Tranexamic acid administration to pediatric trauma patients in a combat setting: the pediatric trauma and tranexamic acid study (PED-TRAX). Eckert MJ, Wertin TM, Tyner SD, Nelson DW, Izenberg S, Martin MJ. J Trauma Acute Care Surg. 2014 Dec;77(6):852-8; discussion 858.
This is a retrospective study of pediatric trauma patients seen at a NATO combat hospital in Afghanistan from 2008 to 2012, which evaluated the comparative effectiveness of no intervention vs. a standard protocol of tranexamic acid (TXA) in trauma patients ≤ 18 years of age. Using a quasi-experimental design with both multivariable logistic regression and 3:1 propensity matching, the investigators examined 766 patients, of which 66 (9%) underwent therapy with TXA. TXA therapy was at the discretion of the treating team, and included standard dosing with 1 g of TXA intravenously within 3 hours of injury, then re-dosed based on clinical assessment. Dosing for small children was not discussed. The TXA group suffered more severe injuries, more penetrating injuries, had lower GCS on arrival, and higher base deficits. 76% of patients underwent surgery and the overall mortality rate was 9%.
Although the TXA group had higher unadjusted mortality (15% vs. 8%, p=0.063), their adjusted mortality was lower in both the regression (OR=0.27, p=0.03) and propensity (12% vs. 15%, p=NS) analysis. Other outcomes including brain injury (based on GCS at discharge) and pulmonary injury (based on need for mechanical ventilation at discharge) also favored TXA. These results are fairly compelling, despite significant potential for bias inherent in a retrospective study, lack of adequate sample size, and multiple hypothesis testing. Questions remain about the efficacy in various subgroups, particularly injury severity and age groups; however, the study has clearly demonstrated the need for a randomized prospective study to evaluate the efficacy and optimal dosing of this intervention in children.
Early surgical intervention for blunt bowel injury: The Bowel Injury Prediction Score (BIPS). McNutt MK, Chinapuvvula NR, Beckmann NM, Camp EA, Pommerening MJ, Laney RW, West OC, Gill BS, Kozar RA, Cotton BA, Wade CE, Adams PR, Holcomb JB. J Trauma Acute Care Surg. 2015 Jan;78(1):105-11.
McNutt et al performed a retrospective review at a single institution in order to determine risk factors associated with blunt bowel and mesenteric injury (BBMI). The authors described a laparotomy as being a “gold standard,” however this was not applied to every patient. Thus, their findings also demonstrated that non-operative management patients do well without the “gold standard” as it was Attending dependent. They identified three specific risk factors associated with BBMI. This included specific CT findings (mesenteric contusion or hematoma with associated bowel wall thickening or an interloop fluid collection or active vascular or oral contrast extravasation or pneumoperitoneum), a white blood cell count of 17.0 or greater, and abdominal tenderness. Patients with two out of three of the above risk factors had a 19 times higher odds of having a BBMI than patients with less than two risk factors. However, the sensitivity and specificity of this cut-off was only 85.7% and 76.2% respectively. This identifies the concern that its use would still lead to a substantial number of false negatives (almost 15%) and false positives. One may feel this is too high if the idea is to limit missed injury. In this retrospective study, this model has yet to be validated and caution must raised if it were to be adopted prematurely. In table 3, on closer review, it describes the median age of the no bowel injury group was 4 years. It may be 44 years of age (?). The author’s findings support the concept that CT findings of mesenteric hematoma, bowel wall thickening, fluid collection, or similar ominous findings plus abdominal pain or leukocytosis, the patient is likely to have an injury sufficient for exploration. Ultimately, the risk score was not validated in a separate cohort of patients, the authors do conclude that prospective validation is necessary.
Do trauma stomas ever get reversed? Godat L, Kobayashi L, Chang DC, Coimbra R. J Am Coll Surg. 2014 Jul;219(1):70-77.
Godat et al performed a review of the California Office of Statewide Health Planning and Development patient database in order to determine the incidence of stoma creation following hollow viscous injury, as well as the incidence of ostomy reversal (both during the initial admission and after discharge). Furthermore, they identified factors associated with both initial admission and post-discharge ostomy reversal. They found that hollow viscous injury was associated with an ostomy creation rate of roughly 11%. 6.4% of these patients had an ostomy reversal during their initial hospitalization while 63% of patients discharged with a stoma had it reversed eventually. The majority of patients (57%) who were discharged with a stoma had their stoma reversed at a different institution than where it was created. Most stomas were reversed within 2 years of their creation. Factors associated with increased odds of stoma reversal during the initial hospital stay included Black race and a Charlson Comorbidity Index of 3 or greater. Among patients discharged with a stoma, a second operation, smoking status, alcohol use, obesity, spinal cord injury, Black race, and treatment at a teaching hospital were significantly associated with a reduced likelihood of undergoing stoma reversal. Conversely, insurance status (either Medicare/Medicaid or private insurance) was associated with an increased likelihood of stoma reversal as compared to not being insured. Based on the author’s findings, investigators would miss more than 50% of the patients if they were not linked to Medicare or have some method of following these patients who went to other hospitals. Using the Cox model to account for competing risk of death was very important. Without this, anytime a patient died, they would have been included in the non-reversed group. This potential error in design would have made the actual rate of reversal appear much worse than was actually demonstrated. This study provides important information regarding what patients are at risk for not having their stoma reversed following diversion secondary to abdominal trauma. Utilizing this knowledge to develop resources to decrease this disparity could ultimately increase the overall rate of stoma reversal, thereby potentially maximizing quality of life.
Obesity and clotting: Body mass index independently contributes to hypercoagulability after injury. Kornblith LZ, Howard B, Kunitake R, Redick B, Nelson M, Cohen MJ, Callcut R. J Trauma Acute Care Surg. 2015 Jan;78(1):30-6; discussion 37-8.
This prospective, observational study examined the relationship of BMI to laboratory measures of coagulation function after level I traumas in a consecutive sample of 377 patients at a level 1 trauma center in California. The investigators measured a number of assays that may be related to coagulation including INR, PTT, platelet count, fibrinogen, factors II, V, VII, VIII, IX, and X, ATIII, protein C, D-dimer, functional fibrinogen level (FLEV), and thromboelastography (TEG) parameters at numerous time points out to 5 days from arrival to the trauma center. They then used multivariable linear regression to evaluate the correlation of these parameters to increasing BMI after controlling for a number of factors including age, sex, injury severity score (ISS), and blood products administered. Finally, multivariable logistic regression was used to evaluate the independent association of BMI with clinical thromboembolic events (CVA, DVT, PE).
The cohort was mostly males (81%), suffering blunt injuries (61%) with mean ISS of 18. 42% had normal BMI, 32% were overweight (BMI 25-30), and 26% were obese (BMI ≥ 30). The multivariable linear regression demonstrated several parameters that may indicate a hypercoaguable state in obese patients including elevated factor IX activity, D-dimer, clot strength, FLEV, and platelet count. In addition, the logistic regression showed an OR of 1.85 (95% CI: 1.12-3.08) for thromboembolic events for every 5 kg/m2 increase in BMI. This study provides an interesting first exploration into the mechanisms behind hypercoaguable state in obesity. Although the study promotes its use of adjustment for multiple hypothesis testing, these adjustments were applied only to specific areas of the analysis and may be inadequate given the number of tests being run. In addition, the investigators did not adjust for mechanism or type of injury, arguably, this may have had an effect on outcomes. Finally, DVT prophylaxis appears independently associated with thromboembolic events, which may be related to unmeasured confounders and may call into question the other results of this analysis. Despite these significant limitations, this study is an excellent, initial exploration of an important topic that clearly merits further evaluation.
Below-the-knee arterial injury: the type of vessel may be more important than the number of vessels injured. Scalea JR, Crawford R, Scurci S, Danquah J, Sarkar R, Kufera J, O'Connor J, Scalea TM. J Trauma Acute Care Surg. 2014 Dec;77(6):920-5.
Scalea et al performed a single institution review of patients who suffered a below the knee arterial injury (BKAI) (n=122) following either blunt (68%) or penetrating trauma (32%) in order to determine how the anatomical type of arterial injury was associated with eventual need for an amputation. From a study of 122 patients, they found that a single below knee vessel injury was associated with an amputation rate of 18%, while a two or three vessel injury was associated with an amputation rate of 29% and 50% respectively. Upon multivariable logistic regression, injuries to the anterior tibial artery were associated with significantly increased rates of amputation than injuries to the peroneal or posterior tibial arteries. The authors suggest that blunt anterior tibialis arterial injury is associated with much higher amputation rates than peroneal or posterior tibial artery injury. In addition, they recommend further studies take place to identify when operative BKAI intervention is warranted.
The recommendation of further studies to simply eschew the often-quoted surgical dogma of non-operative management (as long as a single vessel can feed the leg/foot) is an important one. Exclusion/inclusion criteria in this investigation were not listed. Patients who had died in the hospital were unclear and thus may have interfered with amputation rates. The logistic regression model performed had 7 predictors and 9 outcomes. Analysts may feel that a 5:1 predictor to outcome of interest ratio is needed. Patient outcomes were adjusted according to ISS rather than extremity AIS. It may be that those patients with anterior (compartment) tibial artery injuries had more severe injuries vs those who suffered posterior (compartment) tibial and peroneal artery injuries. Still, identifying these metrics regarding severity of injury are challenging and elusive and simply bear mentioning, and may not be a critical deficiency in this manuscript. The large confidence intervals secondary to predictor: outcome ratio leaves us unclear on the significance of these results. Arguably, a C-statistic may be helpful with the fitness of the predictive model. Lastly, the n was small as only 60 patients had a below knee popliteal injury. This raises the concern of a type II error. Still, this study includes more patients to date than similar studies. In summary, although the statistical analysis could have been performed differently, Scalea et al, raise an important consideration and subsequent investigations may change the approach to BKAI’s. Preservation and possibly reconstruction of the anterior tibial artery could be a consideration despite a patent peroneal or posterior tibial artery as future research will help determine outcomes in blunt and penetrating injury.