All posts by TheTraumaPro

Best Of AAST 2022 #11: Trauma And The Gut Microbiome

You know I don’t usually write about animal studies. I’m going to break that rule today to review an abstract that addresses what I think is an under-appreciated contributor to outcomes in trauma. The gut microbiome describes the collection of all genomes from microorganisms found in a particular environment. These genomes include bacteria, viruses, and fungi and can be found on all external surfaces of humans.

And I use the term “external” loosely. It includes the areas of the human body that are obviously exposed to the environment, but also areas where our body is wrapped around yet still separate from the it, such as the aerodigestive tract and vagina.

We are beginning to recognize the importance of the micro-organisms that inhabit these areas. They aid in digestion, fine tune the immune system, and synthesize proteins, amino acids, and vitamins that are essential to our health to name a few key tasks.

Many things can disrupt the microbiome including disease, diet, stress, and antibiotics. Previous work has shown that the microbiome changes throughout the hospital stay after trauma. Beneficial species tend to die out, and the ratio of pathologic vs beneficial species tilts toward the dark side.

The group from the University of Florida studied the effects of trauma and chronic stress in a group of rats to study the impact on the gut microbiome. One group of rats was subjected to a polytrauma model including pulmonary contusion, shock, cecectomy, and femur fractures. Another received the polytrauma treatment plus two hours of restraint stress daily. These groups were compared to an untreated control group. Gut flora were measured at baseline and on days 3 and 7.

Here are the factoids:

  • As expected, the microbiomes were similar across all groups at baseline
  • Polytrauma caused a significant change in bacterial diversity at both days 3 and 7 with both Bacteroides and Enterococcus prevalent
  • Polytrauma plus stress also depleted “good bacteria” and was associated with a switch to predominantly Enterococcus colonization

The authors suggested that the observed transitions to a pathologic microbiome may influence outcomes after severe trauma and critical illness.

Bottom line: I wanted to highlight this simple study because it relates to a similar topic that is exploding in the clinical nutrition field. The gut microbiome is being recognized as a key element of our overall health. However, it is very sensitive to external events and can be “knocked out of whack” by stress, trauma, bad diet, and even a single dose of antibiotics. Its derangement is recognized as a major factor in the development of C Difficile colitis.

This simple little rat study confirms that major trauma and stress negatively impact the animals’ microbiome. It did not examine outcomes, so no associations can be made here. Any such associations would not be directly applicable to humans, anyway. But it should serve to stimulate some thought and additional human studies to continue investigation in this field.

I have been struck by how we mistreat the gut microbiome in hospitalized patients through my own clinical observations over the years. A short course of antibiotics has been shown to severely impact the diversity of gut flora within days, and may require a year or more to recover back to baseline.

Extended fasting exhausts the food supply for the bacteria which may lead to the use of the gut lining for food, creating additional pathology. The composition of the nutritional supplements used in hospital are formulated from cheap ingredients which have been shown to disrupt the microbiome. Then add on trauma and chronic stress. It’s a terrible combination, yet we see it every day in hospitalized patients!

I predict that we will learn to pay more attention to all our various microbiomes in the future. A more thorough understanding may allow us to reduce complications (think C Diff) and might help us recognize some subtle factors that are contributing to overall mortality. 

Here are my comments and questions for the authors / presenters:

  1. The audience will not be familiar with the microbiome diversity measures described in the abstract. Please take a little time to explain it, what is normal, and what happens when it changes.
  2. Were there any obvious outcome correlations observed that were not reported?
  3. Where do you go from here? Any plans for human studies on this topic?

As you can see, I find this area fascinating and believe that it is an underappreciated source of outcome variability in the patients we take care of. Figuring this out will help us tweak and optimize our overall patient care.

Reference: MULTICOMPARTMENTAL TRAUMATIC INJURY AND THE MICROBIOME: SHIFT TO A PATHOBIOME. Plenary paper #54, AAST 2022.

Best Of AAST 2022 #10: REBOA For Pelvic Fractures

Papers for Resuscitative Endovascular Balloon Occlusion of the Aorta (REBOA) use in trauma patients have been accumulating for the past decade or so. There are three zones within the aorta, and the REBOA balloon can be inflated in either Zone II for abdominal vascular injuries or Zone III for pelvic bleeding. Here’s a nice diagram for reference:

The original studies focused on Zone II deployments, but over the past five years or so there has been growing focus on using REBOA in Zone III. Pelvic arterial bleeding can be quite problematic, and if the patient is hypotensive in the emergency department, it is not permitted to take them to interventional radiology for management. The only choice is a trip to OR for preperitoneal packing or some less effective procedure.  With the use of Zone III REBOA, it became possible to stabilize vital signs and then allow transport to IR or hybrid OR for angioembolization.

After the initial rush of very positive research, more recent papers are a bit more tempered with results that are not as cut and dried. This abstract from Hartford Hospital attempts to add a bit more information about the use of Zone III REBOA in patients who required a hemorrhage control procedure. The authors performed a retrospective review of four years of TQIP data. They compared outcome data in hypotensive adult patients with pelvic fractures requiring some type hemorrhage control procedure. The authors divided patients into REBOA+ and REBOA- groups, and looked at mortality, blood utilization, lengths of stay, and REBOA complications.

Here are the factoids:

  • Of the 4,453 patients who met inclusion criteria, only 139 patients underwent Zone III REBOA
  • The REBOA+ patients had lower BP and GCS and ISS was higher prompting the use of propensity matching
  • In order to equalize comparisons patients were propensity matched the the variables listed above, yielding 121 pairs for study
  • In-hospital and 24-hour mortality were double in the REBOA+ group (50.5% vs 25% and 31% vs 14.3%, respectively)
  • Blood transfusion was also higher in the REBOA+ group (median 4L vs 1.75L)
  • AKI was higher in the REBOA+ group (16% vs 7%) but the amputation rates were the same (no numbers given)

The authors concluded that Zone III REBOA appears to have worse outcomes and suggest that more prospective studies are indicated.

Bottom line: I have been a REBOA skeptic for some time so I have to be careful not to feed my confirmation bias. Many of the previously published positive papers include authors who have a relationship with one of the major REBOA device manufacturers. Papers from centers without any conflicts of interest are generally less positive.

Despite my own bias I also have some major questions about this abstract. The biggest thing is that I can’t make the statistics work. Granted, the entire analysis is not in the abstract. But the mortality rates given in percentages don’t yield integers when multiplied by the 121 patients in each group.

This makes me worry that we are not seeing all the statistics and we are somehow getting relative risk rather than absolute risk. I’m also confused about why the SBP, ISS, and GCS matched groups would have a difference in AKI rates in addition to the mortality numbers. Are there other significant variables affecting morbidity and mortality that were not identified or controlled?

Here are my comments and questions for the authors / presenter:

  1. Review the mortality calculations for us. How can you have a hospital mortality rate of 50.5% with 121 patients (= 61.105 people)? Provide the absolute mortality numbers so we can do the math.
  2. Why are the transfusion numbers so much higher in the REBOA+ group? Isn’t this device supposed to reduce bleeding? It seems unlikely that REBOA is making them bleed more. Is there something else going on?
  3. Similarly, why would AKI be higher in REBOA+ patients? The balloon is located below the kidneys and they should benefit from better perfusion.
  4. Could there be other factors not analyzed that contributed to the poorer outcomes in the REBOA+ group? What might they be?

I suspect the snapshot that the TQIP data allows may not be enough to let us see the entire picture here. I am looking forward to additional information during the presentation to help clarify these issues.

Reference: DOES THE USE OF REBOA IMPROVE SURVIVAL IN PATIENTS WITH PELVIC FRACTURES REQUIRING HEMORRHAGE CONTROL INTERVENTION? AAST Plenary paper #46, AAST 2022.

Best Of AAST 2022 #9: Management Of Low Grade Blunt Thoracic Aortic Injury

It has been interesting to watch the evolution of the treatment of blunt thoracic aortic injury (BTAI). In my training, patients with an abnormal mediastinal contour on chest x-ray were whisked away to interventional radiology for a thoracic aortic angiogram. Yes, there was no such thing as a CT scan for anything but the head! Here’s a sample chest x-ray:

I wrote a post 12 years ago describing the various findings of a blunt aortic injury, many of which are present above. You can the post by clicking here.

In the 1-2% where the angio was positive (yes, you read that right; we did a lot of negative angiograms) the patient was then whisked off to the OR for an open aortic graft via thoracotomy. The advent of CT scans of the torso and improving resolutions allowed us to be more selective with angiography. And finally as CT angiography matured, aortic angiography became a thing of the past.

Then came TEVAR about 20 years ago. There were some growing pains as we refined the technology, but now this endovascular procedure is the standard of care for most aortic injuries. While there was also a place for nonoperative management, it was really just maintaining a reasonably low blood pressure to protect the vessel while getting the patient into good enough condition to tolerate anoperation.

Now we are beginning to slice and dice treatment based on grade. Many centers have recognized that an intimal injury (Grade I) is only a very minor disruption to the inside of the vessel and does not make it more susceptible to rupture. Grade II management has been less clear. Here is a diagram of the various grades.

It makes sense that an invasive procedure may be less helpful for injuries that do not disrupt the layer of the vessel that provides its strength. But remember, common sense isn’t always the truth.

The group at Dell Seton Hall reviewed all patients with low grade BTAI (Grades I and II) in the Aortic Trauma Foundation Registry for a six year period. Their hypothesis was that these injuries could be successfully treated with medical management alone. They reviewed the data for mortality, complications, vent days, and lengths of stay.

Here are the factoids:

  • A total of 880 patients were enrolled and 274 had low grade injuries; 5 were then excluded when their lesion progressed and they underwent TEVAR
  • Of the 269 remaining patients, 81% were treated with medical management (81% Grade I, 19% Grade II) and the remainder with TEVAR (20% Grade I, 80% Grade II)
  • Rates of thoracotomy, craniectomy, and sternotomy were the same in both groups, but TEVAR patients were more likely to have a laparotomy (31% vs 15%)
  • Mortality was significantly higher in the TEVAR group (18% vs 8%) but the mortality from the aorta was not quite significant (4% vs 0.5%)
  • Complications (DVT and ARDS) were also significantly higher in the TEVAR group
  • Vent days and lengths of stay were equivalent

The authors concluded that medical management alone is safe and appropriate, with a lower mortality and decreased complications compared to routine TEVAR.

Bottom line: Hmm, color me skeptical. Remember, this is a registry study, so information tends to be limited outside the usual demographics and data points that are very pertinent to the purpose of the registry. The most important concept is that the patients in each group must be identical in every way except for the intervention.

We can try to make them as identical as possible by matching pairs or subgroups of patients. But the biggest problem is that the number of patient characteristics that might be important to match may not be available for analysis in the database.

When the patients were initially treated at contributing centers, there were no specific rules that the individual surgeons had to follow to decide between medical management and TEVAR. They could pick and choose based on their own experience. Could the surgeon have recognized some patients as higher risk and opted for TEVAR to make sure the aorta would not become an issue in conjunction with their other injuries? And unfortunately, perhaps that higher risk issue is what ultimately killed them and not the aorta. It’s basically a form of unhealthy user bias.

This abstract is an interesting tidbit that should push us to question whether medical management is better, at least in some subsets of low grade aortic injury patients. Then someone can perform a more robust study to confirm or refute its safety. Unfortunately, this may never happen due to the low incidence of this injury. It took six years to accumulate only 269 eligible patients in this registry!

Here are my questions for the authors and presenter:

  • What data points are actually in the registry? Specifically, is there fine detail about the other injuries the patient had? Could these have contributed to TEVAR mortality, or the selection of the patient for TEVAR to try to reduce the perceived mortality risk?
  • When you stated that there was “no difference in demographics or mechanism of injury” what were these, exactly?
  • What did patients actually die from in the Grade I and Grade II groups? Be more specific than “aortic-related” or not.
  • Do you have any worries about the five patients whose lesions progressed? When should patients be re-scanned to identify those who might benefit more from TEVAR?

This is a thought provoking abstract, and I am very interesting in hearing what the next steps should be.

Reference: MEDICAL MANAGEMENT IS THE TREATMENT OF CHOICE FOR LOW GRADE BLUNT THORACIC AORTIC INJURIES. Plenary paper #45, AAST 2022.

Best of AAST 2022 #8: Financial Vulnerability Of Trauma Centers

Here’s another abstract on trauma center / system finance. Trauma centers are part of the safety net in the healthcare systems of many countries. The way they are funded varies tremendously. In the US, health insurance pays most of the bill for patient care. Unfortunately, not all patients are covered, so there is financial risk to the center based on how many underpaying patients present for care.

The group at Scripps Mercy in San Diego performed a financial health analysis of all ACS-verified trauma centers in the US. They applied a Financial Vulnerability Score metric (FVS), although I could not locate anything on this via an internet search. They analyzed the RAND Hospital Financial Database, which is based on information obtained from the CMS Healthcare Cost Report Information System. Using this data they calculated the FVS for each center. They sub-grouped the hospitals into high, medium, and low vulnerability and compared them.

Here are the factoids:

  • A total of 617 trauma centers were identified and analyzed: 194 Level I, 278 Level II, and 145 Level III
  • Level III trauma centers made up 59% of the high financial risk centers
  • The majority of Level I and II centers were in the middle or low risk categories
  • Characteristics of high risk centers were lower number of beds, negative operating margin, and less cash on hand
  • Low risk centers had greater asset to liability ratios, lower outpatient shares, and 3x less uncompensated care
  • The largest proportion of HFR hospitals were in New England and East North Central regions
  • Non-teaching centers had significantly higher financial risk than teaching hospitals (46% vs 29%)

The authors concluded that about 25% of Level I and II trauma centers are at high financial risk and that factors such as payor mix and outpatient status should be targeted to reduce this risk.

Bottom line: This is a fascinating abstract but leaves a lot to the imagination. The databases used have not been used in previous papers, and the information contained in them is proprietary. The FVS is also new and I have not been able to obtain any details.

Nevertheless, if the data and analysis are sound it may provide some new information to trauma centers and perhaps some insight on what factors to address to lessen their financial vulnerability. This is a lot of ifs. Hopefully the authors will enlighten us during the presentation so we can appreciate the real world value of the analysis.

Here are my questions and comments for the authors/presenter:

  • Please explain both the dataset used and the new FVS metric. Most readers and listeners are unfamiliar. We need to see how the data and analysis apply to trauma center financials and how the FVS has been validated.
  • How can the vulnerability factors be addressed? Payor mix is based on patient coverage and their socioeconomic status. It would seem to be difficult to manipulate by the trauma center. Outpatient vs inpatient status is also difficult to change and not fall afoul of CMS rules. What were other factors that were identified that could help centers reduce their financial vulnerability?

This could be an interesting abstract, but there was not enough room in the abstract to reveal all the details. Hopefully all will become obvious during the presentation.

Reference: FINANCIAL VULNERABILITY OF TRAUMA CENTERS: A NATIONAL ANALYSIS, Plenary paper #41, AAST 2022.

Best Of AAST 2022 #7: Funding State Trauma Systems

As of about 10 years ago, every state in the US has implemented a state trauma system. The way these systems are funded and their sources vary widely. Most receive their funds via some combination of a line item in the budget, grants, or fees levied on state residents. The total per capita funding amount can be calculated based on the total dollars received divided by the state population.

The per capita funding across the US varies widely. Some provide very little assistance while others are much more generous. I have visited trauma centers in nearly every state and have witnessed the difficulties many state trauma systems face due to funding shortfalls.

Does financial support have an impact on trauma care? A better question is, do state systems have an impact on trauma care? The literature on that last question tells us the answer is a resounding yes. We assume that having some money improves what the system can do. But that’s just an assumption. Until now.

The group at Northwestern University did a deep dive into the financial support of state trauma systems to address this question. Their goal was to measure the impact of it on something easily measurable: mortality. They selected five states to study based upon their broad geographic and demographic variability and the availability of comprehensive data for analysis.

Two states (MA, NY) allocated $0.00 per capita trauma funding, and three provided between $0.09 and $1.80 per capita.  The authors analyzed data from comprehensive patient encounter databases from these states. Their primary goal was to compare mortality differences between the two levels of support. They were also able to link ED and inpatient encounters and analyze triage decisions for appropriateness.

Here are the factoids:

  • Nearly a quarter of a million patients with ISS > 15 were analyzed over a study period of two years
  • Median age was 72 and ISS was 17 (older with significant injuries)
  • Trauma mortality was significantly decreased in the states that provided financial support compared with those that did not (odds ratio 0.75)
  • All triage levels were associated with a similar decrease in mortality (appropriate triage to a Level I or II center, undertriage admitted to a Level III or IV center, and re-triage (emergent transfer from an ED to a Level I or II center))
  • The best improvement in mortality was seen in re-triaged patients in funded states (odds ratio of 0.63)

The authors concluded that state system funding was associated with lower mortality in severely injured patients and that increasing funding may facilitate it.

Bottom line: Several papers have been written showing the benefit of having a trauma system, and all of the US has bought into this concept. However, funding of these systems varies widely by state. The assumption (up to now) has been that just having a system is enough. This abstract goes a step further and suggests that actually paying for it is important, too. 

The authors performed quite of bit of data manipulation and used some sophisticated statistical tools that I am not familiar with. However, the authors are certainly qualified, and I am going to assume that the analysis was sound. 

The one important point to keep in mind is that this paper is one of the first of its kind but can only show association, not causation. Nevertheless, it is an important contribution to the trauma system literature. It will definitely prompt additional research which will hopefully corroborate the results. This line of work should give state systems the ammunition to approach their legislatures to open the pocketbook and improve the outcomes of their residents. Mortality reductions of 25-37% cannot be ignored!

Here are my questions and comments for the authors/presenters:

  1. Please explain the data sources you used and the analysis (briefly). These are very comprehensive databases and appear to contain a wealth of data. Tell us why you chose them, how you linked them, and explain your analyses.
  2. Why limit your analyses to these five states? Could you have found some with even higher levels of support that would drive your point home even better?
  3. Be sure to provide more detail on your analysis of triage status. There are some important lessons to be learned here regarding state triage criteria vs mortality.
  4. Are you planning any additional similar studies? This is important work that will help state systems obtain a more appropriate level of support.

This is an intriguing abstract that should prompt additional research regarding state trauma system funding. 

Reference: THE IMPACT OF STATE TRAUMA FUNDING ON TRIAGE AND MORTALITY OF TRAUMA PATIENTS. Plenary paper #38, AAST 2022.