All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
Thank you for your careful and comprehensive revision. I have reviewed your responses to the previous reviewer’s comments, and I am satisfied that you have addressed all substantive points:
1) You clarified the exact parameters used in calculating the Cleveland Clinic Score and aligned them with the original Thakar et al. (2005) model.
2) You strengthened the statistical analysis by adding a bootstrapped ROC with confidence intervals and providing sensitivity and specificity values for the overall cohort and subgroups.
3) You adequately justified the retrospective design limitations, including the unavailability of urine output data, and clearly acknowledged these constraints in the manuscript.
4) You clarified the purpose and scope of the study, emphasizing that the aim was to evaluate CCS performance rather than develop a new predictive model.
As one reviewer did not provide feedback in this round, I conducted an independent evaluation and find your revisions satisfactory.
I am pleased to confirm that your manuscript is now accepted for publication. Congratulations on your work
[# PeerJ Staff Note - this decision was reviewed and approved by Celine Gallagher, a PeerJ Section Editor covering this Section #]
Title & Abstract
The title is informative and conveys the core ideas of the study, i.e. the investigation (CCS), disease (AKI) and the setting (Elective cardiac surgeries) in a retrospective manner.
The abstract is well structured and the authors have now added sensitivity and specificity to the results section. The p values are reported correctly. No further changes are suggested.
Introduction
The introduction has a logical flow, starting with the description of AKI, moving on to how cardiac surgeries predispose to AKI and then discussing the risk assessment scores. The CCS and its subsequent modifications have also been explained well and relevant studies have been cited. No further changes are suggested.
Figures & Tables
The figures and table are clear and free from unnecessary modification.
Material and Methods
The materials and methods section is clear. In the revision comments, the authors have clearly explained the limitations of using urine output as a parameter, which is plausible. Further, they have also explained their basis of calculation of CCS. I think there are no more corrections to be done in the methods section.
Results
The results are written nicely in the previous as well as current version of the manuscript. The authors have added sensitivity and specificity data in the revised version, including that for the CABG and AVR/AVP. There are no further changes suggested in the results section.
Discussion
The discussion looks complete in the revised version. They have also discussed the implications of high sensitivity and low specificity found in their results. They have confirmed the AUC for CCS via bootstrap analysis, addressing the concerns of the anonymous reviewer. No further changes are suggested.
Conclusion
The conclusions capture the main findings of the study well, and describe them in sufficient detail without going into more granular details. There is a clear message on improving the predictive accuracy of the CCS. No further changes are suggested.
**PeerJ Staff Note:** Please ensure that all review, editorial, and staff comments are addressed in a response letter and that any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.
Title & Abstract
The revised title has incorporated my suggestions and now conveys properly the key messages from the study, and also mentions the retrospective nature of the study.
The abstract has also been corrected, the p-values are reported correctly, and a separate conclusion section can now be seen, which is accurate as per the findings of the study.
Introduction
The authors have provided a detailed description of the CCS (lines 62-67), which now makes the introduction more reader-friendly. They have also added the context to explain how the CCS has been modified, as well as cited the suggested reference. The use of standard KDIGO terminology – AKI – is also appreciated. However, they do not need to expand the abbreviation AKI (line 54) once they have already done so.
Figures & Tables
The figures and tables are clear and free from unnecessary modification.
Material and Methods
The materials and methods section is clear, and the authors have incorporated the suggested changes, such as how they used the CCS for the patients and the justification of urine output criteria not used for defining the AKI due to the potential effect of postoperative diuretic use. The abbreviations LVEF and CCS have already been expanded in the introduction and should not be expanded again in the methods section.
Results
The results were written nicely in the previous version of the manuscript, and only one change was suggested regarding the mechanistic explanation of the findings, which has now been added. AKI and eGFR have already been expanded earlier and may not be expanded again.
Discussion
The discussion is well written, and only one change was suggested regarding the addition of the clinical implications of the findings. The authors have added these at the end of the discussion. However, they do not need to expand the abbreviations already used earlier in the manuscript.
Conclusion
The authors have replaced the conclusion as suggested, and no further changes are required.
-
-
-
The authors need to respond to the initial review comments
**PeerJ Staff Note:** Please ensure that all review, editorial, and staff comments are addressed in a response letter and that any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.
**Language Note:** When you prepare your next revision, please either (i) have a colleague who is proficient in English and familiar with the subject matter review your manuscript, or (ii) contact a professional editing service to review your manuscript. PeerJ can provide language editing services - you can contact us at [email protected] for pricing (be sure to provide your manuscript number and title). – PeerJ Staff
Title & Abstract
Although the title, "Cleveland Clinic Score as a predictor of acute kidney injury after cardiac surgery," is succinct, it is vague about the study's objective, which is to assess CCS in relation to various surgical procedures. This aspect would be better captured by a title that is more descriptive, like "Evaluation of the Cleveland Clinic Score for Predicting Acute Kidney Injury Across Different Elective Cardiac Surgeries."
The abstract provides a comprehensive overview, detailing the background (AKI as a serious complication), methods (610 patients, retrospective study, KDIGO criteria), and results (27.2% AKI incidence, AUC 0.630 overall, significant for CABG and AVR/AVP). However, it does not explicitly contain a distinct "Conclusions" section. Instead, it integrates concluding remarks within the results section, stating that the CCS "showed a good prognosis for the occurrence of renal replacement therapy (RRT) in AKI, but for individual groups the scale should be modified by adding several new factors" (lines 35-37). The phrase "good prognosis for the occurrence of renal replacement therapy (RRT) in AKI" (line 35) may overstate findings, as an AUC of 0.630 is moderate, not "good" (typically, AUC >0.7 is acceptable, >0.8 is good). Rephrasing to "moderate predictive ability for AKI (AUC 0.630) and particularly useful for predicting RRT" would enhance accuracy.
The authors should also write a separate “conclusion” section for the abstract.
The p-value notation "p=<0.001" (line 32, 136) is incorrect and should be "p<0.001."
Introduction
The introduction provides a robust background on AKI as a serious postoperative complication, its risk factors, and the CCS as a widely tested predictive tool (lines 59-63). It clearly identifies the research gap: the lack of validation of CCS across different cardiac surgery types, justifying the study’s purpose. It does not, however, provide a concise explanation of the CCS components, which would help readers who are not familiar with the tool. Furthermore, the context would be strengthened and the study would be positioned within current research trends if it included recent studies (post-2020) on CCS improvements, such as the addition of biomarkers or clinical factors.
I suggest clarifying the tool on lines 62-64 with or without the table as “The Cleveland Clinic Score (CCS), developed by Thakar et al. (2005), is a clinical tool to predict AKI risk after cardiac surgery, incorporating factors such as female gender, congestive heart failure, left ventricular ejection fraction (LVEF) <35%, preoperative intra-aortic balloon pump (IABP) use, chronic obstructive pulmonary disease (COPD), insulin-dependent diabetes, previous cardiac surgery, emergency surgery, type of surgery, and preoperative creatinine levels, with scores ranging from 0 to 17 points.” Furthermore, mention recent evidence suggesting CCS improvements to contextualize the study. Mention on line 66 “Recent studies have proposed enhancing the CCS by incorporating additional predictors such as baseline hemoglobin, estimated glomerular filtration rate (eGFR), and glycosylated hemoglobin (HbA1c), which may improve its predictive accuracy across diverse patient populations {Vives et al., 2024, PMID 37997304}.”
The introduction uses "acute renal failure" (ARF) interchangeably with AKI, which is outdated per KDIGO guidelines. Standardizing to "AKI" ensures alignment with modern nomenclature.
Figures & Tables
The figures and tables are clear and free from unnecessary modification.
Material and Methods
The methods section is clear and detailed, describing a retrospective study of 610 patients aged ≥18 undergoing elective cardiac surgery in 2023 at a single center in Krakow, Poland. Inclusion/exclusion criteria are well-defined, and data collection is thorough. AKI is defined using KDIGO creatinine-based criteria (lines 89-94). Statistical methods align with the objective of assessing CCS’s predictive ability.
The authors can clarify how CCS was used, as it was the main focus of the study. They may consider adding “The CCS was calculated retrospectively for each patient based on preoperative data extracted from medical records, using the standard formula outlined by Thakar et al. (2005) (Table 1)" on Page 2, after line 85.
The authors should address Urine Output limitation and revise lines 94-95 to: "Due to data limitations and the potential effect of postoperative diuretic use, urine output criteria were not included in the KDIGO definition of AKI. This may have led to underdiagnosis of AKI, particularly in cases where urine output was reduced without significant creatinine changes, as urine output is a sensitive early indicator of AKI.”
Results
The study’s novelty lies in evaluating CCS across different surgery types (e.g., CABG, AVR/AVP, MIDCAB), unlike prior studies that focused on general cardiac surgery populations (lines 179-181). This contributes to understanding procedure-specific AKI risk, a gap noted in the introduction (lines 22-23). The data are plausible, with AKI incidence (27.2%) within the reported range and consistent with studies like Robert et al. (line 148). The AUC values are consistent with previous CCS validations (e.g., Wong et al.). Known risk factors are supported by Table 2 data (e.g., lower eGFR, higher creatinine in AKI patients).
The authors should provide a mechanistic explanation for findings, enhancing interpretability. Consider adding on line 138 “The better performance of CCS for CABG and AVR/AVP may be due to shared pathophysiological mechanisms, such as atherosclerosis in coronary artery disease and aortic stenosis, which are captured by CCS components like preoperative creatinine and LVEF (Table 1)”.
Discussion
The discussion correlates well with results, comparing the 27.2% AKI incidence to the literature and discussing CCS’s moderate predictive ability. It addresses the study’s objectives by evaluating CCS across surgery types and suggesting modifications, which are relevant to improving AKI prediction. The discussion of atherosclerosis as a shared mechanism for CABG and AVR/AVP is insightful, but could emphasize clinical implications more. Add on Page 5, after line 190: "Our findings suggest that clinicians can use CCS to identify patients at higher risk of AKI requiring RRT, particularly for CABG and AVR/AVP, enabling targeted preoperative optimization. For milder AKI stages or other surgeries, alternative or refined risk models may be necessary.”
Conclusion
Please replace the current text with “Our study found that the Cleveland Clinic Score was a moderate predictor of acute kidney injury and was especially useful in identifying patients who needed renal replacement therapy, especially those undergoing CABG and AVR/AVP procedures. According to our findings, the predictive accuracy of the CCS may be enhanced for different stages of AKI and cardiac surgery types by taking into account variables like advanced age, preoperative estimated glomerular filtration rate, and comorbidities like diabetes and hypertension, which demonstrated strong correlations with AKI.”
The investigators need to be very clear about exactly what parameters are put into the Cleveland Clinic score, is there are a number of modifications of this, and it's not clear from their manuscript exactly which factors have been added.
ROCs of less than 0.7 may be statistically significant, but the question is whether they are functionally useful. The investigators need to clearly indicate how are seeds of .65 would be useful. They also need to indicate how they determined there was any difference between the ROCs of .65 and .629. I would recommend these be bootstrapped.
The investigators need to clearly indicate precisely what they put in the model and what they mean by the Cleveland Clinic statistic, as this has been presented in a modified form, and it is unclear precisely what they were referring to. They should also, as a matter of design, try their own data in a regression model.
The investigators have provided the data in the form of a spreadsheet, and this is, in fact, very laudable. However, they should clearly specify, as I mentioned above, what specifically went into the various models. As I mentioned above, they should try building their own and comparing it to the Cleveland Clinic score model.
As it stands, I'm unclear on what basis they think this is a good predictor of renal outcomes. They need to be clear about what they consider to be good accuracy and a reasonable false positive, false negative, true positive true negative rates.
The investigators wish to determine whether the Cleveland Clinic cardiac score could be useful to determine the relative risk differences of different types of cardiac surgery. In the study, they used creatinine criteria only as an assessment of KDIGO stage AKI and did not include urine output, as apparently this was not available, as part of their retrospective analysis. They acknowledge this limitation. However, this is a significant limitation as there are a number of patients who may get early KDIGO AKI based on urine output but not have much in the way of a rise in serum creatinine. They also should discuss the limitations of the diagnosis of AKI in patients undergoing cardiac surgery, where patients generally receive large volumes of IV fluid, and in fact, if there's a failure to reduce the serum creatinine, this already suggests that the patients may be undergoing some degree of renal impairment.
The investigators need to be very clear about exactly what parameters are put into the Cleveland Clinic score, is there are a number of modifications of this, and it's not clear from their manuscript exactly which factors have been added.
The problem with the Cleveland Clinic score and the Euroscore is that they are highly dependent on preoperative serum creatinine, with this being the most influential factor; this needs to be addressed. The usual ROCs for these scoring systems, which you are attempting to predict AKA I, are usually around .65 and certainly less than .7, and hence not a great value. The data presented here shows very little difference between the different types of surgery, and I doubt there's any significant difference between the RFCs of.65 and .629. If the investigators present this very clearly and I would recommend a bootstrap system with clear confidence intervals, although I sincerely doubt that these will be significantly different. That said the data clearly supports the use of the CSS score it is not a particularly good predictor in their patient cohort and confirms that It is not a particularly good predictor. I would also recommend that it be useful to present the sensitivity and specificity of the different measures, not just the simple ROCs. The reason for this is that sometimes it's more important to have a differentiation between false positives or false negatives, and it would be interesting to see what the investigators think is the most important outcome here. Specifically, it would be worth stating just how many false positives and false negatives there would be to give readers an idea of the value of the test.
In their conclusion, they state did their results confirm that both the total score of the scale and its individual stages are ineffective prognostic tools, especially in the context of identifying patients requiring postoperative RT. I think they need to make the data that supports this conclusion clearer. Generally speaking RCS of greater than .8 would need to be observed before people would say this was a good predictor. That said, what is more important in terms of false positives, false negatives, true positives, and true negatives is clearly an issue here, and this could be explained in a bit more detail.
I would encourage the investigators to use their own data collected to try and construct a regression tool that might outperform the CSS score. With this in mind, I would recommend they use a generalised regression equation with a LASSO approach, and they may find that they can get ROCS closer to 0.7, which well not ideal, but may well be a significant improvement.
It would also be nice to know if, at all possible, the timing of the diagnosis of AKI and the duration of the AKI if available.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.