Does a Preoperative Carbohydrate Drink Reduce Postoperative Inflammation? A Systematic Review and Meta-Analysis
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
I would like to thank the authors for allowing me to review this SRMA looking at the post-surgical inflammatory response after the exposure of either carbohydrate drink (CHD) or fasting for more than 6 hours. Ten RCTs are identified, all with unknown or high risk of bias. Inflammation was reduced with CHD, with IL-6 at 24 hours as the primary outcome. Despite this favouring CHD, the secondary outcomes process to be insignificant. The limitations are well documented, as is the lack of registration of the methodology (PROSPERO or otherwise). I believe this is an adequate SRMA with a clear message.
Author Response
Reviewer 1:
Comment: I would like to thank the authors for allowing me to review this SRMA looking at the post-surgical inflammatory response after the exposure of either carbohydrate drink (CHD) or fasting for more than 6 hours. Ten RCTs are identified, all with unknown or high risk of bias. Inflammation was reduced with CHD, with IL-6 at 24 hours as the primary outcome. Despite this favouring CHD, the secondary outcomes process to be insignificant. The limitations are well documented, as is the lack of registration of the methodology (PROSPERO or otherwise). I believe this is an adequate SRMA with a clear message.
Author response: We thank you for your valuable time and your kind comments.
Reviewer 2 Report
Comments and Suggestions for Authors
* Thanks for opportunity to review.
* Overall study should present both biomarker data (IL-6, CRP, Albumin, insulin sensitivity) and clinical data for relevance to policy-makers, clinicians and patients.
* Similar meta-analyses of CHO, placebo or fasting preoperatively include discussion of clinically meaningful outcomes such as morbidity including surgical infections, hospital LOS, nausea and vomiting, insulin sensitivity or rates of aspiration (see Ricci et al 2021, Cheng et al 2021, Tong et al 2022). This should be part of analysis and how this correlates with biochemical data.
* Specific comments on paper:
- 65 - Previous trials have included analysis of inflammatory markers and clinical outcomes. This would be the appropriate for this study as well.
- 85 - No evidence of literature search of grey literature such as clinicaltrials.gov, conference proceedings or experts in the field.
- Table 1 - why exclude diabetic patients - no rationale is described in the paper.
- Figure 1 - exclusion of studies with absent means and SD. Was there an attempt to contact the primary authors and obtain the data to analyse?
- Table 2 - final column describes relevant outcomes. Presented in a confusing manner. Would be appropriate to list which outcomes were measured. The relevant cohorts column allows for understanding of which primary and secondary outcome was measured.
- 178-183 - Why did all trials have unclear risk of bias in D5? This is not described and is dissimilar to other reviews on the same topic.
- Overall discussion has limited structure and is convoluted. This needs significant revision to make succinct and relevant comments on the conclusions of the study in the bigger picture of published literature.
- Discussion : The clinical significance of the measured values of biomarkers reduction and albumin increase should be elaborated on. What does a -21.6 pg/mL mean? What does a 2 g/L increase in albumin concentration mean?
- Discussion : The strength of the conclusion that there is reduced inflammation is not necessarily supported by this study. It does show reduction in IL-6, but not a significant reduction in CRP.
- Discussion : 345 - Elevated IL-6 levels post operatively are associated with increased morbidity, however interventions that reduce IL-6, eg preoperative CHO loading need clinical outcome measurement to demonstrate reduction in morbidity.
- Discussion - Aspiration is a rare event. These studies will not be powered to look for a difference in outcomes. This should be noted when discussing safety conclusions.
Author Response
Reviewer 2:
Comment 1: - 65 - Previous trials have included analysis of inflammatory markers and clinical outcomes. This would be the appropriate for this study as well.
Author response: We wanted specifically to focus on postoperative inflammatory markers. The very recent systematic reviews included clinical outcomes, but we felt they didn’t adequately demonstrate the most up-to-date information concerning inflammatory markers. Furthermore, given the protocol for this review was created before conducting the review, it would be inappropriate for us to now include other outcome measures in this study. Thank you for this comment.
Comment 2: - 85 - No evidence of literature search of grey literature such as clinicaltrials.gov, conference proceedings or experts in the field.
Author response: It’s true that we didn’t include these things. Since RCTs represent the best form of evidence, we decided to restrict out search to RCTs. Conference proceedings therefore would be excluded as is done for most systematic reviews. As for lack of search on clinicaltials.gov, this was again deliberate as most often results were not available and most studies were still recruiting. Results posted on clincialtrial.gov are also not peer reviewed and hence would attract criticism. As for consulting experts in the field, again this is not peer reviewed and hence all of these inclusions would be inappropriate for a systematic review with clear inclusion/exclusion criteria. Thank you for this comment.
Comment 3: - Table 1 - why exclude diabetic patients - no rationale is described in the paper.
Author response: A reason for why diabetic patients were excluded has been added to the inclusion and exclusion criteria section (See lines 97-99).
Comment 4: - Figure 1 - exclusion of studies with absent means and SD. Was there an attempt to contact the primary authors and obtain the data to analyse?
Author response: Yes, every author was contacted multiple times with numerous email addresses, but none of them responded and provided the requested figures. A sentence to this effect has been added in lines 110-112.
Comment 5: - Table 2 - final column describes relevant outcomes. Presented in a confusing manner. Would be appropriate to list which outcomes were measured. The relevant cohorts column allows for understanding of which primary and secondary outcome was measured.
Author response: Upon reflection, we agree that the relevant outcomes column is both confusing to the reader and also unnecessary given the relevant cohorts column already provides half of this information. We have changed the column to just include which inflammatory markers are measured and the notes below have been removed.
Comment 6: - 178-183 - Why did all trials have unclear risk of bias in D5? This is not described and is dissimilar to other reviews on the same topic.
Author response: The reason we felt they all had some concerns is because they didn’t mention if what they actually collected was from a pre-determine protocol. Thus, despite in other aspects of D5 being unbiased, it is not possible to award them with a green bias score. A justification has been added in lines 191-194.
Comment 7: - Overall discussion has limited structure and is convoluted. This needs significant revision to make succinct and relevant comments on the conclusions of the study in the bigger picture of published literature.
Author response: We have deleted some text and improved the flow of the discussion as per this comment.
Comment 8: - Discussion : The clinical significance of the measured values of biomarkers reduction and albumin increase should be elaborated on. What does a -21.6 pg/mL mean? What does a 2 g/L increase in albumin concentration mean?
Author response: We mentioned in the introduction that serum albumin is reduced and IL-6 is increased when inflammation increases and described the reasoning for this. However, we are not making the argument that these changes are clinically significant according to conventional guidelines. As mentioned in the discussion, the case for the benefit of preoperative carbohydrate drinks has already been established and is already what the guidelines recommend, so clinicians should already be giving them to patients. This study wanted to explore the benefit from a different perspective, to see if from an inflammatory point of view, the there are significant changes. The Il-6 change however, in theory, could reduce the patients from a high postoperative level to a lower level, but it isn’t larger than the range of an entire level.
Comment 9: - Discussion : The strength of the conclusion that there is reduced inflammation is not necessarily supported by this study. It does show reduction in IL-6, but not a significant reduction in CRP.
Author response: The discussion and conclusion has been changed. We no longer suggest that there is a reduced inflammation, rather there may be a reduced inflammation given the reduction in some inflammatory markers.
Comment 10: - Discussion : 345 - Elevated IL-6 levels post operatively are associated with increased morbidity, however interventions that reduce IL-6, eg preoperative CHO loading need clinical outcome measurement to demonstrate reduction in morbidity.
Author response: We agree and now have added a sentence in lines 522-523 to reflect this.
Comment 11: - Discussion - Aspiration is a rare event. These studies will not be powered to look for a difference in outcomes. This should be noted when discussing safety conclusions.
Author response: We feel our comment about the lack of risk of aspiration is appropriate to emphasise that preoperative drinks are not harmful. This has been demonstrated through studies looking at stomach emptying and clinically there isn’t evidence that preoperative drinks have increased aspiration. Thank you for this comment. We have however, added lines 542-543 to draw the reader to the caution mentioned here by the reviewer in terms of powering the study adequately to make definitive conclusions about aspiration.
Round 2
Reviewer 2 Report
Comments and Suggestions for Authors
Line 433 and 434 needs a minor modification to make sense.
Author Response
Reviewer comment: Line 433 and 434 needs a minor modification to make sense.
Author response: We have amended lines 433-436 in response to the reviewer comment.
