11.05 Tranexamic Acid is Associated with Increased Mortality in Patients with Physiologic Levels of Fibrinolysis

Posted on January 25, 2017

H. B. Moore1, B. Huebner1, T. L. Nydam1, G. Settler1, G. Nunns1, C. C. Silliman1, A. Sauaia1, E. E. Moore1  1University Of Colorado Denver,Surgery,Aurora, CO, USA

Introduction: Utilization of tranexamic acid (TXA) in trauma remains debated.  While European guidelines recommend empiric TXA in hypotensive trauma patients, many trauma centers in the United States question this practice.  Recent appreciation of the spectrum of fibrinolysis acutely after injury has identified an associated protective effect of a moderate level of fibrinolysis.  There are concerns that TXA may harm this patient population.  TXA administration at our trauma center is goal directed on rapid thrombelastography (rTEG) LY30 results, although clinicians can empirically administer TXA if they believe it is indicated.   We hypothesize that this is a futile intervention and poses the risk of increased mortality in patients with physiologic fibrinolysis levels.

 

Methods: Trauma activations from 2015-2016 with blood samples obtained in the ambulance or emergency department were analyzed with rTEG.  Patients included in the analysis had an LY30 between 0.8 and 2.9 (previously defined as physiologic fibrinolysis).  Demographics, clinical variables, and blood product utilization were collected by prospectively by trained research assistants.  The primary outcome of interest was in hospital mortality contrasted between patients who received TXA and no TXA. Confounders(age, NISS, systolic blood pressure(SBP), Glasgow Coma Score(GCS), RBC transfusion in the first 2 hours from injury) were adjusted with multivariate logistic regression and cox regression analysis.

 

Results:  Fourtny nine percent of patients (141/291) were identified to have physiologic levels of fibrinolysis  The median NISS was 27 and mortality rate was 6% (significantly less than hyperfibrinolysis 20% and shutdown 16% p=0.004). Patients with physiologic phenotype were given TXA 5% of the time and delayed delivery(>3 hours) occurred in 38% of patients. NISS was higher but not significant in patients given TXA (48 vs 27p=0.334), while SBP (108 vs 118 p=0.325) and GCS were similar (15 vs 15 p=0.779). TXA patients received more RBC units at hour 1 and 2 during resuscitation (1 vs 0 p<0.001 and 2 vs 0 p=0.001). Mortality was significantly higher in the TXA group 38% vs 4% (p=0.004). After adjusting for confounders TXA was significantly associated with increased mortality in logistic regression analysis (p=0.024) and cox regression (HR 14.5 p=0.042). In patients with hyperfibrinolysis there was no differnce in survival with TXA use before (p=0.521) and after adjustment (p=0.531).

 

Conclusion:Patient's with physiologic levels of fibrinolysis that receive TXA have increased mortality compared to patients who did not receive this medication.  While the TXA patients had an overall higher requirement of blood products and were given this medication based on clinician gestalt, there was no observed benefit. These data support the continued concerns of empiric utilization of TXA, and has identified a potential danger of giving this medication to patients who present to the hospital with physiologic levels of fibrinolysis. 

 

11.04 Early Thromboprophylaxis With Low Molecular Weight Heparin In Patients With Pelvic Fractures Is Safe

Posted on January 25, 2017

F. Jehan1, K. Ibraheem1, A. Azim1, A. Tang1, T. O’Keeffe1, N. Kulvatunyou1, L. Gries1, G. Vercruysse1, R. Friese1, B. Joseph1  1University Of Arizona,Trauma,critical Care, Burn And Emergency Surgery/Department Of Surgery,Tucson, AZ, USA

Introduction:
Early initiation of thromboprophylaxis is highly desired in patients with pelvic fractures but it is often delayed due to fears of re-bleeding and hemorrhage. The aim of our study was to assess the safety profile of early initiation of venous thromboprophylaxis in patients with pelvic trauma.

Methods:
Three year (2010-2012) retrospective study of trauma patients with pelvic fractures presenting at single level-I trauma center was performed. Patients who received thromboprophylaxis with low molecular weight heparin (LMWH) during their hospital stay were included. Patients were stratified in two groups based on timing of initiation of prophylaxis; early (initiation within first 24 hours) and late (initiation after 24 hours) initiation. Signs of bleeding or hemorrhage were defined as presence of pelvic hematoma, free fluid, or blush on CT scan. Decrease in hemoglobin (Hb) was defined as difference between admission Hb level and lowest post-prophylaxis Hb level. Our primary outcome measures were decrease in Hb levels, pRBC units transfused, and need for hemorrhage control (operative or angioembolization) after initiation of prophylaxis. Secondary outcome measures were hospital and ICU length of stay. Multivariate regression analysis was performed.

Results:
 

255 patients were included (158 in early and 97 in late group). Mean±SD age was 48.2±23.3 years, 50.6% were male, and mean±SD number of pRBC units was 0.62±1.59. After adjusting for confounders, there was no difference in the decrease in Hb levels (b= 0.087, 95% [CI]=[-0.253 – 1.025], p=0.23) or pRBC units transfused (b= -0.005, 95% [CI]= [-0.366 – 0.364]; p=0.75) between the two groups. Only one patient required hemorrhage control after initiation of thromboprophylaxis and belonged to the late group. There was no difference in the hospital LOS (b=0.120, 95% [CI]= -0.165 – 4.929; p=0.67). ICU length of stay was significantly shorter in early prophylaxis group (b= 0.206, 95% [CI]= 0.206 – 4.762; p=0.03).

On sub-analysis of patients with signs of bleeding or hemorrhage (n=52), there was no difference in decrease in Hb levels (b= 0.131, 95% [CI]= -1.411 – 2.586; p=0.55) or pRBC units transfused (b= -0.007, 95% [CI]= -1.588 – 1.518; p=0.96) between the two groups

Conclusion:
Our study shows no difference in pRBC transfusion requirements, drop in hemoglobin levels, or need for hemorrhage control between early and late initiation of thromboprophylaxis. We conclude that fear of hemorrhage with early thromboprophylaxis is not substantiated in patients with pelvic fractures

11.03 Early Versus Delayed Prophylactic Anticoagulation In Adult Trauma Patients With Pulmonary Contusions

Posted on January 25, 2017

M. B. Linskey1, A. B. Podany1, A. S. Kulaylat1, A. L. Lauria1, S. R. Allen1,2, J. D. Chandler1,2, R. M. Staszak1,2, S. B. Armen1,2  1Penn State University College Of Medicine,Department Of Surgery,Hershey, PA, USA 2Penn State University College Of Medicine,Division Of Trauma, Acute Care & Critical Care Surgery,Hershey, PA, USA

Introduction: Pulmonary contusions (PC) lead to morbidity and mortality in trauma patients, placing them at increased risk for mechanical ventilation, acute respiratory distress syndrome, and pneumonia. Tissue injury and hemorrhage in PC result in inflammation, edema, atelectasis, and intrapulmonary shunting even in the uninjured lung. We hypothesized that early prophylactic anticoagulation (pAC) would be associated with worsened respiratory outcomes in patients with PC.

Methods: A retrospective cohort study identified patients with PC from a rural Level I trauma center’s institutional registry. Patients with severe traumatic brain injury, prior use of therapeutic anticoagulation or antiplatelet therapy, and those who did not receive pAC were excluded. The cohort was stratified into those receiving early or delayed pAC, within or after 48 hours of admission, respectively. Outcomes including 30-day mortality, 30-day venous thromboembolism (VTE) rate, retained hemothorax, and pneumonia were modeled using multivariable logistic regression to control for patient and injury characteristics. Propensity score matching was then used to isolate two groups with similar comorbidities and injuries. Univariate statistics were performed to compare nadir oxygen saturation levels and supplemental oxygen requirements between the two groups before and after administration of pAC.

Results: 356 patients met inclusion criteria; 195 in the early and 161 in the delayed groups. The groups did not differ with respect to age, sex, race, mechanism, pulmonary comorbidities, number of rib fractures, or proportion with flail chest. The group receiving delayed pAC had lower admission GCS scores (12.0 vs 14.1, p<0.001) and higher injury severity scores (27.7 vs 20.0, p<0.001), and was significantly more likely to have bilateral PC (41.3% vs 28.4%, p<0.05), concomitant solid organ injury (42.2% vs 12.8%, p<0.001), intracranial or spinal hematoma (35.4% vs 5.64%, p<0.001), or other organ space hematoma (28.0% vs 14.9%, p<0.01). After controlling for differences between the groups, initiation of pAC within 48 hours of injury in patients with PC did not significantly increase the odds of 30-day mortality. Similarly, early pAC was not significantly associated with retained hemothorax or pneumonia. Delayed pAC was also not associated with VTE. Of the propensity score-matched groups, those with early pAC had a decrease between their pre- and post-pAC nadir oxygen saturation levels while those with delayed pAC had a slight increase (93.2% to 90.1% among early vs 90.9 to 92.1% among delayed, p<0.001). Changes in oxygen requirements before and after pAC, however, did not differ between the two groups (37% to 28% among early vs 36% to 25% among delayed, p=0.401).

Conclusion: In this study, early vs delayed pAC did not significantly impact outcomes in patients with PC, suggesting that other clinical factors should guide timing of pAC in adult trauma patients.

11.02 Putting the Pieces Together: A Principal Component Analysis of Acute Traumatic Coagulopathy in Kids

Posted on January 25, 2017

C. M. Leeper1,2, M. D. Neal2, C. McKenna1, T. Billiar2, B. A. Gaines1  1Children’s Hospital Of Pittsburgh Of UPMC,Pediatric Surgery,Pittsburgh, PA, USA 2University Of Pittsburgh Medical Center,General Surgery,Pittsburgh, PA, USA

Introduction:
Injured children commonly present with acute traumatic coagulopathy (ATC) defined by elevated international normalized ratio (INR). ATC is associated with poor outcome, though these patients usually are not clinically coagulopathic. INR, therefore, is not always a therapeutic target but rather a marker of complex systemic dysregulation. Our goal is to evaluate multiple coagulation parameters that encompass the broader hemostatic system and identify patterns after injury that may be associated with clinical outcomes.

Methods:
We performed principal components analysis (PCA) on prospectively collected data from children with highest trauma activation in our pediatric center from June 2015-June 2016. Admission labs included INR, platelet count and thromboelastography (TEG) parameters (clotting factors (ACT), fibrinogen (K), platelet function (MA) and fibrinolysis (LY30)). Variables were reduced to principal components (PC) and PC scores were generated for each subject for use in logistic regression. Outcomes included mortality, disability (based on functional independence measure score or discharge to rehabilitation facility), venous thromboembolism (VTE; screening ultrasound for high-risk or symptomatic patients), and blood transfusion in the first 24 hours.

Results:
133 subjects were included with median(IQR) age =10(5-13), median(IQR) ISS =17(9-25), 73.5% male, 70.8% blunt trauma. The rate of mortality was 5.6% (n=7), disability was 23.9% (n=28), early blood transfusion was 26.3%(n=35) and VTE was 10.3%(n=11). PCA identified 3 significant PCs accounting for 75.0% of overall variance. PC1 identified clot strength (platelets and fibrinogen); PC2 identified abnormal fibrinolysis, both hyperfibrinolysis and fibrinolysis shutdown (LY30 and INR); and PC3 identified global clotting factor depletion (INR and K). PC1 score was associated with increased mortality (odds ratio [OR] =1.63; p<0.001) and early transfusion (OR 1.36; p=0.002). PC2 score was correlated with ISS (rho 0.4; p<0.001) and associated with VTE (OR 1.84; p=0.034), functional disability (OR 1.66; p=0.017), increased mortality (OR 2.07; p=0.003) and early blood transfusion (OR 2.79; p<0.001). PC3 score was associated with increased mortality (OR 1.92; p=0.007) and early transfusion (OR 1.25; p=0.075).

Conclusion:
PCA detects three distinct patterns of coagulation dysregulation using widely available laboratory parameters: 1) abnormalities in clot strength; 2) abnormalities in fibrinolysis, and 3) clotting factor depletion. All were associated with poor outcomes; however, fibrinolytic dysregulation is associated with more severely injured patients and portends particularly poor outcome including increased mortality, DVT, disability and need for transfusion.
 

11.01 Massive Transfusion Protocol is Associated with Higher Rate of Venous Thromboembolism

Posted on January 25, 2017

N. K. Dhillon1, E. J. Smith1, A. Ko1, M. Y. Harada1, K. Patel1, M. Scheipe1, G. Barmparas1, E. J. Ley1  1Cedars-Sinai Medical Center,Los Angeles, CA, USA

Introduction:  Massive Transfusion Protocol (MTP) is often initiated in patients who are unstable secondary to hemorrhagic shock. Thrombotic events have been associated with MTP, however the risk factors for the development of venous thromboembolism (VTE) within this patient population is unknown.

Methods:  A retrospective review was conducted by examining the electronic medical records of all trauma patients admitted to a Level I trauma center who had MTP initiated from 2011 to 2015. Data was collected on patient demographics, mechanism of injury, injury severity scores, quantity of packed red blood cells (PRBC) transfused during MTP activation, incidence of VTE, ICU length of stay (LOS), hospital LOS, and ventilator days.

Results: Of the 63 patients identified who had MTP activated, 11 (17.5%) developed a VTE during their hospital admission. One patient was diagnosed with a pulmonary embolus. Patients who developed VTE were compared to those who did not. Age (40 (22-62) vs. 42.5 (25.5-54) years, p=0.94), sex (46% vs. 73% male, p=0.09), and mechanism of injury (59% vs. 64% blunt, p=1.0) were similar. ICU LOS, hospital LOS, and ventilator days were longer in the patients who were diagnosed with a VTE (Table 1). Multivariable analysis revealed an increase in the odds for developing a VTE with each unit of PRBC transfused (AOR=1.17, p=0.011).

Conclusion: Patients who received PRBC after MTP activation were at higher risk for developing VTE. Clinicians may need a higher suspicion for the presence of VTE within this patient population.

 

10.20 Challenges of Implementing Trauma Registries in Low and Middle Income Countries

Posted on January 25, 2017

K. Bommakanti1, I. Feldhaus1, R. Dicker1, C. Juillard1  1University Of California – San Francisco,Center For Global Surgical Studies,San Francisco, CA, USA

Introduction:  Trauma registries are an essential part of injury surveillance and trauma quality improvement programs in high-income countries (HICs), where they have been used to decrease trauma morbidity and mortality. In low- and middle-income countries (LMICs), estimated to have the greatest burden of injury, trauma registries are increasingly being implemented. While some LMICs have managed to successfully implement registries, many others have dealt with challenges in adapting the models used by HICs. We sought to analyze the barriers to trauma registry implementation faced by LMICs in order to inform development of sustainable trauma registry models.

Methods:

A structured review of published literature was performed. Relevant abstracts were identified using the PubMed, Embase, and CINAHL databases. The search terms included were: “implement registry”, “trauma registry”, wounds and injuries”, and “injury registry” combined with "LMIC", “developing countries”, and different world regions. Articles including relevant information on trauma registry implementation were reviewed in full and details were abstracted.

Results:

Thirty articles were identified that addressed the challenges of implementing trauma registries. Overall data quality was cited by seventeen articles as the most significant barrier to success. Difficulty with administrative duties and hospital organization were reported by five studies, while three reported on lack of technology and other infrastructure. Two studies identified insufficient prehospital care as the primary concern and the remaining three cited a lack of human resources, lack of trauma education, and unfavorable health care policies. Solutions to identified barriers were proposed by nine articles and included increasing trauma education, simplifying trauma scoring tools, and transitioning to electronic medical systems. All thirty studies acknowledged that the presence of at least one local trauma registry improved injury surveillance and promoted better outcomes.

Conclusion:

Many LMICs still face unique challenges to implementation that they must overcome to create sustainable trauma databases. Understanding these barriers and their proposed solutions, which include trauma education programs and efforts to standardize injury scoring, may facilitate improved trauma registry implementation in LMICs to further improve trauma care models and have a lasting impact on the development of future programs.

 

10.19 Trends in Unintentional Firearm Related Injuries in Rural Pediatric Trauma Settings

Posted on January 25, 2017

I. I. Maizlin1, G. F. Smith1, R. T. Russell1  1University Of Alabama at Birmingham,Pediatric Surgery,Birmingham, Alabama, USA

Introduction:  Gunshot wounds (GSWs) are the second most common cause of traumatic pediatric deaths nationwide. Since most evaluations of pediatric GSWs were performed in urban settings, we utilized our institution’s data from a mostly rural catchment area to identify trends in circumstances and clinical outcomes resulting from unintentional pediatric GSWs and determine consequent geographical and age-based foci for most effective intervention.  

Methods: A retrospective review of our institution trauma registry identified all pediatric patients (≤18 years old) admitted for unintentional GSWs between January 2000 and December 2015. Patients were stratified into 4 age groups: ≤5 years, 6-9 years, 10-14 years, 15-18 years. Demographics, patterns of injury, and outcomes were analyzed by χ2 and ANOVA. Incidence rates were collated by location and compared to the 2010 US Census data to create geographic representations (“heat maps”) of cases per 100,000 residents throughout the catchment area.

Results: 194 children (79.4% male, 72.0% African American) sustained unintentional GSWs, with mean age of 10.7 ±4.6 years. The most common utilized firearm was a handgun (54.2%), followed by a pellet gun (24.7%) and hunting rifle (5.3%). As compared to the overall national trauma mortality rate of 2.7%, unintentional GSWs in our facility resulted in mortality rate of 4.1%. Mean hospital stay was 4.7 days, with median Injury Severity Score (ISS) of 5 (Range 1-13) and most common injury site being the extremity. Young teenagers (10-14 years) were most likely to experience an unintentional GSW, with rates of injury being more than twice as likely as other age groups. Younger age groups were associated with higher mean ISS at presentation (15.8 vs. 10.8 vs. 9.0 vs. 5.7, p=0.011), though no difference in mortality rates (p=0.898) or length of hospitalization (p=0.449) were observed. The youngest patients were more likely to suffer injuries to head/neck and abdominal cavity, while young teenagers were more commonly shot in the extremities and older teenagers were most likely to experience thoracic trauma. Evaluation of geographic trauma locations (Figure 1), demonstrated that unintentional GSW rates increased as one approached urban centers of population.

Conclusion:  We identified demographic and geographic trends regarding pediatric unintentional GSWs. Individuals 10-14 years of age are most likely to be victims of unintentional GSWs, with handguns as likeliest injury mechanism and probability of such traumas increased in the peri-urban areas. Utilizing consequent heat maps of the catchment area, we were able to determine the most effective areas and populations to be targeted for preventative intervention.

 

10.18 4-FACTOR PCC USE FAILS TO IMPROVE OUTCOMES WHEN COMPARED TO FFP IN MILD TO MODERATE TBI

Posted on January 25, 2017

W. Readdy2, I. Farooqi2, A. Rayner2, R. Gupta1, J. Hanna1  1Rutgers-Robert Wood Johnson Medical School,Division Of Acute Care Surgery,New Brunswick, NJ, USA 2Rutgers Robert Wood Johnson Medical School,New Brunswick, NJ, USA

Introduction:
Four Factor Prothrombin Complex Concentrate (4F-PCC) has become an accepted standard of care for reversal of pharmacologic coagulopathy in the setting of traumatic brain injury (TBI). However, the benefit with regard to morbidity and  mortality remains unknown. We hypothesize that 4F-PCC use will result in improved outcomes when compared to fresh frozen plasma (FFP).

Methods:
A prospectively maintained database at a level one trauma center was queried for patients with acute mild or moderate TBI, an Injury Severity Score (ISS) < 25, with a therapeutic INR on pharmacologic anticoagulation who received either FFP or 4F-PCC between 2011 and 2016. A retrospective chart review was performed to identify admission characteristics, in-hospital interventions, and outcomes.

 

Results:

Twenty-one and thirty-three patients who met criteria were identified who  received FFP or 4F-PCC respectively. Admission characteristics including age and ISS were similar between the groups.  Furthermore, GCS, brain injury, and initial INR were similar between the groups as shown in Table 1. There was no significant difference in discharge GCS, morbidity or mortality between groups (13.17 ± 3.68 vs 13.05 ± 4.48, 42.4% vs 38.1%, 21.2% vs. 14.3%, 4F-PCC vs FFP respectively), although time to reversal was shorter (360min vs 1737min, p<0.0001). Cost for pharmacologic therapy was significantly higher in the 4F-PCC cohort ($5493 vs $313, p<0.0001).

 

Conclusion:

4F-PCC has become the preferred method for rapid reversal of coagulopathy in TBI in many centers.  However, these data suggest that despite a significant increase in cost, no clear benefit is evident with regard to discharge GCS, morbidity or mortality in mild to moderate TBI. We hypothesize that while emergent initiation of reversal is important, time to completion may not be as critical as once thought. Further study is needed to clarify the role of 4F-PCC in the management of TBI patients.

10.17 OUTCOMES OF PEDIATRIC FIREARM INJURIES AT ADULT AND PEDIATRIC LEVEL-1 TRAUMA CENTERS

Posted on January 25, 2017

E. J. Onufer1, P. M. Choi1, C. M. Courtney1, M. Wallendorf1, A. M. Vogel1, M. S. Keller1  1Washington University,General Surgery,St. Louis, MO, USA

Introduction:  Controversy exists regarding optimal trauma center qualifications for management of children injured by firearms. We sought to determine if outcome differences exist for these patients if managed at adult vs pediatric, American College of Surgeons (ACS)-verified Level-1 trauma centers.

Methods:  We conducted a retrospective review of the 2013-2014 National Trauma Databank. We included all patients aged < 18 years who were injured by a firearm and admitted to an ACS-verified Level-1 trauma center. Patients who died on arrival to the Emergency Unit or were transferred were excluded. Centers were classified as freestanding Pediatric Trauma Centers (PTC), Adult Trauma Center with Pediatric qualifications (ATC/PTC), and Adult Trauma Centers (ATC). Patients were grouped to ≤ 14 years and 15-17 years of age.

Results: 1866 children met inclusion criteria. Younger patients were treated more commonly at a PTC(Table1).  Across all centers, both age groups demonstrated demographic and injury severity differences. After controlling for these differences, children ≤14 years admitted to an ATC/PTC had a higher adjusted odds ratio of blood transfusions (OR 2.77; 95%CI 1.13-6.8); and laparotomies (OR 3.57; 95%CI 1.14-11.24) compared those admitted to a PTC.  Children 15-17 years of age, managed at either an ATC (OR 8.87; 95%CI 3.14-25.05) or ATC/PTC (OR 9.38; 95%CI 3.15-27.97), also had a greater adjusted odds ratio for laparotomy than those managed at a PTC. There were no differences in mortality, complications, computed tomography, thoracotomies, length of stay (LOS), ICU LOS, or ventilator days.

Conclusion: After accounting for demographic and injury severity differences, there were no differences in outcome variables between the PTC, ATC/PTC and ATC. These data support the management of children with firearm related injuries, typically considered an “adult pattern injury”, at PTC.

 

10.16 Is There a “Weekend Effect” in Emergency General Surgery?

Posted on January 25, 2017

D. Metcalfe1, O. A. Olufajo6, A. J. Rios-Diaz5, C. K. Zogg4, R. Chowdhury2, J. M. Havens2, A. Haider2, A. Salim2,3  6Washington University School Of Medicine,Department Of Surgery,St Louis, MO, USA 1University Of Oxford,Kadoorie Centre For Critical Care Research,Oxford, OXFORDSHIRE, United Kingdom 2Brigham And Women’s Hospital,Center For Surgery And Public Health,Boston, MA, USA 3Harvard Medical School,Boston, MA, USA 4Yale University School Of Medicine,New Haven, CT, USA 5Thomas Jefferson University Hospital,Department Of Surgery,Philadelphia, PA, USA

Introduction:  

Weekend admission is associated with increased mortality across a range of patient populations and healthcare systems. However, it is unknown whether this “weekend effect” exists in emergency general surgery (EGS). The aim of this study was to determine whether weekend admission is independently associated with serious adverse events (SAE), in-hospital mortality, or failure to rescue (FTR) in an EGS population.

 

Methods:  

An observational study using the Nationwide Inpatient Sample (NIS) 2001-2011; the largest all-payer inpatient database in the United States, which represents a 20% stratified sample of hospital admissions. The inclusion criteria were all inpatients with a primary ICD-9-CM diagnosis of acute appendicitis, abdominal cavity hernia (obstructed or strangulated), intestinal obstruction, or peritonitis. Outcomes were SAE, in-hospital mortality, and FTR (in-hospital mortality in the population of patients that developed an SAE). Logistic multivariable regression models were used to adjust for patient- (age, sex, race, payer status, Charlson comorbidity index) and hospital-level (trauma designation, hospital bed size) characteristics.

 

Results: 

There were 758,915 individual patient records (3.7 million weighted admissions). The overall rate of SAE was 10.6% (10.9% weekend, 10.5% weekday, p<0.001), in-hospital mortality 1.4% (1.4% weekend, 1.4% weekday, p=0.048), and FTR 8.7% (8.7% weekend, 8.7% weekday, p=0.819). Within logistic regression models, weekend admission was an independent risk factor for development of SAE (1.04, 1.02-1.06) but not for FTR (0.98, 0.91-1.05) or in-hospital mortality (1.01, 0.95-1.07).

 

Conclusion: 

This study did not find any evidence of increased mortality for EGS patients admitted at the weekend.

10.15 Increased Rate of C. Difficile in Patients Following TPIAT. What are the Risk Factors?

Posted on January 25, 2017

J. E. Taylor1, K. A. Morgan1  1Medical University Of South Carolina,Gastrointestinal And Laparoscopic Surgery,Charleston, Sc, USA

Introduction: Patients with chronic pancreatitis often have associated medical comorbidities prior to total pancreatectomy with islet cell autotransplantation (TPIAT).  Predicting which patients may have an increased risk of certain complications is difficult.  Long-term outcomes evaluating insulin use and quality of life have been evaluated in these patients.  Little information has been published about the postoperative complications and the potential long-term implications.

Methods: A prospectively maintained database of 169 patients who underwent TPIAT at our institution from March 2009 to April 2016 was retrospectively reviewed.  Interest was given to development of Clostridium difficile infection (CDI) following surgery and the determination of time to diagnosis after surgery.  Pre-operative risk factors were analyzed.  Charts were reviewed for postoperative complications, including pneumonia, urinary tract infection (UTI), biliary leak, portal vein thrombosis, as well as long-term mortality. 

Results: A total of 17 patients (10.1%) developed C. difficile infection following TPIAT.  A binary logistic regression analysis of pre-operative factors was performed on the patients who developed C difficile infection.  Of the factors in the analysis, history of alcohol abuse (P = 0.025), diabetes (P = 0.016), previous emergency department visits (P = 0.004), previous hospitalizations (P = 0.009), enteral nutrition requirement (P = 0.046), and daily morphine use (P = 0.043) were found to be statistically significant (P < 0.05).  Postoperative complications found to be associated with the development of C. difficile infection include greater ICU length of stay and pneumonia. 

Conclusion: Following TPIAT at our institution, there is an increased rate of C. difficile infection compared to the overall rate of development within our hospital.  Several pre-operative risk factors in the patient population have been associated with the contraction of C. difficile postoperatively.  Knowledge of these factors may aid in the development of new protocols preoperatively to help minimize the risk that these elements have on this patient population.

 

10.14 External Validation of Clinical Criteria for Obtaining Maxillofacial Computed Tomography in Trauma

Posted on January 25, 2017

A. W. Harrington1, K. Pei1, R. Assi1, K. A. Davis1  1Yale University School Of Medicine,Department Of Surgery,New Haven, CT, USA

Introduction:
Patients sustaining multisystem trauma are at risk for oral and maxillofacial fractures.  To date, no externally validated criteria is available to guide the clinician in obtaining additional cross sectional imaging to evaluate possible facial fractures.  Our aim was to externally validate the University of Wisconsin’s Criteria and to report modern practice patterns at a tertiary, academic, Level 1 trauma center.

Methods:
A retrospective case study was performed of all patients who had computed tomography of the facial bones (CT face) at a tertiary, academic, Level 1 trauma center over the 6-month period ending June 30, 2015. The electronic medical record was reviewed for the five University of Wisconsin criteria (bony step off or instability, periorbital ecchymosis, malocclusion, tooth absence, and GCS).  Final interpretation of CT face findings (facial fractures, intracranial hemorrhage, and cervical spine injury) were also captured. Our modeling was similar to that described by Sitzman, et al. Sensitivity, specificity, negative and positive predictive values with 95% confidence intervals were evaluated.  A p<0.05 was considered significant.

Results:
The presence of any one or more of the five criteria identified on physical exam resulted in an 81% sensitivity for any facial fracture which is lower than the sensitivity initially described (97%). The absence of all five physical examination criteria had a negative predictive value of 60%, again lower than that initially described (81%).

Conclusion:
We were unable to validate the University of Wisconsin criteria for predicting facial fractures. These criteria may be institutionally specific and not generalizable to other trauma centers. Further research to refine the criteria for CT of the face is needed to improve resource allocation. 
 

10.13 Influence of Pain Regimen on Pain Scores and Length of Stay on Blunt Thoracic Trauma Patients

Posted on January 25, 2017

L. Williams1, C. Talley1, D. Davenport1, K. Brumagen1, J. S. Roth1  1University Of Kentucky,Surgery,Lexington, KY, USA

Introduction: Pain control in blunt thoracic trauma (BTT) patients with multiple rib fractures is of critical importance to assure adaquate ventilation and avoid possible complications such as atelectasis, pneumonia, and subsequent respiratory compromise. The objective of our study was to compare epidural analgesia (EPID), intravenous patient-controlled analgesia (IV-PCA), & oral analgesia (PO) with outcome measures of visual analog pain scale (VAS) scores, length of stay (LOS), and complication rates.

Methods: We analyzed the trauma registry of 348 isolated non-intubated blunt thoracic trauma patients between May 2012 and December 2013 at our Level 1 trauma center. We studied those with age > 18 years, ≥3 rib fractures, a chest abbreviated injury scale (AIS) ≥3 and excluded non-chest AIS ≥4, emergent surgery, and LOS <24hrs. Patient demographics, mechanism and extent of injury, complications including intubations, VAS scores, and LOS were collected for each patient.  VAS scores were averaged in increments of 0-6, 6-12, 12-24, 24-48, and 48-72 hours from admission.  Estimated VAS scores as well as LOS were adjusted for age, gender, number of rib fractures, injury severity score (ISS), and chest-AIS.

Results: Demographics showed similarity among the groups except for a higher BMI in the IV-PCA group.  Admission VAS was higher for EPID (7.5 vs 7.1 vs 6.3) compared to IV-PCA and PO. EPID patients had more median rib fractures (8 vs 6.5 vs 5, p<.001) and more flail chests (17.2% vs 5.0% vs 4.0%, p=0.01) than IV-PCA and PO. EPID led to significantly lower pain at 24-72 hours compared to IV-PCA and PO. Median LOS is higher in both IV-PCA and EPID compared to PO (6 vs 5 vs 3 days, p<.001). Thirteen patients required intubation after admission, 4 had IV-PCA and 9 had PO. No patients requiring intubation had an EPID.

Conclusion: Significant pain control for seriously injured blunt chest trauma patients can be achieved with EPID but with increased LOS. There were no intubations after EPID was started.  IV-PCA however may increase LOS without decreasing VAS as significantly as EPID demonstrating diminished utility in the treatment of isolated blunt chest trauma.

 

10.12 INFLUENCE OF HEMOSTATIC RESUSCITATION ON THE RATE OF DAMAGE CONTROL LAPAROTOMIES

Posted on January 25, 2017

N. W. Kugler1, T. Spees1, C. Pinkertson1, J. Paul2, O. Kaslow1, E. Hanko1, T. Carver1  1Medical College Of Wisconsin,Milwaukee, WI, USA 2University Of New Mexico HSC,Albuquerque, NM, USA

Introduction:  Hemostatic resuscitation (HR) focusing on balanced ratios of blood products and minimizing crystalloids has been used to improve outcomes in patients with exsanguinating blood loss from trauma. Our institution implemented a Massive Transfusion protocol (MT) to facilitate HR and participate in prospective trials. Over five years four distinct time periods were noted; (1) Pre- HR, (2) the PROMMTT study, (3) Transition and (4) the PROPPR study. The aim of this study was to determine whether the implementation and utilization of an MT protocol and change in resuscitation patterns affected our use of Damage Control Laparotomy (DCL) in management of trauma patients with exsanguinating blood loss.

Methods:  An IRB approved retrospective chart review of adult trauma patients identified through the trauma registry and operative logs whom underwent exploratory laparotomy within the first 24-hours of their index admission between January 1, 2008 and December 31, 2013 was conducted. Demographics, mechanism of injury, and an ABC Score for Massive Transfusion was calculated. Prehospital, trauma room, and intraoperative crystalloid and blood product administration were recorded. Exploratory laparotomies were scored as primary closure or DCL under several defined categories. Four distinct time periods were defined as above by differences in resuscitation strategies. 

Results: A total of 755 patients underwent exploratory laparotomy over the study period. 27 were excluded for delayed laparotomy, 27 excluded due to intraoperative death, and 18 excluded due to missing data resulting in 683 patients. The overall rate of DCL was 17.7%. A total of 131 patients with an ABC score of two or three were identified comprising the study group: median age 31 years (IQR 24,44), majority penetrating (90.1%), ISS 13 (IQR 9,25), and AIS abdomen 3 (IQR 3,3). Analysis demonstrates no significant differences in the overall rate of DCL (35%, 40%, 40.9%, 40%) or when controlling DCL due to injury pattern (18.8%, 25%, 30.4%, 25%) for the defined time periods.

Conclusion: Changes in resuscitation through MT and maximizing hemostatic resuscitation does not appear to impact the rate of DCL, even when controlling specific injury patterns.

 

10.11 Time of Day Is Not Associated with the Performance of Damage Control Laparotomy in Trauma Patients

Posted on January 25, 2017

K. Asi1, M. George1, M. D. Swartz1, M. McNutt1, L. Moore1, C. Wade1, B. Cotton1, J. Holcomb1, J. Harvin1  1University Of Texas Health Science Center At Houston,Department Of Surgery, Division Of Acute Care Surgery,Houston, TX, USA

Introduction: Damage control laparotomy (DCL) in severely injured patients is associated with improved survival in the setting of acidosis, coagulopathy and hypothermia. The decision to leave a patient’s abdomen open after the initial laparotomy is multifactorial and can depend on resources and assistance available, which varies based on time of day even at major trauma centers. We hypothesized that fewer immediately available resources on nights, weekends and holidays would increase the rate of DCL.

 

Methods: All trauma patients from 2011-2015 who underwent emergent laparotomy were included. Emergent laparotomy in this patient population was defined as admission from the emergency department (ED) directly to the operating room (OR). Intraoperative deaths were excluded from analysis. Patients were grouped by the time of ED arrival: 1) Monday through Friday 7:00 AM to 5:00 PM and 2) Monday through Friday 5:00 PM to 7:00 AM, weekends, and institutional holidays. The groups were compared in a univariate fashion to determine differences in baseline characteristics. A purposeful multivariate logistic regression model was constructed using variables selected a priori (mechanism, Injury Severity Score [ISS], ED and OR transfusions) as well as those found to be both statistically and clinically significant on univariate analysis. Continuous variables are presented as: median with 25th and 75th inter-quartile ranges [IQR].

 

Results: From 1/1/2011-12/31/2015 there were 23,410 trauma admissions. 1,058 underwent emergent laparotomy – 687 (65%) definitive laparotomy, 325 (31%) DCLs, and 46 intraoperative deaths (4%). On weekdays 203 (19%) emergent laparotomies were performed, with 121 (60%) being definitive, 72 (35%) DCL, and 10 intraoperative deaths (5%). On nights, weekends and holidays, 855 (81%) emergent laparotomies were performed, with 566 (66%) being definitive, 253 (30%) DCL, and 36 (4%) intraoperative deaths. The table summarizes characteristics of patients surviving the initial emergent laparotomy. Controlling for age, injury mechanism, and arrival physiology, multivariate analysis demonstrated that Injury Severity Score (OR 1.04, 95% CI 1.02-1.06, p<0.001) and ED and OR blood transfusions (OR 1.23, 95% CI 1.17-1.28, p<0.001) were associated with an increased odds of DCL. Weeknight, weekend, and holiday laparotomy were not associated with DCL (OR 1.30, 95% CI 0.81-2.09, p=0.0.284).

Conclusions: Patients undergoing emergent trauma laparotomy on weeknights, weekends, and holidays are not more likely to be managed with DCL than those patients cared for during the weekdays. Weeknight, weekend, and holiday resources appear sufficient to not factor into the decision to perform damage control.

10.10 Repeat Computed Tomography After Intracranial Bleeding Is Not Needed In Patients Taking Aspirin

Posted on January 25, 2017

M. Lew1, W. Snyder1, B. Daley1  1University Of Tennessee Medical Center – Knoxville,Surgery,Knoxville, TN, USA

Introduction:
Repeat Computed Tomography of the Head (RCTH) is used to evaluate progression of traumatic brain injury (TBI) in patients on anticoagulants, even without neurologic change.  We hypothesized patients with aspirin (ASA) therapy alone do not benefit from routine RCTH.

Methods:
Adult patients who were not anticoagulated (NAC) or on ASA alone with a TBI at a Level I trauma center were studied. Patients who underwent emergent neurosurgical intervention were excluded. Initial and RCTH, demographics, and Glasgow Coma Score (GCS) were reviewed. Primary outcome was delayed neurosurgical intervention. Secondary outcomes were progression of TBI, hospital length of stay, ventilator days, discharge GCS, and disposition.

Results:
From July 2012 to June 2015, 324 NAC and 110 ASA patients were identified. Mean age was 54, 61% were male, 39% were falls, and 55% were vehicular crashes. Initial GCS (14 ASA and 12 NAC) and subsequent GCS (14 ASA and 13 NAC) were better in ASA patients. While Injury Severity Scale was different (21 ASA and 25 NAC; p = 0.02), Head Abbreviated Injury Scale was not (3.8 both ASA and NAC). Progression was similar in ASA (22%) and NAC (29%) (not signficiant). No ASA and only 6 NAC patients (1.8%) had a neurosurgical intervention after RCTH (not significant). 4 of 54 patients (7%) with a decline in GCS and 5 of 46 patients (11%) with a subsequent GCS < or = 8 had neurosurgical intervention. Mortality was similar for ASA (7%) and NAC (6%). Hospital length of stay (2.4 to 3.8; p < 0.01) and ventilator days (0.7 to 1.9; p < 0.01) were decreased in the ASA group. Both groups had a discharge GCS of 15, but ASA discharge to home was less frequent (p = 0.004).

Conclusion:
RCTH is not warranted for patients with traumatic brain injuries in patients treated with aspirin. RCTH should be performed and is useful for those with a decline in neurologic status or GCS < or = 8.
 

10.09 Trauma Patients Meeting Both CDC Definitions for VAP Had Worse Outcomes Than Those Meeting Only One.

Posted on January 25, 2017

D. Younan1, R. Griffin1, T. Swain1, B. Camins2  1University Of Alabama at Birmingham,Acute Care/ Surgery,Birmingham, Alabama, USA 2University Of Alabama at Birmingham,Infectious Diseases/ Medicine,Birmingham, Alabama, USA

Introduction:  The Centers for Disease Control and Prevention replaced its old definition for ventilator associated pneumonia (VAP) with the Ventilator associated events (VAEs) algorithm in 2013. We sought to compare the outcome of trauma patients meeting the two definitions.

Methods: Trauma patients with blunt or penetrating injuries and at least 2 ventilator support days were identified using the trauma registry from 2013-2014. Ventilator associated pneumonia (VAP) was determined by using two

Methods: (1) pneumonia as defined by the “Old”, clinically based CDC definition and (2) pneumonia as defined by the updated “New” CDC definitions of Ventilator-Associated Events. For the current analysis, the latter definition of pneumonia included infection-related ventilator associated conditions (IVACs) and possible VAPs (PVAPs). Cohen’s Kappa statistic was determined to compare the two definitions for VAP. To compare demographic and clinical outcomes the chi-square and Kruskal-Wallis were used for categorical and continuous variables, respectively. 

Results:From 2013-2014, there were 1170 trauma patients admitted for treatment who had at least 2 days of ventilator support. Of all, 144 (12.3%) met the “New” CDC definition for IVAC or PVAP, 358 (30.6%) met the “Old” definition of VAP, and 101 (8.6%) met “Both” definitions. The kappa statistic between pneumonia as currently defined under the “New” and “Old” definitions was 0.28 (95%CI 0.22-0.33). Between the definitions categories of VAP, there were no differences in age, gender, race, or injury severity score. Those meeting “Both” definitions had longer ventilator support days (p<0.0001), ICU length of stay (LOS) (p<0.0001), and hospital LOS (p<0.0001) when compared to those meeting only one definition. There was no difference in mortality for those meeting “Both” or just one definition for VAP. 

Conclusion:There was no difference in mortality between patients meeting the “Old” and “New” definitions; those who met “Both” definitions had longer ventilator support days, ICU and hospital length of stay.  
 

10.08 Maldistribution of Trauma Centers: Impact on Patient Care and Resource Utilization

Posted on January 25, 2017

P. J. Parikh1, B. Guthrie1, T. Erskine2, M. McCarthy1, P. P. Parikh1  1Wright State University,Dayton, OH, USA 2Ohio Department Of Public Safety,Emergency Medical Services,Columbus, OH, USA

Introduction:
Previous studies suggest that geographic distribution of trauma centers correlates with injury-related mortality. However, the impact of distribution of trauma centers on system performance in the state, including statewide resource utilization and patient care is unclear.  

Methods:
All trauma and emergency medical services (EMS) data for 2008-2012 were obtained from the Ohio Department of Public Safety (ODPS), which included 34,494 unique patient records. Overtriage (OT) error was defined as the proportion of patients with ISS≤15 transported to a Level 1/2 trauma center; undertriage (UT) error referred to patients ISS>15 transported to a Level 3 or non-trauma center, where OT indicates resource utilization, and UT and mortality directly impact patient care. Proportions of patients experiencing OT and UT errors, and subsequent mortality across all the homeland security regions of Ohio were compared. 

Results:
Over- and under-triage errors showed specific patterns by region (Table 1).  For instance, Regions 7 and 8 had the highest UT (4.6% and 12.9%, respectively) and lowest OT (24.4% and 3.2%, respectively) errors.  Because there are no L1/L2 trauma centers in these regions, triage patients are likely to be transferred from the scene to the nearest L3 or non-trauma facilities. Similarly, Regions 2 and 5 have the least UT errors (1.8% and 1.5%, respectively) and the highest OT errors (49.4% and 61.5%, respectively), probably because of the ease of accessible L1/L2 centers prompting even less severely-injured patients to be transferred there. Although mortality did not vary substantially between the regions, it was the highest in Region 8 (6.2%).  

Conclusion:
The location of trauma centers in the region can directly affect quality care access and resource utilization in any trauma system. A model or tool that could help the state identify the optimal distribution of trauma centers could improve equity of care among these regions while optimizing trauma resources. 
 

10.07 Vacuum-Assisted Fascial Closure– Continued Success With Decreased Time to Closure

Posted on January 25, 2017

A. N. Hildreth1, P. R. Miller1, M. C. Chang1, F. Marayati2, J. W. Meredith1  1Wake Forest University School Of Medicine,Department Of Surgery,Winston-Salem, NC, USA 2Princeton University,Princeton, NJ, USA

Introduction: Damage control laparotomy following trauma remains common, and management of the open abdomen presents substantial challenges.  In 2004, our group reported a technique of vacuum-assisted fascial closure (VAFC) after open abdomen, with a closure rate of 88% and a mean time to closure of 9.5 days. After 15 years of using this technique, we examine our recent experience for durability of the results.

Methods:   The trauma registry at our level 1 trauma center was queried over a 21-month period to identify patients who underwent laparotomy for trauma.  Patients requiring management with an open abdomen after initial laparotomy were selected, and chart review was performed to determine fascial closure rate, time to closure, and related complications. 

Results:  From January 2013 to March 2016, 336 patients underwent laparotomy for trauma; 96 (28.5%) of these required management of the open abdomen.  Mean injury severity score for this group was 29; 55 (57.2%) survived until abdominal closure or Vicryl mesh placement.  VAFC was used in all patients to attempt closure.  Closure rate was 89% (49), with mean time to closure of 4.7 days (range 1-19 days).  Mean number of laparotomies prior to closure was 1.3.  Complications in closure group included four ventral hernias, two enterocutaneous fistulae, and three episodes of fascial dehiscence. 

Conclusion:  VAFC continues to result in high rates of successful abdominal closure.  Since the inception of this technique at our facility, the majority of patients with an open abdomen have successful closure, with consistent rates over a period of 15 years. Although closure after several weeks remains possible, time to closure is considerably decreased with increased experience.

 

 

10.06 Temporal Variations in Pediatric Trauma: Rationale for Altered Resource Utilization

Posted on January 25, 2017

E. Groh1, P. Feingold2, B. Hashimoto3, L. McDuffie1, T. Markel1  1Indiana University School Of Medicine,Surgery,Indianapolis, IN, USA 2National Cancer Institute,Bethesda, MD, USA 3American University Of Sharjah,Sharjah, , United Arab Emirates

Introduction: An understanding of temporal variations in pediatric trauma may have significant implications for appropriate allocation of hospital resources and optimal timing of community prevention and education programs. The goal of this study was to characterize the temporal trends in pediatric trauma activations from a Pediatric Trauma Center. 

Methods: Prospectively collected data from all trauma patients who were evaluated at an ACS-certified Level 1 Pediatric Trauma Center between the years 2011-2015 were retrospectively reviewed after IRB approval was obtained. Demographics, date and time of injury, type of injury (blunt vs. penetrating), and post emergency department disposition were reviewed. All patients requiring admission between 2011-2015 were further analyzed to assess temporal trends. A database consisting of all consecutive hours during the entire study period was constructed including total trauma counts per-hour. To assess temporal trends, heat maps were constructed and a Poisson regression model was used to assess statistical significance.  A p value less than 0.05 was significant.

Results:The average number of trauma admissions per year was relatively consistent throughout the 5-year analysis period. Trauma admissions from both blunt and penetrating injuries occured at a higher rate in months with warm weather as compared to months with cold weather. Analysis of average trauma counts across the 168 hours in each week revealed that there were specific time points, most commonly between 6pm and midnight, where trauma admissions occur at a statistically higher frequency (Figure 1). This held true as a trend for blunt-trauma specific admissions. Analysis of penetrating trauma admissions revealed no statistically significant time point, likely related to the overall low number of subjects in this category.    

Conclusion:Temporal variations in trauma do occur with most trauma admissions occurring more frequently in the summer months. Trauma admissions occur more frequently between the hours of 6pm and midnight. These data can provide useful information for hospital resource utilization. The emergency department, operating rooms and intensive care units should be prepared for increased trauma related volume during these times. Additionally, community prevention and education programs for children should be increased during summer months of the year when trauma is most prevalent. 

 

« Previous Page
Next Page »