Association between Menstrual Hygiene Management and School Performance among the School-Going Girls in Rural Bangladesh

Round 1

Reviewer 1 Report

This is a potentially useful addition to the literature as few studies have looked at the association of poor MHM and educational performance.  However (as I mentioned in a previous review of this paper for a different journal), I am concerned about the categorisation of MHM as it looks like no girls using disposable pads were categorised as having good MHM and I think this may be an artefact of your definition requiring "drying of re-usable cloths". Please could you check the coding carefully to ensure that if a girl does not use re-usable cloths, she can still be categorised as having good MHM.

The English language needs revising throughout.

Specific comments:

Abstract
It would be good to give the age range in the abstract (not just mean age) as international readers may not know what Grades 7-10 means.
Please clarify what you mean specifically by 'sanitary napkins'. Is this manufactured disposable pads? Or include manufactured reusable pads? Or home-made pads?
How was "poor", "average" and "good" MHM defined?

Please don’t use the term “chance” to describe the odds ratio

Introduction
Some of the references are specific studies and could be replaced by systematic reviews and international reports which are more generalisable.  For example, the statement "One in ten school-age African girls could not attend school during menstruation or dropped out at puberty due to lack of cleanliness and separate toilet facilities for female students at schools" is based on one study from Ethiopia and can be generalised to all African school age girls.

Methods
The authors should clarify eligibility criteria (“included” is not specific). Being unwilling to participate is not an eligibility criteria – students can be eligible but not enrolled because they are unwilling to participate.  The proportion refusing is important to inform selection bias.

How was the random selection of girls done?  Were the same number of girls selected within each school?  This would be appropriate if the sampling of schools was proportional to their size.  But it is confusing because the authors also say that all adolescent girls were invited to participate in the study (line 107).

What was the design effect of 1.5 based on?

How is “good school performance” defined?  This is the primary outcome but I can’t see a definition in the paper.

How could you be sure that parents were fully informed and gave consent if the team did not directly meet with the parents?

Are the examinations standardised across schools?

The strategy for building up the multivariable model was not very clear. How did you decide what a confounder was in this analysis - was it based on changes in the association between the main exposure and outcome (which would be appropriate given you have one main exposure), or associations with the outcome.

It would be good to do some sensitivity analyses around the cutoff for "good" performance and see how robust the results are. Did you also consider including performance as a linear variable?

Results
The response rate is quite poor (78%) and should be mentioned.  Do you know any basic characteristics of girls who were unwilling to participate?

There is no need to give the chi-square statistic in the text (or the table).

Table 1 is clear, but the analysis (chi-squared test) does not take into account clustering.  It would be better to use GEE logistic regression here. Similarly, Table 2 does not take the clustering into account - here you would need ordinal logistic GEE regression.

You also don’t need to highlight the associations with P<0.05 as this is an arbitrary cutoff.

Table 2: Should “source of information” be “main source”?  I.e. only one category per participant?

Also in Table 2, it looks like no girls who use sanitary napkins or new cloths have good MHM - this seems very odd. Do you think it is to do with your definition of good MHM e.g. if you include 'drying of re-usable cloths' how do you categorise people who do not use reusable cloths? I am concerned about your definition based on the results presented in Table 2. This also affects the questions about buying/disposing of pads.

Table 3: There are some discrepancies between the ORs shown in the text and the tables.  Also the p-values here should be one per variable, not one per exposure level.

Discussion
When discussing the MHM literature, please cite either systematic reviews, or studies specific to the Indian sub-continent. You need to be clear which references are systematic reviews (e.g. Sumpter & Torondel) and give these more weight than individual studies.

Please also discuss your findings in the context of the MHM framework given by Hennegan (2019)

I am surprised you mention HIV in the discussion as I think the prevalence of HIV is very low in Bangladesh general population - and as far as I know there is very little evidence of fomite transmission of HIV.

Author Response

Response file attached.

Author Response File: Author Response.docx

Reviewer 2 Report

The authors investigated the relationship between menstrual hygiene practices and academic performance among rural Bangladeshi adolescent girls. They surveyed 499-secondary school-going adolescent girls attending either public or private institutions in rural Bangladesh. The findings suggest that girls’ school academic performance can be improved by practicing good menstrual hygiene practices. School-based menstrual hygiene management education could be helpful. 

The article has a fluid narrative and adequate contextualization. In addition, methods and procedures were carefully addressed to assure robustness to the study.

Honestly, the results were as expected: better Menstruation hygiene management (MHM) influences positively school academic performance. As the authors state, in conclusion, education campaigns and the establishment of adequate facilities at school may improve the reproductive health of the school-going girls in Bangladesh.  However, these measures are poorly explored. Therefore, I ask the authors to give more on the policies to improve MHM in the discussion section.

Also, the conclusion section is just a paragraph. Please extend it to at least three paragraphs, showing to the readers (in the last one) the further steps of this research.

Minor comments:

- Please revise the guidelines to rescale the font size of Figure 1;

- Although I consider the writing style is good, I ask the authors to perform a complete revision in the paper (grammar). It seems the article lacks a final refinement on this matter. Please send to a final proofreading service or use English grammar software to perform these final adjustments.

Author Response

Response file attached.

Author Response File: Author Response.docx

Reviewer 3 Report

Kia ora.  This is an interesting study.  It requires  more work on sentence structure and clarifying language. I have made a number of suggestions below - and then just identified where more changes are required.  iI have mafe the suggestions by line number.  My comment about the overall design of the paper is that I wonder if there should be more analysis, discussion about the link between access  to materials and resources/income and girl's ability to practice good hygiene.  i.e - if girls can't afford products and live in poor conditions, they can't mange their periods as wealthier girls can.  

 

33 improved not better

40 restricts

44 - first use of limited - use compromised instead

48 the second part of your sentence is a new idea.  Put it in a new sentence

52.  Suggest using this as the overarching sentence for all of the studies you are discussing - ie. from line 45

in general - be careful you are not repeating yourself from different studies - put the same ideas together in the same place.  

54, or have toilets that are not...

55 could is the wrong word here - use can?

Choose one word to use - menstruation period - tell us why and use it all the way through - don't mix and change - ditto with the word you use for toilet - decide on toilet or latrine or tell us the difference between the two and be explicit about why you use one word or the other. 

57 Menstrual hygiene management does not have capital letters except in the brackets (MHM). Same for wash etc

60-61 you've said this before - combine with previous

61 - what is the pupil latrine ratio?  Tell us about this

62 - locks/toilet paper

63 present,

64.  New paragraph from WHO

68 limited knowledge of

70 all inhibit the implementation of (no capital letters

76 barriers -  new paragraph

81 on,

84 delete the before academic

91 who were -  not was

92/93.  Delete who are.  Participate or (not and)

118 delete correctly after response options

119 study teams or the study team

129 an individual

131 are not is (outcomes)

144-148 restructure sentence for clarity

156 what is turtile?

180 - were not was

181 participants - not they, delete also

183 withdraw from

208 A majority

210 use sanitary napkins during their period

all the rest needs tidying up in a similar way.  pieces of cloth for example

259-289 there are a lot of ideas in the paragraph.  Maybe you culd divide the ideas up into a few paragraphs with similar ideas in each one.  it would make reading it easier.  

I think this is a very interesting study.  Thank you for doing this research.  

 

 

 

Author Response

Response file attached.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The manuscript is improved but unfortunately there are several issues which have not been adequately resolved as raised in my previous comments, as follows:

Major points

1. The authors have clarified the definition of good MHM, but it does not make sense that girls can only have good MHM if they fulfill all 3 parts of the definition because the 3^rd part of the definition is drying re-usable cloths. This is not applicable if they use disposable pads so none of the girls using sanitary napkins are classified as having good MHM.  Surely for girls who use solely disposable pads, they only need to fulfill the other 2 parts of the definition to be classified as having good MHM?  (And "average" would mean fulfilling one part of the definition)

 

2. The discussion of the literature in both the Background and Discussion is still not adequate. We need a more nuanced interpretation of the data on school attendance and menstruation. For example, the authors still do not cite the systematic review on MHM and school attendance (Hennegan & Montgomery, 2016 https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0146985), or include a more comprehensive list of references (e.g. Kansiime et al, 2020 is not included; https://pubmed.ncbi.nlm.nih.gov/32024786/).  They mention the systematic review by Sumpter & Torondel but not in a meaningful way.  Also, references 11 and 12 are not from the World Bank, as the authors state – and as far as I can see they do not provide evidence that girls miss upto 4 days of school each month because of their periods (some of the 4 days will be weekends etc) – and other studies have shown no effect (e.g. ref 10). 

There is a similar issue in the Discussion – for example the first paragraph of the discussion mentions individual studies, and a systematic review, but doesn’t specify which is the systematic review or what it found (reference 21).

Further, the authors have still not used the framework defined by Hennegan (2019) – they do now cite this paper (reference 1) but do not mention the framework or anything substantive in the paper, and their response to my comment on this point cites different papers which are not relevant.

 

3. The statistical methods are still not clear or appropriate. Specifically, the responses to my previous comments are not adequate e.g.:
   1. The sampling method is still not clear to me – how is the sampling PPS? What is the lottery method?
   2. The rationale for the design effect is still not clear (the authors say this is based on an ICC of 0.01, but what is the rationale for ICC=0.01?)
   3. The strategy for building the model is still not clear – e.g. stepwise regression cannot be used to assess confounding because there is no statistical test for confounding based on p-values – it is based on the change in the OR.
   4. Table 3 still shows one p-value per exposure category not one p-value per variable.

 

Minor points

1. The authors should use the term “odds” and not “likelihood” in the abstract when describing the odds ratios (the wording is correct in the results)

 

2. Please report the number (%) who practiced poor, average & good MHM in the abstract.

Author Response

Please see attachment.

Author Response File: Author Response.docx