Comparative efficacy and acceptability of treatment options for depressive disorders in children and adolescents: protocol of a network meta-analysis Authors: Jing Guo, Zhe Lu, Yuyanan Zhang, Yaoyao Sun, Zhewei Kang, Junyuan Sun, Guorui Zhao, Xiaoyang Feng, Rui Yuan, Yunqing Zhu, Tong Yu, Yang Yang, Huan Chen, Weihua Yue Published: 2026 ABSTRACT Abstract AbstractIntroductionThe incidence of depression among children and adolescents has been increasing in recent years, posing significant challenges to public health and clinical care. A variety of treatments, including pharmacotherapy, psychotherapy and physical interventions, are widely used in clinical practice. However, a comprehensive synthesis of the evidence on the efficacy and acceptability of all these treatment modalities is currently lacking. This study aims to use network meta-analysis (NMA) to compare the efficacy and acceptability of all available treatments for depression in children and adolescents, offering valuable insights to inform clinical decision-making and guide future research in this critical area.Methods and analysisWe will include randomised controlled trials evaluating active interventions for depressive disorders in children and adolescents. Seven electronic databases (PubMed, Embase, the Cochrane Library, Web of Science, PsycINFO, Scopus and ClinicalTrials.gov) were searched from inception to 2 July 2024 and updated on 2 November 2025. Two of four investigators will independently screen studies, extract data from eligible articles and assess the risk of bias using the Cochrane Risk of Bias 2.0 tool. The primary outcome will be the change in depressive symptoms. Secondary outcomes will include acceptability (all-cause discontinuation), response rate, remission rate and overall functioning. Pairwise and Bayesian NMA will be conducted. Small-study effects and publication bias will be assessed. The certainty of the evidence will be evaluated according to the Confidence in Network Meta-Analysis approach.Ethics and disseminationAs this review involves secondary analysis of previously published studies, ethical approval is not required. The findings will be disseminated through publication in peer-reviewed journals.PROSPERO registration numberPROSPERO-ID CRD42024557384.Keywords: Depression & mood disorders, Child & adolescent psychiatry, Network Meta-Analysis, Treatment Outcome STRENGTHS AND LIMITATIONS OF THIS STUDY A Bayesian network meta-analysis will combine direct and indirect evidence to compare multiple interventions within a single analytical framework. A Bayesian network meta-analysis will combine direct and indirect evidence to compare multiple interventions within a single analytical framework. A hierarchical analytical strategy across macro-level, class-level and specific-intervention-level analyses improves interpretability across different levels of intervention granularity. A hierarchical analytical strategy across macro-level, class-level and specific-intervention-level analyses improves interpretability across different levels of intervention granularity. Searches across seven databases, updated on 2 November 2025, will maximise the completeness of eligible randomised evidence. Searches across seven databases, updated on 2 November 2025, will maximise the completeness of eligible randomised evidence. A prospectively defined protocol and adherence to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standards enhance transparency and reproducibility. A prospectively defined protocol and adherence to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standards enhance transparency and reproducibility. Clinical and methodological heterogeneity across interventions, populations and study designs may limit comparability. Clinical and methodological heterogeneity across interventions, populations and study designs may limit comparability. Introduction Depression in children and adolescents is a prevalent and debilitating mental health disorder, with significant implications for their development, academic performance, social functioning and overall quality of life.1,3 The global burden of disease attributable to mental health disorders, including depression, is substantial, and early onset depression is often associated with a more severe and chronic course, increasing the risk of recurrence and other psychiatric comorbidities in adulthood.4,7 Given its significant impact on the development, mental health and overall functioning of children and adolescents, establishing the most effective and acceptable treatment options is critical for improving long-term prognosis. Current treatment approaches for depression in children and adolescents include pharmacological interventions, psychotherapies, physical therapies, behavioural interventions and complementary treatments.8,10 In 2015, Zhou et al conducted a network meta-analysis (NMA) comparing the efficacy and acceptability of various psychotherapies.11 In 2016, Cipriani et al focused on antidepressant medications.12 A subsequent analysis in 2020 evaluated psychotherapy, pharmacotherapy and their combination.13 However, existing NMAs have primarily focused on limited categories of interventions and have not captured the full spectrum of treatment strategies currently available or under investigation. As treatment options continue to evolve, a more comprehensive synthesis is needed to reflect current clinical practice. Recognising this limitation, we propose a systematic review and NMA to compare all currently available interventions for paediatric depression, including pharmacotherapy, psychotherapy, physical therapies, nutritional supplements and their combinations. Unlike pairwise meta-analyses, which only assess direct comparisons between two treatments at a time, the NMA integrates both direct and indirect evidence, enabling simultaneous comparison of multiple interventions and providing a relative ranking based on efficacy and acceptability.14,16 This approach offers a more comprehensive understanding of the comparative value of treatment options, with the goal of supporting evidence-based clinical decision making and identifying gaps for future research. Methods and analysis The protocol was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for protocols and registered on 22 June 2024 with PROSPERO (registration number: CRD42024557384) (online supplemental appendix 1). The reporting of the systematic review and NMA will follow the PRISMA statement for NMAs (online supplemental appendix 2).17 The PROSPERO record was updated on 20 April 2026 to reflect the revised title, search update and review timeline. Any further protocol amendments will be updated in PROSPERO and documented in the final publication. Eligibility criteria We will include randomised controlled trials (RCTs) that investigate any treatment for depression in children and adolescents. The eligibility criteria are defined based on the Population, Intervention, Comparator, Outcome, Study Design (PICOS) framework.18 19 Participants Studies will be eligible if they include participants diagnosed with depression or related conditions (eg, major depressive disorder, persistent depressive disorder, depressive symptoms) based on established diagnostic criteria such as the Diagnostic and Statistical Manual of Mental Disorders or the International Classification of Diseases.20,23 Participants must be aged 19 years or younger (age ≤19 years).24 Studies focusing exclusively on specific populations, such as those with severe intellectual disability or medical conditions where depression is a secondary symptom, will be excluded. Interventions All active treatment modalities for depression in children and adolescents will be included. Interventions will be broadly categorised based on their mechanism of action or other relevant characteristics. These include, but are not limited to: Pharmacotherapy: oral antidepressants (eg, selective serotonin reuptake inhibitors (SSRIs), serotonin-norepinephrine reuptake inhibitors, tricyclic antidepressants), injectable medications (if applicable) and nutritional supplements (eg, omega-3 fatty acids, vitamin C).525,31 Pharmacotherapy: oral antidepressants (eg, selective serotonin reuptake inhibitors (SSRIs), serotonin-norepinephrine reuptake inhibitors, tricyclic antidepressants), injectable medications (if applicable) and nutritional supplements (eg, omega-3 fatty acids, vitamin C).525,31 Psychotherapy: various modalities such as cognitive behavioural therapy (CBT), interpersonal therapy, family therapy, among others.832,34 Psychotherapy: various modalities such as cognitive behavioural therapy (CBT), interpersonal therapy, family therapy, among others.832,34 Physical and neuromodulation therapies: light therapy, repetitive transcranial magnetic stimulation (rTMS), electroconvulsive therapy (ECT) and other non-invasive brain stimulation techniques.35,38 Physical and neuromodulation therapies: light therapy, repetitive transcranial magnetic stimulation (rTMS), electroconvulsive therapy (ECT) and other non-invasive brain stimulation techniques.35,38 Behavioural interventions: structured behavioural approaches including exercise-based interventions.39 40 Behavioural interventions: structured behavioural approaches including exercise-based interventions.39 40 Combined therapies: any combination of the above interventions.41 Combined therapies: any combination of the above interventions.41 Control conditions Eligible control conditions will include: Placebo: administration of an inert substance or sham procedure designed to mimic an active intervention without therapeutic effect. Placebo: administration of an inert substance or sham procedure designed to mimic an active intervention without therapeutic effect. Waitlist: participants receive no active treatment during the study period but may receive treatment after the study concludes. Waitlist: participants receive no active treatment during the study period but may receive treatment after the study concludes. Treatment as usual (TAU): standard care provided in routine clinical practice, which may vary across settings and could include minimal support or non-specific interventions. Treatment as usual (TAU): standard care provided in routine clinical practice, which may vary across settings and could include minimal support or non-specific interventions. Other active comparators: any other active treatment falling within the scope of interventions mentioned above. Other active comparators: any other active treatment falling within the scope of interventions mentioned above. Outcomes We will extract data for both primary and secondary outcomes. For continuous outcomes, such as changes in depression severity or functioning scores, change scores from baseline to endpoint will be prioritised. For dichotomous outcomes, such as response, remission or dropout rates, the number of participants meeting the criteria in each group will be extracted. The primary outcome will be efficacy, measured by the reduction in depression symptom severity using validated depression rating scales. These include, but are not limited to, the Children’s Depression Rating Scale-Revised (CDRS-R),42 43 Hamilton Depression Rating Scale (HAMD),44 45 Montgomery-Åsberg Depression Rating Scale (MADRS),46 Beck Depression Inventory (BDI),47 and Children’s Depression Inventory (CDI).48 Secondary outcomes will include the following: Acceptability, assessed by treatment discontinuation (for any reason) and the incidence of adverse events.49,51 Acceptability, assessed by treatment discontinuation (for any reason) and the incidence of adverse events.49,51 Additional efficacy measures, including response rate (eg, a ≥50% reduction in symptom severity from baseline, or as defined by the study authors) and remission rate (eg, achieving scores below clinical cut-offs on validated depression scales).49 Additional efficacy measures, including response rate (eg, a ≥50% reduction in symptom severity from baseline, or as defined by the study authors) and remission rate (eg, achieving scores below clinical cut-offs on validated depression scales).49 Overall functioning, assessed by global functioning scales such as the Children’s Global Assessment Scale (CGAS) and the Clinical Global Impression-Severity (CGI-S) scale.52 53 Overall functioning, assessed by global functioning scales such as the Children’s Global Assessment Scale (CGAS) and the Clinical Global Impression-Severity (CGI-S) scale.52 53 Study design Only RCTs comparing relevant treatments for depression in children and adolescents will be included. Both parallel-group and crossover RCTs are eligible; for crossover trials, only data from the first phase will be extracted to avoid carry-over effects. There will be no restrictions on sample size, follow-up duration or country of origin.54 Information sources and search strategy We will conduct a comprehensive search across multiple electronic databases to identify all relevant RCTs. The search will not be restricted by publication status or time period but will be limited to English-language publications to ensure accurate assessment of study content and data extraction. The following databases will be searched from inception to 2 July 2024: PubMed, Embase, the Cochrane Library, Web of Science, PsycINFO, Scopus and ClinicalTrials.gov. To ensure that the evidence base remains up to date, the search will be updated before completion of the final systematic review and NMA. Search terms will be developed based on the PICOS framework, using a combination of controlled vocabulary (eg, Medical Subject Headings [MeSH] for PubMed, Emtree for Embase) and free-text keywords related to depression, children and adolescents and various treatment modalities. The search strategy has been informed by previously published NMAs11,1355 and will be adapted to the indexing structure of each database to ensure optimal retrieval. The full PubMed search strategy is provided in online supplemental appendix 3. In addition to electronic database searching, we will manually screen the reference lists of all included studies and relevant systematic reviews or meta-analyses to identify additional eligible studies. Duplicates will be removed using EndNote reference management software. Study selection and data collection Study selection All identified records will be imported into reference management software (eg, EndNote) and duplicates will be removed. Two independent reviewers (JG and ZL) will screen titles and abstracts against the predefined eligibility criteria. Full-text articles will be retrieved for studies deemed potentially relevant. Subsequently, the same two reviewers will independently assess the full texts for inclusion. Reasons for exclusion at the full-text level will be documented and presented in the PRISMA flow diagram and a supplementary table. Discrepancies at any stage will be resolved through discussion or, if necessary, consultation with a third reviewer (WY). Data extraction Data from included studies will be extracted independently by two reviewers using a standardised, piloted extraction form in Microsoft Excel. Extracted information will include study design, participant characteristics, intervention and control details and all relevant outcome data. Disagreements will be resolved through discussion or arbitration by a third reviewer. For continuous outcomes, we will prioritise change scores from baseline to endpoint. If SDs are not reported, they will be calculated from other available statistics (eg, standard errors, CIs, p values) using formulas provided in the Cochrane Handbook.58 For dichotomous outcomes, we will extract the number of events and total participants in each group. If data are incomplete or unclear, we will attempt to impute missing values based on established statistical approaches. The following data items will be extracted: Study characteristics: first author, year, country, study design (eg, parallel-group or crossover RCT) and duration of intervention. Study characteristics: first author, year, country, study design (eg, parallel-group or crossover RCT) and duration of intervention. Participant characteristics: sample size, age range (or mean age with SD), gender distribution, diagnostic criteria. Participant characteristics: sample size, age range (or mean age with SD), gender distribution, diagnostic criteria. Intervention characteristics: type and details of pharmacotherapy (drug name, dose, duration), psychotherapy (modality, format, number/duration of sessions), physical therapies (type, parameters) and combined treatments. Control conditions (placebo, waiting list, TAU) will also be described. To ensure consistent classification of interventions, we will meticulously extract information on all concomitant non-pharmacological interventions in each study arm from primary publications, supplementary materials and trial registrations. For pharmacological trials, if more than 10% of participants in a given arm receive concomitant psychotherapy or behavioural intervention, that arm will be classified as combination therapy; otherwise, it will be classified as pharmacotherapy alone. This threshold-based classification strategy provides a transparent and objective criterion while preserving the clinical interpretability of our network structure. Intervention characteristics: type and details of pharmacotherapy (drug name, dose, duration), psychotherapy (modality, format, number/duration of sessions), physical therapies (type, parameters) and combined treatments. Control conditions (placebo, waiting list, TAU) will also be described. To ensure consistent classification of interventions, we will meticulously extract information on all concomitant non-pharmacological interventions in each study arm from primary publications, supplementary materials and trial registrations. For pharmacological trials, if more than 10% of participants in a given arm receive concomitant psychotherapy or behavioural intervention, that arm will be classified as combination therapy; otherwise, it will be classified as pharmacotherapy alone. This threshold-based classification strategy provides a transparent and objective criterion while preserving the clinical interpretability of our network structure. Outcome data: means and SDs for continuous outcomes; event counts and denominators for dichotomous outcomes. Outcome data: means and SDs for continuous outcomes; event counts and denominators for dichotomous outcomes. Risk of bias assessment Two of the four independent reviewers (JG and ZL, ZK and GZ) will assess the risk of bias for each included RCT using the revised Cochrane Risk of Bias tool (RoB 2.0).59 The RoB 2.0 tool evaluates the risk of bias at the outcome level across five domains: (1) bias arising from the randomisation process; (2) bias due to deviations from intended interventions (based on the effect of assignment to intervention); (3) bias due to missing outcome data; (4) bias in measurement of the outcome; (5) bias in selection of the reported result. Each domain will be rated as ‘low risk,’ ‘some concerns’ or ‘high risk’ of bias according to responses to signalling questions. An overall risk of bias judgement will be derived for each outcome. Discrepancies between reviewers will be resolved through discussion or consultation with a third reviewer (WY). Results will be summarised in both tabular and graphical formats. Data synthesis We will perform a Bayesian NMA to synthesise direct and indirect evidence from the included RCTs. The NMA will be conducted using the gemtc package in R, which implements hierarchical Bayesian models via just another Gibbs sampler.60 A random-effects consistency model will be used, as we anticipate clinical and methodological heterogeneity across studies. The model will assume a normal likelihood and identity link for continuous outcomes and a binomial likelihood with a logit link for dichotomous outcomes. Model estimation will be performed using Markov chain Monte Carlo simulation. We will run multiple chains with sufficient adaptation, thinning and sampling iterations to ensure convergence. Convergence will be assessed using Gelman-Rubin diagnostics and inspection of trace plots and density plots. Treatment comparisons will be presented as posterior median estimates with 95% credible intervals. Treatment ranking will be assessed using the surface under the cumulative ranking curve values and rankograms derived from posterior samples. Effect sizes For continuous outcomes, such as changes in depression symptom scales (eg, CDRS-R, HAMD, MADRS, BDI, CDI) and functional measures (eg, CGAS, CGI-S), we will calculate the standardised mean difference (SMD) as the effect size. The SMD standardises effect sizes across different scales, enabling comparison across trials using different instruments. A positive SMD will indicate greater improvement in the intervention group compared with the comparator. For dichotomous outcomes, including treatment response, remission, dropout due to any cause and adverse events, we will use ORs derived from the posterior distributions. For adverse outcomes, an OR <1 will favour the intervention in terms of better acceptability, while for efficacy outcomes (eg, response/remission), an OR >1 will indicate superiority. Network geometry A network diagram will be constructed to visualise the structure of available evidence. In this diagram, nodes will represent different interventions, and edges will represent direct comparisons between interventions derived from the included trials. The size of each node will be proportional to the total number of participants allocated to that intervention, while the thickness of each edge will reflect the number of studies contributing to the corresponding direct comparison. We plan to construct two types of networks based on the classification of interventions. The first is a broad-category network, in which interventions are grouped into major treatment classes, such as pharmacotherapy, psychotherapy, physical therapy and combined therapy. This approach provides a high-level overview of comparative effectiveness and acceptability across core treatment domains (see figure 1, illustrative example). The second is a mechanism-based network, where interventions are classified with finer granularity based on their pharmacological mechanisms, therapeutic modality or treatment combination (eg, SSRIs, CBT, rTMS). This network is intended to support more detailed clinical comparisons and is conceptually illustrated in figure 2. Illustrative network diagram for the primary analysis. Nodes represent distinct interventions, and edges represent direct comparisons between pairs of interventions based on available studies. Node size is proportional to the total number of participants receiving each intervention, and edge thickness reflects the number of trials informing each comparison. This diagram is intended as a theoretical example and does not reflect the final network structure or data. Illustrative mechanism-based network diagram. Nodes represent specific interventions categorised by their mechanisms of action or therapeutic modality, and edges represent direct comparisons between interventions from the included studies. Node size reflects the total number of participants assigned to each intervention, while edge thickness corresponds to the number of trials informing each comparison. This diagram is for illustrative purposes only and does not reflect the final network structure or dataset. … … represents additional interventions not shown in the figure. BLT, bright light therapy; CBT, cognitive behavioural therapy; IPT, interpersonal psychotherapy; NMDA, N-methyl-D-aspartate receptor antagonist; rTMS, repetitive transcranial magnetic stimulation; SNRI, serotonin-norepinephrine reuptake inhibitor; SSRI, selective serotonin reuptake inhibitor; TcAs, tricyclic antidepressants. The final classification schemes and corresponding network structures will depend on the specific interventions and comparisons identified in the included studies. Adjustments to node definitions and grouping strategies may be made as needed to ensure clinical interpretability and statistical feasibility of the NMA. Transitivity assumption The validity of indirect comparisons in NMA relies on the transitivity assumption, which assumes that effect modifiers are similarly distributed across treatment comparisons. We will assess the plausibility of transitivity by examining key clinical and methodological variables across trials, including age, sex, baseline depression severity, diagnostic criteria, treatment duration, risk of bias and blinding. Summary tables and visual inspection will be used to explore imbalances. If substantial differences are observed, we will explore them through subgroup analyses or meta-regression. Assessment of heterogeneity Statistical heterogeneity across studies will be modelled through a common between-study variance (τ²) in the random-effects model. We will evaluate the magnitude of τ² and compare it with empirical distributions from similar meta-analyses. Visual inspection of trace plots, posterior summaries of τ² and posterior predictive checks will be used to interpret heterogeneity.61 If substantial heterogeneity is detected, subgroup analyses or meta-regression will be conducted to explore potential effect modifiers. A priori, we plan to consider key participant-level and study-level characteristics commonly reported in this literature, including age, sex distribution, baseline depression severity, treatment duration, diagnostic criteria and overall risk of bias, subject to data availability. Any additional analyses performed beyond these prespecified variables will be clearly identified as exploratory. Assessment of incoherence Incoherence (or inconsistency) refers to disagreement between direct and indirect evidence for the same comparison. It will be assessed using both local and global approaches. For local incoherence, we will apply the node-splitting method, which separately models direct and indirect evidence for specific comparisons and tests for statistical inconsistency. For global incoherence, we will use the posterior deviance information criterion and posterior mean residual deviance to compare the consistency model to alternative models (eg, inconsistency or unrelated mean effects models). Forest plots and deviance residual plots will be used to visually inspect inconsistency patterns. If substantial incoherence is observed, we will consider analytical strategies such as presenting direct and indirect evidence separately or reclassifying interventions into more homogeneous categories. Sensitivity analyses To evaluate the robustness of the primary findings, we will perform a series of sensitivity analyses. These will include the exclusion of studies assessed as having a high risk of bias, as well as the exclusion of studies with unclear or implied randomisation procedures. Given potential regional variations in study conduct and reporting quality, we will also conduct analyses excluding studies from specific geographical regions, such as mainland China, if relevant. In addition, sensitivity analyses will be performed using alternative definitions or measurement instruments for key outcomes, to examine whether the results are consistent across different operationalisations of the same construct. To assess the potential impact of missing data handling, we will apply different imputation strategies and compare the results. Finally, we will explore the influence of different network structures, particularly in the mechanism-based classification, by testing alternative ways of grouping interventions. The results of all sensitivity analyses will be compared with the primary analysis to assess the stability and consistency of the conclusions. Any substantial discrepancies will be investigated and discussed in terms of their methodological and clinical implications. Small-study effects and reporting bias We will explore small-study effects and publication bias using comparison-adjusted funnel plots for the NMA and Egger’s test in pairwise comparisons (where ≥10 studies are available). The risk of selective outcome reporting will be assessed through the RoB 2.0 tool, and we will evaluate whether small studies or studies at high risk of bias disproportionately influence the results. Confidence in the evidence The confidence in the evidence will be evaluated for the primary outcome (ie, symptom reduction) and for acceptability (measured as dropouts due to any reason) using the Confidence in Network Meta-Analysis (CINeMA) framework.62 This assessment will be performed independently by two reviewers (JG and ZL), with discrepancies resolved through discussion or consultation with a third reviewer (WY), following the same process as described for study selection and risk of bias assessment. Each treatment comparison will be rated as having high, moderate, low or very low confidence based on judgments across the six CINeMA domains. The ratings will be summarised in both a narrative synthesis and a summary table and will be used to inform the strength of the conclusions and clinical recommendations. Ethics and dissemination This study is a systematic review and NMA based exclusively on data from previously published studies and publicly available trial records; therefore, formal ethics approval is not required. No individual participant data will be collected, and no patients will be directly involved in the conduct of the review. The findings will be disseminated through publication in a peer-reviewed journal and presentation at relevant academic conferences. The protocol has been registered in PROSPERO (CRD42024557384), and relevant amendments have been updated in the registry accordingly. Where permitted, relevant materials, including search strategies and analytical code, will be made available to enhance transparency and reproducibility. Patient and public involvement Patients and members of the public were not involved in the design, conduct, reporting or dissemination plans of this protocol because the study is based on previously published aggregate data and does not involve direct participant recruitment. Discussion This NMA protocol outlines a comprehensive approach to synthesise the evidence on the comparative efficacy and acceptability of all available treatments for depression in children and adolescents. To our knowledge, this will be the first NMA to include such a broad range of interventions, encompassing pharmacotherapy (oral and injectable), psychotherapy, physical therapies (including light therapy, rTMS and ECT) and nutritional supplements, as well as various combined treatment approaches. This comprehensive scope addresses a significant gap in the existing literature, which has largely focused on more limited sets of interventions. One of the key strengths of this study lies in its comprehensive search strategy, which will cover seven major electronic databases and aim to capture all relevant evidence on current treatment modalities for depression in children and adolescents. This systematic and inclusive approach is intended to ensure a broad and representative evidence base for analysis.14 15 Furthermore, our intention to classify interventions not only into broad categories but also based on their mechanisms of action (eg, specific drug classes, psychotherapy modalities) is a novel aspect that will enhance the clinical utility of our findings. Previous meta-analyses have often focused on specific drug types, which, while valuable, may have limited generalisability due to the relatively small number of studies for each specific drug. By grouping treatments based on their mechanisms, we anticipate increasing the number of studies within each comparison, thereby improving the robustness and clinical relevance of our results. Another important contribution of this NMA is the inclusion of overall functioning as a secondary outcome, measured by scales such as CGAS and CGI-S. While efficacy in symptom reduction is crucial, assessing improvements in social and global functioning provides a more holistic understanding of treatment benefits, which is highly relevant for the long-term well-being of children and adolescents with depression. Additionally, our detailed assessment of acceptability, including dropout rates and adverse events, will provide critical information for shared decision-making between clinicians, patients and their families. We anticipate several potential challenges and limitations. The heterogeneity across studies in terms of participant characteristics, diagnostic criteria, intervention protocols and outcome measures is a common challenge in systematic reviews of mental health interventions.63 64 We will address this by employing a random-effects model for our NMA and by conducting extensive subgroup and sensitivity analyses to explore the sources of heterogeneity and assess the robustness of our findings.64 The potential for publication bias and small-study effects will also be carefully assessed using appropriate statistical methods and visual inspections of funnel plots. The certainty of the evidence will be transparently evaluated using the CINeMA approach, which will provide a clear indication of the confidence in our results.56 Another limitation is the restriction to English-language publications, which may introduce language bias and lead to the exclusion of some relevant studies. Future updates of this review may consider the inclusion of non-English studies, should adequate multilingual resources or validated translation workflows become available. Despite these challenges, the findings from this NMA are expected to have significant implications for clinical practice and future research. By providing a comprehensive ranking of all available treatments for depression in children and adolescents based on both efficacy and acceptability, this study will offer valuable guidance for clinicians in selecting the most appropriate interventions.15 65 It will also highlight areas where the evidence base is weak or where further high-quality RCTs are needed, particularly for newer or less-studied interventions. Ultimately, this research aims to contribute to improving the outcomes and quality of life for children and adolescents affected by depression. Supplementary material Footnotes Funding: This study was supported by the STI2030-Major Projects (2021ZD0200702); Beijing Municipal Health Commission Research Ward Programme (third batch); National Natural Science Foundation of China (82301687). This work was also supported by China Postdoctoral Science Foundation (2024M760141); Peking University Medicine Sailing Programme for Young Scholars’ Scientific & Technological Innovation, the Fundamental Research Funds for the Central Universities (BMU2025YFJHPY044); Grant BX20240029 from the National Postdoctoral Programme for Innovative Talents (ZL). Prepublication history and additional supplemental material for this paper are available online. To view these files, please visit the journal online (https://doi.org/10.1136/bmjopen-2025-108754) Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Patient and public involvement: Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research. Data availability statement Data sharing not applicable as no datasets were generated and/or analysed for this study. All data relevant to the study are included in the article or uploaded as supplementary information. References Supplementary Materials Data Availability Statement Data sharing not applicable as no datasets were generated and/or analysed for this study. All data relevant to the study are included in the article or uploaded as supplementary information.