 2009;78(6):364–71. pmid:19738402
- 25. Fazel M, Patel V, Thomas S, Tol W. Mental health interventions in schools in low-income and middle-income countries. Lancet Psychiatry. 2014;1(5):388–98. pmid:26361001
- 26.
England MJ, Butler AS, Gonzalez ML. Psychosocial interventions for mental and substance use disorders: a framework for establishing evidence-based standards. Washington (DC): National Academies Press (US). 2015. http://www.ncbi.nlm.nih.gov/books/NBK305126/
- 27. Weisz JR, Kuppens S, Ng MY, Eckshtain D, Ugueto AM, Vaughn-Coaxum R, et al. What five decades of research tells us about the effects of youth psychological therapy: A multilevel meta-analysis and implications for science and practice. Am Psychol. 2017;72(2):79–117. pmid:28221063
- 28. World Bank. Low and middle income. Data. 2022. Accessed 2024 April 9. https://data.worldbank.org/income-level/low-and-middle-income
- 29. Cochrane. LMIC Filters | Cochrane Effective Practice and Organisation of Care. 2021. Accessed 2024 April 9. https://epoc.cochrane.org/lmic-filters
- 30. Sterne JAC, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366:l4898.
- 31. Sattar R, Lawton R, Panagioti M, Johnson J. Meta-ethnography in healthcare research: a guide to using a meta-ethnographic approach for literature synthesis. BMC Health Serv Res. 2021;21(1):50. pmid:33419430
- 32. El-Khani A, Cartwright K, Maalouf W, Haar K, Zehra N, Çokamay-Yılmaz G, et al. Enhancing Teaching Recovery Techniques (TRT) with parenting skills: RCT of TRT + parenting with trauma-affected syrian refugees in lebanon utilising remote training with implications for insecure contexts and COVID-19. Int J Environ Res Public Health. 2021;18(16):8652. pmid:34444403
- 33. Ahmadi SJ, Musavi Z, Samim N, Sadeqi M, Jobson L. Investigating the feasibility, acceptability and efficacy of using modified-written exposure therapy in the aftermath of a terrorist attack on symptoms of posttraumatic stress disorder among afghan adolescent girls. Front Psychiatry. 2022;13:826633. pmid:35463492
- 34. Ioannidis JPA, Trikalinos TA. The appropriateness of asymmetry tests for publication bias in meta-analyses: a large survey. CMAJ. 2007;176(8):1091–6. pmid:17420491
- 35. Rücker G, Schwarzer G, Carpenter JR, Binder H, Schumacher M. Treatment-effect estimates adjusted for small-study effects via a limit meta-analysis. Biostatistics. 2011;12(1):122–42. pmid:20656692
- 36. Chen Y, Shen WW, Gao K, Lam CS, Chang WC, Deng H. Effectiveness RCT of a CBT intervention for youths who lost parents in the Sichuan, China, earthquake. Psychiatr Serv. 2014;65(2):259–62. pmid:24492904
- 37. Akhtar A, Malik A, Ghatasheh M, Aqel IS, Habashneh R, Dawson KS, et al. Feasibility trial of a brief scalable psychological intervention for Syrian refugee adolescents in Jordan. Eur J Psychotraumatol. 2021;12(1):1901408. pmid:34868475
- 38. Barron IG, Abdallah G, Smith P. Randomized control trial of a CBT trauma recovery program in palestinian schools. J Loss Trauma. 2013;18(4):306–21.
- 39. Barron I, Abdallah G, Heltne U. Randomized control trial of teaching recovery techniques in rural occupied palestine: effect on adolescent dissociation. J Aggress Maltreat Trauma. 2016;25(9):955–73.
- 40. Barron I, Freitas F, Bosch CA. Pilot randomized control trial: efficacy of a group-based psychosocial program for youth with PTSD in the Brazilian favelas. J Child Adolesc Trauma. 2020;14(3):335–45. pmid:34471452
- 41. Bella-Awusah T, Ani C, Ajuwon A, Omigbodun O. Effectiveness of brief school-based, group cognitive behavioural therapy for depressed adolescents in south west Nigeria. Child Adolesc Ment Health. 2016;21(1):44–50. pmid:32680363
- 42. Betancourt TS, Newnham EA, Brennan RT, Verdeli H, Borisova I, Neugebauer R, et al. Moderators of treatment effectiveness for war-affected youth with depression in northern Uganda. J Adolesc Health. 2012;51(6):544–50. pmid:23174463
- 43. Betancourt TS, McBain R, Newnham EA, Akinsulure-Smith AM, Brennan RT, Weisz JR, et al. A behavioral intervention for war-affected youth in Sierra Leone: a randomized controlled trial. J Am Acad Child Adolesc Psych. 2014;53(12):1288–97. pmid:25457927
- 44. Bolton P, Bass J, Betancourt T, Speelman L, Onyango G, Clougherty KF, et al. Interventions for depression symptoms among adolescent survivors of war and displacement in northern Uganda: a randomized controlled trial. JAMA. 2007;298(5):519–27. pmid:17666672
- 45. Bryant RA, Malik A, Aqel IS, Ghatasheh M, Habashneh R, Dawson KS, et al. Effectiveness of a brief group behavioural intervention on psychological distress in young adolescent Syrian refugees: a randomised controlled trial. PLoS Med. 2022;19(8):e1004046. pmid:35960704
- 46. Daryabeigi M, Pourmohamdreza-Tajrishi M, Dolatshahi B, Bakhshi E. The effectiveness of cognitive-behavioral therapy on internalizing problems of children with externalizing disorders. Iran Rehabil J. 2020;18(2):193–202.
- 47. Dorsey S, Triplett N, Akiba C, Lucid L, Wasonga A, Whetten K. Lay-counselor delivered trauma-focused CBT (TF-CBT) for bereaved African children: outcomes and acceptability. J Am Acad Child Adolesc Psych. 2021;60(10 Supplement):S313.
- 48. Ede MO, Igbo JN, Eseadi C, Ede KR, Ezegbe BN, Ede AO, et al. Effect of group cognitive behavioural therapy on depressive symptoms in a sample of college adolescents in Nigeria. J Rat-Emo Cognitive-Behav Ther. 2019;38(3):306–18.
- 49. Getanda EM, Vostanis P. Feasibility evaluation of psychosocial intervention for internally displaced youth in Kenya. J Ment Health. 2020.
- 50. Gordon JS, Staples JK, Blyta A, Bytyqi M, Wilson AT. Treatment of posttraumatic stress disorder in postwar Kosovar adolescents using mind-body skills groups: a randomized controlled trial. J Clin Psychiatry. 2008;69(9):1469–76. pmid:18945398
- 51. Jacob J, De Guzman RG. Effectiveness of taking in the good based-bibliotherapy intervention program among depressed Filipino female adolescents. Asian J Psychiatr. 2016;23:99–107. pmid:27969089
- 52. Jordans MJD, Brown FL, Kane J, Taha K, Steen F, Ali R, et al. Evaluation of the Early Adolescent Skills for Emotions (EASE) intervention in Lebanon: a randomized controlled trial. Compr Psychiatry. 2023;127:152424. pmid:37748283
- 53.
Kaesornsamut P, Sitthimongkol Y, Williams R, Sangon S, Rohitsuk W, Vorapongsathron T. Effectiveness of the BAND intervention program on thai adolescents’ sense of belonging, negative thinking and depressive symptoms. 2012;16(1):29–47.
- 54. Kalantari M, Yule W, Dyregrov A, Neshatdoost H, Ahmadi SJ. Efficacy of writing for recovery on traumatic grief symptoms of Afghani refugee bereaved adolescents: a randomized control trial. Omega (Westport). 2012;65(2):139–50. pmid:22953510
- 55. Khalid A, Haqqani S, Williams C. Guided self-help Urdu version of the living life to the full intervention for secondary school adolescents with low mood and anxiety in Pakistan: a feasibility study. Heliyon. 2022;8(7):e09809. pmid:35859639
- 56. Khan A, Malik TA, Ahmed S, Riaz A. Translation, adaptation and implementation of coping cat program with Pakistani children. Child Youth Care Forum. 2019;49(1):23–41.
- 57. Layne CM, Saltzman WR, Poppleton L, Burlingame GM, Pasalić A, Duraković E, et al. Effectiveness of a school-based group psychotherapy program for war-exposed adolescents: a randomized controlled trial. J Am Acad Child Adolesc Psychiatry. 2008;47(9):1048–62. pmid:18664995
- 58. Li J, Yuan L, Zhou Y, Zhang W, Qu Z. The acceptability, feasibility, and preliminary effectiveness of group cognitive behavioral therapy for Chinese children with posttraumatic stress disorder: a pilot randomized controlled trial. Psychol Trauma Theory Res Pract Policy. 2022.
- 59. Li J, Li J, Zhang W, Wang G, Qu Z. Effectiveness of a school-based, lay counselor-delivered cognitive behavioral therapy for Chinese children with posttraumatic stress symptoms: a randomized controlled trial. Lancet Reg Health West Pac. 2023;33:100699. pmid:36785644
- 60. O’Callaghan P, McMullen J, Shannon C, Rafferty H, Black A. A randomized controlled trial of trauma-focused cognitive behavioral therapy for sexually exploited, war-affected Congolese girls. J Am Acad Child Adolesc Psychiatry. 2013;52(4):359–69. pmid:23582867
- 61. Osborn TL, Venturo-Conerly KE, Arango G S, Roe E, Rodriguez M, Alemu RG, et al. Effect of shamiri layperson-provided intervention vs study skills control intervention for depression and anxiety symptoms in adolescents in Kenya: a randomized clinical trial. JAMA Psychiatry. 2021;78(8):829–37. pmid:34106239
- 62. Peter A, Srivastava R, Agarwal A, Singh AP. The effect of mindfulness-based cognitive therapy on anxiety and resilience of the school going early adolescents with anxiety. J Indian Assoc Child Adoles Mental Health. 2022;18(2):176–85.
- 63. Pityaratstian N, Piyasil V, Ketumarn P, Sitdhiraksa N, Ularntinon S, Pariwatcharakul P. Randomized controlled trial of group cognitive behavioural therapy for post-traumatic stress disorder in children and adolescents exposed to tsunami in Thailand. Behav Cogn Psychother. 2015;43(5):549–61. pmid:24709121
- 64. Saw JA, Tam CL, Bonn G. Development and validation of a school-based cognitive-behavioural therapy (CBT) intervention for Malaysian high school students with depressive symptoms. Asia Pacific J Counsel Psychother. 2019;10(2):171–87.
- 65. Saw JA, Tam CL, Thanzami V, Bonn G. Contextualized school-based cognitive behavioral therapy (CBT) intervention for Malaysian secondary school students. Front Psychiatry. 2020;11:565896. pmid:33408652
- 66. Tol WA, Komproe IH, Jordans MJD, Vallipuram A, Sipsma H, Sivayokan S, et al. Outcomes and moderators of a preventive school-based mental health intervention for children affected by war in Sri Lanka: a cluster randomized trial. World Psychiatry. 2012;11(2):114–22. pmid:22654944
- 67. Tol WA, Komproe IH, Jordans MJD, Ndayisaba A, Ntamutumba P, Sipsma H, et al. School-based mental health intervention for children in war-affected Burundi: a cluster randomized trial. BMC Med. 2014;12:56. pmid:24690470
- 68. Ugwu GC, Ugwuanyi CS, Okeke CIO, Uzodinma UE, Aneke AO. Efficacy of rational emotive behavior therapy on depression among children with learning disabilities: implications for evaluation in science teaching. J Rat-Emo Cognitive-Behav Ther. 2021;40(2):313–33.
- 69. Zafar H, Khalily MT. Didactic therapy for management of stress and co-morbid symptoms of depression and anxiety in Pakistani adolescents. Pak J Psychol Res. 2015;30(1):131–49.
- 70. Healy EA, Kaiser BN, Puffer ES. Family-based youth mental health interventions delivered by nonspecialist providers in low- and middle-income countries: a systematic review. Fam Syst Health. 2018;36(2):182–97. pmid:29902035
- 71. Das JK, Salam RA, Lassi ZS, Khan MN, Mahmood W, Patel V. Interventions for adolescent mental health: an overview of systematic reviews. J Adolesc Health. 2016;59(4):S49-60.
- 72. Ribeiro WS, Grande AJ, Hoffmann MS, Ziebold C, McDaid D, Fry A, et al. A systematic review of evidence-based interventions for child and adolescent mental health problems in low- and middle-income countries. Compr Psychiatry. 2023;121:152358. pmid:36508775
- 73. Singla DR, Waqas A, Hamdani SU, Suleman N, Zafar SW, Zill-E-Huma, et al. Implementation and effectiveness of adolescent life skills programs in low- and middle-income countries: a critical review and meta-analysis. Behav Res Ther. 2020;130:103402. pmid:31146889
- 74. Uppendahl JR, Alozkan-Sever C, Cuijpers P, de Vries R, Sijbrandij M. Psychological and psychosocial interventions for PTSD, depression and anxiety among children and adolescents in low- and middle-income countries: a meta-analysis. Front Psychiatry. 2020.
- 75. Bockting CLH, Williams AD, Carswell K, Grech AE. The potential of low-intensity and online interventions for depression in low- and middle-income countries. Glob Ment Health (Camb). 2016;3:e25. pmid:28596893
- 76. Patel V, Chowdhary N, Rahman A, Verdeli H. Improving access to psychological treatments: lessons from developing countries. Behav Res Ther. 2011;49(9):523–8.
- 77. Pacini A, Shrestha P. Editorial perspective: adapting western psychological interventions for children and adolescents in LMICs: lessons from Nepal. J Child Psychol Psychiatry. 2024;65(1):116–9. pmid:37606149
- 78. Campbell OLK, Bann D, Patalay P. The gender gap in adolescent mental health: A cross-national investigation of 566,829 adolescents across 73 countries. SSM - Popul Health. 2021;13:100742.
- 79. Thabet AA, Vostanis P, Karim K. Group crisis intervention for children during ongoing war conflict. Eur Child Adolesc Psychiatry. 2005;14(5):262–9. pmid:15981138
- 80. Grande AJ, Hoffmann MS, Evans-Lacko S, Ziebold C, de Miranda CT, Mcdaid D. Efficacy of school-based interventions for mental health problems in children and adolescents in low and middle-income countries: a systematic review and meta-analysis. Front Psychiatry. 2023.
- 81. Lasater ME, Flemming J, Bourey C, Nemiro A, Meyer SR. School-based MHPSS interventions in humanitarian contexts: a realist review. BMJ Open. 2022;12(4):e054856. pmid:35450900
- 82.
Yalom ID, Leszcz M. The theory and practice of group psychotherapy. Basic Books; 2008.