Bowel Bubble Formation with Oral Sulfate Solution for Colonoscopy Bowel Preparation Using Same-Day and Split-Day Regimens: The B-BOSS Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript addresses a relevant topic in digestive endoscopy, but the current methodology is insufficient to support the conclusions. The issues related to endoscopic quality metrics, statistical adjustment for confounding factors, and incomplete adverse event reporting are substantial. If the authors can provide a revised analysis addressing the points above, the study may become suitable for publication. However, if the methodological flaws (particularly the selection bias and lack of standardized endoscopic protocols) cannot be resolved, the manuscript should be rejected in favor of a prospective design or a more rigorously controlled retrospective cohort.

1. Methodology: Definition and Quality of Endoscopic Intervention/Diagnosis
The authors describe the endoscopic procedure in the methods section, but there is a lack of standardized criteria regarding the quality indicators. For instance, if this involves polypectomy or lesion detection, the authors do not specify the withdrawal time, bowel preparation quality (Boston Bowel Preparation Scale), or whether the procedures were performed by trainees versus experts. Without stratifying for these confounders, the reported outcomes (e.g., adenoma detection rate or complication rates) are not reliable. The authors must provide a detailed breakdown of endoscopic quality metrics.

2. Statistical Analysis: Handling of Confounding Variables (Selection Bias)
This appears to be a retrospective, single-center study. There is a significant risk of selection bias regarding which patients underwent the novel endoscopic technique versus the conventional standard. The authors used univariate analysis to compare groups but failed to perform multivariate logistic regression to adjust for baseline imbalances (e.g., lesion size, location, fibrosis, or patient comorbidities). 

3. Endoscopic Imaging/Pathology Correlation
As a reviewer for a digestive endoscopy journal, I find the correlation between the endoscopic appearance and the final pathological diagnosis insufficient. The manuscript states that lesions were resected endoscopically, but it does not clarify whether en bloc resection was achieved or whether the pathological evaluation included deep margins (submucosal invasion depth). For lesions suspected of early neoplasia, the absence of a detailed pathological assessment (e.g., Sm1, Sm2, or according to the Vienna/Paris classification) undermines the validity of the “curative” resection claims.

4. Adverse Events and Follow-up
The reporting of adverse events is incomplete. The authors mention “bleeding” and “perforation” but do not adhere to the ASGE (American Society for Gastrointestinal Endoscopy) lexicon for adverse events. There is no distinction between intraprocedural and delayed events, nor is there a standardized grading of severity (mild, moderate, severe). Furthermore, the follow-up period is heterogeneous, and the rate of loss to follow-up (if any) is not reported, which compromises the assessment of long-term recurrence or delayed perforation.

 

Author Response

Reviewer 1

The manuscript addresses a relevant topic in digestive endoscopy, but the current methodology is insufficient to support the conclusions. The issues related to endoscopic quality metrics, statistical adjustment for confounding factors, and incomplete adverse event reporting are substantial. If the authors can provide a revised analysis addressing the points above, the study may become suitable for publication. However, if the methodological flaws (particularly the selection bias and lack of standardized endoscopic protocols) cannot be resolved, the manuscript should be rejected in favor of a prospective design or a more rigorously controlled retrospective cohort.

 

Answer: We appreciate your valuable comments. We apologize our insufficient explanation of the study inclusion criteria. We amended it as follows: The inclusion criteria of this study were as follows: 1. Colonoscopy performed for screening, surveillance, or symptomatic evaluation; and 2. patients with an estimated glomerular filtration rate (eGFR) ≥30 mL/min who agreed to OSS use for SAR and SPR bowel preparation. SPR was reserved only for patients with a prior history of sodium picosulphate-related abdominal pain or inadequate bowel preparation with a same-day preparation using 1 L high-concentrated polyethylene glycol (H-PEG) or 480 mL OSS (SAR), based on our previously published protocols [5,8].

Consequently, the inclusion criteria differed substantially between the two groups. However, the aim of this study was not a direct comparison between SAR and SPR, but to describe the prevalence of bowel bubbles and other colonoscopic findings in real-world practice. Accordingly, no statistical adjustment for confounding factors was performed. P-values were presented descriptively to illustrate differences in baseline characteristics and colonoscopic findings, not for inferential comparison. This has been clarified in the Methods, Results and Limitations sections to avoid potential misinterpretation. All other comments are addressed below.

 

Method section

The inclusion criteria of this study were as follows: 1. Colonoscopy performed for screening, surveillance, or symptomatic evaluation; and 2. patients with an estimated glomerular filtration rate (eGFR) ≥30 mL/min who agreed to OSS use for SAR and SPR bowel preparation. SPR was reserved only for patients with a prior history of sodium picosulphate-related abdominal pain or inadequate bowel preparation with a same-day preparation using 1 L high-concentrated polyethylene glycol (H-PEG) or 480 mL OSS (SAR), based on our previously published protocols [5,8].

 

Given that inclusion criteria differed substantially between the SAR and SDR groups, no direct statistical comparison between regimens was intended. Rather, the aim of this study was to characterize the real-world prevalence of bowel bubbles and associated colonoscopic findings in each regimen independently. Accordingly, no adjustment for confounding factors was performed.

 

Results

Due to the inclusion criteria difference, p-values were presented descriptively to illustrate differences in baseline characteristics and colonoscopic findings, rather than to support inferential comparison.

 

P-values were presented descriptively only for showing the differences in colonoscopic findings in the two groups.

 

Limitation section 

Fourth, the primary aim of our study purpose was not a direct comparison between the SAR and SPR groups; accordingly, no statistical adjustment for confounding variables was performed and multivariate logistic regression was not conducted for between group comparisons.

 

1. Methodology: Definition and Quality of Endoscopic Intervention/Diagnosis

The authors describe the endoscopic procedure in the methods section, but there is a lack of standardized criteria regarding the quality indicators. For instance, if this involves polypectomy or lesion detection, the authors do not specify the withdrawal time, bowel preparation quality (Boston Bowel Preparation Scale), or whether the procedures were performed by trainees versus experts. Without stratifying for these confounders, the reported outcomes (e.g., adenoma detection rate or complication rates) are not reliable. The authors must provide a detailed breakdown of endoscopic quality metrics.

 

Answer: Thank you for your valuable comments. As the reviewer pointed out, we added the explanation and about inclusion criteria and the objective definition for withdrawal time, bowel preparation, and endoscopists experiences. Bowel preparation was assessed using the Aronchick Bowel Preparation Scale according to the recommendations of the Japan Gastroenterological Endoscopy Society guidelines. However, bowel preparation quality in each bowel location is not included in the scale. We mentioned it in the Limitation section. Regarding the selection bias of our study, we answered it in a previous answer described above. The inclusion criteria differed substantially between the two groups. However, the aim of this study was not a direct comparison between SAR and SPR, but to describe the prevalence of bowel bubbles and other colonoscopic findings in real-world practice. Accordingly, no statistical adjustment for confounding factors was performed. P-values were presented descriptively to illustrate differences in baseline characteristics and colonoscopic findings, not for inferential comparison. This has been clarified in the Methods, Results and Limitations sections to avoid potential misinterpretation. All other comments are addressed below.

 

Method section

Additionally, cases receiving cold snare polypectomy for benign lesions less than 10 mm in size were included.

 

Total procedure time included insertion time, withdrawal time, and polypectomy time.

 

All colonoscopy were performed only by experts.

 

Bowel preparation quality was assessed using the Aronchick Bowel Preparation Scale in accordance with the Japan Gastroenterological Endoscopy Society guidelines [13].

 

Limitation section

Fifth, bowel preparation quality was not assessed in each by colonic segment, as the Aronchick Bowel Preparation Scale yields a global rather than segmental score.

 

2. Statistical Analysis: Handling of Confounding Variables (Selection Bias)

This appears to be a retrospective, single-center study. There is a significant risk of selection bias regarding which patients underwent the novel endoscopic technique versus the conventional standard. The authors used univariate analysis to compare groups but failed to perform multivariate logistic regression to adjust for baseline imbalances (e.g., lesion size, location, fibrosis, or patient comorbidities).

 

Answer: Thank you for your comment. As the reviewer pointed out, our study was a retrospective setting and included a significant risk of selection bias. We mentioned it in the Limitation section. As we mentioned in the previous answer described above, our main purpose was not the comparison between the SAR and SPR groups. Thus, we did not perform multivariate logistic regression analysis for comparing the two groups. We also mentioned it in the Limitation section.

 

Limitation section

First, this was a retrospective study with a relatively small sample size, limiting statis-tical power. Second, patient enrollment was non-consecutive and bowel preparation regimens were selected at the discretion of individual physicians, introducing a potential risk of selection bias. Third, the 480 mL OSS regimen was evaluated only in a Japanese population, whose body habitus is generally smaller than that of Western populations, which may limit generalizability. Fourth, the primary aim of our study purpose was not a direct comparison between the SAR and SPR groups; accordingly, no statistical adjustment for confounding variables was performed and multivariate logistic regression was not conducted for between group comparisons.

 

3. Endoscopic Imaging/Pathology Correlation

As a reviewer for a digestive endoscopy journal, I find the correlation between the endoscopic appearance and the final pathological diagnosis insufficient. The manuscript states that lesions were resected endoscopically, but it does not clarify whether en bloc resection was achieved or whether the pathological evaluation included deep margins (submucosal invasion depth). For lesions suspected of early neoplasia, the absence of a detailed pathological assessment (e.g., Sm1, Sm2, or according to the Vienna/Paris classification) undermines the validity of the “curative” resection claims.

 

Answer: We appreciate your comment. In our study, we only enrolled total colonoscopic examination including cold snare polypectomy for benign lesions including SSL and adenoma. Patients receiving endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD). Thus, T1 cancer was not resected in this study. Histopathological assessment was performed according to WHO classification. We mentioned it in the Method section.

 

Method section

Cases in which cold snare polypectomy was performed for benign lesions less than 10 mm in size were also included [9].

 

Exclusion criteria were as follows: 1. use of H-PEG for bowel preparation; 2. in-complete clinical data, 3. incomplete colonoscopy and 4. procedures involving endoscopic mucosal resection and endoscopic submucosal dissection.

 

Histopathological diagnosis of lesions, including adenomas and SSLs, was made according to the 2019 WHO classification [18].

 

4. Adverse Events and Follow-up

The reporting of adverse events is incomplete. The authors mention “bleeding” and “perforation” but do not adhere to the ASGE (American Society for Gastrointestinal Endoscopy) lexicon for adverse events. There is no distinction between intraprocedural and delayed events, nor is there a standardized grading of severity (mild, moderate, severe). Furthermore, the follow-up period is heterogeneous, and the rate of loss to follow-up (if any) is not reported, which compromises the assessment of long-term recurrence or delayed perforation.

 

Answer: Thank you for your comment. We only included cases receiving colonoscopy and colonoscopy including cold snare polypectomy in this study. There were no periprocedural and post-procedural severe adverse events in the two groups. We mentioned the definition of adverse events in the Method section and these results in the result section. Long term follow-up about the recurrence was not examined in this study. We wrote it in the Limitation section.

 

Method section

Regarding adverse events, severe periprocedural and delayed bleeding and perforation were recorded. Delayed bleeding was defined as post-procedural bleeding requiring endoscopic hemostasis or a decrease in hemoglobin of more than 2 g/dL within 30 days after procedure [17]. Delayed perforation was defined as the presence of free air detected by computed tomography within 14 days of the procedure, in the absence of an intraprocedural perforation [17].

 

Results section

There were no periprocedural and postprocedural adverse events in the two groups.

 

Limitation section

Sixth, long-term outcomes including the recurrence of the resected lesion were not examined.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear author,
congratulations on your paper. I believe it is an original idea that can be helpful in clinical practice.

I would like to make some suggestions for improving your paper:

- I advise you to revise the paper to make it easier to read. Some sentences are very long and difficult to grasp. Also, please check for some spelling errors.

- Regarding the limitations of the study:

Recognize that the patient selection method can lead to selection bias.

The comparison of the two groups must be done with caution because the selection method does not follow clear guidelines.

- Is the classification of the bubbles validated? Internally/externally? If not, I think it would be advisable to mention this in the paper.

- Recognizing the weaknesses of the paper means that the conclusions must be commented on with caution because they could be the result of an interpretation based on significant biases. I believe this paper could be the start of a line of research aimed at improving the performance and efficiency of colonoscopies, and is therefore very interesting.
Congratulations!

Comments on the Quality of English Language

I believe the work has no language problems, but I am not an expert on the subject.

Author Response

Reviewer 2

Dear author,

congratulations on your paper. I believe it is an original idea that can be helpful in clinical practice.

 

I would like to make some suggestions for improving your paper:

 

- I advise you to revise the paper to make it easier to read. Some sentences are very long and difficult to grasp. Also, please check for some spelling errors.

 

Answer: Thank you for your comments. We received Englich correlation for the whole paper by a native speaker; Hardesh Dhillon who supported our study and amended our paper. Accordingly, we would like to add this person as a co-author.

 

- Regarding the limitations of the study:

 

Recognize that the patient selection method can lead to selection bias.

 

The comparison of the two groups must be done with caution because the selection method does not follow clear guidelines.

 

Answer: Thank you for your comment. We revised the limitation section referring the selection bias and the comparison between the two groups.

 

Limitation section

Several limitations of this study warrant acknowledgement. First, this was a retrospective study with a relatively small sample size, limiting statistical power. Second, patient enrollment was non-consecutive and bowel preparation regimens were selected at the discretion of individual physicians, introducing a potential risk of selection bias. Third, the 480 mL OSS regimen was evaluated only in a Japanese population, whose body habitus is generally smaller than that of Western populations, which may limit generalizability. Fourth, the primary aim of our study purpose was not a direct com-parison between the SAR and SPR groups; accordingly, no statistical adjustment for confounding variables was performed and multivariate logistic regression was not conducted for between group comparisons.

 

- Is the classification of the bubbles validated? Internally/externally? If not, I think it would be advisable to mention this in the paper.

 

Answer: We appreciate your comment. The validation of bubbles scores was not examined in the current study. We performed it in the previous study [reference 8]. We mentioned it in the Method section and Limitation section.

 

Method section

Our previous study showed that the interobserver agreement rate of the BBS-C among three physicians was 75% (75/100) in an internal validation [8].

 

Limitation section

Finally, formal external validation of bubble score was not performed in the current study; prospective validation in larger, independent cohorts is warranted.

 

- Recognizing the weaknesses of the paper means that the conclusions must be commented on with caution because they could be the result of an interpretation based on significant biases. I believe this paper could be the start of a line of research aimed at improving the performance and efficiency of colonoscopies, and is therefore very interesting.

Congratulations!

 

Answer: Thank you for your comment. I changed the conclusion in the abstract and the main text with caution.

 

Abstract

Conclusions: Severe bowel bubble formation was associated with longer cecal insertion and procedure times and appeared to be more frequent with a longer preparation-to-colonoscopy interval.

 

Main text

Conclusions

Severe bowel bubble formation was associated with longer cecal insertion and procedure times and appeared to be more frequent with a longer preparation-to-colonoscopy interval. Optimizing colonoscopy scheduling relative to the completion of bowel preparation may help reduce severe bubble burden and improve procedural efficiency.

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have been responsive to previous feedback and have significantly improved the manuscript's transparency regarding its design limitations. The research question is valid, and the finding linking the preparation-to-colonoscopy interval with bubble formation is clinically relevant.

However, the fundamental contradiction between the stated descriptive aim and the comparative presentation of data remains a major flaw that must be rectified. Additionally, the omission of withdrawal time data is a significant weakness.

Given that these issues can be addressed without further data collection (by revising the text, tables, and discussion), the manuscript is close to being acceptable.

1. Contradictory Framing of the Study: Descriptive Aim vs. Comparative Presentation

The authors have rightly clarified in their response and in the revised text that the study was not intended to be a direct comparison between the SAR and SPR groups due to significant selection bias. This is an appropriate and honest framing. However, the manuscript's structure, tables, and presentation still strongly imply a comparison.

• Tables 1 and 2: These tables present p-values directly comparing the SAR and SPR groups for multiple variables (e.g., insertion time, total procedure time, cleansing time). Presenting a p-value of 0.032 for insertion time, for instance, inherently suggests a statistically significant difference between the two independent groups, which contradicts the authors' statement that the p-values are merely "descriptive." This is confusing for the reader and scientifically unsound. If the groups are not meant to be compared, no inferential statistics should be used between them.
• Recommendation: To align the manuscript with its stated descriptive goal, all p-values comparing the SAR and SPR groups in Tables 1 and 2 should be removed. The text should also be revised to avoid comparative language (e.g., instead of "cleansing time was longer in the SPR group (p=0.021)," state "The median cleansing time was 2.0h in the SAR group and 2.7h in the SPR group."). This will make the manuscript more robust and less open to misinterpretation.

2. Insufficient Detail on Critical Endoscopic Quality Metrics

While the authors have clarified that all procedures were performed by experts and that total procedure time includes withdrawal time, a critical quality metric is missing: the documented withdrawal time itself. Adenoma Detection Rate (ADR) is highly dependent on withdrawal time.

• Withdrawal Time: The standard quality benchmark is a mean withdrawal time of at least 6 minutes in screening/surveillance colonoscopies. Without reporting the mean/median withdrawal time for each group (or for the cohort as a whole), the reported ADRs are difficult to interpret and their validity is diminished. The fact that severe bubbles were associated with a longer total procedure time without affecting ADR could simply be because endoscopists spent more time suctioning bubbles, thus artificially normalizing the withdrawal/inspection time.
• Recommendation: The authors should report the mean or median withdrawal time for the cohorts. If this data is unavailable, it must be stated as a major limitation, as it is a critical confounder for the ADR analysis.

3. Need for Justification and Broader Context for the Bubble Scoring System (BBS-C)

The authors use their own "Bubble Scoring System of Colonoscopy (BBS-C)." While they cite an internal validation study with 75% inter-observer agreement, this level of agreement is only moderate. Furthermore, other validated scoring systems, such as the Colon Endoscopic Bubble Scale (CEBuS), have been published.

• Justification: The manuscript would be strengthened if the authors briefly explained the rationale for developing their own scale. What limitations of existing scales does the BBS-C address?
• External Validity: The lack of external validation should be more prominently featured as a limitation. Readers should understand that the reported prevalence of "severe bubbles" is specific to this instrument and may not be comparable to rates reported using other scales.
• Recommendation: Add a sentence to the Methods section justifying the use of the BBS-C over other existing scales. In the Discussion/Limitations, explicitly state that the BBS-C has not been externally validated and that the reported prevalence rates should be interpreted in this context.

4. Typographical Errors and Minor Data Presentation Issues

There are several errors in the manuscript that undermine its professionalism and require correction.

• Study Period: The study period is listed as "between October 2024 and December 2025." This is in the future and must be a typographical error. This needs immediate correction.
• Table 4 Title: The title is "Risk factors of bowel bubble in the proximal colon." However, the table includes outcomes that are consequences, not risk factors (e.g., insertion time, total procedure time). A more accurate title would be "Comparison of Clinical and Procedural Characteristics Based on the Presence of Severe Proximal Bubbles."
• Table 6 Data: In Table 6, the header for the "Adverse Events" column seems to contain summary data (20/19, 8.84) that is confusing and doesn't align with the sub-columns. This should be reformatted for clarity.
• Recommendation: The authors should carefully proofread the entire manuscript to correct these and any other typographical errors. The title of Table 4 should be revised, and the formatting of Table 6 should be improved.

Author Response

Reviewer 1

The authors have been responsive to previous feedback and have significantly improved the manuscript's transparency regarding its design limitations. The research question is valid, and the finding linking the preparation-to-colonoscopy interval with bubble formation is clinically relevant.

However, the fundamental contradiction between the stated descriptive aim and the comparative presentation of data remains a major flaw that must be rectified. Additionally, the omission of withdrawal time data is a significant weakness.

Given that these issues can be addressed without further data collection (by revising the text, tables, and discussion), the manuscript is close to being acceptable.

 

1. Contradictory Framing of the Study: Descriptive Aim vs. Comparative Presentation

The authors have rightly clarified in their response and in the revised text that the study was not intended to be a direct comparison between the SAR and SPR groups due to significant selection bias. This is an appropriate and honest framing. However, the manuscript's structure, tables, and presentation still strongly imply a comparison.

Tables 1 and 2: These tables present p-values directly comparing the SAR and SPR groups for multiple variables (e.g., insertion time, total procedure time, cleansing time). Presenting a p-value of 0.032 for insertion time, for instance, inherently suggests a statistically significant difference between the two independent groups, which contradicts the authors' statement that the p-values are merely "descriptive." This is confusing for the reader and scientifically unsound. If the groups are not meant to be compared, no inferential statistics should be used between them.

Recommendation: To align the manuscript with its stated descriptive goal, all p-values comparing the SAR and SPR groups in Tables 1 and 2 should be removed. The text should also be revised to avoid comparative language (e.g., instead of "cleansing time was longer in the SPR group (p=0.021)," state "The median cleansing time was 2.0h in the SAR group and 2.7h in the SPR group."). This will make the manuscript more robust and less open to misinterpretation.

Answer: Thank you for your comment. As the reviewer pointed out, we deleted p-value in Table 1 and Table 2. The text about the comparison was also removed.

 

2. Insufficient Detail on Critical Endoscopic Quality Metrics

While the authors have clarified that all procedures were performed by experts and that total procedure time includes withdrawal time, a critical quality metric is missing: the documented withdrawal time itself. Adenoma Detection Rate (ADR) is highly dependent on withdrawal time.

Withdrawal Time: The standard quality benchmark is a mean withdrawal time of at least 6 minutes in screening/surveillance colonoscopies. Without reporting the mean/median withdrawal time for each group (or for the cohort as a whole), the reported ADRs are difficult to interpret and their validity is diminished. The fact that severe bubbles were associated with a longer total procedure time without affecting ADR could simply be because endoscopists spent more time suctioning bubbles, thus artificially normalizing the withdrawal/inspection time.

Recommendation: The authors should report the mean or median withdrawal time for the cohorts. If this data is unavailable, it must be stated as a major limitation, as it is a critical confounder for the ADR analysis.

 

Answer: Thank you for your comments. However, the data about the withdrawal time is not available. We stated it in a limitation section.

 

Limitation section

Withdrawal time was not assessed in this study, although it has been reported to be significantly associated with ADR [27]. Therefore, the ADRs reported here should be interpreted with this limitation in mind.

Reference

27. Barclay RL, Vicari JJ, Doughty AS, et al. Colonoscopic withdrawal times and adenoma detection during screening colonoscopy. N Engl J Med. 2006; 355: 2533–2541.

 

3. Need for Justification and Broader Context for the Bubble Scoring System (BBS-C)

The authors use their own "Bubble Scoring System of Colonoscopy (BBS-C)." While they cite an internal validation study with 75% inter-observer agreement, this level of agreement is only moderate. Furthermore, other validated scoring systems, such as the Colon Endoscopic Bubble Scale (CEBuS), have been published.

Justification: The manuscript would be strengthened if the authors briefly explained the rationale for developing their own scale. What limitations of existing scales does the BBS-C address?

External Validity: The lack of external validation should be more prominently featured as a limitation. Readers should understand that the reported prevalence of "severe bubbles" is specific to this instrument and may not be comparable to rates reported using other scales.

Recommendation: Add a sentence to the Methods section justifying the use of the BBS-C over other existing scales. In the Discussion/Limitations, explicitly state that the BBS-C has not been externally validated and that the reported prevalence rates should be interpreted in this context.

 

Answer: Thank you for your comment. We added the sentence as the reviewer pointed in the Method section and Discussion section.

Method section

Although the existing scale defines severe bubbles as involvement of ≥50% of the visual field, this threshold may include cases that do not interfere with the examination [10]. Therefore, in our BBS-C, we applied a stricter definition, setting severe bubbles as involvement of ≥70% of the visual field to capture only those that could potentially impair the procedure.

Discussion section

The BBS-C has not been externally validated, and therefore the prevalence rates of bubbles and other analysis reported in this study should be interpreted with this limitation in mind.

 

4. Typographical Errors and Minor Data Presentation Issues

There are several errors in the manuscript that undermine its professionalism and require correction.

Study Period: The study period is listed as "between October 2024 and December 2025." This is in the future and must be a typographical error. This needs immediate correction.

Table 4 Title: The title is "Risk factors of bowel bubble in the proximal colon." However, the table includes outcomes that are consequences, not risk factors (e.g., insertion time, total procedure time). A more accurate title would be "Comparison of Clinical and Procedural Characteristics Based on the Presence of Severe Proximal Bubbles."

Table 6 Data: In Table 6, the header for the "Adverse Events" column seems to contain summary data (20/19, 8.84) that is confusing and doesn't align with the sub-columns. This should be reformatted for clarity.

Recommendation: The authors should carefully proofread the entire manuscript to correct these and any other typographical errors. The title of Table 4 should be revised, and the formatting of Table 6 should be improved.

 

Answer: Thank you for your comment. The study period is correct. This study was approved by our IRB as a part of the large-scale prospective and retrospective endoscopic study. We included this in the Methods section to prevent potential misunderstanding by readers. Additionally, as the reviewer suggested, we changed the title of Table 4 as “Comparison of clinical and procedural characteristics based on the presence of severe proximal bubbles.” We also amended Table 6.