Trauma & Respiratory II (Nov 2015)
Research: Holcomb et al. Massive Transfusion PROPPR. JAMA 2015
Research: Gupta et al. Selective Use CT vs Whole Body Imaging in Blunt Trauma. Annals EM 2011
AIR Module: Respiratory Part 2
Research: Holcomb et al. Massive Transfusion PROPPR. JAMA 2015
Research: Gupta et al. Selective Use CT vs Whole Body Imaging in Blunt Trauma. Annals EM 2011
AIR Module: Respiratory Part 2
15 comments
Holcomb: Confirms what many have suspected based on prior research that 1:1:1 massive transfusion protocol improves outcomes. It’s a shame that it’s compared to 1:1:2 since that’s pretty aggressive as well. I guess there’s enough mounting evidence that the IRB likely would not approve of a comparison to “standard care” transfusion with crystalloids or blood products by clinician gestalt and lab values. It’s interesting that they did not need consent for this.
Lastly, keep in mind despite its robust methods, The study was only powered to detect an absolute difference of 10% in mortality. Therefore, a smaller effect on mortality may have been present but unable to be detected.
Gupta et al:
REALLY interesting paper I read a few years back when this came out. It’s nice to hear about the trauma surgeons’ perspective. However, they keep talking about not wanting to change their approach because the patients had “good outcomes”. But it’s impossible to tell if these outcomes would change without those extra CT’s. Certainly in terms of radiation risks, long-term outcomes from radiation-induced cancers were not considered. Not to mention cost.
I personally believe that EM physicians have a broader perspective since we see the sickest traumas, as well as the minor ones that we never consult surgery on. It’s reasonable to feel nervous about missing injuries when all you see are the sickest trauma patients from surgery’s perspective.
Only a prospective randomized trial of selective vs. pan scan (or perhaps CT ordering by EM physicians vs. trauma surgeons) measuring patient outcomes will get to the bottom of this. I can imagine that’s really hard to do.
Gupta et al.
strengths: Important that this study was collaborated on both trauma surgeons, and emergency physicians. The relative discussions from each specialty service or insightful and highlight the difference in perspective and interpretation of the data.
Limitations: Only was a single sites that he, but weakness that was not mentioned in the paper was that the public ability to other emergency departments sitting in different patient population medics and having providers with different risk thresholds and interpretations of what findings represent a clinically significant findings make its applicability difficult.
Main outcome: Trauma surgeons and this study believes that “definition of all injuries at presentation remains a standard of care until a prospective evaluation determines which of these injuries are indeed innocuous.” emergency physicians in the study believe that many CTs are unnecessary and the CTs which were deemed necessary were largely unrevealing of clinically significant injuries or injuries which “could have been managed without initial CT.”
future area research: Collaborative efforts to determine what findings represent clinically significant findings in order to uses criteria for future research.
Holcomb et al.
1. What are some strengths of the study?
Large, multicenter, RTC. Nice blinding design.
2. What are the limitations?
Like Alex said, study was powered to detect 10% and 12% absolute difference in 24-hour and 30-day mortality, respectively. Also, the “catching up” phenomenon is a potential limitation. Would the paper’s conclusions have changed if different mortality time end-points were measured (for example, 3-hour mortality)?
3. What are the main outcomes of the study?
Absolute percentage group differences for 24-hour and 30-day mortality.
4. Does reading this article change your practice? If so, how? My threshold to order platelets will be lower. It is interesting to compare this paper’s recommendations to our Massive Transfusion Protocol. The BMC algorithm can be found here: http://internal.bmc.org/sicu/documents/massivetransfusionflowchart2011_000.pdf
Holcomb et al.
1. What are some strengths of the study?
Large prospective RCT.
2. What are the limitations?
Like Alex and Andrew said, underpowered, may be missing smaller but still significant variation (especially since there is a non-statistically significant 4.3% difference in 24-hr mortality). Also, the ratios were only the intended ratios; they did not correspond exactly with what each group got during the intervention. Plus, as Andrew also mentioned, the ratios were irrelevant once the intervention ended, at which point the control group got a lot more plasma and platelets and started to catch up to the other arm.
3. What are the main outcomes of the study?
Main outcomes were 24-hr mortality and 30-d mortality (all-cause). Not significantly different in the 1:1:2 v 1:1:1 transfusion groups. Exsanguination at 24h was decreased and hemostasis was increased in the 1:1:1 group. No difference in complications.
4. Does reading this article change your practice? If so, how?
This is not an arena that I’ve had a lot of experience in, but that said, I think it generally supports the idea that 1:1:1-ish is a reasonable and safe goal.
Gupta et al.
1. What are some strengths of the study?
I think this study is really fascinating. Biggest strength is the joint involvement from both trauma and EM, particularly in authorship and interpretation.
2. What are the limitations?
As the authors admit, single-center, prospective & hypothetical. Also, while their interpretations definitely align within specialties, important to remember that these are only 5 trauma surgeons involved and only 3 of them are authors, so you can’t really generalize that to an entire field.
3. What are the main outcomes of the study?
Primary outcome was how often an undesired scan had abnormal results that resulted in a predefined critical action — which was very infrequent. Most interesting outcome to me was that no one could decide what actually mattered.
4. Does reading this article change your practice? If so, how?
Reinforces the appropriateness of using clinical judgment to determine imaging. A good reminder to ask our surgical colleagues how results will actually change care.
Holcomb et al.
1) strengths: large, multi Center RCT
2) limitations: once blood probably boxes open, providers are not blinded to the therapy. 24-hour and 30 the mortality are important points, but I think many readers would be interested in mortality differences in the acute phase of management, i.e. the first few hours.
3) main outcome: no difference in 24-hour, or 30 day mortality, or consultations observed. 1:1:1 group achieved faster hemostasis.
4) change practice?: not much. no one really knows what the optimal ratio is right now. In the ED phase of management, there is a lot of thought about the ratio, just took the first batch IN.
Gupta et al.
* What are some strengths of the study?
People mentioned that a strength of this study was that it involved both Surgery and EM perspectives, but I was also impressed with how they detailed their methods, they really went into specifics about who was surveyed, and how they defined all of their measures (except for what was a “clinically important” finding). They made a real effort to predefine what would constitute a clinically important injury and management change, but had a hard time coming to an agreement.
* What are the limitations?
They were never able to really agree on what findings on CT were clinically significant, which was their primary outcome. For example, it was never really agreed upon whether ICU admission should be counted as a clinically important management change.
* What are the main outcomes of the study?
The primary outcome was how often a an undesired scan had abnormal findings that resulted in a critical action. The secondary outcome was how often an undesired scan resulted in abnormal findings, regardless of need for critical action.
* Does reading this article change your practice? If so, how?
This reading does not especially change my practice, because the most interesting aspect was the differing perspectives of the surgeons and EM physicians. What it does inspire me to do, however, is be more thoughtful with our surgical colleagues about ordering pan-scans, and having more of a discussion about selectively CTs.
Holcomb et al
* What are some strengths of the study?
As has already been said, the main strength is that this article was a very large, multicenter randomized controlled trial.
* What are the limitations?
I think a major limitation is the fact that they did not power it to look at small differences in survival, and that many of the patients in the 1:1:2 group eventually caught up to the 1:1:1 group by receiving plasma and platelets later on in their hospital course. The study dismisses the bias from this effect by arguing that many prior studies have shown that death often occurs in the first 2-3 hrs, but I think that the potential effect from this should have been controlled for, or at least explored further.
* What are the main outcomes of the study?
The primary outcome was absolute differences in 24hr and 30-day mortality. Secondary outcomes were more focused on safety measures, and included time to hemostasis, amount of blood delivered, and hospital/ICU days, inpatient procedures, and complications.
* Does reading this article change your practice? If so, how?
This article does not really change my practice, because they found no real significant difference in their measures. The bottom line is that right now we do not know what the exact optimal ratio is, but it seems like the most important intervention for survival is getting blood (in some form) to the patient early.
Gupta et al.
* What are some strengths of the study?
I love this study. It’s a prospective, observational paper that asks a very important and clinically meaningful question and gets multiple providers’ input on their interpretation of the results. It’s a paper and an editorial all rolled into one!
* What are the limitations?
Single center study so it is subjected to all biases attributable to their center. This study is hypothetical because it just asks the physicians what they would’ve ordered had there not been a pan CT requested by surgons, (“believe me this is hypothetical” – office space anyone?)
* What are the main outcomes of the study?
The primary outcome was “how often an undesired scan had abnormal findings that resulted in a critical action.” This outcome occurred in 3/992 undesired scans (0.3%). There were disagreements on abnormal scans that did not lead to critical actions and they had an interesting EM and TS editorial.
* Does reading this article change your practice? If so, how?
Not particularly. I think it’s interesting the % change of physicians that would want to know about abnormal findings found on undesirable scans with ranges from 0-26% for EM MDs and 48-88% for trauma surgeons. Most of these findings did not change overall management…but would YOU want to know if you had a small SDH? I don’t know the answer to that. I agree with Ellie and will be more thoughtful about my panscans though it is humbling about some of their findings.
Holcomb et al. PROPPR
1. What are some strengths of the study?
Randomized – controlled, multiple trauma centers,
2. What are the limitations?
Unblinded, excluded a lot of patients because they did not receive blood in prehospital setting or within 1h of arrival, only followed up pts to 30d.
3. What are the main outcomes of the study?
24h and 30d mortality – no significant difference.
Time to hemostasis, 23 pre-defined complications, hospital-, ventilator-, and Icu-free days – no sig difference.
death by exsanguination in 1st 24h was decreased in intervention (1:1:1) group.
4. Does reading this article change your practice? If so, how? I haven’t been in many traumas requiring massive transfusion but I would ask for a 1:1:1 ratio, if need be. Important to remember that blood is preferable to crystalloid fluids in trauma patients. Thanks to Andrew for linking to our protocol!
But don’t take my word for it…(author interview below)
http://emcrit.org/podcasts/proppr/
Holcomb et al.
– What are some strengths of the study?
First time the damage control resuscitation has been examined via large scale, randomized methods. High degree of compliance at 24h and 30h benchmarks. Pt’s were rapidly enrolled which is critical for this type of study (although they weren’t consented).
– What are the limitations?
Sample size too small to get enough power to detect a difference > 10%. Also physicans weren’t blinded too blood products. Would’ve been interesting to see more outcomes w/I the first few hours as recent trauma resuscitation studies have demonstrated that most early deaths due to hemorrhage occur within 2-3 hours.
– What are the main outcomes of the study?
No significant differences in 24h and 30h mortality when comparing 1:1:1 vs 1:1:2. However, 1:1:1 use was found to be a/w faster hemostasis and fewer deaths secondary to exsanguination.
– Does reading this article change your practice? If so, how?
Well I don’t have a practice right now when it comes to trauma resuscitation. However, looking at this paper and the PROMMTT trial(and with the paucity of evidence out) there, I will likely aim to deploy 1:1:1 products ASAP.
Gupta et al.
-What are some strengths of the study?
I appreciate the authors’ willingness to tackle head on the practicing differences and thoughts b/w EM and surgery physicians. I thought the 2 “perspective” pieces at the end gave the reader’s greater insight and as a result the study didn’t appear to pick sides.
-What are the limitations?
Prospective observational study. Single center. No agreement with authors as to what makes up a critical intervention. The hypothetical nature of the study (ie, would the EM physician have been more conservative if the trauma surgeon was not there to order more imaging?) is also a limitation. It also would’ve been great to stratify the decision making by years of experience.
-What are the main outcomes of the study?
EM docs were more willing to forgo elements of pan scan vs trauma surgeons (35% vs 7%). Ten percent of ‘unnecessary scans’ had abnormalities and 0.3% would require a critical action. Would’ve been nice to include disposition related outcomes (ie, ICU, etc) since those decisions play major role in terms of course, costs, etc.
-Does reading this article change your practice? If so, how?
Probably not for now. This article will likely make me think twice about reflexively ordering a pan-scan. And while I probably will do so regardless, I’ll at least think to myself if certain imaging is unnecessary. I do think however that a .03% miss rate requiring intervention when using selective scanning is pretty good. Again, the hypothetical nature of the study needs to be taken into account but a the low failure rate is promising and suggests the EM docs generally have a good handle on the issue vis-a-vis their surgical colleagues.
Holcomb et al.
What were some of the strengths of the study:
– Large study, multi-center, prospective. Well defined clinical outcomes.
What were some of the limitations:
– As the authors mentioned: the physicians involved in these serious trauma cases could not be blinded to which patients were receiving the 1:1:1 versus 1:1:2 protocol once blood products were administered.
What were the outcomes associated with this trial:
– In patients sustaining severe trauma the administration of plasma, platelets and RBCs in a 1:1:1 ratio reduced the number experiencing death due to exsanguination at 24 hours compared to the 1:1:2 group.
Would the findings of this study influence or change my practice:
– the 1:1:1 protocol seems like it makes sense in severe trauma cases if you’re looking to reduce the chance of bleeding out in the first 24 hours. In the traumas I was involved with in Ireland our massive transfusion protocol used the 1:1:2 ratio as a guide. Clearly the thinking has evolved on this issue.
Gupta et al.
What are some of the strengths of this article?
– This prospective observtional single center study is creative in that it compares the expert opinion of trauma surgeons with that of ED attendings with regards to the necessity of “pan scans” and then attempts to look back at scans deemed unnecessary for positive findings.
What are some of the limitations?:
This article was conducted in a single center which the authors admit received lower acuity trauma compared to some other centers. Additionally it is difficult to control for the bias resulting from the knowledge of the attendings that this study is a hypothetical exercise.
What are the outcomes of this study?
– This study set out to determine whether there were certain components of the pan scan that can be selectively omitted without missing clinically important injuries. While ED attendings were more likely to omit some aspects of the scan compared to the surgeons (35% versus 7%) some injuries were missed. The debate between attendings as to what represents a significant injury was the aspect of this study which was difficult to quantify.
Will this study influence or change my practice?
– Clearly no provider wants to miss significant findings when evaluating a patient….particularly in the ED. It’s difficult to balance the concerns of radiation exposure and cost with the concern for an occult injury that could cause significant morbidity and mortality. I can understand how the ED attendings in this study were more likely than the trauma attendings to omit portions of the pan scan. They deal in severe injury while as ED physicians we see then entire spectrum of injury. I think at the end of the day, if you stick to your clinical guidelines and consensus you will make the appropriate choice the vast majority of the time.