Inequalities in Enrollment in Nepal’s National Health Insurance Program: An Intersectional Analysis of Nepal Demographic and Health Survey 2022

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The article evaluates the enrollment intensity of different population categories in Nepal in the national health insurance program. The idea of investigating this topic is timely and interesting from several perspectives, as the authors argue in the manuscript. The paper is appealing for its use of mathematical methods to examine relationships among components of the national health survey, which the authors treat as a secondary data source for calculations and to inform their recommendations for the national health system.

However, the manuscript has several serious drawbacks that do not allow it to be published in its current form:

1. In line 109, the authors mention “the key gap in the literature,” but I did not find a sufficiently comprehensive literature review earlier in the manuscript to substantiate this claim. Only a few studies are mentioned, without an in-depth analysis of existing research on the topic.
2. The authors should comment on the significant disparity in the number of men (4,913) and women (14,845) included in the analyzed survey data. How does this imbalance affect the results and conclusions of the study? Could this represent a serious limitation of the survey? What is the reason for such a large difference in the number of respondents by gender?
3. As I understand it, the manuscript relies on the analysis of an extensive survey document that covers multiple aspects. However, the section “Program and Policy Implications” lacks a clear description of concrete actions to address the identified inequalities. I also do not see any benchmarking analysis or references to countries that have successfully addressed similar issues.
4. The authors should thoroughly revise the manuscript to better demonstrate the research's novelty. In its current form, the study appears to be primarily an analysis of existing data using mathematical tools, without a clearly structured research framework or well-developed practical recommendations grounded in the calculated results and graphical analysis.

Author Response

Reviewer 1

The article evaluates the enrollment intensity of different population categories in Nepal in the national health insurance program. The idea of investigating this topic is timely and interesting from several perspectives, as the authors argue in the manuscript. The paper is appealing for its use of mathematical methods to examine relationships among components of the national health survey, which the authors treat as a secondary data source for calculations and to inform their recommendations for the national health system.

Response: We would like to thank you for your assessment.

However, the manuscript has several serious drawbacks that do not allow it to be published in its current form:

1. In line 109, the authors mention “the key gap in the literature,” but I did not find a sufficiently comprehensive literature review earlier in the manuscript to substantiate this claim. Only a few studies are mentioned, without an in-depth analysis of existing research on the topic.

Response: Thank you for your evaluation and feedback. We have nor revised and added the literature gap that was missing in the previous version of manuscript.

1. The authors should comment on the significant disparity in the number of men (4,913) and women (14,845) included in the analysed survey data. How does this imbalance affect the results and conclusions of the study? Could this represent a serious limitation of the survey? What is the reason for such a large difference in the number of respondents by gender?

Response:

Thank you for your important comment regarding the sample size difference between men (4,913) and women (14,845). This imbalance is due to the design of the Nepal Demographic and Health Survey (NDHS) 2022, from which our secondary data was extracted. In the NDHS, all eligible women aged 15-49 in selected households are interviewed, while men of the same age are interviewed in only half of the selected households. Therefore, the sample size for women is intentionally larger.

To ensure the analysis being robust and representative we have taken following steps.

1. Separate analysis: We conducted the analysis separately for men and women. This approach was chosen because gender was not a sampling stratum. The strata were based on province and urban–rural residence. Consequently, the sample imbalance is less likely to bias the results or conclusions when each group is analysed independently.
2. Application of sampling weight: Sampling weights, which are provided with the NDHS dataset to account for the complex survey design and non-response, were applied in the analysis. This ensures that the results are representative of the national population.

We acknowledge that the smaller sample size for men may slightly reduce the precision of the male-specific estimates compared to those for women. This clarification, along with the survey design details, has been incorporated into the methodology section. Further details on the survey methodology can also be found in the S1 File or through the DHS program website: https://dhsprogram.com/pubs/pdf/FR379/FR379.pdf.

 

1. As I understand it, the manuscript relies on the analysis of an extensive survey document that covers multiple aspects. However, the section “Program and Policy Implications” lacks a clear description of concrete actions to address the identified inequalities. I also do not see any benchmarking analysis or references to countries that have successfully addressed similar issues.

Response: Thank you for your critical observation. We have revised the program and policy implication section based on your critical feedback considering the references from different countries.

 

1. The authors should thoroughly revise the manuscript to better demonstrate the research's novelty. In its current form, the study appears to be primarily an analysis of existing data using mathematical tools, without a clearly structured research framework or well-developed practical recommendations grounded in the calculated results and graphical analysis.

Response: Thank you for your suggestions. We have CORRECTED accordingly.

Reviewer 2 Report

Comments and Suggestions for Authors

General comments :

This study is interesting for national health insurance  programs, especially in terms of understanding the inequality of  health insurance enrolment. The article is concise and clearly describes the inequality in health insurance enrolment. The study was written in the correct order and was in line with the title, aims, methods, and results. The study also reported the findings systematically. This article is also equipped with good tables and graphics.

However, I would like to make a few comments to improve the understanding of this paper by readers.  

In the Introduction

Please provide the relevance and importance of this study with respect to the health insurance policy  in Nepal, since the authors discuss it in theDiscussionn section.

In the Method

Why  did the authors use OR in the Cross-sectional study design to determine the association? If this study is not a case-control study, how can the odds of the two groups' comparison be determined? Please see at https://pmc.ncbi.nlm.nih.gov/articles/PMC7850067/

The authors have explained this as a limitation of this study, as the OR  used in this study cannot establish temporality or causality. Therefore, I suggest using the B Coefficient to detect the association instead of the OR as a risk factor value.

Please clarify this in the Results section.

In the Conclusion

Please make the conclusion  concise and short.  Plase focus on the answer for the aim of this study is to explore how multiple and overlapping inequities influence health insurance enrolment in Nepal and their impact on the national health program insurance.

Author Response

Reviewer 2

General comments:

This study is interesting for national health insurance programs, especially in terms of understanding the inequality of health insurance enrolment. The article is concise and clearly describes the inequality in health insurance enrolment. The study was written in the correct order and was in line with the title, aims, methods, and results. The study also reported the findings systematically. This article is also equipped with good tables and graphics.

Response: Thank you for your observation and evaluation.

However, I would like to make a few comments to improve the understanding of this paper by readers.  

In the Introduction

Please provide the relevance and importance of this study with respect to the health insurance policy in Nepal, since the authors discuss it in the discussion section.

Response: Thank you for your critical feedback. We have received similar feedback from other reviewers as well. Based on the feedback, we have revised the last paragraph of introduction section accordingly.  

 

In the Method

Why did the authors use OR in the Cross-sectional study design to determine the association? If this study is not a case-control study, how can the odds of the two groups' comparison be determined? Please see at https://pmc.ncbi.nlm.nih.gov/articles/PMC7850067/

The authors have explained this as a limitation of this study, as the OR used in this study cannot establish temporality or causality. Therefore, I suggest using the B Coefficient to detect the association instead of the OR as a risk factor value.

Please clarify this in the Results section.

Response: Thank you for this important comment. Odds ratios (ORs) estimated using logistic regression are commonly applied in cross-sectional studies to examine associations between an outcome and explanatory variables. Although ORs are often used in case–control studies, they are also appropriate for cross-sectional analyses when the outcome variable is binary. In this study, logistic regression was used to assess the association between explanatory variables and the outcome, and the OR represents the odds of the outcome occurring in one group compared with the reference group.

We agree that, due to the cross-sectional nature of the data, the OR cannot establish temporality or causality. Therefore, the results should be interpreted only as associations rather than causal relationships. While the regression coefficient (B) reflects the change in log odds, ORs are reported because they are easier to interpret and are commonly presented in epidemiological studies.

We have clarified this explanation in the Results and Limitations sections to avoid misinterpretation of the findings.

 

In the Conclusion

Please make the conclusion concise and short.  Please focus on the answer for the aim of this study is to explore how multiple and overlapping inequities influence health insurance enrolment in Nepal and their impact on the national health program insurance.

Response: Thank you for your critical feedback. We have revised and CORRECTED the conclusion accordingly.

Reviewer 3 Report

Comments and Suggestions for Authors

CLARIFICATION OF OBJECTIVE AND HYPOTHESIS

The main objective is clearly implied, but it would benefit from being explicitly stated in a concise objective statement inside the Introduction. Also explicit hypotheses to improve analytical clarity.

INTERSECTIONAL VARIABLE CONSTRUCTION

Clear explanation about dichotomization of wealth, education, and ethnicity:

• A justification for collapsing quintiles into binary categories.

• Sensitivity analysis using full quintiles.

• A brief theoretical explanation of why additive counting (triple/double/single disadvantage) appropriately operationalizes intersectionality.

ECONOMETRIC SPECIFICATION

The logistic regression is appropriate, and survey design corrections. However:

• Model fit statistics (e.g., pseudo R², goodness-of-fit test) are not reported.

• No robustness checks are presented.

• Potential omitted variable bias (e.g., health status, prior service utilization, disability status) is not discussed.

• Explanation about how this method is better than other one.

If logistic regression was used as better model, include additional diagnostics (discrimination tests as KS, confusion matrix, etc.), these would increase scientific rigor.

CONCENTRATION INDEX METHODOLOGY

• The mathematical formula of RCI should be included.

• It should be clarified whether correction for binary outcome (e.g., Wagstaff correction) was applied.

• Confidence interval estimation method should be specified.

INTERPRETATION OF URBAN-RURAL FINDINGS

The adjusted result showing lower odds among urban women requires more theoretical discussion, as it contradicts typical access patterns.

Comments on the Quality of English Language

The manuscript is readable but requires careful grammatical revision (e.g., agreement, article usage, phrasing inconsistencies).

Author Response

Reviewer 3

Comments and Suggestions for Authors

CLARIFICATION OF OBJECTIVE AND HYPOTHESIS

The main objective is clearly implied, but it would benefit from being explicitly stated in a concise objective statement inside the Introduction. Also explicit hypotheses to improve analytical clarity.

INTERSECTIONAL VARIABLE CONSTRUCTION

Clear explanation about dichotomization of wealth, education, and ethnicity:

• A justification for collapsing quintiles into binary categories.
• Sensitivity analysis using full quintiles.
• A brief theoretical explanation of why additive counting (triple/double/single disadvantage) appropriately operationalizes intersectionality.

Response: Thank you for this important comment. In this study, we constructed a composite disadvantage measure using education, wealth status, and ethnicity to examine intersectional inequalities. Wealth quintiles were collapsed into two categories (poorer vs. richer), education into lower vs. higher, and ethnicity into advantaged vs. disadvantaged groups. This dichotomization was primarily done to address small sample sizes when combining multiple dimensions of disadvantage and to ensure stable estimates in the intersectional categories.

 

To maintain transparency, the results using the original wealth quintiles and other uncategorized variables are presented in the supplementary files. These additional analyses show patterns consistent with the main findings.

 

The additive approach (single, double, and triple disadvantage) was used to operationalize intersectionality in a quantitative framework. This approach allows us to examine how overlapping social disadvantages may compound inequalities in health outcomes and is widely used in public health research when applying intersectionality with survey data.

 

We had already included some explanation in the limitation section; however, as suggested, we have now added clearer justification for these analytical choices. We also acknowledge in the limitations section that although this study applies an intersectionality framework using quantitative survey data, the concept of intersectionality originates from qualitative traditions, and this approach may not fully capture the complex lived experiences of multiple forms of marginalization. Further qualitative research is recommended to better understand these intersecting dimensions of disadvantage.

 

ECONOMETRIC SPECIFICATION

The logistic regression is appropriate, and survey design corrections. However:

• Model fit statistics (e.g., pseudo R², goodness-of-fit test) are not reported.
• No robustness checks are presented.
• Potential omitted variable bias (e.g., health status, prior service utilization, disability status) is not discussed.
• Explanation about how this method is better than other one.

If logistic regression was used as better model, include additional diagnostics (discrimination tests as KS, confusion matrix, etc.), these would increase scientific rigor.

Response: Thank you for these valuable comments. In addition to the multicollinearity test (VIF) already described in the methodology, we assessed model fit using STATA’s postestimation goodness-of-fit test, and we have now added this information to the methodology section.

Our main objective was to examine inequalities using wealth quintile, education, and caste/ethnicity as intersectional dimensions of marginalization; therefore, these were the primary predictors in the models. Other important variables, such as health status, prior service utilization, and disability, are reported in the supplementary tables.

We used logistic regression because the outcome is binary, and this method appropriately estimates odds ratios while accounting for the survey design. We acknowledge that additional diagnostics (e.g., KS statistics, confusion matrix) and robustness checks could further strengthen model assessment but given the study’s focus on intersectional inequalities and survey-based estimates, these were not performed.

This explanation has now been clarified in the methodology and limitations sections.

 

 

CONCENTRATION INDEX METHODOLOGY

• The mathematical formula of RCI should be included.
• It should be clarified whether correction for binary outcome (e.g., Wagstaff correction) was applied.
• Confidence interval estimation method should be specified.

Response:  Thank you for these important comments. We acknowledge that the original methodology did not include the mathematical formula for the Relative Concentration Index (RCI) or clarify whether corrections for binary outcomes were applied. Although RCI has been widely used in previous studies, based on your suggestion, we have now revised the methodology to use the Erreygers Normalized Concentration Index (ENCI), which is more appropriate for binary outcomes.

In the revised analysis, we report the standard errors, and the formula and calculation procedure for ENCI are now clearly described in the methodology section. This revision ensures that the concentration index estimates are statistically robust and interpretable for binary health outcomes.

These clarifications and methodological updates have been added to both the methods and results sections.

 

INTERPRETATION OF URBAN-RURAL FINDINGS

The adjusted result showing lower odds among urban women requires more theoretical discussion, as it contradicts typical access patterns.

Response:  Thank you for your feedback. The finding shows lower odds among rural women.  We have discussed this in the revised version.

Comments on the Quality of English Language

The manuscript is readable but requires careful grammatical revision (e.g., agreement, article usage, phrasing inconsistencies).

Response:  Thank you for your feedback. We have revised the grammar throughout the manuscript.

Reviewer 4 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review this manuscript. I read it with great interest, as the topic aligns closely with my area of interest. Overall, the paper addresses an important issue and offers valuable insights into inequalities in enrollment under the Nepal Health Insurance Program (NHIP). Below are my specific comments and suggestions.

Abstract

Lines 41–42 and 44–46: The statements regarding lower/higher enrollment among the triply disadvantaged groups convey similar meanings. These sentences could be combined to improve clarity and avoid redundancy.

The recommendation for targeted interventions, including subsidized enrollment, is appropriate. However, it would be helpful to clarify that Nepal already has provisions for subsidized enrollment for low-income populations under the Nepal Health Insurance Program (NHIP).

I recognize that space in the abstract is limited. The authors have appropriately addressed this issue in the Discussion section by highlighting gaps in the current subsidy mechanisms and acknowledging prior studies that emphasize the need to strengthen systems for accurately identifying and enrolling low-income and vulnerable populations. Briefly signaling the existence of these subsidies in the abstract, while leaving the detailed discussion to the main text, may improve clarity and policy context.

Introduction

The background section is well written and provides useful context. Since the paper focuses on inequalities, it would strengthen the manuscript to include more detailed discussion of the NHIP’s equity-oriented provisions, such as premium subsidies for ultra-poor households, older adults, persons with disabilities, and individuals with specific chronic conditions. Highlighting these features would better situate the study within the policy framework intended to address inequalities.

The use of intersectional analysis to examine social determinants of enrollment is a notable strength of the study.

Methodology

Under the outcome measure, when describing the primary outcome, it would be helpful to clarify that the Nepal Demographic and Health Survey (NDHS) treats NHIP enrollment as an individual-level characteristic, even though enrollment is operationally conducted at the household level.

This distinction has important implications and could also be discussed further in the Discussion section, particularly in relation to household decision-making dynamics.

Results

Line 249 (Section 3.3: Inequality by geographical location): The finding that Sudurpaschim Province demonstrates relatively equitable enrollment is particularly interesting. Previous studies suggest that enrollment uptake has increased among subsidized groups. The Nepal Government provides 100% premium subsidies for ultra-poor households and certain high-vulnerable groups, including people living with HIV, whose prevalence is relatively high in Sudurpaschim.

It would strengthen the analysis if the authors could explore whether higher enrollment among lower wealth quintiles in Sudurpaschim is driven by these fully subsidized groups. If disaggregated data are unavailable, this could at least be acknowledged as a plausible explanatory factor.

Discussion

Line 340: The role of Enrollment Assistants (EAs) in community-based outreach could be elaborated. Evidence from prior studies suggests that EAs may be more likely to enroll fully subsidized groups who are exempt from premium payments. Additionally, healthcare providers are often cited as trusted sources of health information. The authors may consider discussing the potential benefit of establishing information desks or enrollment support booths within healthcare facilities to improve outreach and enrollment.

The suggestion to integrate insurance education into school curricula or school-based programs is great. However, it would be helpful to support this recommendation with relevant evidence or examples from other settings.

The relatively high enrollment in Koshi Province is interesting. The authors might consider discussing whether this could be related to greater media exposure, higher internet use, or better geographical accessibility to first points of service contact.

The comparatively lower enrollment in Madhesh Province warrant further exploration and authors have done good job in pointing various possible reasons. One possible explanation could also be cross-border healthcare utilization in neighboring India, where facilities do not accept NHIP coverage. A brief review of relevant literature could help contextualize this finding.

Author Response

Reviewer 4

Thank you for the opportunity to review this manuscript. I read it with great interest, as the topic aligns closely with my area of interest. Overall, the paper addresses an important issue and offers valuable insights into inequalities in enrolment under the Nepal Health Insurance Program (NHIP). Below are my specific comments and suggestions.

Response: Thank you for your observation and evaluation.

Abstract

Lines 41–42 and 44–46: The statements regarding lower/higher enrolment among the triply disadvantaged groups convey similar meanings. These sentences could be combined to improve clarity and avoid redundancy.

Response: Thank you for your feedback. We have now revised the paragraph, combined the sentences to reduce redundancy.   

The recommendation for targeted interventions, including subsidized enrolment, is appropriate. However, it would be helpful to clarify that Nepal already has provisions for subsidized enrolment for low-income populations under the Nepal Health Insurance Program (NHIP).

I recognize that space in the abstract is limited. The authors have appropriately addressed this issue in the Discussion section by highlighting gaps in the current subsidy mechanisms and acknowledging prior studies that emphasize the need to strengthen systems for accurately identifying and enrolling low-income and vulnerable populations. Briefly signalling the existence of these subsidies in the abstract, while leaving the detailed discussion to the main text, may improve clarity and policy context.

Response: Thank you for your feedback. We have now revised the abstract to add the subsides policy targeted for poor households.

Introduction

The background section is well written and provides useful context. Since the paper focuses on inequalities, it would strengthen the manuscript to include more detailed discussion of the NHIP’s equity-oriented provisions, such as premium subsidies for ultra-poor households, older adults, persons with disabilities, and individuals with specific chronic conditions. Highlighting these features would better situate the study within the policy framework intended to address inequalities.

The use of intersectional analysis to examine social determinants of enrollment is a notable strength of the study.

Response: Thank you for your feedback. We have now added additional information on subsides for targeted population through NHIP.

 

Methodology

Under the outcome measure, when describing the primary outcome, it would be helpful to clarify that the Nepal Demographic and Health Survey (NDHS) treats NHIP enrollment as an individual-level characteristic, even though enrollment is operationally conducted at the household level. This distinction has important implications and could also be discussed further in the Discussion section, particularly in relation to household decision-making dynamics.

 

Response: Thank you for your feedback. We have described above that NHIP considers family as a unit while DHS consider the survey at individual level. We have discussed this in discussion and limitation

Results

Line 249 (Section 3.3: Inequality by geographical location): The finding that Sudurpaschim Province demonstrates relatively equitable enrollment is particularly interesting. Previous studies suggest that enrollment uptake has increased among subsidized groups. The Nepal Government provides 100% premium subsidies for ultra-poor households and certain high-vulnerable groups, including people living with HIV, whose prevalence is relatively high in Sudurpaschim.

It would strengthen the analysis if the authors could explore whether higher enrollment among lower wealth quintiles in Sudurpaschim is driven by these fully subsidized groups. If disaggregated data are unavailable, this could at least be acknowledged as a plausible explanatory factor.

 

Response: Thank you for your comments and feedback. The reasons behind relatively equitable enrollment in Sudurpaschim province need to be explored further, however our dataset has this limitation. One plausible mechanism, as suggested by a reviewer, is the national policy of providing 100% premium subsidies for ultra-poor households and highly vulnerable groups, such as People Living with HIV, whose prevalence is noted to be higher in this region. However, service data from health Insurance Board indicates that the number of direct beneficiaries from this specific provision is low and unlikely to explain this observed pattern. Thus, further qualitative research is needed to understand the community-level dynamics and policy implementation factors that contribute to these variations in enrolment.

 

Discussion

Line 340: The role of Enrollment Assistants (EAs) in community-based outreach could be elaborated. Evidence from prior studies suggests that EAs may be more likely to enroll fully subsidized groups who are exempt from premium payments. Additionally, healthcare providers are often cited as trusted sources of health information. The authors may consider discussing the potential benefit of establishing information desks or enrollment support booths within healthcare facilities to improve outreach and enrollment.

Response: Thank you for your comments and feedback. We have added the information desk or support booths in health facilities. Thank you for your feedback.

 

The suggestion to integrate insurance education into school curricula or school-based programs is great. However, it would be helpful to support this recommendation with relevant evidence or examples from other settings.

 

Response: Thank you for your feedback. We have added relevant reference to support our argument here.

 

The relatively high enrolment in Koshi Province is interesting. The authors might consider discussing whether this could be related to greater media exposure, higher internet use, or better geographical accessibility to first points of service contact.

The comparatively lower enrolment in Madhesh Province warrant further exploration and authors have done good job in pointing various possible reasons. One possible explanation could also be cross-border healthcare utilization in neighbouring India, where facilities do not accept NHIP coverage. A brief review of relevant literature could help contextualize this finding.

 

Response: We thank the reviewer for these thoughtful observations and constructive suggestions. We agree that exploring the underlying drivers of both high enrolment in Koshi Province and low enrolment in Madhesh Province adds important depth to the analysis. In response, we have conducted additional analyses and revised the manuscript accordingly.

 

Regarding Koshi Province: To investigate whether media exposure, internet use, and geographical accessibility are associated with higher enrolment, we have now incorporated these variables into our regression analysis. The updated results, which include indicators for media access and internet use, are presented in Supplementary File.  Additionally, we examined health facility density as a proxy for geographical accessibility to first points of service contact. This supplementary analysis, detailed in Supplementary File allows us to assess whether proximity to NHIP-enrolled facilities is associated with enrolment rates at the provincial level.

 

Regarding Koshi Province: Thank you for your suggestion and feedback. We have added the health-seeking pattern in neighbouring India after your critical feedback.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I would recommend that the authors more carefully address the comment provided during the revision process and further elaborate on the study's novelty. At present, the manuscript does not sufficiently highlight the research's original contribution relative to existing studies. I suggest expanding the literature review beyond the currently cited works and further developing the implications section.

 

Author Response

Reviewer 1

Comments Version 1

The article evaluates the enrolment intensity of different population categories in Nepal in the national health insurance program. The idea of investigating this topic is timely and interesting from several perspectives, as the authors argue in the manuscript. The paper is appealing for its use of mathematical methods to examine relationships among components of the national health survey, which the authors treat as a secondary data source for calculations and to inform their recommendations for the national health system.

Response: We would like to thank you for your assessment.

However, the manuscript has several serious drawbacks that do not allow it to be published in its current form:

1. In line 109, the authors mention “the key gap in the literature,” but I did not find a sufficiently comprehensive literature review earlier in the manuscript to substantiate this claim. Only a few studies are mentioned, without an in-depth analysis of existing research on the topic.

Response: Thank you for your evaluation and feedback.

Thank you for your evaluation and feedback. We have revised the manuscript to address the identified gap in the literature. While previous studies have utilized national survey data; such as the Demographic and Health Survey (DHS) 2022[1–3]  and the Nepal Multiple Indicator Cluster Survey 2019 [4, 5]. However, none of the previous studies have applied an intersectionality framework to analyze the data; instead, they have focused on single dimensions of disadvantage. In this paper, we aim to bridge this gap in literature by applying the concept of intersectionality.

The revised text reads as follows (page 3)

“Although national representative studies on health insurance have explored individual dimension of disadvantage, such as wealth, ethnicity, or educational attainment, in relation to health insurance enrolment [13,16–19], they have not applied the intersectionality lens in assessing the disparity in health insurance enrolment.”

 

1. The authors should comment on the significant disparity in the number of men (4,913) and women (14,845) included in the analysed survey data. How does this imbalance affect the results and conclusions of the study? Could this represent a serious limitation of the survey? What is the reason for such a large difference in the number of respondents by gender?

Response: Thank you for your comments.

A total of 4,913 men and 14,845 women aged 15-49 were interviewed face-to-face by trained interviewers using structured questionnaires. In the NDHS 2022, all eligible women aged 15–49 in selected households were interviewed, while men aged 15–49 were interviewed only in half of the selected households. Interviewing all women and half of the men is a standard practice in DHS conducted in over 90 countries [6]. Thus, the large difference in number of respondents by gender is due to survey design which is a standard practice adopted by DHS surveys.

The analysis was conducted separately for men and women using different datasets available. Thus, the imbalance in sample size between men and women is less likely to affect the results and conclusions.  The gender imbalance in sample size is by design in DHS surveys and does not compromise the validity of our gender-specific analyses. We applied complex survey analysis and gender-specific weights to ensure accuracy.

All findings are presented separately for men and women. We did not pool the data or make direct statistical comparison between genders. We also did not perform gender-interaction tests. Because of this approach, the sample size imbalance has no impact on the validity or interpretability of the results.

The previous studies have also applied the same dataset and conducted analysis separately for men and women [1–3, 5, 7]. Likewise, the previous studies utilizing Nepal Multiple Indicator Cluster Survey 2019 have also analysed separately for 14,769 women and 5,491 men [5].

1. As I understand it, the manuscript relies on the analysis of an extensive survey document that covers multiple aspects. However, the section “Program and Policy Implications” lacks a clear description of concrete actions to address the identified inequalities. I also do not see any benchmarking analysis or references to countries that have successfully addressed similar issues.

Response: Thank you for your critical observation. We have revised the program and policy implication section based on your critical feedback considering the references from different countries.

The revised manuscript reads as follows

“This study reveals several actionable policy implications to enhance NHIP enrolment and equity, advancing the pathway to UHC goals. First, the compounded disadvantages for triple-disadvantage groups (poor, illiterate and disadvantaged ethnicity) results in low enrolment rate. To address this, policies should adopt targeted approaches based on the intersectionality framework for groups facing multiple marginalization [56]. The most marginalized population that is derived from health insurance should be identified first, and targeted interventions should be implemented to increase their access to services. Such interventions may include targeted awareness campaigns, integrated outreach pro-grams that combine other health literacy or awareness programs with health insurance, and shift from current blanket subsides for the targeted population towards a multi-dimensional mechanism that identifies populations facing compounded societal barriers to enrolment in NHIP[57].”

 

We have added other references as appropriate in program and policy implication as suggested by the reviewers.

 

1. The authors should thoroughly revise the manuscript to better demonstrate the research's novelty. In its current form, the study appears to be primarily an analysis of existing data using mathematical tools, without a clearly structured research framework or well-developed practical recommendations grounded in the calculated results and graphical analysis.

Response: Thank you for your constructive feedback. We agree that the initial version did not sufficiently explain the novelty of the paper. We have now substantially revised the manuscript to address the reviewer comments. The novelty of this paper relies on applying the intersectionality lens in enrolment on NHIP of Nepal. Furthermore, we used the ENCI and concentration curve to assess inequality, which, to our knowledge, has not been examined in this sector using recent NDHS date. The revised text reads as follows:

 

“Previous studies they have not applied the intersectionality lens in assessing the disparity in health insurance enrolment. Intersectionality provides a critical framework for examining how overlapping social identities intersect to generate unique experiences of advantages and disadvantages [20–22]. The barrier in access to health insurance faced by women of poor household can be different from those women who are poor as well as illiterate and come from a socially backward ethnic group. Understanding this compounded barrier is essential for effective implementation of health insurance policy in Nepal. The country’s health insurance pol-icy aims to provide universal health coverage ensuring the access to healthcare to general public [23]. The policy also aims to provides subsides for targeted population including the poor households [24]. However, such blanket policy approach may fail to adequately capture the most marginalized populations by other social determinants like caste and education attainment. The existing policies have yet to incorporate intersectionality as a tool for identifying and reaching the most vulnerable segments of population within the poor households.

 

In contrast to previous studies that have examined the socio-economic inequality of health insurance in isolation, this study provides novelty in literature by applying intersectionality as an analytical framework to quantify how multiple levels of disadvantages (poor, illiterate and backward ethnic identity) affects in health insurance enrolment in Nepal. In doing so, it provides empirical evidence that highlights the need for targeted strategies to reach populations facing multiple overlapping deprivations existing in society. Furthermore, this research directly supports the “leave no one behind” principle, a central commitment of the Sustainable Development Goals (SDGs) by identifying the most marginalized groups in access to health insurance coverage [25]. It also aligns with SDG goals 10 (reduce inequalities), offering critical insights for assessing and addressing persistent disparities in health equity [26].”

 

Regarding the research framework

 

We have adopted the intersectionality framework/method rather than routine analysis of survey data. We have described how intersectionality of multiple social identities is essential for uncovering the unseen disparities that is not shown by single-dimension analysis. This provides a clear theoretical framework which is the novelty in analysing the inequality of health insurance enrolment in Nepal.

 

Regarding the Results and Discussion/policy implication

 

We have now expanded the discussion and policy implications. We have suggested programmatic implications for triple-disadvantaged groups as presented above.  

We believe these revisions and explanation can address the reviewer concerns.

 

Reviewer 1:

Comments Version 2

I would recommend that the authors more carefully address the comment provided during the revision process and further elaborate on the study's novelty. At present, the manuscript does not sufficiently highlight the research's original contribution relative to existing studies. I suggest expanding the literature review beyond the currently cited works and further developing the implications section.

 

Response: We have addressed the reviewer comments in the above explanation.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Overall, the revised manuscript addresses the key methodological and conceptual concerns identified in the first review.

Comments on the Quality of English Language

No comments.

Author Response

Overall, the revised manuscript addresses the key methodological and conceptual concerns identified in the first review.

Response: We would like to thank you for your assessment and review of previous version of our manuscript. We are pleased that the previous revisions that we had made after your valuable review and feedback were satisfactory.  

 

Round 1: Figures and tables can be improved

Thank you for the feedback. We have tried our best to improve the tables. Likewise, we have uploaded separate png files for the figures that we have kept in the original manuscript.

Round 2: The English could be improved to more clearly express the research.

 

Thank you for your feedback. We have tried to improve the English as suggested.

Author Response File: Author Response.pdf