Propensity Score Matching: Identifying Opportunities for Future Use in Nursing Studies

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

 

 

Thank you for the opportunity to review this manuscript. The topic of propensity score matching (PSM) in nursing research is relevant and potentially valuable, as observational studies in this field could benefit from robust statistical methods to address confounding. However, the manuscript presents several conceptual and methodological concerns that need to be addressed.

The manuscript interchanges the terms "proximity score matching" and "propensity score matching", which could lead to confusion. Propensity score matching (PSM) is a well-established statistical method used to adjust for confounding by estimating probabilities from covariates. "Proximity score matching" does not exist as a recognized method—the authors may be referring to "proximity matching", but this needs clarification.

Terms like "near-empty synthetic review" are vague and not commonly recognized in systematic review methodologies. Their validity should be justified with appropriate references.

Selection bias in the search strategy: the authors assume that nursing research is published only in nursing journals, potentially underestimating the number of relevant records. This issue should be addressed.

Arbitrary selection of the study period (2020–2024). The authors do not justify why this timeframe was chosen, which raises concerns about potential bias in the data collection.

The "Near-Empty Synthetic Review (SNER)" is an unfamiliar concept that lacks a clear definition or validation in the literature.

"Synthetic Thematic Analysis (STA)" is vaguely described and does not appear to differ significantly from existing text-mining techniques.

The reproducibility of the study is problematic. The bibliographic search is conducted using Web of Science and Scopus, but the inclusion/exclusion criteria are not clearly defined.

The use of VOSviewer for bibliometric mapping is interesting, but the methodology behind keyword clustering is not well explained.

The paper fails to address why PSM is underutilized in nursing research. A possible explanation is that many nursing studies are qualitative or have small sample sizes, making PSM impractical.

Without discussing these limitations, the argument remains incomplete and lacks critical engagement with the subject matter.

 

Comments on the Quality of English Language

Could be improved

Author Response

I would like to thank  the reviewer for  constructive comments. The point to point answers are listed bellow 

The manuscript interchanges the terms "proximity score matching" and "propensity score matching", which could lead to confusion. Propensity score matching (PSM) is a well-established statistical method used to adjust for confounding by estimating probabilities from covariates. "Proximity score matching" does not exist as a recognized method—the authors may be referring to "proximity matching", but this needs clarification.
Proximity score matching has been corrected to Propensity score matching

Terms like "near-empty synthetic review" are vague and not commonly recognized in systematic review methodologies. Their validity should be justified with appropriate references.
Parts of the introduction and methodology sections have been rewritten

Selection bias in the search strategy: the authors assume that nursing research is published only in nursing journals, potentially underestimating the number of relevant records. This issue should be addressed.
Study limitations have been added

Arbitrary selection of the study period (2020–2024). The authors do not justify why this timeframe was chosen, which raises concerns about potential bias in the data collection.
The study period explanation was added

The "Near-Empty Synthetic Review (SNER)" is an unfamiliar concept that lacks a clear definition or validation in the literature.

Parts of introduction and methodology sections have been rewritten

"Synthetic Thematic Analysis (STA)" is vaguely described and does not appear to differ significantly from existing text-mining techniques.

Parts of the introduction and methodology sections have been rewritten. The references where more information about  Synthetic knowledge synthesis can be found were added. Aditionaly,  the paper is a Communication with the main aim of informing the nursing community about PSM and the possible opportunities where it can be successfully used.

The reproducibility of the study is problematic. The bibliographic search is conducted using Web of Science and Scopus, but the inclusion/exclusion criteria are not clearly defined.
There were no additional inclusion/exclusion criteria 

The use of VOSviewer for bibliometric mapping is interesting, but the methodology behind keyword clustering is not well explained.  
The references where more information about  Synthetic knowledge synthesis can be found were added. Aditionaly,  the paper is a Communication with the main aim of informing the nursing community about PSM and the possible opportunities where it can be successfully used.

The paper fails to address why PSM is underutilized in nursing research. A possible explanation is that many nursing studies are qualitative or have small sample sizes, making PSM impractical.
Some reasons were added to the Conclusion section

Reviewer 2 Report

Comments and Suggestions for Authors

This paper presents the use of propensity score matching in the field of nursing. 
The article is quite simple and may not provide much information to professionals who are not familiar with propensity score methodology. It would be useful to explain what the propensity score is and to briefly explain the different approaches proposed for calculating the propensity score. In fact, two terms appear in  figure 1, “propensity score matching” and “propensity score” and the authors use them as synonyms when propensity score matching is one of several solutions presented for the propensity score.  Furthermore, the references do not include any relevant review articles that explain what a propensity score is and its different solutions.
The inclusion of an explanation of the propensity score methodology would improve the article. In addition, a synthetic review of the different approaches to the propensity score used in the field of nursing could also be made. The inclusion of these explanations would make the article more interesting for nursing professionals.

Author Response

Thank you for your constructive comments. The point to point response to the comments can be found bellow. The changes in the manuscript are marked in yellow
This paper presents the use of propensity score matching in the field of nursing. 
The article is quite simple and may not provide much information to professionals who are not familiar with propensity score methodology. It would be useful to explain what the propensity score is and to briefly explain the different approaches proposed for calculating the propensity score. In fact, two terms appear in  figure 1, “propensity score matching” and “propensity score” and the authors use them as synonyms when propensity score matching is one of several solutions presented for the propensity score.  Furthermore, the references do not include any relevant review articles that explain what a propensity score is and its different solutions.
The inclusion of an explanation of the propensity score methodology would improve the article. In addition, a synthetic review of the different approaches to the propensity score used in the field of nursing could also be made. The inclusion of these explanations would make the article more interesting for nursing professionals.

The introduction has been expanded with a more detailed description of PSM and PS, and referencing review papers on PSM use in medicine

Reviewer 3 Report

Comments and Suggestions for Authors

Very interesting and potentially useful article for the scientific community. However, the introduction lacks clarity: it is unclear what is meant by proximity score matching and how it differs from propensity score matching. Additionally, the object of the study is not clearly explained. The research objective could be simplified, as it currently appears confusing and difficult to interpret.

The data collection and analysis methods are poorly described and present significant methodological challenges, which should be addressed in more detail. Furthermore, what justifies the chosen inclusion period, considering only 76 articles were included from the initial 104?

How was the content analysis performed? The results are poorly supported and difficult to understand. Additionally, the discussion section is missing. Finally, the number of references doesn't correspond to the included articles

Comments on the Quality of English Language

It might be useful to improve the language

Author Response

Thank you for your constructive comments. The point to point response to the comments can be found bellow. The changes in the manuscript are marked in yellow

 

Very interesting and potentially useful article for the scientific community. However, the introduction lacks clarity: it is unclear what is meant by proximity score matching and how it differs from propensity score matching. Additionally, the object of the study is not clearly explained. The research objective could be simplified, as it currently appears confusing and difficult to interpret.

The introduction jas been rewritten to address the reviewers comments

The data collection and analysis methods are poorly described and present significant methodological challenges, which should be addressed in more detail. Furthermore, what justifies the chosen inclusion period, considering only 76 articles were included from the initial 104?

The methodology section has been expanded and partially rewritten. The period 2020 – 2024 was selected to analyse the latest trends, consequently the number of paper was reduced from 107 to 76

How was the content analysis performed? The results are poorly supported and difficult to understand. Additionally, the discussion section is missing. 

The methodology and the result section were partially rewritten. The paper is a Communication paper, with the main aim of informing the nursing community about PSM and where it can be successfully applied 

Reviewer 4 Report

Comments and Suggestions for Authors

First, I would like to express my gratitude for the opportunity to expand my understanding by reading and analyzing this insightful manuscript. The topic addressed—examining the use of propensity score matching (PSM) in nursing research—holds significant potential for advancing methodological approaches within the field. This study contributes to the ongoing development of evidence-based nursing practices by exploring its applications, limitations, and future opportunities.

However, I suggest revising certain aspects of the manuscript to enhance its clarity, methodological transparency, and scientific rigor. Strengthening these elements could improve the overall quality of the study and maximize its impact on nursing research, particularly in the context of robust data analysis and causal inference.

 

Title

The title should accurately reflect the study's focus, methodology, and contribution to nursing. The “Search for New Opportunities" is not very clear and broad. 

Abstract: The description of the methods lacks sufficient clarity, because synthetic analyses is not specify the techniques employed for the study review. To enhance methodological rigor, it is essential to explicitly state the databases consulted, as well as the inclusion and exclusion criteria applied. Identifies ten prolific themes in observational nursing research, these are not specified, which reduces the impact and applicability of the findings.

The conclusion asserts that the extended use of PSM in nursing research could enhance consistency, relevance, and applicability to clinical practice. However, this claim is not sufficiently substantiated by the findings.

Introduction

It is suggested that the introduction include more data to justify the gap in the use of Propensity Score Matching (PSM) in nursing research. The study objectives should be reformulated to enhance clarity and methodological rigor, clearly distinguishing the different analytical steps and ensuring a well-structured research framework.

Materials and Methods

The choice of Synthetic Near Empty Review (SNER) and Synthetic Thematic Analysis (STA) is not adequately justified. Although these methods are described, the article does not explain why they were selected over more common systematic or bibliometric approaches. The methodology does not specify the inclusion and exclusion criteria used to select the articles, to enhance the study's transparency, reproducibility, and scientific rigor.

Discussion

The article review does not include the discussion section, which limits the understanding of the results and restricts the interpretation of the findings, making it difficult to critically analyze the impact and implications of the results within the context of the research.

Conclusions

The conclusion lacks depth. It should explain in more detail how PSM can improve the the research and nursing practice, as how PSM can be specifically advantageous in typical nursing research scenarios. Suggest directions for future research that could further explore the use of PSM in nursing. Highlight the limitations and propose directions for future research that could explore the use of PSM in nursing.

Comments on the Quality of English Language

No comments

Author Response

Thank you for your constructive comments. The point to point response to the comments can be found bellow. The changes in the manuscript are marked in yellow

 

The title should accurately reflect the study's focus, methodology, and contribution to nursing. The “Search for New Opportunities" is not very clear and broad. 

 

The title has been changed to: Propensity Score Matching: Identifying Opportunities for its Future Use in Nursing Studies.

 

 

Abstract: The description of the methods lacks sufficient clarity, because synthetic analyses is not specify the techniques employed for the study review. To enhance methodological rigor, it is essential to explicitly state the databases consulted, as well as the inclusion and exclusion criteria applied. Identifies ten prolific themes in observational nursing research, these are not specified, which reduces the impact and applicability of the findings.The conclusion asserts that the extended use of PSM in nursing research could enhance consistency, relevance, and applicability to clinical practice. However, this claim is not sufficiently substantiated by the findings.

 

The abstract, Introduction and results were partially rewritten

 

It is suggested that the introduction include more data to justify the gap in the use of Propensity Score Matching (PSM) in nursing research. The study objectives should be reformulated to enhance clarity and methodological rigor, clearly distinguishing the different analytical steps and ensuring a well-structured research framework.

 

The Introduction and results sections were partialy rewritten

 

The choice of Synthetic Near Empty Review (SNER) and Synthetic Thematic Analysis (STA) is not adequately justified. Although these methods are described, the article does not explain why they were selected over more common systematic or bibliometric approaches. The methodology does not specify the inclusion and exclusion criteria used to select the articles, to enhance the study's transparency, reproducibility, and scientific rigor.

 

The Introduction and results were partially rewritten and extended with future research

 

The article review does not include a discussion section, which limits the understanding of the results and restricts the interpretation of the findings, making it difficult to critically analyze the impact and implications of the results within the context of the research. The conclusion lacks depth. It should explain in more detail how PSM can improve the the research and nursing practice, as how PSM can be specifically advantageous in typical nursing research scenarios.

The conclusion has been extended

 

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The study has improved but I think the introduction to the propensity score methodology is still insufficient. The propensity score was defined by Rosenbaum and Rubin (1983) as the probability of treatment assignment conditional on baseline covariates. There are several methods of propensity score and one of them is propensity score matching. For example, other methods are: “stratification”,” weighted by the inverse of the probability of treatment” and “adjustment for covariates”. In addition, there are several ways to estimate the propensity score, one of them is by logistic regression, but it can also be done by recursive partitions or regression tree methods.

Therefore, what I mean is that this information should appear in the article so that the reader knows that the propensity score is not just the propensity score matching and that propensity score is not synonym of propensity score matching. I think that the review  has included more methods, not just propensity score matching .  In figure 1 two terms appear: “propensity score matching” and propensity score”, and my question is when the articles use only the term propensity score, those articles are using the matching procedure or a different method. In fact, the authors include the paper of Cea-Soriano (2022) where they use two procedures of propensity score. I don't want the authors to focus on the statistical aspects of this procedure, but readers of paper should be clear that the concept of propensity score is broader than the propensity score matching method and therefore that in the analyzed articles that simply use the term propensity score, other propensity score procedures other than matching may have been used.

Author Response

We thank the reviewer for valuable comments

 

Comment 1

Therefore, what I mean is that this information should appear in the article so that the reader knows that the propensity score is not just the propensity score matching and that propensity score is not synonym of propensity score matching.

We agree and we extended the introduction to make the distinction clearer

 

Comment 2

I think that the review  has included more methods, not just propensity score matching .  In figure 1 two terms appear: “propensity score matching” and propensity score”, and my question is when the articles use only the term propensity score, those articles are using the matching procedure or a different method. In fact, the authors include the paper of Cea-Soriano (2022) where they use two procedures of propensity score. I don't want the authors to focus on the statistical aspects of this procedure, but readers of paper should be clear that the concept of propensity score is broader than the propensity score matching method and therefore that in the analyzed articles that simply use the term propensity score, other propensity score procedures other than matching may have been used.

We agree and clarified this issue in study limitations