Risk Factors Associated with Postoperative Nausea and Vomiting After Esophagogastroduodenoscopy
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsRisk Factors Associated with Postoperative Nausea and Vomiting after Esophagogastroduodenoscopy.
The manuscript under review presents a single-center, retrospective study.
Abstract. The abstract is concise, yet providing sufficient information to understand the study question, main results, and conclusions.
Introduction. The introduction of the manuscript is concise and well-written, providing sufficient background information to understand the importance of the study question.
Methodology. Observational and retrospective studies are expected to comply with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines. It is stated in the methodology that the STROBE guidelines were followed. Ethical approval was obtained.
Line 98: Although the endoscopic procedures were performed by a single gastroenterologist, no information is provided on the anesthesia / anesthesiologist involved or sedation protocols, which may represent an additional source of variability.
The description of the methodology is well written and provides sufficient information to ensure the reproducibility of the study.
The patient selection is described and includes both inclusion and exclusion criteria.
A concern regarding patient selection is that reliance solely on ASA physical status may introduce confounding, as this classification does not capture clinically relevant differences between comorbidities. For instance, an ASA II patient with hypertension differs from an ASA II patient with diabetes, the latter being associated with an increased risk of delayed gastric emptying.
Results. The results are presented clearly, though the table may benefit from improved formatting or organization.
In table 3: Complaint of nausea vomiting >> complaint of nausea or vomiting.
In the results section, it would be valuable to include statistical analyses stratified by specific comorbidities that may influence the risk of PONV, such as diabetes or chronic kidney disease.
Additionally, it would be advisable to analyse BMI as a categorical variable rather than as a continuous one. For instance, patients with BMI > 40 (morbid obesity) may constitute a distinct clinical subgroup, often classified as ASA III.
Discussion and conclusion. Only 152 out of 520 patients (~29%) were included due to missing data. This raises serious concerns about external validity. Patients with complete records may differ systematically: this limitation is acknowledged, but its impact is underestimated. Moreover, the absence of data on the anesthesiologist involved may have introduced selection bias, as the completeness of records could depend on individual documentation practices.
It is stated in line 229 that “this finding suggests that the Apfel score, originally developed for surgical populations, may also have applicability in patients undergoing EGD under sedation”. However, this interpretation should be approached with caution. The incidence and risk profile of postoperative nausea and vomiting (PONV) may differ substantially between procedures performed under general anaesthesia and those conducted under sedation. General anaesthesia, particularly when involving volatile agents and perioperative opioids, is associated with a higher risk of PONV, whereas sedation-based techniques—especially those using propofol—are generally linked to a lower incidence due to their intrinsic antiemetic properties. Importantly, most established risk prediction models, including the Apfel score, were developed and validated in surgical populations undergoing general anaesthesia. Consequently, their applicability to endoscopic procedures performed under sedation remains uncertain. Therefore, although the present retrospective study found a higher incidence of PONV in patients with higher Apfel scores, this observation should not be directly extrapolated to broader clinical settings without further prospective validation.
The limitation of the follow-up period may lead to an underestimation of true incidence.
Author Response
We thank the reviewer for the careful evaluation of our manuscript and for the constructive comments and suggestions. In response to these valuable recommendations, the manuscript was extensively revised to improve methodological transparency, statistical interpretation, clarity of reporting, and overall discussion of the findings.
Comment 1: The manuscript under review presents a single-center, retrospective study. The abstract is concise, yet provides sufficient information to understand the study question, main results, and conclusions. The introduction of the manuscript is concise and well-written, providing sufficient background information to understand the importance of the study question. Observational and retrospective studies are expected to comply with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines. It is stated in the methodology that the STROBE guidelines were followed. Ethical approval was obtained.
Response 1: We thank the reviewer for the positive and encouraging evaluation of our manuscript. We also appreciate the acknowledgment regarding adherence to STROBE guidelines and ethical approval procedures.
Comment 2: Line 98: Although the endoscopic procedures were performed by a single gastroenterologist, no information is provided on the anesthesia / anesthesiologist involved or sedation protocols, which may represent an additional source of variability.
Response 2: We thank the reviewer for this important comment. Additional information regarding the anesthesiologist involved, anesthesia evaluation, monitoring, oxygen supplementation, and sedation protocol has been added to the Materials and Methods section. During the study period, all procedures were performed by a single gastroenterologist and sedation was administered by a single anesthesiologist (the corresponding author), as no other anesthesiology staff were involved in endoscopic procedures at that time. This ensured a high level of consistency in both procedural and anesthetic management and reduced variability related to anesthetic practice and documentation. In addition, all eligible patients within the study period were included based on predefined criteria, without selection based on outcomes. These clarifications have now been incorporated into the revised manuscript (Materials and Methods section, Lines 103–115).
Comment 3: The description of the methodology is well written and provides sufficient information to ensure the reproducibility of the study. The patient selection is described and includes both inclusion and exclusion criteria.
Response 3: We thank the reviewer for this positive assessment of the methodology and patient selection process. In accordance with the reviewer’s additional comments, the description of anesthesia practice and sedation management was further expanded and clarified in the Materials and Methods section to improve methodological transparency and reproducibility.
Comment 4: A concern regarding patient selection is that reliance solely on ASA physical status may introduce confounding, as this classification does not capture clinically relevant differences between comorbidities. For instance, an ASA II patient with hypertension differs from an ASA II patient with diabetes, the latter being associated with an increased risk of delayed gastric emptying.
Response 4: We thank the reviewer for this important observation. We agree that ASA physical status alone may not fully reflect clinically relevant differences between comorbidities. Therefore, additional subgroup analyses evaluating specific comorbidities and medication use were incorporated into the revised manuscript. Associations between PONV and diabetes mellitus, thyroid disease, chronic kidney disease, oral antidiabetic drug use, insulin use, and antihypertensive drug use were analyzed separately and added to the Results and Discussion sections. In addition, Table 1 and Table 4 were revised to include these newly analyzed clinical variables and medication-related data. The Abstract, Discussion, and Conclusion sections were also updated accordingly to reflect these additional analyses and findings (Results section, Lines 199- 209; Discussion section, Lines 270–295 Tables 1 and 4).
Comment 5: he results are presented clearly, though the table may benefit from improved formatting or organization.
Response 5: We thank the reviewer for this valuable suggestion. The tables were revised and reorganized to improve clarity, readability, and overall presentation. Additional subgroup analyses and newly evaluated clinical variables were also incorporated into the revised tables to provide a more comprehensive presentation of the findings.
Comment 6: In Table 3: “Complaint of nausea vomiting” should be revised to “complaint of nausea or vomiting”.
Response 6: We thank the reviewer for identifying this issue. The wording in Table 3 was revised from “Complaint of nausea vomiting” to “Complaint of nausea or vomiting” for improved clarity and accuracy.
Comment 7: In the results section, it would be valuable to include statistical analyses stratified by specific comorbidities that may influence the risk of PONV, such as diabetes or chronic kidney disease.
Response 7: We thank the reviewer for this valuable suggestion. Additional subgroup analyses evaluating specific comorbidities and medication-related variables potentially associated with PONV were incorporated into the revised manuscript. Associations between PONV and diabetes mellitus, thyroid disease, chronic kidney disease, oral antidiabetic drug use, insulin use, and antihypertensive drug use were analyzed separately and added to the Results section. These findings were also incorporated into the Discussion, Abstract, and Conclusion sections, and the relevant data were added to Table 1 and Table 4. (Results section, Lines 200-209)
Comment 8: Additionally, it would be advisable to analyse BMI as a categorical variable rather than as a continuous one. For instance, patients with BMI > 40 (morbid obesity) may constitute a distinct clinical subgroup, often classified as ASA III.
Response 8: We thank the reviewer for this important suggestion. In addition to continuous BMI analysis, BMI was also evaluated as a categorical variable in the revised manuscript. Patients were stratified according to BMI <30 and BMI ≥30, and the association between obesity and PONV was analyzed separately. A separate analysis for morbid obesity (BMI >40) could not be performed because of the limited number of patients within this subgroup. These results were added to the Results section and incorporated into Table 4. In addition, the potential relationship between obesity and PONV was further discussed in the Discussion section. (Results section, Li,nes 195-198; Discussion section , Lines 255-269)
Comment 9: Only 152 out of 520 patients (~29%) were included due to missing data. This raises serious concerns about external validity. Patients with complete records may differ systematically: this limitation is acknowledged, but its impact is underestimated. Moreover, the absence of data on the anesthesiologist involved may have introduced selection bias, as the completeness of records could depend on individual documentation practices.
Response 9: We thank the reviewer for this important and thoughtful comment. The Limitations section was substantially revised to more explicitly emphasize the potential impact of incomplete data, selection bias, and restricted external validity. We also clarified that all procedures were performed by a single gastroenterologist and sedation was administered by a single anesthesiologist during the study period, thereby reducing variability related to anesthetic practice and documentation habits. In addition, we emphasized that the high rate of incomplete data likely reflects real-world limitations in the routine assessment and documentation of PONV-related risk factors in endoscopic units rather than systematic exclusion based on patient outcomes. These limitations are now discussed in greater detail in the revised manuscript. (Limitations section, Lines 341-351)
Comment 10: It is stated in line 229 that “this finding suggests that the Apfel score, originally developed for surgical populations, may also have applicability in patients undergoing EGD under sedation”. However, this interpretation should be approached with caution. The incidence and risk profile of postoperative nausea and vomiting (PONV) may differ substantially between procedures performed under general anaesthesia and those conducted under sedation. General anaesthesia, particularly when involving volatile agents and perioperative opioids, is associated with a higher risk of PONV, whereas sedation-based techniques—especially those using propofol—are generally linked to a lower incidence due to their intrinsic antiemetic properties. Importantly, most established risk prediction models, including the Apfel score, were developed and validated in surgical populations undergoing general anaesthesia. Consequently, their applicability to endoscopic procedures performed under sedation remains uncertain. Therefore, although the present retrospective study found a higher incidence of PONV in patients with higher Apfel scores, this observation should not be directly extrapolated to broader clinical settings without further prospective validation.
Response 10: We thank the reviewer for this insightful and valuable comment. We agree that the applicability of the Apfel score to endoscopic procedures performed under sedation should be interpreted with caution, as the score was originally developed and validated in surgical populations undergoing general anesthesia. In response to this comment, the Discussion section was revised to provide a more balanced interpretation of our findings. Additional discussion regarding the differences between general anesthesia and propofol-based sedation, including the intrinsic antiemetic properties of propofol and the potentially different PONV risk profile in sedated endoscopic procedures, was incorporated into the revised manuscript. We also emphasized that the applicability of the Apfel score in patients undergoing EGD under sedation remains uncertain and requires further prospective validation before broader clinical extrapolation. (Discussion section, Lines 296-306, 319-322)
Comment 11: The limitation of the follow-up period may lead to an underestimation of true incidence.
Response 11: We thank the reviewer for this important comment. The limitation related to the follow-up period was emphasized more explicitly in the revised manuscript. We clarified that PONV assessment was limited to the immediate postoperative period in the PACU and that delayed symptoms occurring after discharge were not evaluated. We also expanded the Discussion and Limitations sections to address the potential underestimation of the true incidence of PONV and the clinical relevance of PDNV.
Reviewer 2 Report
Comments and Suggestions for AuthorsThis manuscript investigates the incidence and associated factors of postoperative nausea and vomiting (PONV) in patients undergoing esophagogastroduodenoscopy (EGD) under sedation. The topic is clinically relevant, particularly given the increasing volume of ambulatory endoscopic procedures and the impact of PONV on patient comfort and discharge efficiency.
The study provides real-world data and identifies several factors associated with PONV, including female sex, higher body mass index, preoperative nausea, and Apfel score. However, despite these strengths, the manuscript has significant methodological and reporting limitations that currently preclude publication in its present form.
1) All findings are based on univariate comparisons. This is a critical limitation. Without multivariate analysis: no independent predictors can be identified and reported “risk factors” are not adjusted for confounders. The authors should: perform logistic regression analysis OR and clearly state that findings represent associations only.
2) PONV was assessed only during the PACU stay. This excludes delayed PONV (0–24 h) and likely underestimates true incidence. This limitation should be explicitly emphasized and discussed in relation to existing literature.
3) PONV definition is binary and based on retrospective records. lack of severity grading, potential underreporting and variability in documentation are problem. Authors should clarify how nausea was recorded and whether any scoring system was used.
4)There should have been a section on the methods part regarding the anesthesia preparation process.
5) There are almost no references from the last 5 years. Your references could be more up-to-date.
6)The study appears to have been registered after data collection was completed, which raises concerns regarding transparency and potential selective reporting.
7)The effect of the anesthetic agents used in the discussion section on PONV could have been understood in more detail.
8)Regarding the limitations, the following should be added high data loss (520 → 152), selection bias → should be emphasized more strongly, and PACU-only evaluation → early PONV.
9)The discussion section is too long and repetitive, the focus is weak, and the statistical limitations are not discussed thoroughly enough. Some comments are overly assertive
Author Response
We thank the reviewer for the careful evaluation of our manuscript and for the constructive comments and suggestions. In response to these valuable recommendations, the manuscript was extensively revised to improve methodological transparency, statistical interpretation, clarity of reporting, and overall discussion of the findings.
Comment 1: All findings are based on univariate comparisons. This is a critical limitation. Without multivariate analysis: no independent predictors can be identified and reported “risk factors” are not adjusted for confounders. The authors should: perform logistic regression analysis OR and clearly state that findings represent associations only.
Response 1: We thank the reviewer for this important and insightful comment. We agree that the absence of multivariate analysis represents an important limitation of the study. However, due to the relatively small number of PONV events and the presence of complete separation, with all PONV cases occurring in female patients, a reliable multivariate logistic regression analysis could not be performed. In response to this comment, the manuscript was revised to avoid overinterpretation of the findings. Throughout the revised manuscript, the terminology was modified to emphasize that the identified variables represent associations with PONV rather than independent predictors or causal risk factors. In addition, this limitation was explicitly discussed in the Limitations section. ( Results section, Lines 231-235; Limitations section, Lines 363-366)
Comment 2: PONV was assessed only during the PACU stay. This excludes delayed PONV (0–24 h) and likely underestimates true incidence. This limitation should be explicitly emphasized and discussed in relation to existing literature.
Response 2: We thank the reviewer for this important comment. The limitation related to the follow-up period was emphasized more explicitly in the revised manuscript. We clarified that PONV assessment was limited to the immediate postoperative period in the PACU and that delayed PONV occurring after discharge within the first 24 hours was not evaluated. The Discussion and Limitations sections were expanded accordingly to address the potential underestimation of the true incidence of PONV and the clinical relevance of PDNV in relation to the existing literature. In addition, the importance of extended follow-up in future studies evaluating PONV after endoscopic procedures was emphasized. (Limitations section, Lines 352-359)
Comment 3: PONV definition is binary and based on retrospective records. lack of severity grading, potential underreporting and variability in documentation are problem. Authors should clarify how nausea was recorded and whether any scoring system was used.
Response 3: We thank the reviewer for this important comment. The Methods section was revised to clarify how PONV was assessed and recorded in the present study. PONV was defined as patient-reported nausea and/or documented vomiting during the PACU stay based on routine postoperative clinical records. Due to the retrospective design of the study, no validated nausea severity scoring system was consistently used, and symptoms were recorded as a binary outcome according to routine clinical documentation. In response to this comment, the potential impact of underreporting, lack of standardized severity grading, and variability in documentation practices was also emphasized more explicitly in the Limitations section. (Material and methods section, Lines 147-149; Limitations, Lines 349-351)
Comment 4: There should have been a section on the methods part regarding the anesthesia preparation process.
Response 4: We thank the reviewer for this valuable suggestion. In response to this comment, the Materials and Methods section was expanded to provide a more detailed description of the anesthesia preparation and sedation process. Additional information regarding pre-procedural anesthesia evaluation, patient monitoring, oxygen supplementation, anesthetic management, and the sedation protocol was incorporated into the revised manuscript to improve methodological transparency and reproducibility. (Materials and methods section, Lines 103-115)
Comment 5: There are almost no references from the last 5 years. Your references could be more up-to-date.
Response 5: We thank the reviewer for this valuable suggestion. The reference list was revised and updated to include more recent literature published within the last five years. Several contemporary studies, systematic reviews, and updated consensus guidelines related to PONV, PDNV, propofol-based sedation, thyroid-related gastrointestinal dysmotility, and metformin-associated gastrointestinal adverse effects were incorporated into the revised manuscript to strengthen the scientific background and discussion.
Comment 6: The study appears to have been registered after data collection was completed, which raises concerns regarding transparency and potential selective reporting.
Response 6: We thank the reviewer for this important comment and apologize if the date format caused confusion. The study was designed as a retrospective observational analysis of patients who had already undergone EGD procedures between June and November 2023. Ethical approval for retrospective data analysis was obtained on 7 December 2023 in accordance with institutional procedures. ClinicalTrials.gov registration was subsequently completed to improve transparency and prospective accessibility of the study protocol.
Comment 7: The effect of the anesthetic agents used in the discussion section on PONV could have been understood in more detail.
Response 7: We thank the reviewer for this valuable suggestion. The Discussion section was expanded to provide a more detailed evaluation of the potential effects of anesthetic agents and sedation techniques on PONV. Additional discussion regarding the intrinsic antiemetic properties of propofol, the lower incidence of PONV associated with propofol-based sedation compared with volatile anesthetics and opioid-based techniques, and the limited variability in anesthetic exposure within our cohort was incorporated into the revised manuscript. (Discussion section, Lines 302-318)
Comment 8: Regarding the limitations, the following should be added high data loss (520 → 152), selection bias → should be emphasized more strongly, and PACU-only evaluation → early PONV.
Response 8: We thank the reviewer for this important comment. In response to this suggestion, the Limitations section was substantially revised. The high rate of excluded patients due to incomplete data (520 → 152) and its potential impact on selection bias and external validity were emphasized more explicitly. In addition, we clarified that PONV assessment was limited to the immediate postoperative period in the PACU, meaning that the reported incidence primarily reflects early PONV and may underestimate the true overall incidence due to the absence of post-discharge follow-up. (Limitations section, Lines 341-359)
Comment 9: The discussion section is too long and repetitive, the focus is weak, and the statistical limitations are not discussed thoroughly enough. Some comments are overly assertive
Response 9: We thank the reviewer for this constructive comment. In response to this suggestion, the Discussion section was carefully revised to improve clarity, focus, and overall flow. Repetitive statements were reduced, and several sections were reorganized to provide a more concise and balanced interpretation of the findings. In addition, the statistical limitations of the study, including the retrospective design, limited number of PONV events, absence of multivariate analysis, and potential selection bias, were discussed more explicitly. The wording throughout the manuscript was also revised to avoid overly assertive interpretations and to emphasize that the findings represent associations rather than causal relationships.
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsDear author, first of all, thank you for implementing my comments.
The article is clearer and more engaging for readers.
I have some minor suggestions.
Sincerely,
You haven't included the abbreviation EGD in the introduction.
There's no need to say Postoperative nausea and vomiting (PONV) in the methods section.
Both the methods and results sections use "visual analog scale (VAS)".
Please review these abbreviations throughout the article.
Author Response
Comment 1: You haven't included the abbreviation EGD in the introduction.
There's no need to say Postoperative nausea and vomiting (PONV) in the methods section.
Both the methods and results sections use "visual analog scale (VAS)".
Please review these abbreviations throughout the article.
Response 1: We thank the reviewer for these helpful minor suggestions. In response, the manuscript was carefully reviewed and revised for abbreviation consistency throughout the text. The abbreviation “EGD” was clarified in the Introduction, and redundant repeated definitions of previously introduced abbreviations were removed where appropriate. In addition, the abbreviations list was updated accordingly. We also performed an additional careful review of the manuscript and tables to correct minor language, spelling, grammar, and formatting inconsistencies.
