Factors Influencing the Implementation of Intimate Partner Violence Screening by Perinatal Obstetric Healthcare Providers in China: A Social–Ecological Perspective

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors of “Perinatal Obstetric Providers’ Perspectives on Implementing Intimate Partner Violence Screening: A Social Ecological Perspective” interviewed 20 obstetricians at a tertiary hospital in Suzhou and organized their findings using Bronfenbrenner’s social-ecological model. The paper elicits several important provider observations about screening: they regard regular prenatal attendance as evidence against abuse; physicians delegate screening to nurses as a “non-medical” task; and provider safety concerns deter screening. The ecological framework fits the question. The manuscript has issues that should be addressed before publication, which I will describe below.

Major Comments

1. The method description is internally inconsistent. The paper identifies the approach as directed content analysis (Elo & Kyngäs) — a deductive method that codes to a pre-existing framework — but then describes open, axial, and selective coding, which belong to grounded theory. If the ecological model drove coding from the top down, the grounded-theory language should come out. If coding was more inductive with the framework applied afterward, the write-up should say so.
2. Inter-rater agreement procedures should be elaborated. Two coders and a third adjudicator are mentioned, but there is no inter-rater agreement, member checking, audit trail, or negative-case analysis. A short section addressing these would help.
3. Interpretive statements appear in Results. Lines such as “This systemic gap fundamentally diminishes their initial motivation to conduct screening” should be in the discussion. Moving such asides out of the results would sharpen both sections.
4. Generalizability claims should be narrow. Twenty providers from one tertiary hospital in Suzhou, an affluent coastal city, support some cultural and policy claims but not others. I wondered whether statements about “Chinese cultural education,” “China’s traditional patriarchal social structure,” and national enforcement gaps in the Anti-Domestic Violence Law go beyond what a single site can support. I leave it to the authors to decide whether they believe such a large, diverse country could have such singular, narrow elements. The limitations mentions this in one sentence, but the discussion still speaks in broad, national terms. Softening those claims to “consistent with prior literature on” framing would help.
5. Participant characteristics need more detail. Table 1 omits sex, which matters given the paper’s arguments about gendered dynamics and its central finding about physicians. The sex breakdown of the 20 participants, particularly of the 8 physicians, would help. Please also clarify whether midwives were included and how the purposive sampling matrix was constructed.
6. Quote density varies across themes. Some subthemes receive a single quote from one participant; the “Perception of Safety Risks” section cites only N17. A participant-count column in Table 2, or a note about how widely each theme was endorsed, would let readers weigh the findings.
7. A quantification inconsistency in results. The text reports “12 out of the 18 respondents (66.7%)” but the sample is 20. Beyond the arithmetic, reporting percentages in descriptive qualitative work is a choice worth thinking through, since it is not as meaningful as it is in quantitative analyses.
8. The introduction frames IPV as mostly physical and sexual violence. The sentence, “The health impacts of IPV on victims are profound, extending beyond immediate physical trauma to include psychological sequelae...” implies that all abuse is physical or sexual. The statement logically does not apply to emotional IPV, which would not have direct physical impacts. Several of the paper’s findings — provider skepticism about whether non-physical behaviors count as IPV, participant quotes about women misreading coercive control as “common marital conflicts” — also act as if emotional IPV is not IPV. The authors certainly don’t want to imply this, and should ensure that both the introduction and discussion do not imply that, even through omission.
9. The Introduction does not engage the screening-measurement literature. Participants raise measurement questions directly: whether a single cross-sectional screen is adequate, whether pregnancy produces false positives, whether detection is accurate. The psychometric literature on IPV screeners speaks to these questions. Screeners developed against ICD-11/DSM-5 criteria, with sensitivity and specificity above .90 in general-population samples, exist (see Heyman et al., 2021 for this work and a table that summarizes qualities of IPV screeners). Orienting readers to that literature in the Introduction and connecting participant concerns to what is known about screener performance would make the paper stronger.
10. The “ideal victim” finding could be developed further. Providers inferring that women attending prenatal visits cannot be experiencing serious IPV is an important cognitive bias. Regular prenatal attendance may be one of the few safe, routine contacts available to an woman experiencing IPV. The paper approaches this point before retreating into “narrowly utilitarian understanding of screening value.”
11. Conclusions go beyond the evidence. Recommendations for nationwide changes to medical education, cultural attitudes, and policy require broader grounding than a 20-provider single-site study supplies. The authors should consider scaling recommendations to the evidence, while framing larger changes as directions for further research.

Minor Comments

1. Several typos should be caught before resubmission: “revolted around” for revolved around, “andscreen”, “stable adn thus”, “Limites”, “invention” for intervention in the conclusion. The highlights have odd commas.
2. “Social ecosystem theory” is used throughout; social-ecological theory or Bronfenbrenner’s ecological systems theory would be more conventional.
3. Saturation is asserted but not demonstrated. A sentence noting at what interview number new themes stopped emerging would help.

 

 

References

Heyman, R. E., Baucom, K. J. W., Xu, S., Slep, A. M. S., Snarr, J. D., Foran, H. M., Lorber, M. F., Wojda, A. K., & Linkh, D. J. (2021). High sensitivity and specificity screening for clinically significant intimate partner violence. Journal of Family Psychology, 35(1), 80–91. https://doi.org/10.1037/fam0000781

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript addresses a highly important and sensitive public health issue—intimate partner violence (IPV) screening during pregnancy—through the perspectives of obstetric healthcare providers in China. The use of a social-ecological framework to explore multilevel determinants is appropriate and conceptually strong. The study contributes valuable contextual insights into barriers and facilitators of IPV screening within maternal healthcare systems.

However, while the topic is significant and the study has clear merit, there are methodological, analytical, and reporting limitations that reduce the rigor and clarity of the manuscript in its current form. Substantial revisions are required to enhance transparency, strengthen methodological justification, and align interpretations with the presented data.

Major Strengths

1. Relevance and Significance

The manuscript addresses a critical gap in maternal healthcare, particularly in low-implementation settings such as China, where IPV screening is not routinely standardized . The focus on perinatal care is especially important given the heightened vulnerability during pregnancy.

2. Appropriate Theoretical Framework

The application of Bronfenbrenner’s social-ecological theory provides a strong conceptual structure to examine multilevel influences (micro-, meso-, exo-, and macro-systems) . This enhances the depth and interpretability of findings.

3. Rich Qualitative Data

The manuscript includes detailed participant quotations that effectively illustrate key themes such as:

• Low self-efficacy
• Role ambiguity
• Cultural stigma
• System-level barriers

These enhance the credibility and contextual richness of the findings.

4. Clear Identification of Multilevel Barriers

The categorization into:

• Individual competence
• Interdisciplinary gaps
• Organizational constraints
• Sociocultural barriers

is coherent and logically presented (see Table 2, page 7) .

Major Concerns

1. Methodological Transparency and Rigor

a. Limited Detail on Reflexivity

The manuscript does not adequately address researcher reflexivity, despite interviewers being embedded in clinical settings (e.g., one interviewer was a nursing student) . This raises concerns about:

• Power dynamics
• Social desirability bias
• Interpretive influence

➡️ Recommendation: Include a reflexivity statement discussing researcher positioning and its impact on data collection and interpretation.

b. Directed Content Analysis Requires Stronger Justification

The study uses directed content analysis guided by social-ecological theory , but:

• It is unclear how inductive vs. deductive coding were balanced.
• Risk of confirmation bias is not addressed.

➡️ Recommendation: Clarify:

• How new categories emerged beyond the framework
• How bias toward predefined categories was minimized

c. Sampling Limitations

The study is conducted in a single tertiary hospital (Suzhou) with 20 participants .

➡️ Issues:

• Limited generalizability
• Lack of diversity in institutional contexts

➡️ Recommendation: Explicitly acknowledge transferability limitations and justify sample adequacy beyond “data saturation.”

2. Insufficient Depth in Data Analysis

Although themes are identified, the analysis remains largely descriptive rather than interpretive.

For example:

• The “attitude–practice gap” is described but not theoretically explored.
• Cultural norms are reported but not deeply analyzed in relation to power structures or gender theory.

➡️ Recommendation: Strengthen interpretive depth by:

• Linking findings more critically to existing literature
• Expanding conceptual insights beyond categorization

3. Overextension of Conclusions

The manuscript proposes system-level interventions and policy implications, including:

• Curriculum reform
• Institutional restructuring
• Legal enforcement improvements

However:

• These recommendations are not directly derived from empirical data alone
• They reflect author interpretation rather than participant consensus

➡️ Recommendation:
Reframe conclusions to distinguish:

• Empirical findings vs. author-derived implications

4. Integration of Qualitative Evidence Needs Clarification

While participant quotes are included, the manuscript does not clearly explain:

• How quotes were selected
• Whether translation procedures were used
• How representative they are of the dataset

➡️ Recommendation:
Add a brief section on:

• Quote selection criteria
• Translation/validation (if applicable)

5. COREQ Reporting is Incomplete

Although COREQ is mentioned , several key elements are missing or underreported:

• Researcher characteristics
• Interview context (privacy, setting influence)
• Data saturation justification
• Coding audit trail

➡️ Recommendation: Ensure full adherence to COREQ checklist.

Minor Concerns

1. Language and Clarity

• Multiple grammatical issues (e.g., repetition: “Inclusion criteria comprised…” appears twice)
• Inconsistent phrasing and punctuation throughout

➡️ Requires professional language editing

2. Terminology Consistency

• “Social ecosystem theory” vs. “social-ecological theory” used interchangeably
  ➡️ Standardize terminology

3. Table Presentation

• Table 2 contains formatting inconsistencies and typographical errors (e.g., “Facotors,” “Limites”)

4. Ethical Reporting

• Ethics approval is mentioned but lacks:
  • Approval number
  • Institutional name
    ➡️ Needs completion for publication standards

Section-Specific Feedback

Abstract

• Generally clear but contains minor language errors
• Recommendation: tighten wording and avoid redundancy

Introduction

• Strong background, but somewhat overly descriptive
• Could be shortened and focused more clearly on the research gap

Methods

• Needs significant improvement in:
  • Transparency
  • Rigor
  • Justification of analytical choices

Results

• Well structured by ecological levels
• However:
  • Some overlap between themes
  • Limited analytical depth

Discussion

• Insightful but:
  • Occasionally speculative
  • Not always tightly linked to findings

Conclusion

• Appropriate direction, but:
  • Claims should be moderated

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

COMMENT 1: The paper provides a much needed socio-ecological perspective on IPV screening, with the introduction to the study adequately addressing contemporary literature on the topic and with a compelling rationale for the study being provided. Well done!  I did however feel that the authors could have expanded a bit on the specific eco-systemic framework employed in the study. Was it Bronfenbrenner’s or someone elses?

COMMENT 2 (line 186): “revolted” should be “revolved”

COMMENT 3: Discussion section: This section is well handled and adequately addresses the implications of the study findings. However, I believe that the discussion could be enhanced if the authors were to more comprehensively address study implications for future research.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The authors explored the factors hindering the implementation of intimate partner violence (IPV) screening during pregnancy within the framework of social-ecological theory by conducting semi-structured interviews with 20 obstetric healthcare professionals (8 physicians and 12 nurses) working at a tertiary hospital in Suzhou, China. The topic is clinically important; the findings—particularly the paradox of “awareness without competence,” physicians’ tendency to narrow their role boundaries, intra-institutional safety concerns, and the axis of cultural shame—offer meaningful contributions within the Chinese context. The use of COREQ, the resolution of disagreements between two independent coders by a third researcher, and the attention paid to participant diversity are strengths of the study.

Nevertheless, I must note that the study raises serious concerns regarding the consistency of its theoretical framework, methodological transparency, reference accuracy, and the proportionality of its interpretations to the supporting evidence. Unless some of these concerns are addressed, it is difficult for me to consider the manuscript ready for publication. My section-by-section evaluation is presented below.

Introduction

1. Zhang Kudon et al., 2024 (line 79). This source, which is cited for CDC NISVS data, is not included in the reference list.
2. Multiple different “Li et al., 2024” studies are cited throughout the Introduction and Discussion (e.g., lines 73, 93, 105, 107, 721). In the reference list, however, I could not identify any entry other than Quanlei Li that could correspond to “Li et al., 2024.” This means that either different statements are being supported by citation to the same source, or multiple distinct studies are being cited under the same label. In either case, clarification is required. Please make it possible to distinguish which citation refers to which Li study (e.g., Li QL, 2024 / Li JY, 2024). Bronfenbrenner’s ecological systems theory (micro-, meso-, exo-, macrosystem) and McLeroy et al.’s (1988) ecological model of health promotion (intrapersonal, interpersonal, institutional, community, public policy) are different theoretical constructs. In the manuscript, however, both are cited simultaneously without distinction. Yet in the Results section (Table 2, lines 256–257), Bronfenbrenner’s labels are used, whereas the content appears closer to McLeroy’s levels: under “Microsystem,” individual HCP factors are placed (whereas in Bronfenbrenner’s model the microsystem refers to the person’s immediate environments such as family, work, or school, not individual cognitive characteristics); under “Mesosystem,” communication between departments is included (whereas in Bronfenbrenner’s model the mesosystem refers to interactions between two microsystems); under “Exosystem,” hospital-level factors are listed (whereas in Bronfenbrenner’s model the exosystem refers to settings in which the individual is not directly involved but is nevertheless affected by—whereas the hospital may be a microsystem for staff and an exosystem for patients; this conceptual ambiguity is not explained). My recommendation is as follows: either adhere solely to Bronfenbrenner and operationalize the levels in accordance with his original definitions, or adopt McLeroy’s model and revise the labels accordingly (“intrapersonal, interpersonal, institutional/organizational, community, policy”). The current hybridized mixture of the two models is not theoretically defensible.

Methods
3. Inconsistency in the inclusion/exclusion criteria (lines 163–167 and Table 1). The exclusion criteria state “physicians and nursing staff in training status”; however, Table 1 includes 4 resident doctors. In the Chinese specialist training structure, could “resident doctor” indicate training status? If so, there is a contradiction; if not, the definition of “in training status” should be clarified (e.g., trainee nurses/healthcare students in standardized training programs vs. physicians in postgraduate specialty training).

4. Gynecology-related ambiguity (line 164). The manuscript states, “Worked in the clinical position of obstetrics and gynecology”—yet the study concerns obstetric care/IPV screening during pregnancy. Were any participants recruited from gynecology outpatient clinics? The eligibility criteria appear misaligned with the study design. If only personnel providing obstetric care were included, this should be stated explicitly.

Discussion
5. Zhang et al., 2024 (line 664, cited twice); Lam et al., 2020 (line 699); Madsen & Mikkelsen, 2024 (line 779). These three references are not included in the bibliography. A full, end-to-end review of reference-citation consistency is essential.

6. (lines 451–482, 744–770). Expressions such as “Traditional Chinese patriarchal structure,” “Chinese cultural education has long emphasized that women should learn to endure,” and “husband is paramount”—particularly when supported by participant quotations such as N2, N5, and N7—risk portraying Chinese culture as being uniquely conducive to IPV. This may attract criticism both academically and from an international readership. Adding a comparative perspective—namely, that similar norms of family privacy and patriarchal structures have been documented across many societies—would be both more accurate and more balanced.
7. An important rationale for perinatal IPV screening is its impact on neonatal and early childhood outcomes: low birth weight, prematurity, shorter breastfeeding duration, impaired mother-infant bonding, and behavioral/emotional problems in the developing child. These are briefly mentioned in lines 93–97, but the citations to Doswell et al. (2025) and Goldstein et al. (2025) rely only on a single pilot study and one review; this level of evidence is not sufficient to support claims of such weight. Moreover, these child-related outcomes are weakly integrated as a rationale for intervention—the value that obstetric healthcare professionals attribute to screening remains focused primarily on maternal outcomes (particularly postpartum depression). The relationship between IPV screening and neonatologists, pediatricians, and child health stakeholders is not addressed at all in the Discussion, although this could in fact be considered part of the mesosystem level under the heading of “interdisciplinary collaboration.” Addressing this gap would strengthen the manuscript’s clinical relevance.
8. Various spelling and typographical errors are evident throughout the text (e.g., line 61 “invention” → “intervention”; line 186 “revolted” → “revolved”; line 326 “adn” → “and”; line 404 “andscreen” → “and screen”; line 675 “Deterrrent” → “Deterrent”). In addition, Chinese punctuation marks (、, （ ）) are used within English sentence structure, and double periods (..) appear in multiple places. Given the frequency and variety of these errors, the manuscript should undergo a comprehensive language review and correction by a professional English-language editor.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 5 Report

Comments and Suggestions for Authors

-After reading this article, I suggest the authors reconsider the title of this article to reflect the primary aim of the study. SUGGESTION: factors influencing healthcare professionals’ engagement in intimate partner violence screening in ...(insert the study context): A social Ecological Perspective.

-In line 44, delete .....directed content analysis for data analysis, it is already covered in line 48.

-Introduction:
The first statement should indicate that IPV is one of the serious public health issues that endangers.................

-The gaps are clearly articulated in lines 101-114, running through to line 123.

-In line 124, the authors writes that 'Research indicates ..' Please contextualize this. Research where?

-In the study design, please delete information on data collection and analysis. Describe the design and how it helped achieve the objectives of this study.

• The basic tenets of the theory have been adequately utilized in analyzing the data and this spills over to the discussion section.
• Generally, the article is well written just minor issues that the authors can relook at

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

            I appreciate the thoroughness of the author’s revision of this manuscript. Two important issues remain.

            First, the response on methodological rigor mentions the absence of inter-rater agreement, member checking, an audit trail, and negative-case analysis. Of these, the lack of inter-rater agreement on a sample of the coding is the check that research consumers look for to gauge the amount of noise in the coding. Methodological guidance for coded semistructured interview transcripts emphasizes establishing intercoder reliability or agreement in the development and application of a coding scheme, and reporting standards for qualitative interview studies expect transparency regarding the number of data coders.

            First, the response on methodological rigor mentions the absence of inter-rater agreement, member checking, an audit trail, and negative-case analysis. Of these, the lack of inter-rater agreement on a sample of the coding is the check that research consumers look for to gauge the amount of noise in the coding. Methodological guidance for coded semistructured interview transcripts emphasizes establishing intercoder agreement, as do reporting standards for qualitative coding (Campbell et al., 2013; Halpin et al., 2025; Tong et al., 2007). Because the analysis has already been completed, I am not suggesting that the authors redo the entire coding process. Instead, I strongly suggest that they have a second coder independently code a subsample of transcripts using the existing codebook, report the resulting inter-rater/intercoder agreement, and briefly describe how discrepancies were handled.

            Second, the new Introduction paragraph on screener performance stands somewhat apart from the participants’ own doubts about detection. Participants questioned whether antenatal screening yields false positives and whether a single cross-sectional screen suffices; the screener literature addresses both concerns. IPV screeners carry imperfect specificity, and at the low base rates typical of routine antenatal care, the positive predictive value falls, and a notable share of positive screens prove false (Feltner et al., 2018; Rabin et al., 2009; Zapata-Calvente et al., 2022). BUT THAT IS THE POINT OF TWO-STAGE SCREENING! Screening in such settings should always be followed by an interview, as the Heyman article on specificity and sensitivity the authors cited indicates. Many women disclose only when asked again and repeated assessment across pregnancy detects substantially more cases than a single screen. This is why major guidelines advise multiple screenings (i.e., first prenatal visit, each trimester, postpartum; American College of Obstetricians and Gynecologists, 2012; McFarlane et al., 1992; U.S. Preventive Services Task Force, 2018). A sentence or two in the Discussion linking the participants’ doubts to these findings would strengthen the paper.

Author Response

Please see the response letter.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Accept in present form

Author Response

Response to Reviewer 2:

We sincerely thank Reviewer 2 for the positive evaluation and the recommendation to “Accept in present form.” We also note that the reviewer confirmed that all figures and tables are clear and well‑presented.

We are very encouraged by this recognition and appreciate the reviewer’s time and constructive assessment. No specific revisions were requested by this reviewer; all changes made in the revised manuscript were in response to the comments from the other reviewers.

We are grateful for Reviewer 2’s support and valuable feedback.

Reviewer 4 Report

Comments and Suggestions for Authors

I have carefully evaluated the authors’ responses to the reviewer comments and the corresponding revisions made in the manuscript. The authors have addressed the previously raised concerns in a clear and satisfactory manner. The revisions have significantly improved the overall quality, clarity, and scientific rigor of the paper.

In my assessment, the manuscript is now suitable for publication in its current form.

Author Response

We sincerely thank Reviewer 4 for the thorough re‑evaluation of our revised manuscript and for the very positive assessment.

We are especially encouraged by the reviewer’s comment that our responses have “addressed the previously raised concerns in a clear and satisfactory manner” and that the revisions have “significantly improved the overall quality, clarity, and scientific rigor of the paper.”

We also note that Reviewer 4 has recommended acceptance in the current form and did not raise any further revision requests. All of our revisions were made in response to the comments from the other reviewers, and we are grateful that Reviewer 4 has recognized our efforts.

We greatly appreciate Reviewer 4’s time, expertise, and constructive feedback throughout the review process.