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.
I have changed Lines 260-262 in the conclusions to the following:
"In the nationwide US practice, the weekend AF hospitalizations appear to have lower rates of cardioversion utilization and lower hospitalization cost. Further studies are required to identify the differences and explore the opportunities to improve AF weekend care."
The authors have addressed most of the previous comments adequately. Minor errors remain that require further correction:
Please rephrase the incomplete sentence at line41-42.
The word 'ablation' remains in line 33, 43, 49 & 50.
There is no need to spell out standard deviation in line 45-6.
Your manuscript has improved significantly. However, there are several areas that require further attention:
1. In your abstract, you have not included all the results for the secondary outcomes within 'results' section.
2. Multivariate analysis for mortality was performed and shown in Table 3. However, this was not done for any of the secondary outcomes.
3. I agree with reviewer 2 regarding the usage of 'life-saving' in line 59.
4. Lines 112-3: "We found 90,701 patients corresponding to a national estimate of 453,505 patients (after applying the weights provided by the HCUP-AHRQ) with AF as the primary diagnosis" This statement is ambiguous. Are you referring to 90,701 patients accounting for 453,505 unique hospitalizations?
5. Expression of results appears cumbersome in many instances e.g. there is no need to spell out 'standard deviation'. Also, decimal point needs rationalization - e.g. for age, one decimal point should suffice.
6. Figure 1 appears redundant.
7. It is unclear whether ablation was included in the analysis with cardioversion as there are quite inconsistently reported or discussed throughout.
8. The extremely low rates of anticoagulation require further investigation/explanation.
9. I am not sure if elective admissions were included in your analysis (e.g. cardioversion, ablation).
10. Conclusions (Lines 225-9): the reference to improving rates is not justified here. Please state your findings only and avoid comparing to previous papers.
11. The last NIS data was from 2008. There is an almost 4-fold increase in the number of hospitalizations. I suspect this is due to increasing number of procedures (largely) elective for atrial fibrillation. If this is the case, the present paper requires further clarifications.
12. The discussion section could be further improved.
No comment
No comment
No comment
The authors have satisfactorily responded to my comments and suggestions, but I would recommend that a few details of now the NIS is generated (as mentioned in the rebuttal) is included in the actual manuscript for readers that are not familiar with it.
Firstly, I would like to thank the authors for their extensive revision which has improved the manuscript considerably. However, there are a number of areas that still require addressing as outlined below.
• I think it is excessive to refer to cardioversion and ablation for AF as ‘life saving procedures’. Whilst a cardioversion may be urgently indicated in the setting of acute haemodynamic instability, I do not think it is reasonable to compare this to other procedures such as the use of primary PCI for acute myocardial infarction.
• Tables 2 and 3 would benefit from renaming to assist with comprehensibility for readers. If I am interpreting your data correctly, Table 2 refers to differences between weekend and weekday hospital admissions for AF and Table 3 reflects the multivariate model used to predict in hospital mortality for AF admissions. Please clarify this and consider renaming the tables to reflect the data shown in them.
• In Figure 1 and Table 2 you refer to ‘time to cardioversion’. Is this time to cardioversion and ablation or just cardioversion?
• Figure 1 adds little to the manuscript as the same data is presented in Table 2.
• I see little value in the combination of cardioversion and ablation in to one outcome. These are totally different entities and whilst cardioversion may be used acutely, I cannot think of any indication for a life saving ablation for AF. The most common scenario for AF ablation is it’s use as an elective procedure and as this is much more likely to occur on weekdays, this could skew your results. It would be of much greater benefit to present cardioversion and ablation separately.
• In your multivariate model for prediction of in hospital mortality, were any patient characteristics considered e.g. age, comorbidities such as heart failure, hypertension, prior stroke/TIA, IHD, appropriate use of OAC? You have also stated in your methods that you adjusted for ‘disease severity’. Could you please outline what this means and how this was undertaken?
• Are you able to present the factors examined in univariate analysis?
• I’m not sure if I am interpreting your results correctly, but the result concerning anticoagulation on weekend vs weekday admissions states that approximately 17-19% of the sample were treated with anticoagulation. This seems an extraordinarily low number, even for a random sample. Is it possible there is some data missing here concerning prescription or use of OAC from other providers?
Overall, the revision you have undertaken has improved the manuscript. Further methodological detail and improvements in reporting of your data would further improve this study.
No issues
No issues
No issues
The revision is a vastly improved manuscript.
Your manuscript has received 3 independent peer reviews and several areas of ambiguity have been raised. You are invited to response to these comments and further improve this work.
Additional comments:
- Introduction section is too abbreviated. Please expand.
- Methods: please state clearly primary and secondary outcomes.
- Statistics: factors adjusted for for multivariate analysis need to be included.
- Please add a paragraph on study limtiations before 'conclusions'
- Table 1 was merely referred to in results section without further explanation. this is insufficient. especially, there is no way to tell what the p values mean clinically. Please disclose CHA2DS2-VASc score and anticoagulation usage as well as AF risk factors.
- Table 3 is redundant - suggest presenting your univariate and multivariate results instead.
No concerns.
No concerns.
See general comments.
I had the pleasure of reviewing the manuscript submitted by Dr Voruganti and colleagues to PeerJ, entitled “Outcomes of patients admitted with atrial fibrillation-flutter on a weekday versus 2 weekend: an analysis from a 2014 nationwide inpatient sample.”
In this analysis, they used data from the 2014 Nationwide Inpatient Sample to assess differences in in-hospital mortality, length of stay, procedure use, and cost in patients admitted with AF on weekdays compared to weekends. They concluded that there were no significant differences in mortality or LOS, but cardioversion use and cost was lower for patients admitted on weekends.
This is an interesting topic relevant to healthcare systems given the increasing burden of AF hospitalisations, and builds on previous studies on this topic. The main strengths of this paper are the large sample size and the nationwide cohort.
Their paper would be strengthened further by:
- Speculation, if even brief, as to the possible reasons for the contrasting findings to some of the previous studies in this area
- Expansion of the introduction to build more of a background and rationale for the current study
- Provision of further details on the NIS and how it was analysed. For example, in the first few sentences of the results, it is quoted as there being 90,701 hospitalisations, yielding an national estimate of 453,505. How are these figures calculated? If scaled up from a smaller sample size, what implications does this have on the accuracy of the information?
• Please structure your abstract according to author instructions with headings (Background, Methods, Results, Discussion)
• Although your introduction identifies an appropriate gap in the literature, it would benefit from further additions. There are numerous studies that have identified differences in outcomes between weekday and weekend admissions in other cardiovascular conditions e.g. MI and it may be helpful to add in a small amount of background information concerning this. Furthermore, there are other studies in this field related to AF that you have not included e.g.:
- Comparison of weekday vs. weekend admission on inpatient mortality and total hospital charge on atrial flutter and fibrillation: a nationwide analysis
Lee, Shawn ; Lemor, Alejandro ; Gholitabar, Farid ; Gongora, Carlos A ; Dominguez, Abel Casso ; Kroner, Paul T ; Hurtado, Carolina ; Mehta, Davendra ; Castaneda, Daniel
Circulation, Nov 8, 2016, Vol.134(19)
• I am unclear what this sentence at line 53 ‘but the results were not replicated in a national database’ is referring to. Could you please reference this, or re-phrase for improved comprehension?
• Please state the dates in 2014 for which this analysis was conducted.
• In your abstract, please provide values related to number of days for length of stay, instead of just a p value.
• Could you please clarify how you have moved from 90,701 hospitalisations with AF as a principal diagnosis, to a ‘national estimate of 453,505’ hospitalisations. Please clarify where this number (453,505) is from.
• In regard to mortality, you report having adjusted for patient and hospital characteristics and disease severity, yet the factors used in your adjustment model are not listed. Please list these variables.
• Similarly, in your secondary outcomes there is no description of variables used in the multivariate model and no table of results related to this. Your results related to greater number of comorbidities does not have an associated p value – please provide. There is no description of how information concerning co-morbidity was obtained – please provide this in the methodology.
• For cost data, there is no text related to absolute cost values. Please provide this. Please state in your methodology how cost data was obtained.
• There is no data related to ablation, which you have stated was one of your intended outcomes.
• It is unclear why you have not presented your multivariate adjusted model in table or figure form.
• The paragraph in your discussion beginning with: ‘We notice a trend in improvement of outcomes in patients with AF over a period of 6 years’ is not supported by your data. You have not undertaken any statistical testing of outcomes between these datasets, and therefore this claim is not substantiated.
• You imply in your discussion that cardioversion is a ‘life saving procedure’. Whilst you state that cardioversion is underutilised at weekends, you have stated that outcomes are not different between weekend and weekday admissions. This could also imply that cardioversion is over utilised, perhaps inappropriately, during weekdays leading to higher cost of weekday admissions. It may be that the opportunity to improve care lies in appropriate use of this procedure at all times and your discussion should refer to this as an alternative possibility.
• In your conclusion, you introduce the concept of ‘improved anticoagulation’ during weekend AF hospitalisations, yet there is no data in your manuscript to support this. Please do not introduce new concepts in your conclusion. This data needs to be stated as an outcome, described in the methodology and have the associated results presented.
Whilst this is an interesting concept, there are numerous methodological gaps in the manuscript at this stage that require rectifying. Significant expansion of the methodology section is required, with appropriate reporting of intended outcomes. The results section requires much greater detail, and appropriate description and presentation of multivariate adjusted models.
The paper is well set out and the spelling and grammar are correct. Table 3 could be more clearly formatted. Table 1 and 2 are satisfactory.
References are satisfactory,
My main query here is with regards to the primary and secondary outcomes of the study.
On page 4 in the methods the authors refer to the primary outcome being a composite of 5 factors: in-hospital mortality, utilisation of cardioversion/ablation procedures, length of hospital stay, time to cardioversion and total hospitalisation charges. However in the results section on the subsequent page the authors refer to secondary outcomes of number of cardioversion procedures, interval to the procedure, length of stay and cost of hospitalisation. The final 2 factors are the same in the primary and secondary outcomes. This reviewer is confused, are these primary outcomes or secondary outcomes?
Otherwise I am also slightly confused by the first line of the results. In the abstract the authors refer to a cohort size of 453,505 hospitalisations, however in the first line of the results the authors refer to 90,701 hospitalisations? I am confused about who these 90,701 patients are? Does this mean that nearly 350,000 hospitalisations were for atrial flutter? I would recommend the authors clarify the cohorts more clearly for the reader.
This is an interesting study. The key result is that presentation on the weekend does not increase mortality, though lower rates of cardioversion are clearly more prevalent on the weekends. This study using electronic data has a large cohort and adds to the current body of literature. The data refutes the "weekend effect" and this is important when governing health policy.
With regards to table 1, presuming the p values refer to weekend vs weekday hospitalisations there are a number of differences between the baseline characteristics (gender, race, finances etc etc), which are not really elaborated on in the results. Can the authors please clarify why there are such differences?
A nice figure showing the key results from table 2 (eg cardioversion % in each group) would improve the paper. I dont feel table 3 adds very much to the paper, i would consider modifying this table into a figure (such as a flow chart or graph of events across time) or removing it.
Nothing further not commented on above
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.