Evaluating the Effects of a Mobile Obstetric Emergency System on Healthcare Providers’ Communication and Relationships in Bong County, Liberia

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1) Brief summary (one short paragraph)

This manuscript evaluates how the Mobile Obstetric Referral Emergency System (MORES)—a WhatsApp-based referral platform—affects inter‑facility communication and professional relationships among providers in Bong County, Liberia, using a mixed‑methods follow‑up (surveys + interviews/FGDs) one year after implementation. The main contributions are (i) quantitative evidence of high perceived trust/teamwork and (ii) qualitative themes showing improvements in communication frequency, timeliness, accuracy, and problem‑solving, alongside relationship qualities such as shared goals, mutual respect, and shared knowledge. Strengths include the mixed‑methods approach, the clear practical focus on referral coordination in low-resource settings, and anchoring interpretation in a relational communication/coordination framework.

2) General observations

Strengths

• The topic is highly relevant (maternal referral delays; bidirectional communication), and the overall narrative from rationale → framework → methods → thematic findings is coherent.

Opportunities to improve

• Counterfactual / causal inference: The lack of a counterfactual (no pre-intervention provider perceptions in this paper and no concurrent control) means conclusions should be framed as perceived associations rather than “effects.” Causal language such as “MORES strengthened…” / “improved…” should be moderated throughout.
• Reference list consistency: Most references appear recent (notably 2022–2025), which is appropriate; however, there is a duplicated reference entry (the Mozambique delays paper appears twice; listed as items 3 and 16). Please de-duplicate and ensure citation numbering maps uniquely.
• Self-citations: Ensure self-citations (if any) are proportionate and clearly justified (this is best checked quantitatively in the final revision pass).

3) Chapter by chapter observations

1. A) Introduction

Strengths

• Provides strong context on maternal mortality burden and referral delays, and explains the operational gap in paper referrals and feedback loops.
• Clearly situates MORES within prior pilot work and identifies a specific knowledge gap: impact on trust/collaboration across facilities.

Opportunities to improve

• Tighten the rationale for why “trust/teamwork” is the primary outcome here and how it is expected to link to clinical outcomes (a short causal pathway would help).
• Resolve theoretical naming inconsistency and operationalization: The manuscript uses the label “relational communication theory,” yet the constructs and citations align with relational coordination theory. Use one term consistently and state explicitly how the framework operationalized analysis (i.e., which constructs mapped to which codes/measures), and ensure the conceptual framing aligns with how themes are presented later (including how Figure 1 relates to the framework).

1. B) Methodology (Materials and Methods)

Strengths

• Mixed methods approach with explicit setting and participant eligibility; inclusion of both RHF and hospital provider perspectives strengthens triangulation.
• Qualitative analysis process is described (iterative codebook, consensus meetings among three researchers), supporting credibility.

Opportunities to improve

• Specify mixed methods design and integration: Clearly state whether this is convergent vs sequential, and add a short “Mixed methods integration” description explaining how survey and qualitative findings were triangulated (a joint display plan is ideal).
• Clarify and correct quantitative analysis reporting (line ~164; Tables 1–2): You state “descriptive statistics and chi-square tests” (line ~164), yet Tables 1–2 report means (SD) with p-values consistent with comparisons of continuous/ordinal outcomes (often requiring t-tests/ANOVA or non-parametric alternatives rather than chi-square). Please specify which tests were used for each variable (continuous vs categorical), justify them, and correct the Methods text accordingly.
• Trust & Teamwork instrument completeness (lines ~125–139; Table 2): The scale is described as a 16-item tool (lines ~125–139), but Table 2 appears to report fewer items (≈10). Confirm whether Table 2 is a subset; if so, justify selection and provide the full scoring approach (item-level vs subscale totals), ideally with the complete item list in an appendix.
• Likert handling and robustness: Reporting means for Likert items can be acceptable, but please (i) acknowledge and quantify potential ceiling effects, and (ii) consider sensitivity reporting (median/IQR or ordinal methods) to support robustness.
• Multiple comparisons (Tables 1–2): With many p-values reported, either adjust for multiple comparisons or explicitly frame inferential tests as exploratory.
• Missing data handling (Table 1; lines ~187–191 and continuation on page 6): Clarify how “prefer not to answer” and other missing responses were treated (pairwise vs listwise deletion; imputation if any), and whether any survey items were excluded.
• Sampling and participation bias (lines ~102–109; Table 1): Recruitment is described via flyers/word-of-mouth at RHFs/district hospitals (lines ~102–109). If available, include participation rates and any known differences between participants and non-participants. Also acknowledge earlier (not only later in Discussion) that the sample is dominated by nurses/midwives with very few physician assistants (Table 1), which limits generalizability across cadres.
• Qualitative rigor details (lines ~155–157; ~165–180):
• Individual interviews are stated as 10–15 minutes (lines ~155–157), which is short for relational constructs—justify adequacy or clarify that interviews were structured/focused with limited scope.
• Add the number of FGDs, their composition and group sizes, and how thematic saturation was evaluated.
• Reflexivity: specify who conducted interviews/FGDs, their relationship to MORES implementation, and what steps were taken to minimize desirability bias (important given highly positive findings).
• WhatsApp-specific governance for reproducibility (lines ~68–76): Because the intervention uses WhatsApp (lines ~68–76), add a brief methods-oriented description of privacy safeguards and governance (e.g., whether patient identifiers were avoided, device/account security expectations, and alignment with local policies), as this materially affects replicability and ethical interpretation.

1. C) Results

Strengths

• Clear presentation of demographic characteristics and a useful snapshot of digital readiness (smartphone/WhatsApp use).
• Qualitative results are well organized into themes/subthemes that align with the framework and include illustrative quotes.

Opportunities to improve

• Ceiling effects in Table 2 (lines ~196–204): All positively framed items average >4. Add an explicit statement acknowledging ceiling effects and what they imply for interpretability; consider including distributions and/or subscale totals to complement means.
• Instrument transparency (Table 2 vs lines ~125–139): Ensure Table 2 reflects the full Trust & Teamwork instrument or clearly label it as a subset and provide the complete instrument/scoring elsewhere (appendix/supplement).
• Quantitative interpretability: If inferential comparisons are retained, add minimal effect descriptors (e.g., mean differences with confidence intervals) rather than relying on p-values alone.
• Mixed methods linkage: Add a short paragraph explicitly showing where qualitative findings explain/nuance specific survey items (e.g., survey “respect/trust” ↔ qualitative “mutual respect/shared goals”), ideally supported by a compact joint display.
• Figure 1 clarity (lines ~228–236): Ensure Figure 1 is legible and strengthen the caption so it explicitly maps the two themes to the theoretical framework and shows how they relate to survey constructs.

1. D) Discussion

Strengths

• Discussion appropriately links findings to prior MORES work and to broader evidence on referral communication interventions in sub-Saharan Africa.
• Includes a limitations subsection and acknowledges generalizability constraints and cadre imbalance.

Opportunities to improve

• Tone down causal claims: Align language with the study design (post-implementation perceptions without a control/comparator). Replace causal verbs (“improved,” “strengthened”) with perception-aligned phrasing (“participants reported…,” “was perceived as…”).
• Expand limitations where most salient: Bring forward and elaborate on (i) selection bias (flyer/word-of-mouth recruitment; engaged users more likely to participate), (ii) social desirability bias (especially with very positive results), (iii) ceiling effects in the survey, and (iv) short interview duration for individual interviews.
• Scalability claims (lines ~413–427): When suggesting scalability, specify enabling conditions required for scale-up (training, supervision, network connectivity, governance, device availability) and whether these conditions exist beyond Bong County.
• “Low cost and sustainable” claims (lines ~30–32; ~448–449): These claims appear in the Abstract conclusions and later in Conclusions. Either support them with data (even basic cost/workload indicators) or moderate phrasing to “perceived as feasible/acceptable” unless measured.
• WhatsApp governance in interpretation: Beyond Methods, discuss how WhatsApp privacy/data governance considerations shape feasibility, acceptability, and policy uptake (without needing to disclose sensitive implementation details).

1. E) Conclusions

Strengths

• Conclusions align with main themes (communication and relationship quality) and identify relevance for policymakers and implementers.

Opportunities to improve

• Temper or support claims about “low cost,” “sustainable,” and “highly feasible” unless cost/resource indicators were measured; otherwise phrase as “perceived as feasible/acceptable” (and keep this consistent with the Discussion framing).
• Add a forward-looking line on what future evaluations should include (comparative designs/counterfactuals, linkage to patient outcomes, cost-effectiveness, and privacy/governance assessments).

 

 

Comments on the Quality of English Language

The English is generally understandable, but there are phrasing/consistency issues (e.g., theory naming consistency; some wording/grammar) that could be tightened to express the research more clearly.

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

• Can the authors clarify how communication quality was objectively measured?
• Were any quantitative outcome measures used to validate provider satisfaction improvements?
• How does MORES differ from other referral communication systems discussed in the literature?
• Can the authors elaborate on the limitations affecting generalizability to other healthcare settings?
• Were there any barriers or challenges observed during MORES implementation?
• Can the authors provide more discussion on the long-term sustainability of the MORES intervention?
• Was there any difference in outcomes between rural and hospital providers?
• Can the authors explain whether training was given before implementing MORES?
• How was patient confidentiality maintained during communication exchange?
• Could the improvement in referral outcomes be influenced by factors other than MORES?
• Can the authors clarify how trust and teamwork scores were validated in this setting?
• Was any comparison made between facilities using MORES and those following conventional referral methods?

Comments on the Quality of English Language

English can be improved

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript addresses an important and timely issue concerning the functioning of referral systems in maternal care in resource-limited countries. The authors analyze the impact of implementing the MORES system, based on WhatsApp, on communication, collaboration, and relationships among medical staff at rural health facilities and hospitals in Liberia. The topic of this study falls within the scope of research on mHealth, the organization of maternal care, and improving the health security of mothers and newborns.

It is particularly noteworthy that the authors did not limit themselves solely to evaluating the effectiveness of the technology but also focused on interpersonal relationships, trust, and collaboration among healthcare workers. This is an aspect often overlooked in studies on digital healthcare support systems.

Despite the study’s significant value, there are several areas requiring improvement:

1. The results are presented almost exclusively in a positive light. It would be worthwhile to discuss more extensively the potential difficulties associated with implementing MORES, such as problems with internet access, device maintenance costs, communication overload for staff, issues regarding the privacy of medical data transmitted via WhatsApp, and the risk of the system’s dependence on employees’ personal phones.

2. The “Strengths and Limitations” section is relatively brief. The authors should discuss in greater detail the possibility of social desirability bias, the influence of participants’ involvement in the pilot project on positive responses, the lack of a control group, and the limited generalizability of the results.

3. In the statistical section, the authors limited themselves mainly to descriptive statistics and p-values. It would be worth considering reporting effect sizes, providing a more detailed justification for the choice of statistical tests, and including information on the reliability of the scale used in the current study (e.g., Cronbach’s alpha).

4. Since the system is based on WhatsApp, a more detailed discussion of patient data protection, communication confidentiality, anonymization procedures, and compliance with local regulations regarding medical data is necessary.

5. In the Discussion section, some of the arguments regarding improved communication and collaboration repeat previous findings. The discussion could be shortened and focused more on interpreting the significance of the results for clinical practice and health policy.

6. It would be beneficial to standardize the notation of abbreviations and the formatting of certain tables.

The manuscript presents a valuable and well-designed study on the use of mobile technologies to improve the referral system for obstetric care in resource-limited settings. The work is of both scientific and practical significance. The results indicate that the use of simple communication tools can improve collaboration between facilities and support continuity of patient care. Despite the noted limitations, the manuscript represents an interesting contribution to the literature on mHealth and obstetric care.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Overall, the manuscript addresses an important & timely topic, focusing on improving communication; Furthermore, a valuable collaboration among healthcare providers in low-resource setting through a mobile-based system. Overall the study is well organized, and the mixed-methods approach is appropriate for the research question. 

Overall, Strengths: 

- Quantitative & qualitative approaches for a more complete analysis
- Measurable outcomes alongside provider experiences
- Communication & professional relationships,  are often underexamined
- Deals with a topic of clear relevance to maternal care in resource-limited settings
- Provides practical insights that could support better coordination & patient care

Overall, improvements:

- Introduction: provides useful background but could more clearly explain the research gap and how this study adds to existing knowledge.
- Methods: clarify how participants were selected and provide more information on how the qualitative data were analyzed e.g. coding process
- Results: It may also help to better connect the qualitative findings with the quantitative results.
- Tables & Figures: please adding clearer titles and more detailed captions to make them easier to interpret on their own.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Overall, the revised manuscript represents a clear and substantial improvement over the initial submission. The authors have taken the previous comments seriously and have addressed the majority of the methodological, conceptual, and reporting concerns in a thoughtful and consistent manner. The paper now reads as more coherent, more transparent in its methods, and more appropriately cautious in its interpretation of findings.

One of the most notable improvements is the clarification of the theoretical framework. The earlier inconsistency between “relational communication theory” and “relational coordination theory” has been resolved, and the manuscript now consistently uses relational coordination theory. More importantly, this is no longer just a nominal correction—the authors have meaningfully integrated the framework into the analysis. The mapping of communication quality and relationship quality onto the constructs of the theory, and the way these are reflected in the qualitative themes and Figure 1, significantly strengthens the conceptual grounding of the study.

The methodological section has also improved considerably. The study design is now explicitly described as a mixed-methods convergent design, which directly responds to previous concerns about the lack of clarity regarding integration. The statistical analysis has been clarified in detail: the distinction between chi-square tests, independent t-tests, and Mann–Whitney U tests is now clearly articulated, and the inclusion of both mean (SD) and median (IQR) for Likert-scale data reflects a much more appropriate handling of ordinal variables. This substantially improves the scientific soundness and reproducibility of the quantitative component.

The reporting of the measurement tool has been strengthened as well. The inclusion of Cronbach’s alpha, along with a brief interpretation of its value in an exploratory context, adds transparency and allows the reader to better judge the reliability of the instrument. Similarly, the handling of missing data is now explicitly described, with “prefer not to answer” responses treated as missing and excluded using pairwise deletion. These additions significantly improve methodological clarity.

The qualitative component has also been meaningfully enhanced. A particularly important correction is the clarification of interview duration, now reported as 45–60 minutes, which addresses a major concern from the initial version regarding insufficient depth. The inclusion of bias-reduction strategies (such as creating a non-judgmental environment and using indirect questioning) further strengthens the credibility of the findings. While some elements could still be expanded—specifically the exact number and composition of focus group discussions and a more explicit statement on data saturation—the qualitative rigor is now acceptable and in line with expectations for this type of study.

Another strong improvement is the inclusion of data governance and privacy considerations related to the use of WhatsApp. The manuscript now explains that no patient-identifiable information was shared, that providers used ID-based communication, and that sensitive clinical data continued to be transmitted via paper records. This clarification is particularly important in the context of digital health research and significantly strengthens both the ethical transparency and the practical applicability of the study.

The results section has been improved in terms of interpretability. The inclusion of median and IQR alongside mean values, as well as the explicit acknowledgment of ceiling effects in the discussion, demonstrates a more nuanced understanding of the data. In addition, the integration between quantitative and qualitative findings is more evident, with survey results being used to support and contextualize qualitative themes. While a formal joint display is still absent, the narrative integration is now sufficiently developed.

Importantly, the tone of the manuscript—especially in the abstract, results, and discussion—has been appropriately moderated. The authors have moved away from causal language and now consistently frame findings as “perceived improvements” or “associations,” which is much more consistent with the study design. Overstatements related to cost, sustainability, and scalability have also been tempered, and the limitations section has been expanded to include key issues such as selection bias, ceiling effects, and the influence of external factors beyond the intervention itself.

The reference list has also been corrected, with the previously duplicated citation removed, and overall the references remain appropriate and relevant to the topic.

There are, however, a few minor points that could still be improved. The manuscript would benefit from a slightly more detailed description of the focus group discussions (number of groups, size, and composition), as well as an explicit statement regarding how thematic saturation was assessed. Additionally, while the English is overall clear and readable, there are small instances of repetition and minor phrasing inconsistencies that could be refined during editorial revision. Finally, while mixed-methods integration has improved, the inclusion of a simple joint display would further strengthen the transparency of the integration process, though this is not essential at this stage.

Taken together, these remaining issues are minor and do not affect the overall validity or contribution of the study. The manuscript is now methodologically sound, ethically transparent, and appropriately interpreted. It addresses a highly relevant topic in maternal health systems and provides practical insights into mHealth-enabled referral communication in low-resource settings.

In conclusion, the authors have successfully addressed the major concerns from the previous review, and no critical issues remain. I would therefore recommend acceptance of the manuscript, with only minor editorial polishing if deemed necessary by the journal.

Comments on the Quality of English Language

The English in the revised manuscript is clear, professional, and fully understandable, and it no longer interferes with the scientific content. Compared to the original version, there is a noticeable improvement in clarity, especially in the Methods and Discussion sections where terminology and phrasing are now more precise.

That said, there are still a few minor stylistic issues that could benefit from light editorial polishing:

• Repetition of key terms, particularly “perceived,” “associated,” and “improvement,” in the Results and Discussion sections. While conceptually appropriate (given the non-causal framing), this creates some stylistic redundancy.
• Occasional awkward phrasing or slight grammatical inconsistencies, typical of multi-author manuscripts (e.g., small article/preposition issues, minor sentence flow problems).
• Some sentences are slightly long or dense, particularly in the Discussion, and could be streamlined for readability.

Importantly, these are purely editorial-level issues:

• they do not affect comprehension,
• they do not impact scientific validity,
• and they can be easily corrected during copyediting.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Authors,

Thank you for the opportunity to re-evaluate the manuscript entitled “Evaluating the effects of a mobile obstetric emergency system on healthcare providers’ communication and relationships in Bong County, Liberia”.

The revised version of the manuscript has been substantially improved. The authors expanded the methodological description, supplemented the information regarding the research instrument and data protection, and justified the statistical analyses used. Particularly commendable is the organization of the qualitative findings based on relational coordination theory, as well as the better contextualization of the results within healthcare organization in low-resource settings.

The manuscript addresses an important and timely issue related to the organization of obstetric care and the use of simple mHealth tools to improve communication between healthcare facilities. The study makes a valuable contribution to the literature on digital interventions in healthcare systems in LMICs.

However, I believe that several minor issues should still be addressed before the manuscript can be accepted for publication.

1. The discussion concerning the implementation limitations of the MORES system remains rather limited. Despite the authors’ response to the reviewer’s previous comment, the manuscript itself still presents the intervention almost exclusively in a positive light. It would be worthwhile to add a brief paragraph referring to potential difficulties associated with the long-term implementation of the system, such as:
   • unstable internet access, dependence on healthcare workers’ personal phones, device maintenance and data transmission costs, potential communication overload among staff, and issues related to the sustainability of the system after the project completion.

2. The “Strengths and Limitations” section has been improved; however, it would still be valuable to more explicitly acknowledge the possibility of social desirability bias. Most participants were involved in the original MORES project, which may have contributed to a more favorable evaluation of the intervention.

3. The authors mention that the improvement in outcomes may also have resulted from factors other than MORES itself; however, the lack of a control group should be explicitly identified as a methodological limitation of the study.

4. Minor language and editorial issues requiring correction during the final proofreading stage still remain in the manuscript.

I believe that the manuscript presents valuable and interesting research findings and, after addressing the above minor revisions, will be suitable for publication.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf