The diagnosis of hollow viscus injury (HVI) following blunt trauma remains challenging, and abdominal seatbelt sign (SBS) has been historically offered as a finding which should raise suspicion for HVI. Previous studies suggest HVI can be effectively excluded in SBS patients who lack intra-abdominal findings on CT, but the prediction of HVI in SBS patients has remained elusive and to date existing scoring systems have poor ability to identify the presence of HVI in at-risk patients.
In a planned analysis of their multicenter, prospective observational study, Santos et al aimed to create a pragmatic prediction model to identify the presence of HVI. They examined 754 adult patients with abdominal SBS and a CT that was performed on arrival and prior to any abdominal operation. Candidacy for the inclusion of any predictor variable was predicated on that variable’s existence in real-time and knowable to the triaging surgeon. Therefore, concerning CT findings (including free fluid, free air, mesenteric hematoma or stranding, etc.), SBP <110mmHg, and tenderness or guarding were identified as potential candidates while variables like injury severity score were not. From this, the authors created a predictive model with a high AUROC (0.94, 95% CI 0.9 to 0.98) that culminated in a whole number scoring system ranging from 0 to 17 points with increasing score signifying a higher probability of HVI.
Based on their findings, the authors advocate for a tiered approach of triage and potential management. Scores from 0-5 have a low HVI risk (0.03% - 5.36%) and may be observed while patients scoring between 10-17 have a substantial risk (58.6% to 99.7%) and should undergo diagnostic laparoscopy. Those falling in between, 6-9 points, have a moderate risk (10.6% to 44.1%) and should be strongly considered for exploration, but not mandated.
This study addresses a common conundrum in trauma patients and possesses several strengths including its pragmatic approach and high predictive model performance. Interestingly, quirks of the scoring system may suggest alterations in management that abut conventional practice, including isolated free air which would confer a 2.41% risk of HVI but would fall in an observation pathway of care. The authors, however, aptly point out that pneumoperitoneum is an imperfect predictor as HVI is not the only mechanism generating free air. In their cohort, no patient with isolated free air had concomitant HVI. Instead, they advocate for a marriage of clinical judgment augmented and supported by a robust scoring system.
A randomized controlled trial evaluating the use of REBOA as an adjunct to standard care in the treatment of exsanguinating traumatic hemorrhage has finally been published. The UK-REBOA trial included 90 patients, approximately half of whom were randomized to either standard care or standard care supplemented with the use of REBOA. Mortality was the primary outcome, based upon which the trial was stopped early due to evidence of harm. Jansen et al, conclude that using REBOA does not reduce, and may increase, mortality in exsanguinating hemorrhage compared to standard care alone. The data published in the literature on the use of REBOA is conflicting, but these findings ostensibly corroborate the findings of national registry reviews such as Joseph et al (1) and the Japanese experience published by Norii et al (2) which similarly suggested that the use of REBOA was associated with mortality.
Does this trial spell the end of REBOA? For now, at least, the study’s conclusions may be far stronger than the limitations of the collected data would suggest. First, the patients included in the REBOA arm were more hypotensive on arrival and had a higher median AIS score for the head region (indicating severe TBI). These findings may suggest that some of these patients were not salvageable irrespective of the use of REBOA. Second, as an intention-to-treat study there is variance in the interventions for individual patients in each arm. As such, only 41% of patients in the REBOA actually received arterial access and had a balloon inflated. This resulted in a grand total of 19 patients (spread out over 16 centers across a period of 5 years) in the REBOA arm, and two patients in the standard care arm, who actually received a REBOA intervention providing an exceedingly low sample size. Consequently (and thirdly), such low intervention rates across centers of which many had never used REBOA prior to this study is damning with respect to individual center volume, capability, and experience with the use of REBOA (despite pre-trial training). I.e. higher volume centers may potentially have had better outcomes. It is similarly unclear who exactly is placing these catheters in the Emergency Room. Fourthly, median prehospital time was 90 minutes. This is a long period of time for hemorrhaging, severely injured patients (median ISS was 41 – as a comparison the median ISS of patients in the PROPPR trial was 26). Similarly, the median time to definitive hemorrhage control in the standard care group was 64 minutes, and 83 (up to 156!) minutes in the REBOA group. The influence of time to hemorrhage control on mortality is incontrovertible and almost certainly influenced the outcomes in this study. Fifthly, it took a median of 32 minutes to insert and inflate the REBOA balloon after patient arrival to the ED, and it stayed inflated for a median of 29 minutes (range of 19-64 minutes). Position statements have declared that REBOA not be used if patients cannot obtain hemorrhage control within 15 minutes, and that ischemia time of >30 minutes is associated with increased mortality. This suggests that REBOA in this context was used sub-optimally.
This discussion helps contextualize the main takeaways. REBOA is a tool to help prevent cardiovascular collapse in the exsanguinating patient en route to hemorrhage control. Minimizing the time to a definitive operation is key – this entails optimizing prehospital capacities and hospital systems, as well as maintaining the skills of individual clinicians through practice at high volume centers.
Oh yes, it’s another whole blood study. Dr. Sperry et al recently published a pragmatic multicenter study looking at the outcomes of whole blood vs components, initially intended to focus on whether there was a benefit in TBI patients. Spoiler alert: there was not. But what they did find was still intriguing - in a subset analysis of patients with an elevated risk of prehospital mortality, they found that those who received low-titer group O whole blood (LTOWB) had a significantly decreased documented mortality. Essentially, they found that while the components group saw a more-or-less direct correlation between predicted and actual mortality, the LTOWB group plateaued. Those with a 50% predicted risk of mortality based on prehospital characteristics (mechanism, vitals, etc.) showed an almost 40% decreased risk of mortality if they received LTOWB over components.
This is not a perfect study. As a pragmatic trial a lot was left up to the discretion of the individual institution - including leukoreduction, titer levels, and indications/contraindications for whole blood transfusion. There was also a good amount of crossover with only 66% of patients at LTOWB sites receiving whole blood during their resuscitation. Additionally, a good number of patients received components in addition to whole blood. In a lot of ways this makes sense - whole blood is a limited resource and not always available (or available in abundance).
I think that regardless of its drawbacks it is still a valuable study in the evaluation of using whole blood for hemorrhagic shock. It argues for the safety of BOTH component and whole blood approaches to resuscitation in hemorrhagic shock - which is important for those institutions with the resources for whole blood as well as the ones who do not. The area that warrants further investigation is those patients with severe trauma and high risk of mortality.
It's Membership Renewal Time -Sign into your Profile to check your renewal status and pay your 2024 dues.
This Literature Review is being brought to you by the EAST Manuscript and Literature Review Committee. Have a suggestion for a review or an additional comment on articles reviewed? Please email litreview@east.org. Previous issues available on the EAST website.
