Internal Consistency and Floor/Ceiling Effects of the Gross Motor Function Measure for Use with Children Affected by Cancer: A Cross-Sectional Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Congratulations to the authors for completing the manuscript. In general, analyses focused on the use and testing of GMFM are necessary and welcome. At the same time, it is worth mentioning that the conducted study is interesting in some details. However, I must state at the outset that I propose to reject the current form of the manuscript. I have many reasons for this, which follow:

 

General

The manuscript lacks a clear purpose and objective. In this respect, it is not clear why two sub-issues, namely internal consistency and floor/ceiling effects (FCE), are focused on in one paper and yet are quite different. Moreover, both topics are treated incompletely and inadequately, without the necessary evidence and without appropriate conclusions.

In its current form, the manuscript adds very little value to the reader; it provides very few new insights (both substantive and methodological). Furthermore, the results of the study are difficult to reproduce.

A strong argument for rejecting the paper is the low level of analysis performed. The analysis of internal consistency based only on Cronbach's alpha is inadequate (the problem of internal consistency is much larger than can be reduced in this way); among other things, there is no more thorough analysis of individual items, and there is no adequate correlation analysis between items.

The manuscript makes no mention of other attempts to reduce the large-scale measure. However, there are a number of publications that focus on tests such as the GMFM-66. I leave aside other published attempts and efforts to modify this measure.

The manuscript refers to psychometrics in several places; however, with respect to the actual content of the study, the use of this term is misleading, inaccurate, and deceptive. Psychometrics refers to studies whose content is significantly beyond the scope of the research conducted and this manuscript.

 

Partial comments (should the authors decide to rewrite the manuscript)

The definition of FCE in the paragraph on lines 223-229 is inadequate and also incorrect (see line 228). It is not clear from the text and the tables in the appendix how the definition of FCE was applied in the calculations presented in Table 5 and in the appendix.

It would be appropriate to apply the FCE to the individual dimensions A-E as well.

With respect to lines 136-137, it would be useful to comment on what differences you consider to be significant with respect to the focus of the study.

The analyses and statistical tools used do not provide answers to the hypotheses formulated. Either additional analyses (EFA, CFA, and many others) should be performed or the formulation of the hypotheses should be modified.

It is not clear why two dimensions are reported simultaneously in Tables 2 and 3. It would be appropriate either to present all dimensions (i.e., all 88 items) in one table or to present each dimension in a separate table.

The content of the tables is inadequate; they do not show n, Me, SD for each item. Given the focus of the analyses on the FCE, it would be useful to add skewness and kurtosis.

I would also expect a summary analysis (including FCE) of each dimension.

Instead of the current Table 5, it would be more appropriate to provide a summary table to support the answers to the formulated hypotheses. At this time, it is not clear why the items for Dimension A are listed in the current Table 5 when everything else is listed in the Appendix.

The statements on lines 251-254 should be supplemented with specific references.

Discussion

This section should begin with a summary of the results of the study conducted.

I would expect a comparison of the scores obtained for each dimension with other studies that have focused on validating the properties of this measurement tool.

It would be useful to explain why the FCE occurred in the given cases? Was it due to the inadequate design of the measurement tool, the characteristics of the population, or some other reason?

Conclusions

The above conclusions are only hinted at; the study does not suggest specific actions based on the results obtained in relation to the instrument used; specific items that could be eliminated are not clearly identified.

General statements should be made more specific - see especially lines 349 or 360 (some items) - which items specifically?

Appendix

Consider whether this table format is appropriate and necessary. On the other hand, would it not be more appropriate to add summary tables for each dimension?

Despite the proposed rejection, I would like to support the writing team in their future efforts. I believe that a substantially revised text could be considered in the future with a better result. Thank you and best wishes for your future work.

Sincerely,

Author Response

Comment 1: The manuscript lacks a clear purpose and objective. In this respect, it is not clear why two sub-issues, namely internal consistency and floor/ceiling effects (FCE), are focused on in one paper and yet are quite different. Moreover, both topics are treated incompletely and inadequately, without the necessary evidence and without appropriate conclusions.

 

Response 1: Thank you for your comment, we partially agree with you because the objective seemed clear as well as basic methodology for measuring internal consistency and FCE. We are grateful for your valuable comments, which enabled us to improve our manuscript. We hope you will be able to appreciate our efforts. However, for the objective of our study, we are struggling in specifying further. We wrote “the goal of this multicenter study was to explore for the first time the internal consistency and floor/ceiling effects of the GMFM-88 when used for this population. Concerning the objective, the research group hypothesized that: 1) all the items of the GMFM-88 would positively contribute to determining the total score; 2) some items of the GMFM-88 may not be clinically sensitive for the target population and can show floor-ceiling effect” We hope now is clearer. Please see Page 3 lines 150-155.

 

Comment 2: In its current form, the manuscript adds very little value to the reader; it provides very few new insights (both substantive and methodological). Furthermore, the results of the study are difficult to reproduce.

Response 2: We are really grateful to the effort you have made in giving us so many suggestions to ameliorate our paper. We have followed all your suggestions, and we believe to have improved a lot the actual draft.

 

Comment 3: A strong argument for rejecting the paper is the low level of analysis performed. The analysis of internal consistency based only on Cronbach's alpha is inadequate (the problem of internal consistency is much larger than can be reduced in this way); among other things, there is no more thorough analysis of individual items, and there is no adequate correlation analysis between items.

Response 3: Thank you for your comment, we agree with you that there is additional analysis that can help for understanding if GMFM-88 is an appropriate tool for children with cancer. However, in this paper we are exploring some basic psychometric properties. As expected, GMFM-88 can effectively measure its construct, the gross-motor function in children. This is why, despite expected, we measured internal consistency for each GMFM-88 subscales and item-total correlation. As reported in the paper, having a very high value of internal consistency may indicate that some items are redundant and do not properly work. We added a correlation analysis among GMFM-88 dimensions, using the Pearson correlation coefficient, and found that dimensions are strongly and significantly correlated. Again, this is an expected finding, but there are still some challenges in using GMFM-88 in our clinical practice. We do not believe that GMFM-88 is not suitable for measuring gross motor function, but that it can not be adequate for our population, namely children and adolescents with malignant neoplasm conditions. Therefore we opted to analyze floor and ceiling effects. We added some additional analysis. Please also see the comments below.

Comment 4: The manuscript makes no mention of other attempts to reduce the large-scale measure. However, there are a number of publications that focus on tests such as the GMFM-66. I leave aside other published attempts and efforts to modify this measure.

Response 4: Thank you for pointing this out. We have added this sentence on page 13, lines 463-465: “Other attempts to reduce the number of items of the GMFM-88 have been already performed in the effort to improve the interpretability and clinical usefulness of the GMFM, such as the development of the GMFM-66 (63).” and this one on page 3, lines 129-132: “The GMFM-ALL is a reduced version of the Gross Motor Function Measure (GMFM-88) (29,30), a tool that has been used to evaluate gross motor function in various paediatric populations (31–37), including that of children/adolescents with cancer (38,39). ”.

Comment 5: The manuscript refers to psychometrics in several places; however, with respect to the actual content of the study, the use of this term is misleading, inaccurate, and deceptive. Psychometrics refers to studies whose content is significantly beyond the scope of the research conducted and this manuscript.

Response 5: Agree. We have accordingly modified the sentence on page 4, line 183-185: “The original GMFM-88 developers have been contacted and agreed to the investigation of internal consistency and floor/ceiling effect of the scale for a paediatric population affected by cancer.” and the ones on page 12, lines 425-427: “The value of this study resides in its evaluation of internal consistency and floor/ceiling effect of the GMFM-88 in a population lacking validated gross motor function assessment tools. “.

Comment 6: The definition of FCE in the paragraph on lines 223-229 is inadequate and also incorrect (see line 228). It is not clear from the text and the tables in the appendix how the definition of FCE was applied in the calculations presented in Table 5 and in the appendix.

Response 6: Thank you for your comment. We tried to better specify the FCE definition, using appropriate references. Therefore we modify the definition on page 5, lines 248-253 as follows: “Floor and ceiling effects were evaluated by determining the proportion of patients who achieved the highest and lowest scores, and effects were considered present if 15% of patients obtained either the lowest or highest possible score (45). Floor and ceiling effects have been classified as significant if ≥ 15%, moderate if 10% to < 15%, minor if 5% to < 10%, and negligible if < 5% of participants score the lowest or highest possible score on a measure (45)”.

Comment 7: It would be appropriate to apply the FCE to the individual dimensions A-E as well.

Response 7: Thank you for your input. We agree with you, and we have applied FCE to each GMFM-88 dimension. Our findings revealed floor effect for 3 out 5 dimensions (C, D, E) and ceiling effect for 3 out of 5 dimensions (A, B, C). These findings together with FCE for each item (now in supplementary materials) may suggest inaccuracy of the assessment tool. Please, see table 4.

Comment 8: With respect to lines 136-137, it would be useful to comment on what differences you consider to be significant with respect to the focus of the study.

Response 8: Agree. We have accordingly modified the text on page 3, lines 144-147 to emphasize this point: “Indeed, due to the quantitative nature of the GMFM-88 in measuring gross motor function, this tool can be used with subjects whom gross motor function is compromised in different ways (e.g. by weakness, neuropathy, ataxia, hemiparesis, amputation), being suitable for evaluate a broader population of children and adolescents affected by cancer.

Comment 9: The analyses and statistical tools used do not provide answers to the hypotheses formulated. Either additional analyses (EFA, CFA, and many others) should be performed or the formulation of the hypotheses should be modified.

Response 9: We agree with you that EFA or CFA can guide us to explore structural validity, multidimensionality of the GMFM-88 and highlight if items work properly for measuring the construct of the scale. However, our objective was different, namely measuring internal consistency and FCE. Our findings suggest that some items (and some dimensions) of the GMFM-88 do not accurately measure our target population. We think our results can be quite informative for rehabilitation and healthcare professionals who work in the field.  However, we thank you for your valuable suggestion, which will certainly be addressed in future studies by our working group. We have included in the limitations the need to evaluate factor analysis on page 12, line 443-447: “At the end, in our study a factor analysis was not performed, while it would be useful investigating structural validity and multidimensionality of the scale. Factor analysis can also better explain how GMFM-88 works for children with malignant neoplasm conditions. Future studies should consider these aspects.”. We also slightly modified our hypothesis on page 3, lines 154-155, as follow “some items of the GMFM-88 may not be clinically sensitive for the target population and can show floor-ceiling effect.

Comment 10: It is not clear why two dimensions are reported simultaneously in Tables 2 and 3. It would be appropriate either to present all dimensions (i.e., all 88 items) in one table or to present each dimension in a separate table.

Response 10: Thank you for this comment. According to your input, and those of other reviewers, we modified tables in the manuscript. Now table 2 shows each dimension and each item of the GMFM-88. As recommended (see above) we also added values for correlation among GMFM.88 dimension (table 3) and FCE (table 4), while FCE for each item was moved in supplementary tables.

Comment 11: The content of the tables is inadequate; they do not show n, Me, SD for each item. Given the focus of the analyses on the FCE, it would be useful to add skewness and kurtosis.

Response 11: Thank you for your comment. We added mean (SD) for each item in table 2, reporting also item-total correlation. For what concerns skewness and kurtosis, we opted to report values for GMFM-88 dimension that can be easier to understand for the readers. You can find it on Table 4.

Comment 12: I would also expect a summary analysis (including FCE) of each dimension.

Response 12: Agree. We have accordingly added this sentence on page 10, lines 335-347: “In particular, the two dimensions that revealed floor effect are those that investigate a low functional level related to abilities in “Lying and Rolling” (dimension A) and “Sitting” (dimension B). The abilities investigated in these dimensions are often not allowed during the post-surgical phase or compromised in palliative care settings or in children affected by neurological impairments, being therefore impaired only in a little portion of our sample. “Crawling and kneeling” (dimension C) group a medium functional level of abilities that can present different levels of issues both for younger children and for those affected by muscular weakness, neurological deficits or in subjects with bone tumors who undergo surgeries. It is therefore reasonable that in our population it reached both floor and ceiling effect. “Standing” (dimension D) and “Walking, running and jumping” (dimension E) dimensions contain high level abilities that can be compromised in many subjects during the active phase of treatment, justifying these results in our sample.”

Comment 13: Instead of the current Table 5, it would be more appropriate to provide a summary table to support the answers to the formulated hypotheses. At this time, it is not clear why the items for Dimension A are listed in the current Table 5 when everything else is listed in the Appendix.

Response 13: Thank you for your suggestion. Aforementioned, we modified tables in our manuscript. FCE for each item is reported in supplementary tables, while we inserted a table (table 4) with new analysis reporting FCE, kurtosis and skewness for GMFM-88 dimension.

Comment 14: The statements on lines 251-254 should be supplemented with specific references.

Response 14: Our sincere apologies for this misunderstanding. Probably we struggle in explaining this. This is part of our results section, describing what we found with FCE analysis. We now added a sentence at the top of the table on page 8, lines 280-284 and referred to supplementary materials, so readers can better follow our findings: “The floor-ceiling effect was calculated for both GMFM-88 subscales and for each item of the GMFM-88. In particular, dimension A and B, revealed ceiling effect, dimension C displayed both ceiling and floor affect, while dimensions D and E showed floor effect. Results for GMFM-88 dimensions are synthetized in table 4, while results for each item are summarized in Supplemental Tables 1,2,3,4.”. We have also moved the mentioned sentence on page 11, lines 348-352: “Furthermore, the analysis of the scoring distribution within each item showed that out of eighty-eight items, a floor-ceiling effect was revealed for all scoring categories in eight items, for three of four scoring categories in ten items, for two of them in forty-four items, and for one scoring category in seven items, while only one item did not show a floor-ceiling effect (please see supplementary materials)”. We hope now is more clear.

 

DISCUSSION

Comment 15: This section should begin with a summary of the results of the study conducted.

Response 15: Thank you for this suggestion, we have added the result summary on page 10, lines 302-307: “Internal consistency of the GMFM-88 in a group of 217 Italian children and younger adolescents with cancer showed a high coefficient Cronbach’s α, both for the whole test and for each dimension of the scale. It was found a strong positive linear and significant correlation with GMFM-88 with each subscale and in between each dimension. The analysis of the scoring distribution reveals floor and ceiling effects in several items and in each of the GMFM five dimensions.”.

 

Comment 16: I would expect a comparison of the scores obtained for each dimension with other studies that have focused on validating the properties of this measurement tool.

Response 16: Thank you for your comment. We reported information about the internal consistency of each dimension for those validation studies that performed such analysis. In fact, only few report cronbach’s alpha coefficient. Please see page 10 lines 303-308: “For what concern internal consistency, we found bey high values of Cronbach’s alpha ranging from 0.953 to 0.989. This finding is in line with validation of the modified version of the GMFM-88 for children with both spastic diplegia and visual impairment, reporting internal consistency of dimension scores between 0.97 and 1.00 (47), or with the Korean version (48,49) of the scale but slightly higher than the Indonesian version (alpha range 0.79-0.89 (50). No in all the validation studies is reported the internal consistency analysis. ”

 

Comment 17: It would be useful to explain why the FCE occurred in the given cases? Was it due to the inadequate design of the measurement tool, the characteristics of the population, or some other reason?

Response 17: Thank you for this comment. We added a deeper discussion regarding the FCE in each dimension on page 10, lines 335-347: “In particular, the two dimensions that revealed floor effect are those that investigate a low functional level related to abilities in “Lying and Rolling” (dimension A) and “Sitting” (dimension B). The abilities investigated in these dimensions are often not allowed during the post-surgical phase or compromised in palliative care settings or in children affected by neurological impairments, being therefore impaired only in a little portion of our sample. “Crawling and kneeling” (dimension C) group a medium functional level of abilities that can present different levels of issues both for younger children and for those affected by muscular weakness, neurological deficits or in subjects with bone tumors who undergo surgeries. It is therefore reasonable that in our population it reached both floor and ceiling effect. “Standing” (dimension D) and “Walking, running and jumping” (dimension E) dimensions contain high level abilities that can be compromised in many subjects during the active phase of treatment, justifying these results in our sample.“.

 

CONCLUSIONS

Comment 18: The above conclusions are only hinted at; the study does not suggest specific actions based on the results obtained in relation to the instrument used; specific items that could be eliminated are not clearly identified.

Response 18: Thank you for this comment, we have added this sentence on page 13, lines 458-563: “This aim could be gained using the Content Validity Ratio with the envolvement of an expert panel, in order to define how much each item is relevant for the given population. This process could assist in the definition of a final panel of items that can be judged as clinically important for this specific population and relevant to describe the subject change over time.”.

Comment 19: General statements should be made more specific - see especially lines 349 or 360 (some items) - which items specifically?

Response 19: Agree. We modified the text adding this part on page 12, lines 434-438: “For example, item 46/47 “crawls up 4 steps on hands and knees/feet”/”crawls backwards down 4 steps on hands and knees/feet” are too easy for subjects with a medium level of weakness or for those affected by CIPN, not feasible for subjects in the post-surgical phase or amputees, and of little meaning for school-age children and adolescents”.

 

APPENDIX

Comment 20: Consider whether this table format is appropriate and necessary. On the other hand, would it not be more appropriate to add summary tables for each dimension?

Response 20: Appendix files were modified according to your suggestion. Now, FCE for each item of the GMFM-88 was inserted.

 

Comment 21: Despite the proposed rejection, I would like to support the writing team in their future efforts. I believe that a substantially revised text could be considered in the future with a better result. Thank you and best wishes for your future work.

Response 21: Thank you very much for your revision and for all the comments you gave us to ameliorate the reporting of our research.  

 

REVIEWER 2:

Comment 1: Authors have performed an interesting study about the Internal Consistency and Floor/Ceiling Effects of the Gross Motor Function Measure for use with children affected by cancer.

In my opinion, this manuscript has the quality enough to be published as it is now, as authors have described a great context on the introduction, methods are well described as well as results. 

These results are also well discussed and the also include some strengths and limitations of this study. 

I have no comments for the authors to improve this maunscript, as it is adequate as it is now. 

 

Response 1: Thank you very much for your kind comments. We hope our research could be useful to inform clinical practice in this field.

 

REVIEWER 3:

TITLE

 

Comment 1: The title should be informative about the study design.

 

 

Response 1: Thank you for pointing this out. We have accordingly modified the title adding the study design.

 

INTRODUCTION

Comment 2:

 I suggest that the authors organize the introduction logically into a total of five paragraphs. The content is fine, however it might be better organized. Generally, each paragraph should have specific information until the final one, which contains the study rationale that introduces the study's aim. For example, the first paragraph (lines 64-105) is too long; the second paragraph (lines 106-111) appears disconnected from the others; the last paragraph should summarize the study rationale by including the knowledge gap, the study novelty and the thesis statement.

Response 2: thank you for this comment. We have reworded the entire section to improve it according to your suggestions. The text written in red is the one we have modified. See page 2 and 3.

 

Comment 3:

 The authors should report the statistical analysis that they performed. Further, I suggest describing statistical analysis in a separate paragraph and another one titled “Data collection and procedure”. In this regard, it is unclear how the data were collected.

Response 3: Thank you for your comment. We agree with you. We modified the method sub-section, dividing Data collection and Procedure, and Data analysis. With specific regards to data analysis, we added information on descriptive statistics, internal consistency, correlation across dimensions and floor and ceiling effect. Please see the specific sub-section on page 4, line 199: “2.4. Data collection and Procedures“ and on page 5, lines 234-253: “Socio-demographic information was analyzed with descriptive statistics, using frequency, mean (SD) and median (IQR) (when appropriate). With regards to the first hypothesis, Cronbach’s α was used to evaluate internal consistency of each subscale of the GMFM-88. Our hypothesis anticipated correlation between items of each subscale. As reported by Nunnally (44), a satisfactory index of a scale’s homogeneity should have an α coefficient ≥0.70. We also measured correlation between each GMFM-88 dimension, using Pearson correlation coefficient. The Pearson’s correlation coefficient ranges from 0 (indicating no linear relationship) to 1 (indicating a perfect linear relationship) and was interpreted as follows: <0.3 indicates a weak relationship; 0.3-0.69 indicates a moderate relationship; and ≥0.7 indicates a strong relationship. The correlation values can be either positive or negative, indicating the direction of the relationship (44). With regards to the second hypothesis, the floor-ceiling effect was calculated for each item of the GMFM-88. The floor-ceiling effect describes whether participants have scores that are at, or near, the possible lower or upper limits respectively, preventing measurement of variance above or below a certain level. Floor and ceiling effects were evaluated by determining the proportion of patients who achieved the highest and lowest scores, and effects were considered present if 15% of patients obtained either the lowest or highest possible score (45). Floor and ceiling effects have been classified as significant if ≥ 15%, moderate if 10% to < 15%, minor if 5% to < 10%, and negligible if < 5% of participants score the lowest or highest possible score on a measure (45)”. We hope that now is clearer.

Comment 4:

 The authors reported that “the article was prepared according to the Consensus-based Standards for selecting health Measurement Instruments (COSMIN). However, the COSMIN checklist is used to evaluate the methodological quality of studies on measurement properties, such as in systematic reviews of measurement properties. This study has a different design and should be reported following the appropriate guidelines.

Response 4: Thank you for this comment. The Strengthening the reporting of observational studies in epidemiology (STROBE) checklist for cross sectional studies was used during the revision process to improve the study reporting. We accordingly modified the sentence on page 5, lines 255-256: “This article was prepared according to the Strengthening the reporting of observational studies in epidemiology (STROBE) checklist for cross sectional studies”.

 

RESULTS

Comment 5:

 I understand that the authors included tables with the results; however, the authors should also report the

Response 5: Thank you for pointing this out. We agree with this comment. We have accordingly revised this section adding some comments, please see the “Results” section on page 5-9. In particular, we made these modifications:

-on page 6, lines 259-261: “Male composed 53.3% of the sample, the diagnosis of CNS tumors was reported in 55.8% of the subjects and the more represented treatment phase was to be on treatment (66.8%). “.

-on page 6, lines 268-269: “Item-total statistics for each subscale of the GMFM-88 and each dimension were reported in Table 2.”

-on page 8, lines 271-274: “The analysis of the correlation of the GMFM-88 dimension, as measured with Pearson correlation coefficient, revealed a strong positive linear and significant correlation with GMFM-88 with each subscale (ranging from 0.839-0.937) and in between each dimension (ranging 0.671-0.917). Results are summarized in Table 3”.

-on page 8, lines 280-284: “The floor-ceiling effect was calculated for both GMFM-88 subscales and for each item of the GMFM-88. In particular, dimension A and B revealed ceiling effect, dimension C displayed both ceiling and floor effect, while dimensions D and E showed floor effect. Results for GMFM-88 dimensions are synthesized in table 4, while results for each item are summarized in Supplemental Tables 1,2,3,4“.

on page 9, lines 289-290: “To better understand score distribution of each GMFM-88 dimension and the total score, results are also represented in Figure 1, with evidence of skewness and kurtosis.”

 

 

 

Reviewer 2 Report

Comments and Suggestions for Authors

Authors have performed an interesting study about the Internal Consistency and Floor/Ceiling Effects of the Gross Motor Function Measure for use with children affected by cancer.

In my opinion, this manuscript has the quality enough to be published as it is now, as authors have described a great context on the introduction, methods are well described as well as results. 

These results are also well discussed and the also include some strengths and limitations of this study. 

I have no comments for the authors to improve this maunscript, as it is adequate as it is now. 

 

Author Response

Comment 1: Authors have performed an interesting study about the Internal Consistency and Floor/Ceiling Effects of the Gross Motor Function Measure for use with children affected by cancer.

In my opinion, this manuscript has the quality enough to be published as it is now, as authors have described a great context on the introduction, methods are well described as well as results. 

These results are also well discussed and the also include some strengths and limitations of this study. 

I have no comments for the authors to improve this maunscript, as it is adequate as it is now. 

 

Response 1: Thank you very much for your kind comments. We hope our research could be useful to inform clinical practice in this field.

Reviewer 3 Report

Comments and Suggestions for Authors

Overall, the study seems to be well conducted and described. However, the reporting should be more structured to clarify the contents in the introduction and method sections. The discussions and conclusions are well structured, and the contents are coherent with the findings. Please find attached my specific comments.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

n.a.

Author Response

TITLE

 

Comment 1: The title should be informative about the study design.

 

 

Response 1: Thank you for pointing this out. We have accordingly modified the title adding the study design.

 

INTRODUCTION

Comment 2:

 I suggest that the authors organize the introduction logically into a total of five paragraphs. The content is fine, however it might be better organized. Generally, each paragraph should have specific information until the final one, which contains the study rationale that introduces the study's aim. For example, the first paragraph (lines 64-105) is too long; the second paragraph (lines 106-111) appears disconnected from the others; the last paragraph should summarize the study rationale by including the knowledge gap, the study novelty and the thesis statement.

Response 2: thank you for this comment. We have reworded the entire section to improve it according to your suggestions. The text written in red is the one we have modified. See page 2 and 3.

 

Comment 3:

 The authors should report the statistical analysis that they performed. Further, I suggest describing statistical analysis in a separate paragraph and another one titled “Data collection and procedure”. In this regard, it is unclear how the data were collected.

Response 3: Thank you for your comment. We agree with you. We modified the method sub-section, dividing Data collection and Procedure, and Data analysis. With specific regards to data analysis, we added information on descriptive statistics, internal consistency, correlation across dimensions and floor and ceiling effect. Please see the specific sub-section on page 4, line 199: “2.4. Data collection and Procedures“ and on page 5, lines 234-253: “Socio-demographic information was analyzed with descriptive statistics, using frequency, mean (SD) and median (IQR) (when appropriate). With regards to the first hypothesis, Cronbach’s α was used to evaluate internal consistency of each subscale of the GMFM-88. Our hypothesis anticipated correlation between items of each subscale. As reported by Nunnally (44), a satisfactory index of a scale’s homogeneity should have an α coefficient ≥0.70. We also measured correlation between each GMFM-88 dimension, using Pearson correlation coefficient. The Pearson’s correlation coefficient ranges from 0 (indicating no linear relationship) to 1 (indicating a perfect linear relationship) and was interpreted as follows: <0.3 indicates a weak relationship; 0.3-0.69 indicates a moderate relationship; and ≥0.7 indicates a strong relationship. The correlation values can be either positive or negative, indicating the direction of the relationship (44). With regards to the second hypothesis, the floor-ceiling effect was calculated for each item of the GMFM-88. The floor-ceiling effect describes whether participants have scores that are at, or near, the possible lower or upper limits respectively, preventing measurement of variance above or below a certain level. Floor and ceiling effects were evaluated by determining the proportion of patients who achieved the highest and lowest scores, and effects were considered present if 15% of patients obtained either the lowest or highest possible score (45). Floor and ceiling effects have been classified as significant if ≥ 15%, moderate if 10% to < 15%, minor if 5% to < 10%, and negligible if < 5% of participants score the lowest or highest possible score on a measure (45)”. We hope that now is clearer.

Comment 4:

 The authors reported that “the article was prepared according to the Consensus-based Standards for selecting health Measurement Instruments (COSMIN). However, the COSMIN checklist is used to evaluate the methodological quality of studies on measurement properties, such as in systematic reviews of measurement properties. This study has a different design and should be reported following the appropriate guidelines.

Response 4: Thank you for this comment. The Strengthening the reporting of observational studies in epidemiology (STROBE) checklist for cross sectional studies was used during the revision process to improve the study reporting. We accordingly modified the sentence on page 5, lines 255-256: “This article was prepared according to the Strengthening the reporting of observational studies in epidemiology (STROBE) checklist for cross sectional studies”.

 

RESULTS

Comment 5:

 I understand that the authors included tables with the results; however, the authors should also report the

Response 5: Thank you for pointing this out. We agree with this comment. We have accordingly revised this section adding some comments, please see the “Results” section on page 5-9. In particular, we made these modifications:

-on page 6, lines 259-261: “Male composed 53.3% of the sample, the diagnosis of CNS tumors was reported in 55.8% of the subjects and the more represented treatment phase was to be on treatment (66.8%). “.

-on page 6, lines 268-269: “Item-total statistics for each subscale of the GMFM-88 and each dimension were reported in Table 2.”

-on page 8, lines 271-274: “The analysis of the correlation of the GMFM-88 dimension, as measured with Pearson correlation coefficient, revealed a strong positive linear and significant correlation with GMFM-88 with each subscale (ranging from 0.839-0.937) and in between each dimension (ranging 0.671-0.917). Results are summarized in Table 3”.

-on page 8, lines 280-284: “The floor-ceiling effect was calculated for both GMFM-88 subscales and for each item of the GMFM-88. In particular, dimension A and B revealed ceiling effect, dimension C displayed both ceiling and floor effect, while dimensions D and E showed floor effect. Results for GMFM-88 dimensions are synthesized in table 4, while results for each item are summarized in Supplemental Tables 1,2,3,4“.

on page 9, lines 289-290: “To better understand score distribution of each GMFM-88 dimension and the total score, results are also represented in Figure 1, with evidence of skewness and kurtosis.”

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors,

I would like to inform you that after reviewing the updated manuscript, I have no additional comments. The revisions have met the expectations outlined in my previous feedback and the changes you have made have improved the clarity of the study. I find that the current manuscript has been refined to a level that adequately communicates the research findings and can be accepted.

Sincerely,

Author Response

I would like to inform you that after reviewing the updated manuscript, I have no additional comments. The revisions have met the expectations outlined in my previous feedback and the changes you have made have improved the clarity of the study. I find that the current manuscript has been refined to a level that adequately communicates the research findings and can be accepted.

THANK YOU VERY MUCH FOR ALL YOUR SUGGESTIONS THAT GAVE HAVE US THE POSSIBILITY TO REFLECT ON OUR MANUSCRIPT AND TO AMELIORATE ITE. WE ARE REALLY GREATFULL FOR THE ESSENTIAL HELP YOU GAVE US.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have addressed all my concerns. I have only an additional comment regarding the last version of the manuscript. Please see below:  

Lines 255-256: The following sentence, “This article was prepared according to the Strengthening the reporting of observational studies in epidemiology (STROBE) checklist for cross-sectional studies”, should be reported at the beginning of the “Materials and Methods” section.

Comments on the Quality of English Language

Minor editing of English language required. There are some basic grammatical errors; for example, some commas must be added, and others must be eliminated.  

Author Response

The authors have addressed all my concerns. I have only an additional comment regarding the last version of the manuscript. Please see below:  

Lines 255-256: The following sentence, “This article was prepared according to the Strengthening the reporting of observational studies in epidemiology (STROBE) checklist for cross-sectional studies”, should be reported at the beginning of the “Materials and Methods” section.

AGREE, WE HAVE MOVED THE SENTENCE AT THE BEGINNING OF THE MATERIALS AND METHODS SECTION.

THANK YOU FOR ALL THE REVISIONS YOU SUGGESTED US, WE BELIEVE THAT THESE HAVE REALLY HELPED US TO IMPROVE OUR PAPER.