Use of Haemostatic Devices for the Control of Junctional and Abdominal Traumatic Haemorrhage: A Systematic Review

Round 1

Reviewer 1 Report

Sound systematic review methodology, literature search and the review of the quality of the literature was well done. The author's stated aims and objectives (last paragraph of introduction) overstep the quality of the literature.

Describing the comparative (or any) mortality data is meaningless in the context of the poor quality of the literature reviewed. The iTClamp had zero mortality most likely due to the types of non-lethal injuries to which it was applied. I question whether the iTClamp should be included in the review as it is a very different device to the others examined and is unlikely to halt major junctional bleeding (despite FDA approval for use in the groin or axilla). Perhaps clarify its true role- temporary closure of small, superficial extremity and scalp wounds.

The authors suggest large RCTs be undertaken to improve the evidence. This would be ideal but perhaps unlikely due to challenges in study design, industry involvement, recruitment of enough patients and ethical considerations.

Your conclusions need abbreviating. In the pre-hospital arena (especially on the battlefield) the key variables of transportability, speed of assembly/application and ease with which the patient can be moved to definitive care are critical. These practicality markers are not well described in the literature and this should be highlighted in the conclusions. Again, mentioning the iTClamp mortality and other benefits in the conclusion makes little sense. Stating that these devices "have the potential to prevent many deaths from exsanguination" is supposition and not based on your review of the existing literature. The only real conclusions that you can draw from this small body of poor quality evidence for these devices are that their use in the pre-hospital and military environment is feasible, but better quality data is required to ascertain if any of them change trauma outcomes.

Author Response

Comment: Sound systematic review methodology, literature search and the review of the quality of the literature was well done. The author's stated aims and objectives (last paragraph of introduction) overstep the quality of the literature.

Author response: Thank you for your positive comments. As for the last paragraph of the introduction, this has now been amended.

Comment: Describing the comparative (or any) mortality data is meaningless in the context of the poor quality of the literature reviewed. The iTClamp had zero mortality most likely due to the types of non-lethal injuries to which it was applied. I question whether the iTClamp should be included in the review as it is a very different device to the others examined and is unlikely to halt major junctional bleeding (despite FDA approval for use in the groin or axilla). Perhaps clarify its true role- temporary closure of small, superficial extremity and scalp wounds.

Author response: Despite the reviewer’s reservations about the inclusion of the iTClamp in this systematic review, we believe that it should be included as it has been used to control severe haemorrhage including the groin and axilla and as pointed out by the referee has been approved by the FDA for these uses. As for using mortality data in the poor-quality studies identified in this systematic review, we believe this is still valid. However, we have re-iterated that the poor quality of the studies calls into question any of the conclusion and hence warrants the need for high quality studies in this area (see line 28, Lines 378-379 and lines 431-433).  

 

Comment: The authors suggest large RCTs be undertaken to improve the evidence. This would be ideal but perhaps unlikely due to challenges in study design, industry involvement, recruitment of enough patients and ethical considerations.

Author response: We agree with the reviewer that that are RCT are unlikely due to all of the challenges mention. We have amended lines 419-428 to reflect this.

 

Comment: Your conclusions need abbreviating. In the pre-hospital arena (especially on the battlefield) the key variables of transportability, speed of assembly/application and ease with which the patient can be moved to definitive care are critical. These practicality markers are not well described in the literature and this should be highlighted in the conclusions. Again, mentioning the iTClamp mortality and other benefits in the conclusion makes little sense. Stating that these devices "have the potential to prevent many deaths from exsanguination" is supposition and not based on your review of the existing literature. The only real conclusions that you can draw from this small body of poor quality evidence for these devices are that their use in the pre-hospital and military environment is feasible, but better quality data is required to ascertain if any of them change trauma outcomes.

Author response: We have now amended the conclusions and Lines 415-428, as recommended above.

Reviewer 2 Report

I commend the authors on their work. It must have taken a lot of effort. 

I have significant concerns about the validity of the study. The pooled sample is more like a stew than comparing apples and oranges... I am concerned about the loose inclusion when the PICO is clearly on the efficacy of these devices in the pre-hospital setting. The introduction leads me to believe that the work is on non-civilian trauma and catastrophic civilian trauma ‘under fire’. Then the included studies are predominately non-trauma populations.

While efficacy can be seen by simply applying the Tq to a healthy patient, deploying the device and using an ultrasound to measure distal flow…. It does not take into account the myriad of physiological processes a trauma patient is subject to. Similarity, I am concerned your search has missed some studies, especially for REBOA, where a recent systematic review included 11 observational studies.

I feel the authors need to seriously consider the way they present the data, if they feel strongly that all current studies should be included, the data should be separated by at least the mechanism (volunteer vs traumatic). I would place any cadaveric studies  in with volunteers however. I also emplore the authors to consider using medians for all data, as it is unlikely parametric, and whenever a figure is reported in text, ensure the number of patients that figure is referring to is cited.

I feel this paper is of interest to those in the field, however not at this stage. 

Author Response

Comment: I commend the authors on their work. It must have taken a lot of effort.

Author response: Thank you.

Comment: I have significant concerns about the validity of the study. The pooled sample is more like a stew than comparing apples and oranges... I am concerned about the loose inclusion when the PICO is clearly on the efficacy of these devices in the pre-hospital setting. The introduction leads me to believe that the work is on non-civilian trauma and catastrophic civilian trauma ‘under fire’. Then the included studies are predominately non-trauma populations.

Author response: We have strictly followed our PICO inclusion characteristics for all of the included studies in this systematic review. We have amended the last paragraph of the introduction to define our patient population, which includes non-trauma patients. Basically, we selected all articles that evaluated data related to these haemostatic devices. We have now separated out studies which used junctional tourniquets in military and civilian healthy volunteers (Table 2), perfused human cadavers (Table 3) and civilian and military patients (Table 4) to improve the clarity of the presented data.  

 

Comment: While efficacy can be seen by simply applying the Tq to a healthy patient, deploying the device and using an ultrasound to measure distal flow…. It does not take into account the myriad of physiological processes a trauma patient is subject to. Similarity, I am concerned your search has missed some studies, especially for REBOA, where a recent systematic review included 11 observational studies.

Author response: We agree that efficacy is much more complicated in a trauma patient and hence we have separated the data in the table to include civilian and military trauma patients in Table 4. As for the recent study including 11 studies, they had different inclusion/exclusion criteria; our primary population was the use of REBOA in the prehospital setting and hence our included studies were different.  

Comment: I feel the authors need to seriously consider the way they present the data, if they feel strongly that all current studies should be included, the data should be separated by at least the mechanism (volunteer vs traumatic). I would place any cadaveric studies in with volunteers however. I also emplore the authors to consider using medians for all data, as it is unlikely parametric, and whenever a figure is reported in text, ensure the number of patients that figure is referring to is cited.

Author response: We have now separated the data into volunteer, cadaver and traumatic mechanisms in agreement with the reviewer’s suggestions. Also, when reporting the data, it is difficult to use median as most of the studies report mean values themselves and do not provide the raw data. Even after writing to the first authors, we have been unable to acquire raw data. We realise that this is less than ideal but under the circumstances, this is the best we can do.

 

Comment: I feel this paper is of interest to those in the field, however not at this stage.

Author response: Thank you for pointing out that the stud is of interest, we also think so too. We hope that you like the way we have amended the manuscript based on both reviewer comments and hope that it is now suitable for publication.

Round 2

Reviewer 2 Report

I commend the authors for their revision, but feel there is still work needed to be done.

 This work may be attempting to look at devises and techniques that can help catastrophic abdominal and junctional haemorrhage in the pre-hospital field, but they report data on non-traumatised, healthy, volunteers and then in the next paragraph on severely traumatised military and civilian populations, of which some studies reported a 100% mortality...

The Authors reply to a comment I made raising concern that articles were missing from their included studies, specifically REBOA was met with a reply saying that this work is focused on the pre-hospital arena, which is true for REBOA, but why then include data on devices in volunteers?

I can see why the authors are trying to create a review that outlines the data on these devices to be a sort of ‘one-stop-shop’ for future reference for all of the literature on these devices, but in its current form it seems an unfair comparison.  If the authors wish to include data on all of the devices in both healthy volunteer, cadaveric and traumatic populations, I feel that all studies pertaining to those three categories should be included, provided they meet inclusion criteria. This would then allow readers to compare, if the literature exists, heathy vs healthy, cadaveric vs cadaveric and traumatic vs traumatic. Further to this, van Oostendorp in 2016 broke up his review into regions which may make for easier reading for the reader as some techniques and devices  may not be applicable or have data pertaining to all types of bleeding and it seems wrong to compare a scalp wound needing an iTClamp to a patient with catastrophic  pelvic bleeding needing REBOA.

I also can’t see how all of the patients included in the iTClamp group are worthy of a paper on junctional or abdominal haemorrhage based in Table 6 and wonder how these studies meet the inclusion criteria when the title clearly outlines that the type of haemorrhage being investigated is junctional and abdominal, yet some patients had scalp and chest lacerations. I agree with the other reviewer that the iTClamps role in the world of junctional haemorrhage, despite have FDA approval is likely very limited and the data on this device, as evidence by this review does not support its use either. I feel the title may need to be changed, in which case it opens up many more potential studies and devices or the authors should, as the title dictates, focus on abdominal and junctional catastrophic haemorrhage.

Author Response

Comment: This work may be attempting to look at devises and techniques that can help catastrophic abdominal and junctional haemorrhage in the pre-hospital field, but they report data on non-traumatised, healthy, volunteers and then in the next paragraph on severely traumatised military and civilian populations, of which some studies reported a 100% mortality...

Author response: The of this review was to show that it is possible to use these devices in a pre-hospital setting, not just in the operating room/ ED. However, due to the paucity of studies, further pre-hospital studies would be warranted to improve the existing data pool to be able to compare like-for-like in the future. However, to add clarity to the results we have rationalised the description of the results in this revised version to start in Section 3.2, which details pre-hospital feasibility studies and mentions studies in human cadaver studies and studies in volunteers. Then we bring in the evidence from the three studies which used JTQs in the prehospital trauma scenario (Section 3.3).

Comment: The Authors reply to a comment I made raising concern that articles were missing from their included studies, specifically REBOA was met with a reply saying that this work is focused on the pre-hospital arena, which is true for REBOA, but why then include data on devices in volunteers?

Author comment: The studies related to devices in volunteers were included due to extremely small numbers of patients in the three studies in trauma patients. In addition, the three studies only included 31 patients which is too few to perform a rational systematic review. The aim of the article is to show how JTs can control severe bleeding and for this purpose data on volunteers suffices.

Comment: I can see why the authors are trying to create a review that outlines the data on these devices to be a sort of ‘one-stop-shop’ for future reference for all of the literature on these devices, but in its current form it seems an unfair comparison. If the authors wish to include data on all of the devices in both healthy volunteer, cadaveric and traumatic populations, I feel that all studies pertaining to those three categories should be included, provided they meet inclusion criteria. This would then allow readers to compare, if the literature exists, healthy vs healthy, cadaveric vs cadaveric and traumatic vs traumatic. Further to this, van Oostendorp in 2016 broke up his review into regions which may make for easier reading for the reader as some techniques and devices may not be applicable or have data pertaining to all types of bleeding and it seems wrong to compare a scalp wound needing an iTClamp to a patient with catastrophic pelvic bleeding needing REBOA.

Author response: There were some additional case studies conducted on trauma populations for each technique but were conducted in ED/ operating theatre so were excluded as they did not meet our inclusion criteria (not conducted in the pre-hospital environment).

We have however, included all human cadaveric studies.

Finally, the focus of the systematic review was pre-hospital use, for which there is limited data for all devices. Thus, comparing by region was not possible due to the lack of data across volunteer, cadaveric and trauma populations. We have however, changed the results to compare simulation vs. simulation, and traumatic vs. Traumatic and made changes to the methods to reflect this.

Comment: I also can’t see how all of the patients included in the iTClamp group are worthy of a paper on junctional or abdominal haemorrhage based in Table 6 and wonder how these studies meet the inclusion criteria when the title clearly outlines that the type of haemorrhage being investigated is junctional and abdominal, yet some patients had scalp and chest lacerations. I agree with the other reviewer that the iTClamps role in the world of junctional haemorrhage, despite have FDA approval is likely very limited and the data on this device, as evidence by this review does not support its use either. I feel the title may need to be changed, in which case it opens up many more potential studies and devices or the authors should, as the title dictates, focus on abdominal and junctional catastrophic haemorrhage.

Author response: It is true that our inclusion criteria focussed on abdominal and junctional catastrophic haemorrhage. However, due to the lack of studies and actual raw data in many of the included studies (often presented as a single mean values), we included additional studies that reported use of the device in our region of interest as well as other areas. Hence, some studies mention the use of the device on the scalp too.