Gastrointestinal Bleeding in Patients with SARS-CoV-2 Infection Managed by Interventional Radiology

Round 1

Reviewer 1 Report

Major

 

One of the aims of the study is to examine the risk factors for GI bleeding. However the authors methodology is not designed to evaluate this. They only look at the 34 patients with COVID who had GI bleeding. The correct methodology would be to compare potential risk factors of COVID GI bleeders to COVID patients who did not bleed. As a result, I recommend the authors either change the study to compare these two cohorts or only look at the safety of TAE in COVID patients with a GI bleed that require therapy. This by itself is novel enough and deserves publication.

 

Many studies (some of which the pts cite) recommend conservative management as many of these bleeds will resolve (doi: 10.1016/j.gie.2020.04.028).

In fact in many studies endoscopy was not needed (PMID 32796176 and 33341978 are the largest studies in the literature). Thus it would be helpful to provide a percentage of how many COVID GI bleeders needed TAE (34/XXX).

.

 

 

 

 

Minor

 

Introduction paragraph: The sentence on most common GI COVID symptoms ends with and sporadically GI bleeding. Given the sentence is discussing common issue, split the last part in anew sentence.

 

 

Author Response

 

We changed the title of the paper because it s an evaluation of safety and efficacy of TAE in GIB of Covid 19 pts, as you correctly highlighted.

 

In introduction we added the observation that in more studies endoscopy resulted not necessary and conservative management was enough.

 

In introduction paragraph the form requested changes were done, as sugegsted.

 

English language was revised.

Reviewer 2 Report

This article treats a common theme for the last 2 years regarding the Covid patients. I would have accentuated the cases of gastrointestinal bleeding specific for COVID patients to be outside the well studied pathology of digestive bleeding. 

The introduction has several good statistics. However the phrase were interventional radiology has an important role in hemodynamic instability patients is rather exaggerated than real. Except for several lesions like colonic angiodysplasia the role of interventional radiology for GI bleedings is limited. 

Regarding the study design I am not sure how you can determine risk factors without a control study. Would have been nice to present the whole group of Covid patients (how many COVID patients were hospitalized in 1y and 3 m) and then selected the 34 associated with GIB and also a randomized control group maybe adjusted by some similarities between them. 

Also the definition of GIB in my opinion should not include those with a HGB drop, hemodynamic instability or lower SBP because those are severity factors rather than diagnostic criteria. 

The management protocol for GIB should be presented, before and during pandemics in order to see exactly the differences. Also a more detailed description of the lesions would be a good idea in order to see exactly which one failed more often after endoscopy and which one is more suitable for TAE.  Also, on CT scans how many had arterial bleeding (blush) and how you made the decision that TAE is more suitable compared to re-endoscopy or directly to OR when the patient is unstable. How about for those patients that were already on ventilation (ICU) and developed GIB? 

Technical details of TAE: which approach, which artery more frequently cannulated related to the gastrointestinal lesion, how long was the procedure, maybe compared to endoscopy (and maybe to weight the risk of infection by the duration of the procedure).

The results: tables present basic statistics of risk factors presumed not determined by comparison. The small sample would not be a disadvantage rather the retrospective design and the absence of a control group. Also the mentioned bias of being a multicenter study would not exist if the GIB protocols from all centers would have been presented to see the differences (if there are any). 

Also, would have been nice to see some data about the COVID transmission during endoscopy compared to interventional radiology and a comparison of limitations of both methods of hemostasis.  

Some images during TAE for digestive lesions are usually convincing. 

I am sure that if you change a little bit the design of the study and detail the technical part of endoscopy, TAE, the protocols, it will be a more conclusive article gaining more citations. 

Author Response

TO REVIEWER 2:

 

In introduction the sentence “interventional radiology has an important role in hemodynamic instability patients...” was changed in “In Covid 19 patients to avoid aforementioned problems and in case of persistent bleeding after endoscopy and preexisting hemodinamic instability, interventional radiology could play an important role”, as you suggested.

 

A control group is not available. We are in accordance with your idea about that a correct study design include a control group of non haemorrhagic covid 19 pts. But as mentioned this is a multicenter retrospective observational study. We described the characteristics of these patients presenting with GIB managed with IR.

We included pts with Hb drop, hemodynamic instability or lower SBP because these pts referred to IR service.

In M&M we cited guidelines to which we refer for GIB (in covid 19 and non covid 19 pts).

In Results, we specified that Angiographic bleeding was encountered in 85.3% of our patients; the remanant were pts with a well known site of bleeding revealed by endoscopy but unsolved, for example duodenal ulcer in which GDA is the culprit artery for bleeding.

 

As you suggested , changes were done. We are available to do more corrections, if any. Thanks a lot for your time.

Reviewer 3 Report

This is a retrospective study (in reality an audit) of the demographics, risk factors and outcomes relating to radiological management of GI bleeds in COVID-19 positive patients(who are heparinised). 

Firstly i think this is an important topic, to myself and anyone managing GI bleeding during the pandemic. The risks of aerosolisation related to endocopy has increased the importance of radiological procedures and it would be very helpful to be able to quote risks and benefits to patients in this subset(covid positive bleeding on anticoagulants). It would also be helpful to know how many patients on anticoagulants in this unit developed bleeding. 

In general the paper is well written and readable, other than the use of the word 'pandemia' as opposed to pandemic I would have thought this was written by an advanced speaker of the language. 

I undestand the need to report all the results however there needs to be a bit more selectivity, it looks like one big table. Some results can maybe be moved into supplementary material and some things need to be added(see below) 

1. I am not sure why the authors tested for normality if they had no intention of comparing means. Same with p value? why did you report significance if no statistical test has been carried out?
2. Saying there were'nt enough patients for significance or testing i think is a bit weak. For example there could have been a comparison with the patients who had covid, got heparin and did not bleed. Or a comparison with the 5 who died. Secondly with this data the authors should power test and say how many are needed. 
3. Some abbreviations although maybe obvious are not elucidated - OAT, ASA
4. In general i think the authors need to have a word with a statistician and understand what they are reporting and why certain values are reported as medians and others as means.
5. The conclusion reflects the above and is a bit weak. I am also a bit uncomofrtbale by the comment that we dont have significant data to report - negative results need to be reported too!   

The title reeds a bit of reflection as you are reporting outcomes not risk factors. Even in the abstract you say these 'risks were encroutered more frequently' more frequently than what? the general population? a non bleed control group, were they significantly more frequent?

The first reference is incomplete.

In general this a paper I would love to read and be able to quote the findings of, or include this in a meta analysis when the body of evidence grows. I think the authors needs to analyse and be a bit more selective with their data and beef up their conclusion. 

Author Response

 

• What means the first sentence: tested for normality? what do you refer?

The sentence “SPSS version 25.0 (IBM, Armonk, New York, United States) was used for all statistical analyses; p-values were considered significant when < 0.05” was cancelled: as you noted, p value has no sense because a statistical analysis was not performed.

• We did not collected the population of covid 19 pts treated with heparin not presenting bleed.

The comparison with the 5 deaths resulted not enough and no significant results were found.

• abbreviations missing was included.
• Mean and standard deviation (SD) were provided for normally distributed variables, median and interquartile range (IQR) for non-normally distributed variables, number and percentages for categorical variables: sentence reported in the paragraph “statistical analysis”
• In the paragraph discussing about limitations of the study new proposal based on the weak points of the present study were reported.

We didn’t report that we don’t have enough data, but only that the presented data need future elaboration and comparison with others to make them significant. Are you agree?

 

The title was changed, as you and Reviewer 1 suggested.

We adjusted the sentence in abstract: risk factors were found frequently in our population.

Reference 1 was corrected.

 

Thank you very much for you suggestions and corrections

We are available for more corrections if need.

Best Regards

 

Corresponding Author

Round 2

Reviewer 1 Report

The introduction still lists the aim is to evaluate the risk factors for GI bleeding. The aim needs to be changed. 

 

The limitations state that the number of COVID GI bleeders in this study is the most collected in the literature. This is not true and I would delete this. I listed the two largest studies in the previous review.

Author Response

-In “introduction” and “abstract” we corrected the aim of our study as you correctly suggested.

-We corrected the sentence with “...one of the largest study in literature”, can it fit?

-English language was revised

 

Thanks again for your time and cooperation to improve our paper

We are available for more corrections/suggestions, if any

 

Best Regards

 

Corresponding Author

Reviewer 3 Report

You are still testing for normality, as you acknowleged you dont intend to comapare means, so why assses for normality?

Your results are descriptive because you designed the experiment that way(because you have nothing to compare to) not because of sample size.

This is a 'retrospective study'(reaslistically an audit/case series) you don't always have to test a hypothesis. If you want to discuss the number, power test how big a trial should be but I don't really think this is necessary. 

line 63,64 - Sentence starting with Its does not flow.

Table 1&2 - Variable not variabile

Again refererence one is incomplete. If you use software like endnote it can guide you on how to reference a website. eg. author, title, date of plublishing. 

Acknowledgments - the corrected version is grammatically incorrect. (took part in)

I again iterate that this is an important piece of work but you keep making (in my eyes) uneccessary statistical assertions and comments. There is no shame in presenting case series or audits; covid-19 is a new disease and this is a niche topic. As a practitioner I am happy with percentages, to inform myself and patients of the options. 

Author Response

-This is a multicenter retrospective observational study, so we eliminated the paragraph “statistical analysis”.

-The sentence starting with “its “ was corrected as you suggested

-Table 1 and 2 were corrected: “variables” instead of “variabile”

-Reference 1 is not a paper, so it cannot include authors, ect; this reference refers to WHO website

- Acknowledgments: We corrected “took part to” with “took part in”

-English language was revised

 

 

Thanks again for your time and cooperation to improve our paper

We are available for more corrections/suggestions, if any

 

Best Regards

 

Corresponding Author