Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

General Comments

The paper is a systematic review on whether teacher training programs improve teachers’ ability to detect and report Child Victimization (CV), with the evidence focusing mainly on Child Sexual Abuse (CSA) (abstract lines 8-10). The review included 10 empirical studies, mostly from the United States, with others from Australia, Turkey, Cambodia, and the Philippines. No European studies were included (lines 527-528).

Section 1. Introduction: broadly introduces CV, its estimates from the World Health Organization (WHO), its manifestations according to well-known psychosocial models such as Bronfenbrenner’s ecological model, and the socio-physical adverse consequences associated with it (lines 29-50). Its concealability is also highlighted by the difficulties in interpreting its signs and the obstacles professionals face in detecting and reporting it (lines 51-59). A lack of specialized knowledge is also occurring among teachers (lines 60-69). The Knowledge, Attitudes, and Practices (KAP) model is identified by the authors as one of the most appropriate approaches for determining that behavior is influenced not only by available information and understanding of a given phenomenon, but also by complex, interconnected interactions and background preparation around it (lines 70-80). School, as such an institutional interactional environment, is therefore considered a key observational point for protecting children, both for identifying potential harm and for activating actions to prevent it (lines 85-93).

Child Sexual Abuse (CSA) is then defined and discussed as one of the most severe and complex forms of CV. Trends on perpetrators and the mode of action both highlight the potential concealability of this sad phenomenon of child abuse (lines 106-114). The importance of training programs for improving teachers' specific knowledge of the phenomenon is also highlighted (lines 116-119), but, unfortunately, there is significant fragmentation and variability in methodologies and approaches, which affects the actual effectiveness of these programs (lines 120-122).

From there, the intended scope of this review is to systematically collect evidence on how teacher training can improve the detection and reporting of CSA, identify knowledge gaps, and inform evidence-based action to improve educational policies and tools (lines 123-126).

Starting Section 2. In Materials and Methods, the paper's scope is repetitive, with the listing of the variables considered in this review: a. knowledge, b. attitudes, c. perceived self-efficacy, and d. teachers’ professional practices on CV.

Sub-section 2.2. Methodology illustrates the PRISMA-based selection procedure for deriving the final set of 10 empirical studies from the initial 587 records published over the last 15 years, taking into account the greater international community interest in CSA matters since the 21^st century (lines 172-182) and the consequent expansion of prevention programs. The literature search was conducted from May to August 2025, using well-known electronic databases for the online archiving of academic papers (lines 149-151). Table 1, line 155, shows how the basic AND/OR operators were used to search electronic databases, noting that a manual review also helped the authors retrieve papers that were not found after the initial query (lines 157-161). The Appendix table, lines 574-575, also adds value by systematically mapping the main characteristics of the included studies. However, given that only 10 of the 85 full-text articles assessed for eligibility were ultimately included (lines 192-194), the authors should further clarify the exclusion process, particularly by providing more detailed reasons for exclusion. The exclusion criteria listed in lines 183-185 do not clearly explain why these categories have been excluded. A “narrative study” or a “doctoral dissertation” could constitute a fruitful source for this review. This would strengthen the reproducibility and methodological credibility of the review. In addition, Figure 1, line 198, indicates that 12 studies were included in the review, whereas the studies used are 10 (lines 192-194 and 206-210). Another unclear criterion for the quantitative scoring of studies, attributed by the authors to the selected papers for their review, appears in lines 211-215. Among the Results in Section 3, the authors indicated that the selected data span 2018-2024, that the United States is the most represented country, and that quantitative quasi-experimental designs are predominant (lines 228-235). The study populations in the 10 selected studies for the review have heterogeneous sizes and sociodemographic characteristics, and include teachers and/or large groups of students (lines 244-250). Gender and professional career thresholds of professionals involved in the surveys show variability, with most studies comprising predominantly women (lines 251-257).

 

A particular interest in this review comes from a comprehensive, multifaceted program for teachers' training to prevent CSA, called the Second Step Child Protection Unit (CPU) (lines 263-269). Another program of particular interest is Enough! Preventing Child Sexual Abuse in My School. Some programs, especially the Second Step Child Protection Unit, show broader positive effects, including improvements in school climate and, indirectly, students’ knowledge and ability to recognize, reject, and report abuse.

The authors then describe and comment on other programs, such as J-Safe, noting that teachers’ overall knowledge of CSA improves, even though some observed post-intervention effects indicate low baseline levels of knowledge (lines 343-351). The paper emphasizes an important limitation: better knowledge does not always translate into actual reporting behavior. Some studies found improved awareness but no clear increase in real reports (lines 372-382).

The gap between training and actual reporting appears to be influenced by contextual and organizational factors, such as school climate, institutional support, clarity of protocols, fear of consequences, and teachers’ confidence (lines 391-400; see also a little further in lines 488-497). In comparison, students in the intervention groups demonstrate better understanding of CSA and the ability to act professionally in these situations (lines 413-424).

Section 4. Discussion is rather self-referential to the authors’ review findings, which are mostly secondary and reiterate that increased teacher knowledge is not necessarily associated with a significant increase in case reporting after training (lines 460-465). What in lines 70-80 was called the KAP model is now called the CAP model, adding the advantage of including attitudinal and contextual variables to the study of the topic (lines 465-468). The authors argue that CSA training should not be a one-time intervention. Because some studies show that knowledge and confidence may decline over time, they recommend ongoing reinforcement and refresher training (lines 498-504).

The review’s main limitations are the small number of studies, the focus mainly on CSA rather than other forms of child victimization, limited geographical diversity, and insufficient attention to contextual factors (lines 512-522). The paper concludes that teacher training is a promising and necessary strategy, but future programs should go beyond information delivery by strengthening self-efficacy, addressing emotional and institutional barriers, and adapting training to specific school contexts.

 

 

Amendments

1. Please collect more recent estimated data on Child Victimization (CV) by the World Health Organization (WHO) or any other international authority in this field. The 2024 data on lines 31-33 are too old for the targets of your review study.
2. Please collect more recent estimated data on Child Sexual Abuse (CSA) by the UNICEF or any other international authority in this field. The 2024 data on lines 100-102 are too old for the targets of your review study.
3. The PRISMA-based selection procedure is useful because it reduces selection bias and allows readers to understand how the final set of 10 empirical studies was derived from the initial 587 records. Though the rather unclear exclusion criteria and the number of final selected studies shown in Figure 1 suggest the PRISMA procedure currently lacks sufficient transparency and internal consistency, reducing its reproducibility.  
4. In Section 3 Results, you speak about control/intervention groups as they were part of your own research. Please make it clear to the audience that these are the survey characteristics of some of the 10 studies under review, not yours.
5. Since you claim that teacher training improves knowledge and preparedness, but its real impact on reporting depends heavily on school-level support, clear procedures, and long-term sustained reinforcement, I personally would create a dedicated small section where significant examples of these CSA treatment’s findings are detailed and shown.
6. I personally would compact the long and somewhat repetitive content of the paper’s sections and provide better Tables/Figures in the paper. For instance, very simple statistics on this topic are presented in the paper by Towsend, C. & Haviland, M. (2016). The impact of child sexual abuse training for educators on reporting and victim outcomes: The Texas Initiative. Charleston, S.C., Darkness to Light. Retrieved on July 8, 2026, from https://www.d2l.org/wp-content/uploads/2024/10/The-Impact-of-CSA-Training-for-Educators.pdf. You could derive similar statistics by combining the trends from the selected 10 studies.
7. Several of the included primary studies rely on self-reported knowledge, confidence, or preparedness rather than objective behavioral outcomes. The implications of this methodological characteristic should be acknowledged in the review as a limitation, particularly given the discrepancy between increased knowledge and actual reporting behavior.

There are misprints in some of the authors’ references, such as no. 2 Álvarez et al. ("proffesionals"), no. 32 Kisely et al. ("Chil maltreatment"), and reference no. 5 Bayes et al. (2019), which appears to concern the bioavailability of folate supplementation. Please verify the bibliography carefully, as this reference does not appear relevant to the methodological statement in Section 2.2.4 regarding the quality assessment thresholds. Please also check the automatic spell checker throughout the reference list.

REFERENCES:

Towsend, C. & Haviland, M. (2016). The impact of child sexual abuse training for educators on reporting and victim outcomes: The Texas Initiative. Charleston, S.C., Darkness to Light. Retrieved on July 8, 2026, from https://www.d2l.org/wp-content/uploads/2024/10/The-Impact-of-CSA-Training-for-Educators.pdf

Author Response

Reviewer 1:

The article is a systematic review on whether teacher training programs improve teachers' ability to detect and report Child Victimization (CV), with the evidence mainly focused on Child Sexual Abuse (CSA) (summary lines 8-10). The review included 10 empirical studies, mostly from the United States, with others from Australia, Turkey, Cambodia, and the Philippines. No European studies were included (lines 527-528).

Section 1. Introduction: generally presents CV, its estimates by the World Health Organization (WHO), its manifestations according to psychosocial models known as Bronfenbrenner's ecological model, and the associated adverse socio-physical consequences (lines 29-50). Its concealability is also evident in the difficulties in interpreting its signals and the obstacles faced by professionals in detecting and reporting it (lines 51-59). There is also a lack of specialized knowledge among teachers (lines 60-69). The Knowledge, Attitudes, and Practices (KAP) model is identified by the authors as one of the most suitable approaches to determine that behavior is influenced not only by the information available and understanding of a given phenomenon, but also by complex and interconnected interactions and the surrounding background preparation (lines 70-80). School, as an institutional setting for interaction, is therefore considered a key point of observation to protect children, both to identify possible harm and to trigger actions that prevent it (lines 85-93).

Child sexual abuse (CSA) is then defined and discussed as one of the most serious and complex forms of CV. Trends regarding perpetrators and their methods highlight the potential for this sad phenomenon of child abuse to be hidden (lines 106-114). The importance of training programs to improve teachers' specific knowledge of the phenomenon is also emphasized (lines 116-119), but, unfortunately, there is significant fragmentation and variability in methodologies and approaches, which affects the actual effectiveness of these programs (lines 120-122).

From there, the intended scope of this review is to systematically gather evidence on how teacher training can improve the detection and reporting of child abuse (CSA), identify knowledge gaps, and inform evidence-based actions to improve educational policies and tools (lines 123-126).

Starting Section 2. In Materials and Methods, the scope of the article is repetitive, with the list of variables considered in this review: a. knowledge, b. attitudes, c. perceived self-efficacy, and d. professional practices of CV teachers.

Subsection 2.2. The methodology illustrates the selection process based on PRISMA to derive the final set of 10 empirical studies from the initial 587 records published over the past 15 years, taking into account the growing interest of the international community in CSA issues since the 21st century (lines 172-182) and the consequent expansion of prevention programs. The literature search was conducted from May to August 2025, using well-known electronic databases for the online archive of academic articles (lines 149-151). Table 1, line 155, shows how the basic AND/OR operators were used to search electronic databases, noting that a manual review also helped the authors retrieve articles that were not found in the initial search (lines 157-161). The table in the Appendix, lines 574-575, also adds value by systematically mapping the main features of the included studies. However, since only 10 of the 85 full articles assessed for eligibility were ultimately included (lines 192-194), the authors should clarify the exclusion process further, especially by providing more detailed reasons for exclusion. The exclusion criteria listed in lines 183-185 do not clearly explain why these categories were excluded. A 'narrative study' or a 'doctoral thesis' could be a valuable source for this review. This would strengthen the reproducibility and methodological credibility of the review. Additionally, Figure 1, line 198, indicates that 12 studies were included in the review, while the studies actually used are 10 (lines 192-194 and 206-210). Another unclear criterion for the quantitative scoring of the studies, attributed by the authors to the articles selected for their review, appears in lines 211-215. Among the results in Section 3, the authors indicated that the selected data cover 2018-2024, that the United States is the most represented country, and that quantitative quasi-experimental designs predominate (lines 228-235). The study populations in the 10 studies selected for the review show heterogeneous sizes and sociodemographic characteristics, and include teachers and/or large groups of students (lines 244-250). The gender and professional career thresholds of the professionals involved in the surveys show variability, with most of the studies being predominantly women (lines 251-257).

Response: Thank you for this helpful comment. We have clarified the exclusion process by specifying the number of articles excluded for each reason and have revised the PRISMA flow diagram accordingly. We have also corrected the typographical error in the previous version of the figure, which incorrectly reported 12 included articles instead of the correct number of 10.

 

A particular interest in this review comes from a comprehensive and multifaceted teacher training program to prevent child sexual abuse, called the Second Step Child Protection Unit (lines 263-269). Another program of particular interest is Stop! Preventing Child Sexual Abuse in My School. Some programs, especially the Second Step Child Protection Unit, show broader positive effects, including improvements in the school climate and, indirectly, in students' knowledge and ability to recognize, reject, and report abuse.

The authors describe and discuss other programs, like J-Safe, pointing out that teachers' general knowledge about CSA improves, although some effects observed after the intervention indicate low levels of knowledge at the baseline (lines 343-351). The article highlights a major limitation: better knowledge doesn't always translate into actual reporting behavior. Some studies found increased awareness but no clear rise in actual reporting (lines 372-382).

The gap between training and actual reporting seems to be influenced by contextual and organizational factors, such as school climate, institutional support, clarity of protocols, fear of consequences, and teachers' confidence (lines 391-400; see also a bit later in lines 488-497). In comparison, students in the intervention groups show a better understanding of ASC and the ability to act professionally in these situations (lines 413-424).

Section 4. The discussion is rather self-referential regarding the authors' review findings, which are mostly secondary and reiterate that an increase in teacher knowledge is not necessarily associated with a significant rise in case reporting after training (lines 460-465). What was called the KAP model in lines 70-80 is now referred to as the CAP model, adding the advantage of including attitudinal and contextual variables in the study of the topic (lines 465-468). The authors argue that ASC training should not be a one-time intervention. Since some studies show that knowledge and confidence can decrease over time, they recommend continuous reinforcement and refresher training (lines 498-504).

The main limitations of the review are the small number of studies, the focus mainly on child sexual abuse rather than other forms of child victimization, the limited geographic diversity, and the insufficient attention to contextual factors (lines 512-522). The article concludes that teacher training is a promising and necessary strategy, but future programs should go beyond just providing information, strengthening self-efficacy, addressing emotional and institutional barriers, and tailoring training to specific school contexts.

Amendments

1. Please gather the most recent estimated data on Child Victimization (CV) from the World Health Organization (WHO) or any other international authority in this field. The 2024 data on lines 31-33 are too old for the purposes of your review study.

Response: Thank you for this helpful recommendation. We have updated the manuscript with more recent estimates on child victimization from an international authoritative source, replacing the previously reported data.

 

2. Please gather the most recent estimated data on child sexual abuse (CSA) from UNICEF or any other international authority in this field. The 2024 data on lines 100-102 are too old for the purposes of your review study.

Response: Thank you for this helpful recommendation. We have updated the manuscript with more recent estimates on child sexual abuse from an international authoritative source, replacing the previously reported data.

 

3. The selection procedure based on PRISMA is useful because it reduces selection bias and allows readers to understand how the final set of 10 empirical studies was derived from the initial 587 records. Although the exclusion criteria, which are quite unclear, and the number of final selected studies shown in Figure 1 suggest that the PRISMA procedure currently lacks enough transparency and internal consistency, which reduces its reproducibility.

Response: Thank you for this helpful comment. We have improved the transparency and reproducibility of the review process by adding a new “Synthesis of Results” subsection to the Methods and revising the PRISMA flow diagram to clearly report the number of studies excluded at each stage and the reasons for exclusion.

 

4. In Section 3 Results, you talk about the control/intervention groups as if they were part of your own research. Please make it clear to the audience that these are the survey characteristics of some of the 10 studies under review, not yours.

Response: Thank you for this helpful comment. We have revised the “General Overview” section to clarify that the intervention and control groups described correspond to the characteristics of some of the studies included in the review, rather than to groups from the present study.

 

5. Since you claim that teacher training improves knowledge and preparation, but its real impact on information largely depends on school-level support, clear procedures, and sustained long-term reinforcement, I would personally create a small dedicated section where the findings of these child sexual abuse interventions are detailed and shown with meaningful examples.

Response: Thank you for this helpful recommendation. We have addressed this point by clarifying throughout the manuscript the distinction between improvements in teacher knowledge and preparedness and their translation into actual reporting, as well as the importance of school-level support, clear procedures, and sustained reinforcement. These aspects have been further emphasized in the revised Discussion and Conclusions.

 

6. Personally, I would condense the long and somewhat repetitive content of the article's sections and provide better tables/figures in the work. For example, very simple statistics on this topic are presented in the article by Townsend, C. & Haviland, M. (2016). The Impact of Child Sexual Abuse Training for Educators on Reporting and Victim Outcomes: The Texas Initiative. Charleston, South Carolina, from Darkness to Light. Accessed July 8, 2026, from https://www.d2l.org/wp-content/uploads/2024/10/The-Impact-of-CSA-Training-for-Educators.pdf. Similar statistics could be derived by combining the trends from the 10 selected studies.

Response: Thank you for this helpful recommendation. We have carefully revised and streamlined the manuscript, reducing repetition and improving the conciseness of the different sections in response to the comments received. We appreciate the suggestion regarding the inclusion of additional tables or figures and have taken it into consideration when revising the presentation of the findings.

 

7. Several of the primary studies included rely on self-reported knowledge, confidence, or readiness rather than on objective behavioral outcomes. The implications of this methodological feature should be acknowledged in the review as a limitation, especially given the gap between increased knowledge and actual reporting behavior.

Response: Thank you for this important comment. We have incorporated this issue into the Limitations section, explicitly acknowledging the predominance of self-reported knowledge, confidence, and preparedness measures and the limited evidence regarding objective behavioral outcomes, particularly actual detection and reporting.

Reviewer 2 Report

Comments and Suggestions for Authors

This articles tackles a very important issue of child victimization (CV), and child sexual abuse (CSA) in the school context. The article describes the issue well and focuses on the shortcoming of teacher education. Consequently, teacher training in the prevention, detection, and reporting of CSA is identified as the key strategic priority for intervention. This article has ambitious global approach and the shows that these problems are widespread in the world, although there seems to be glimmer of hope of better awareness spreading in many societies. This article also includes descriptions of several intervention programs in several societies and their positive effects. In other words, there is a plenty to learn from successful intervention programs and the responsibility falls on schools to take seriously the issues and make necessary and permanent changes to teacher training. In other words, this study has major practical recommendations to teacher training backed by sound evidence and research publications. The most important contribution of this article is the raising awareness of this issue and then giving practical advice how to move on to implement intervention programs. In terms of analyzing the issues of CV and CSA this article does not analyze the issues further as the focus is in more practical issues. In my opinion, this article is ready to publish as it is as it has includes important contributions especially for rethinking and improving teacher education.

Author Response

Reviewer 2:

This article tackles a very important topic about child victimization (CV) and child sexual abuse (CSA) in the school context. The article describes the problem well and focuses on the gaps in teacher training. Consequently, teacher training in the prevention, detection, and reporting of CSA is identified as the key strategic priority for intervention. This article has an ambitious global approach and shows that these problems are widespread around the world, although there seems to be a glimmer of hope that better awareness will spread in many societies. The article also includes descriptions of various intervention programs in different societies and their positive effects. In other words, this study includes important practical recommendations for teacher training backed by solid evidence and research publications. The most important contribution of this article is raising awareness about this issue and then offering practical advice on how to move forward with implementing intervention programs. As for the analysis of CV and CSA topics, this article doesn't go into the topics in depth, since the focus is more on practical matters. In my opinion, this article is ready to be published as is, since it includes important contributions, especially for rethinking and improving teacher training.

Response:  We sincerely thank the Editor for the positive and encouraging evaluation of our manuscript. We greatly appreciate the recognition of the relevance of addressing child victimization and child sexual abuse in the school context, as well as the importance of teacher training for prevention, detection, and reporting. We are also grateful for the acknowledgement of the practical implications of the review and the contribution that the findings may make to improving teacher training programmes. We appreciate the Editor’s overall assessment of the manuscript and the opportunity to contribute to raising awareness of this important issue.

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript addresses an important and highly relevant question for education and child protection: whether teacher training improves the detection and reporting of child victimization, particularly child sexual abuse (CSA), in school settings. The review has several methodological strengths, including use of PRISMA 2020, prospective PROSPERO registration, searches across four major databases, independent screening and data extraction by two reviewers, third-reviewer arbitration, and high inter-reviewer agreement.

The review has potential value because it distinguishes between gains in knowledge, attitudes, confidence, and perceived self-efficacy on the one hand, and actual reporting behaviour on the other. This distinction is important, and the manuscript appropriately recognizes that improvements in professional knowledge do not necessarily translate into changes in practice.

However, I believe several methodological issues should be addressed before the conclusions can be considered sufficiently robust.

1. The search strategy requires verification and greater comprehensiveness.

The authors acknowledge that terms such as “child maltreatment,” “sexual abuse,” and “neglect” were not systematically included in all database searches. This is a significant limitation because the included evidence ultimately focuses predominantly on CSA. Broad terms such as “child victimization” or “reporting child abuse” may not retrieve all relevant studies indexed under more specific terminology.

The Boolean structure of some search strings reported in Table 1 also requires clarification, particularly the grouping of AND/OR operators in PsycINFO and Scopus. The authors should verify that the strings reproduced in the manuscript correspond exactly to those executed and, if necessary, update the search using a broader set of controlled and free-text terms related to child maltreatment, CSA, detection, recognition, identification, safeguarding, and reporting.

If the search is updated, any resulting changes in the included study set should be reported transparently.

2. The scope of the review should be clarified: child victimization broadly or CSA specifically.

The title and eligibility framework refer to child victimization, whereas the Abstract and Discussion focus primarily on CSA. The authors themselves acknowledge that the evidence base is overwhelmingly concentrated on CSA.

This does not invalidate the review, but the inferential scope needs to be consistent. Either the manuscript should be clearly framed as a review of child victimization in which the available evidence is largely limited to CSA, or the title, aims, and conclusions should be narrowed accordingly. The current wording risks implying broader generalizability than the evidence supports.

3. The methodological quality assessment should be reported more transparently.

The use of JBI appraisal tools is appropriate, but the manuscript currently explains the assessment procedure without providing a sufficiently clear study-by-study presentation of the results. Readers should be able to see which methodological criteria were met or not met by each included study and how these judgments influenced the synthesis.

I also recommend reconsidering or better justifying the use of global numerical quality thresholds. A simple total score may conceal important differences in risk of bias across critical domains. The methodological source used to justify these thresholds should also be checked carefully; one cited reference, Bayes et al. (2019), concerns folate supplementation rather than JBI methodology.

4. The synthesis method needs to be described explicitly.

The manuscript provides detailed information on search, selection, appraisal, and extraction, but the procedure used to synthesize the findings is not sufficiently explicit. The included studies are heterogeneous in design, intervention content, outcomes, and follow-up, and include RCTs, quasi-experimental studies, and one qualitative study.

A narrative synthesis is entirely reasonable in this context, but the authors should state clearly how studies were grouped, how differences in design and methodological quality were handled, and why meta-analysis was not considered appropriate or feasible.

5. Claims concerning effectiveness should be more carefully calibrated.

The evidence is generally positive, particularly for knowledge-related outcomes. Improvements are also reported in attitudes, confidence, perceived competence, and willingness to report. However, these outcomes should not be treated as equivalent to actual detection or reporting behaviour.

The manuscript itself notes that increases in knowledge did not always lead to increased reporting. For this reason, I would recommend moderating statements such as “teacher training is an effective intervention” and specifying that the strongest evidence concerns proximal outcomes, especially knowledge, whereas evidence for sustained behavioural change and actual reporting remains more limited and heterogeneous.

This distinction should be reflected consistently in the Abstract, Discussion, and Conclusion.

5. The student-focused findings should be positioned more clearly.

The eligibility criteria identify in-service teachers as the main population, but the Results include a section on student outcomes. These findings may be relevant in school-wide programs, but they fall outside the principal PICO as currently described.

The authors should either present student outcomes explicitly as secondary or contextual findings, or revise the review framework to explain prospectively why these outcomes were included.

6. The limitations should acknowledge additional sources of bias.

The current limitations section appropriately notes the predominance of CSA, limited geographical diversity, and possible incompleteness of the search. I recommend adding the exclusion of grey literature and the restriction to English- and Spanish-language publications as potential sources of publication and language bias.

The heterogeneity of intervention content, follow-up periods, outcome measures, and methodological quality should also be acknowledged more explicitly.

A few minor issues also require attention. The PROSPERO identifier should be provided in the final manuscript, and any deviations from the registered protocol should be disclosed. The terminology KAP/CAP should be standardized. The reference list requires careful editing, as several entries contain typographical and formatting inconsistencies, and the in-text year for Dudley et al. should be checked against the reference list.

Overall, this is a relevant and potentially publishable review. The methodological framework is generally sound, and the distinction between improved teacher knowledge and the more difficult translation into actual reporting practice is an important contribution. However, the search strategy, quality assessment reporting, synthesis procedure, and scope of the effectiveness claims need substantial clarification and, in some cases, methodological revision.

 

Author Response

Reviewer 3:

This manuscript tackles an important and very relevant issue for education and child protection: whether teacher training improves the detection and reporting of child victimization, especially child sexual abuse (CSA), in school settings. The review has several methodological strengths, including the use of PRISMA 2020, prospective registration in PROSPERO, searches in four major databases, independent selection and data extraction by two reviewers, arbitration with third reviewers, and a high level of agreement between reviewers.

The review has potential value because it distinguishes between gains in knowledge, attitudes, confidence, and perceived self-efficacy on one hand, and actual reporting behaviors on the other. This distinction is important, and the manuscript appropriately acknowledges that improvements in professional knowledge do not necessarily translate into changes in practice.

However, I think several methodological issues need to be addressed before the conclusions can be considered solid enough.

 

1. The search strategy requires verification and greater thoroughness.

The authors acknowledge that terms like "child maltreatment," "sexual abuse," and "neglect" were not systematically included in all database searches. This is a significant limitation because the evidence included mainly focuses on CSA. General terms like "child victimization" or "reporting child abuse" might not capture all relevant studies indexed under more specific terminology.

The Boolean structure of some search strings reported in Table 1 also needs clarification, especially the grouping of AND/OR operators in PsycINFO and Scopus. The authors should check that the strings reproduced in the manuscript match exactly those that were executed and, if necessary, update the search using a broader set of controlled and free-text terms related to child maltreatment, child sexual abuse, detection, recognition, identification, protection, and reporting.

If the search is updated, any resulting changes in the set of included studies should be reported transparently.

Response: Thank you for this detailed and important comment. We have carefully reviewed the search strategy and clarified the rationale behind the terms and Boolean structure used. The search strategies reported in the manuscript correspond to those actually conducted during the review. The research question was initially framed around child victimization broadly, and therefore we did not restrict the search a priori to specific forms of victimization, such as sexual, physical, or emotional abuse or neglect, as this could have narrowed the scope of the review and excluded evidence addressing other forms of child victimization. The predominance of studies addressing child sexual abuse emerged from the evidence identified rather than from an a priori restriction of the search. Regarding the Boolean structure, we acknowledge that the formulation of some search strings, particularly in PsycINFO and Scopus, could have been more comprehensive and precise. We have therefore clarified the search strategy as reported and acknowledged this aspect as a methodological limitation, while retaining the searches actually conducted to ensure transparency and consistency with the review process.

 

2. The scope of the review needs to be clarified: child victimization in general or child sexual abuse specifically.

The title and eligibility framework refer to child victimization, while the Abstract and Discussion focus mainly on child sexual abuse. The authors themselves acknowledge that the evidence base is overwhelmingly concentrated on CSA.

This doesn’t invalidate the review, but the inferential scope needs to be consistent. Either the manuscript should be clearly framed as a review of child victimization where the available evidence is largely limited to Child Abuse, or the title, objectives, and conclusions should be scaled down accordingly. The current wording risks implying a broader generalization than the evidence supports.

Response: Thank you for this important comment. We have clarified the scope of the review by revising the objective and adding a paragraph to the Introduction to explicitly explain the relationship between child victimization and the predominantly child sexual abuse-focused evidence base. We have also revised the wording throughout the Discussion and Limitations to ensure that the interpretation of the findings does not extend beyond the evidence available.

 

3. The methodological assessment of quality should be reported more transparently.

The use of JBI evaluation tools is appropriate, but the manuscript currently explains the evaluation procedure without providing a sufficiently clear study-by-study presentation of the results. Readers should be able to see which methodological criteria were met or not in each included study and how these judgments influenced the synthesis.

I also recommend reconsidering or better justifying the use of global numerical quality thresholds. A simple total score can hide important differences in the risk of bias across critical domains. The methodological source used to justify these thresholds should also be carefully checked; one cited reference, Bayes et al. (2019), refers to folate supplementation rather than JBI methodology.

Response: Thank you for this detailed and helpful recommendation. We have revised the methodological quality assessment by removing the global quality categories (high, moderate, and low) and the corresponding numerical thresholds, while retaining the individual methodological assessment of each included study. 

 

4. The synthesis method must be explicitly described.

The manuscript provides detailed information on search, selection, appraisal, and data extraction, but the procedure used to synthesize the findings is not sufficiently explicit. The included studies are heterogeneous in design, intervention content, outcomes, and follow-up, and include RCTs, quasi-experimental studies, and one qualitative study.

A narrative synthesis is entirely reasonable in this context, but the authors should clearly indicate how the studies were grouped, how differences in design and methodological quality were handled, and why a meta-analysis was not considered appropriate or feasible.

Response: Thank you for this helpful recommendation. We have added a new “Synthesis of Results” subsection to the Methods section, explicitly describing the narrative synthesis approach, how studies were grouped according to their outcomes and methodological characteristics, how differences in study design and methodological quality were considered, and why a meta-analysis was not considered appropriate given the heterogeneity of the included studies.

 

5. Claims about effectiveness should be made with more caution.

The evidence is generally positive, especially regarding knowledge-related outcomes. Improvements in attitudes, confidence, perceived competence, and willingness to report are also reported. However, these results should not be treated as equivalent to actual detection or reporting behavior.

The manuscript itself notes that increased knowledge does not always lead to increased reporting. For this reason, I would recommend tempering statements like 'teacher training is an effective intervention' and clarifying that the strongest evidence relates to proximal outcomes, especially knowledge, while evidence for sustained behavioral change and actual reporting remains more limited and mixed.

This distinction should be consistently reflected in the Abstract, Discussion, and Conclusion.

Response: Thank you for this important recommendation. We have carefully revised and moderated the statements regarding the effectiveness of teacher training throughout the manuscript, including the Abstract, Discussion, and Conclusion, to emphasize that the strongest evidence concerns proximal outcomes, particularly knowledge, while changes in actual detection and reporting behaviors remain more limited and heterogeneous.

6. Sudent-focused findings need to be positioned more clearly.

The eligibility criteria identify in-service teachers as the main population, but the Results include a section on student outcomes. These findings may be relevant in school programs, but they are outside the main PICO as currently described. 

The authors should explicitly present student outcomes as secondary or contextual findings, or revise the review framework to prospectively explain why these outcomes were included.

Response: Thank you for this helpful comment. We have clarified the position of the student-related findings within the review by renaming the corresponding section as “Secondary Outcomes Related to Students and Their Education on CSA” and adding a paragraph explicitly describing these findings as secondary or contextual outcomes, outside the main PICO. We have also clarified their secondary nature in other sections of the manuscript where these findings are mentioned.

 

7. The limitations should acknowledge additional sources of bias.

The current limitations section properly points out the predominance of CSA, the limited geographic diversity, and the possible incompleteness of the search. I recommend adding the exclusion of gray literature and the restriction to publications in English and Spanish as possible sources of publication and language bias. The heterogeneity of the intervention content, follow-up periods, outcome measures, and methodological quality should also be recognized more explicitly.

Response: Thank you for these helpful recommendations. We have expanded the limitations section to explicitly acknowledge the exclusion of grey literature, the restriction to English- and Spanish-language publications, and the heterogeneity in intervention content, follow-up periods, outcome measures, and methodological quality.

 

Some minor issues also require attention. The PROSPERO identifier must be provided in the final manuscript, and any deviations from the registered protocol must be disclosed.

Response: Thank you for this comment. The PROSPERO identifier was omitted from the manuscript to preserve anonymity during the review process; it will be included in the final version of the manuscript, and any deviations from the registered protocol will be appropriately reported.

 

The KAP/CAP terminology should be standardized. The reference list needs careful editing, as several entries contain typographical and formatting inconsistencies, and the year in Dudley et al.'s text should be checked against the reference list. 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for carefully revising the manuscript and addressing the comments raised in the previous review. The revised version has improved substantially in methodological transparency and evidence interpretation. In particular, the corrected and expanded PRISMA flow diagram now provides the numbers and reasons for full-text exclusions; the quality-assessment procedure has been clarified without relying on unsupported cut-off scores; and the new subsection on synthesis of results better explains the rationale and organization of the narrative synthesis. The distinction between proximal outcomes (knowledge, attitudes, perceived self-efficacy, and confidence) and actual detection/reporting behavior is also much clearer throughout the revised manuscript. The expanded discussion of self-reported outcomes, intervention heterogeneity, publication/language bias, and limited geographical representation further improves the interpretation of the findings.

I have only a few residual points that would benefit from correction before publication. In the Abstract, the expression “consistent evidence” appears well supported for improvements in knowledge, but seems somewhat stronger than the evidence presented for attitudes, perceived self-efficacy, and confidence; the wording could therefore distinguish these outcomes more cautiously. Please also ensure consistency between KAP, used and defined in the Introduction, and CAP, currently used in the Discussion.

A final proofreading would also be useful: “emotional abuse” is repeated in the introductory prevalence statement; “characteristics..” (line 283) contains duplicated punctuation; and “clinical and methodological heterogeneity” in Section 2.2.6 would be more appropriately expressed as “intervention-related and methodological heterogeneity.”

Please also perform a final proofreading and formatting check, as a few typographical and formatting artefacts remain in the revised manuscript (e.g., duplicated punctuation and residual asterisks): **Therefore, no meta-analysis… (line 259); …studies included (**n = 10) (line 263); (e.g., Kantipudi & Chuemchit, 2024) **and RCTs (line 274).

With Best Regards,

Author Response

Thanks for carefully reviewing the manuscript and addressing the comments from the previous review. The revised version has substantially improved in methodological transparency and interpretation of evidence. In particular, the corrected and expanded PRISMA flow diagram now provides the numbers and reasons for exclusions at the full-text stage; the quality assessment procedure has been clarified without relying on unsupported cut-off scores; and the new subsection on results synthesis better explains the logic and organization of the narrative synthesis. The distinction between proximal outcomes (knowledge, attitudes, perceived self-efficacy, and confidence) and actual screening/reporting behavior is also much clearer throughout the revised manuscript.

Response: We sincerely thank the reviewer for this positive assessment of the revised manuscript and for recognizing the improvements made in response to the previous comments.

The extended discussion on self-reported outcomes, intervention heterogeneity, publication/language bias, and limited geographic representation further improves the interpretation of the findings. I only have a few leftover points that could benefit from correction before publication. In the abstract, the phrase 'consistent evidence' seems well-supported for improving knowledge, but it seems a bit stronger than the evidence presented on attitudes, perceived self-efficacy, and confidence; therefore, the wording could differentiate these results more cautiously.

Response: Thank you for this helpful suggestion. We have revised the abstract to better reflect the strength of the evidence across the different outcomes, distinguishing the consistent improvements in knowledge from the more variable findings regarding attitudes, perceived self-efficacy, and confidence.

Please also ensure consistency between KAP, used and defined in the Introduction, and CAP, currently used in the Discussion.

It would also be helpful to make one last correction: 'emotional abuse' is repeated in the introductory prevalence statement; 'Characteristics...' (line 283) contains duplicate punctuation; and 'clinical and methodological heterogeneity' in Section 2.2.6 would be more appropriately expressed as 'intervention-related and methodological heterogeneity.'

Please also do a final check for corrections and formatting, as there are still some typographical and formatting artifacts in the revised manuscript (for example, duplicate punctuation and leftover asterisks): **Therefore, there is no meta-analysis... (line 259); … the studies included (**n = 10) (line 263); (for example, Kantipudi & Chuemchit, 2024) **and RCT (line 274).

 

Best regards,

 

Response: Thank you for these helpful suggestions. We have revised the manuscript to ensure terminological consistency, correct the wording identified, and address the typographical and formatting issues throughout the text. We have also conducted a final proofreading to improve the overall clarity and presentation of the manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

The revised manuscript has improved substantially, and the authors have addressed most of the concerns raised in the previous review in a careful and constructive manner. In particular, the clarification of the review scope, the revised reporting of the JBI appraisal, the explicit description of the narrative synthesis, the distinction between proximal outcomes and actual detection/reporting behaviour, the positioning of student-related findings as secondary outcomes, and the expanded discussion of limitations have strengthened the methodological transparency and interpretation of the review.

Two issues nevertheless require attention before the manuscript can be considered for acceptance.

First, the comprehensiveness of the search strategy remains a methodological concern. As noted in the previous review, terms such as “child maltreatment,” “sexual abuse,” and “neglect” were not systematically included across the database searches. The revised manuscript acknowledges this limitation, but the justification that broader descriptors such as “child victimization” and “reporting child abuse” conceptually encompass these terms does not fully address the possibility that relevant studies indexed under more specific terminology may have been missed. The authors should therefore either update the search using a broader and more sensitive combination of controlled vocabulary and free-text terms related to child maltreatment, child sexual abuse, neglect, detection, recognition, identification, safeguarding, and reporting, or provide a sufficiently robust methodological justification for retaining the original strategy. If the search is updated, the authors should report transparently whether additional eligible studies were identified and revise the PRISMA flow diagram and synthesis if necessary.

Second, the PROSPERO registration number should be provided in the final manuscript. The current version still displays the registration as “ID: XXXXXXXX”. This should be replaced with the actual registration identifier, and any deviations from the registered protocol, if applicable, should be clearly disclosed.

Overall, the manuscript is considerably stronger than in the previous round. The principal methodological and interpretative concerns have been satisfactorily addressed, and the conclusions are now more appropriately calibrated to the available evidence. Provided that the remaining issues concerning the search strategy and PROSPERO registration are adequately resolved, the manuscript would be suitable for publication.

Author Response

The revised manuscript has improved substantially, and the authors have addressed most of the concerns raised in the previous review in a careful and constructive way. In particular, clarifying the scope of the review, revising the JBI assessment, explicitly describing the narrative synthesis, distinguishing between proximal outcomes and actual detection/reporting behavior, positioning findings related to the student as secondary outcomes, and expanding the discussion on limitations have strengthened the methodological transparency and interpretation of the review.

Response: We sincerely thank the reviewer for this positive assessment and for recognizing the improvements made to the manuscript in response to the previous comments.

However, two issues need attention before the manuscript can be considered for acceptance.

First of all, the thoroughness of the search strategy remains a methodological concern. As noted in the previous review, terms like 'child maltreatment,' 'sexual abuse,' and 'neglect' were not systematically included in the database searches. The revised manuscript acknowledges this limitation, but the justification that broader descriptors like 'child victimization' and 'child abuse reporting' conceptually cover these terms does not fully address the possibility that relevant studies indexed under more specific terminology may have been overlooked. Therefore, the authors should update the search using a broader and more sensitive combination of controlled vocabulary and free-text terms related to child maltreatment, child sexual abuse, neglect, detection, recognition, identification, protection, and reporting, or provide a sufficiently solid methodological justification for keeping the original strategy. If the search is updated, the authors should transparently report whether additional eligible studies have been identified and update the PRISMA flowchart and synthesis if necessary.

Nevertheless, we acknowledge that the absence of these specific terms from the search strings may have limited the retrieval of additional studies indexed under terminology not captured by the original strategy. We have clarified the rationale for the search strategy and acknowledged this potential limitation in the revised manuscript.

Response: Thank you for raising this methodological concern. The search strategy was designed to capture the broad construct of child victimization, rather than focusing exclusively on a specific form of victimization, in line with the original scope of the review. Although specific terms relating to child maltreatment, sexual abuse, and neglect were not systematically included in all database search strategies, an examination of the titles and keywords of the retrieved records showed that the search had identified studies using specific terminology related to child sexual abuse (CSA). For example, the term “child sexual abuse” appeared in the titles and keywords of almost all included studies, alongside other relevant terms identified across the records, such as “child maltreatment,” “prevention,” “neglect,” “child neglect,” “child abuse,” “report child abuse and neglect,” “preventing sexual abuse,” and “child sexual abuse prevention.” On this basis, we considered that the original strategy had captured relevant literature within the scope of the review and retained it.

Nevertheless, we acknowledge that the absence of these specific terms from the search strings may have limited the retrieval of additional studies indexed under terminology not captured by the original strategy. We have clarified the rationale for the search strategy and acknowledged this potential limitation in the revised manuscript.

Secondly, the PROSPERO registration number must be indicated in the final manuscript. The current version still shows the registration as "ID: XXXXXXXX." This should be replaced with the actual registration identifier, and any deviations from the registered protocol, if applicable, must be clearly disclosed.

Response: Thank you for this comment. The registration number was anonymized in the previous version to preserve the anonymity of the authors during peer review. We have now included the actual PROSPERO registration number in the manuscript. Any deviations from the registered protocol have been disclosed, where applicable.

Overall, the manuscript is considerably stronger than in the previous round. The main methodological and interpretative concerns have been satisfactorily addressed, and the conclusions are now more appropriately aligned with the available evidence. As long as the remaining issues regarding the search strategy and PROSPERO registration are properly resolved, the manuscript would be suitable for publication.

Response: We thank the reviewer for this encouraging assessment.