55.18 New York City’s Vision Zero Action Plan Reduces Traffic Related Pedestrian Injuries

Posted on January 18, 2018

D. K. Donley1, R. Policherla1, K. Smith1, A. Kelly1, J. Shou1, P. Barie1, R. Winchell1, M. Narayan1  1Weill Cornell Medical College,Trauma, Burns, Critical And Acute Care,New York, NY, USA

Introduction:

Vision Zero (VZ) is a public health approach to road safety. The model proposes that responsibility in the event of injury lies with both the user and the designer, and that better designed traffic systems are the best way to maximize a community’s mobility and safety.  We hypothesize a reduction in traffic related injuries and deaths in the years following the implementation of a VZ Action Plan for New York City.  

Methods:  

A review of the accident and injury statistics from Traffic Accident Management System maintained by the New York City Police Department was performed.  Data were analyzed from 2009 to 2016, the last year for which comprehensive information is available.  The mean number of fatalities and injuries for the preVZ and postVZ periods were calculated.  Means were compared using a two sided t-test.  

Results:

Mean fatality and injury data for the periods under review are shown in the table.  There was a statistically significant (p<.05) reduction in pedestrians injured.  There were no significant reductions in fatalities.    

Conclusion:

Traffic related injuries have fallen with the enactment of the VZ Action Plan in NYC.  This review shows that injuries to pedestrians have been reduced with the VZ Action Plan, but highlight that more work is needed.  Further studies regarding specific traffic design interventions (pedestrian islands, bicycle lanes, etc) are needed to determine relative efficacies of different design elements.  

55.15 Breaking Bounce-backs: Number of Pre-Injury Hospital Admissions Predicts Readmission after Trauma

Posted on January 18, 2018

Z. G. Hashmi1,2, C. K. Zogg3, M. P. Jarman1, A. Salim1, A. H. Haider1  1Brigham And Women’s Hospital,Center For Surgery And Public Health,Boston, MA, USA 2Sinai Hospital Of Baltimore,Surgery,Baltimore, MD, USA 3Yale University School Of Medicine,New Haven, CT, USA

Introduction:  Up to 10% of trauma patients are readmitted after their index hospitalization leading to worse outcomes and increased costs. Understanding patient factors that increase the risk of unplanned readmissions will enable us to develop interventions to mitigate this problem of “bounce-backs”. The objective of this study is to determine if unplanned hospital admissions in the 90-days before the index trauma are associated with increased risk of 30-day unplanned readmissions after trauma.

Methods:  Adult patients (age ≥16y) with blunt/penetrating injuries included in the Healthcare Cost and Utilization Project (HCUP) Nationwide Readmissions Database (NRD) 2014 were analyzed. All unplanned admissions in the 90-days before a trauma admission (pre-trauma admissions, PTAs) and within 30-day after a trauma discharge (trauma readmissions, TRs) were noted. Weighted multivariable logistic regression, with TR as the primary outcome, was performed, adjusting for age, sex, insurance status, injury mechanism, Injury Severity Score, head Abbreviated Injury Scale and All Patient Refined-Diagnosis Related Group severity, to ascertain the association of PTAs with TRs. We also matched patients with and without PTAs using Coarsened Exact Matching for sensitivity analyses. 

Results: A total of 235,978 patients were analyzed. The overall PTA rate was 3.2%(7,646/235,978) and the TR rate was 4.6%(10,924/235,978). Among those readmitted, 10.5%(1,145/10,924) had ≥1 PTA. These patients were older(≥65y: 71.8% vs 68.2%) and sustained more falls(89.1% vs 76.5%) compared to readmitted patients without PTAs(p<0.01 for both). Most PTA patients(81.6%) had a medical hospitalization and were subsequently discharged to a skilled nursing facility(62%) before their trauma visit. Unadjusted TR rate was higher for patients who had a PTA versus those who did not(15% versus 4.3%, p<0.001). Even after adjusting for multiple patient factors, both the presence of a PTA [2.9 (2.7-3.2); OR (95% CI)] and the number of PTAs(Fig1) significantly increased the risk of TR. Sensitivity analyses demonstrated similar findings.

Conclusion: The number of unplanned hospital admissions 90-days prior to a trauma admission is highly associated with 30-day unplanned readmission after trauma. Asking patients/caregivers regarding admissions in the last 90 days can provide a simple way to identify patients who may benefit from interventions to reduce the risk for readmission. These may include ensuring pre-discharge medical optimization, appropriate post-discharge care coordination and fall prevention. This study highlights a target population for interventions to break the self-perpetuating “bounce-back” cycle among frequently admitted trauma patients. 

55.16 CT Completion Times in a MCI: Is There a Difference Between Pre-Registered EMR and Downtime Protocol

Posted on January 18, 2018

R. D. Rampp1, M. Sammer1, S. D. Bhattacharya1  1University Of Tennessee College Of Medicine,Department Of Surgery,Chattanooga, TN, USA

Introduction: Mass Casualty Incidents (MCI) are potentially overwhelming events that can stress the capabilities of even the most prepared facilities. While electronic medical records and computerized physician order entry has significantly improved patient safety and workflow, these tools may also fail to function as intended during an MCI. During a recent pediatric MCI at our hospital, the overflow of patients forced children to be evaluated in our adult and pediatric emergency departments to maximize resources. These two ERs function under separate EMR systems and trauma protocols allowing an opportunity to study this event. At the time of our incident, our adult ER had an electronic, trauma-specific order set that has most commonly used x-rays and CT scans that may be ordered as well as a previously unassigned list of “John Doe” patients that can absorb a bolus of patients. Our Children’s ER functioned without preset trauma orders and every child must be manually entered into the computer system by the ER registrar before orders can be placed or imaging completed.

Methods: After IRB approval, data was collected retrospectively in the weeks following this MCI using patient medical record numbers to collect imaging order times, times images were obtained, and times images were read by a radiologist.

Results: Using Independent Samples t Test, we found that the mean time from imaging ordered to imaging completed differed by 133.74 minutes (p<0.002) and the time from image ordered to image read differed by 268.46 minutes (p<0.002) between the two systems.

Conclusion: EMR preparedness with pre-registered patients prepared for CPOE will significantly improve time to imaging completion and image read times. These improvements will allow for more expeditious triage of a large volume of patients in an MCI event. Continued planning and forethought for mass casualty events will hopefully improve outcomes as these events become more prevalent in non-urban settings across the nation.

 

55.12 Elevated sST2 Levels are Associated with Adverse Outcomes and Mortality in Blunt Trauma Patients

Posted on January 18, 2018

I. Billiar1, J. Guardado1, J. Brown1, O. Abdul-Malak1, Y. Vodovotz1, T. R. Billiar1, R. A. Namas1  1University Of Pittsburgh,Pittsburgh, PA, USA

Introduction: Soluble suppression of tumorigenicity 2 (sST2), a decoy receptor for interleukin (IL)-33, has emerged as a novel biomarker in various disease processes. Recent studies have elucidated on the role of IL-33/sST2 complex in modulating the balance of Th1/Th2 immune response following tissue stress. However, the role of sST2 as a biomarker following traumatic injury remains unclear. To address this, we sought to evaluate serum sST2 correlations with in-hospital outcomes and mortality as endpoints in blunt trauma patients. 

Methods: We retrospectively analyzed clinical and biobank data of 493 blunt trauma victims (472 survivors [mean age: 48±0.9; injury severity score (ISS): 20±0.5] and 19 non-survivors [mean age: 59±4.5; ISS: 23±2]) admitted to the intensive care unit (ICU). Given the confounding impact of age on the inflammatory response, we derived a matched survivor sub-group (n=19; mean age: 60±3; ISS: 23±2) using an IBM SPSS® case-control matching algorithm. Serial blood samples were obtained from all patients (3 samples within the first 24 h and then from day (D) 1 to D7 post-injury). Thirty inflammatory biomarkers were assayed using Luminex™. Two-Way Analysis of Variance was used to compare groups (P<0.05). Spearman rank correlation was performed to determine the association of circulating sST2 levels with in-hospital outcomes and biomarker levels.

Results: Circulating sST2 levels of the non-survivor cohort were statistically significantly elevated at 12 h post-injury and remained elevated up to D7 when compared to the parent survivor cohort. Admission sST2 levels obtained from the first blood draw within 12 h post-injury in the parent survivor cohort correlated weakly but positively with admission base deficit (correlation coefficient [cc]=0.1; P=0.029), the shock index (heart rate/systolic blood pressure; cc=0.2; P=0.0006), creatinine (cc=0.1; P=0.02), INR (cc=0.1, P=0.03), ISS (cc=0.1, P=0.008), and the average Marshall multiple organ dysfunction score between D1 to D7 (cc=0.1, P=0.04). Analysis of biomarker correlations in the matched survivor group showed that sST2 correlates strongly and positively with the type 2 cytokines IL-4 (cc=0.6, P=0.002), IL-5 (cc=0.5; P=0.01), and IL-13 (cc=0.4, P=0.02), as well as IL-21 (cc=0.5; P=0.02), IL-2 (cc=0.5, P=0.02), sIL-2Rα (cc=0.5, P=0.02), and IL-17 (cc=0.5, P=0.03). Interestingly, the sST2 levels in the non-survivor group had negative correlations with MIG (cc= -0.6; P=0.02) and IL-10 (cc= -0.5; P=0.02) with no positive correlations with the other biomarkers.

Conclusion: Elevations in serum sST2 levels are associated with poor clinical trajectories and mortality following blunt trauma. The lack of positive correlations of sST2 with other mediators in non-survivors (unlike the survivors) suggests that the host response becomes dysregulated early in patients that go onto to die.  Thus, sST2 could serve as an early prognostic biomarker in trauma patients.

 

55.13 Predictors of Reintubation After Self-extubation in the Intensive Care Unit.

Posted on January 18, 2018

L. Marcia1, Z. Ashman2, E. Howell2, D. Kim1,2, D. Plurad1,2  1David Geffen School Of Medicine, University Of California At Los Angeles,Los Angeles, CA, USA 2Harbor-UCLA Medical Center,Surgery,Torrance, CALIFORNIA, USA

Introduction: Patient self-extubation can occur in the intensive care unit (ICU). Determining which patients will require reintubation can be challenging. Identification of risk factors for reintubation may facilitate the decision to reintubate.

 

Objective: Identify predictors for reintubation after patient self-extubation.

 

Methods: Single institution retrospective analysis of all adult ICU self-extubation events over a two-year period was performed. Patients requiring reintubation within 24 hours were compared to patients that did not using t-test, chi-square test, and bivariate analysis.

 

Results: Eighty-three patients self-extubated. Sixty-six percent were male, thirty-four percent were female, and the mean age was 54.5 years. Twenty-eight of the 83 patients (34%) required reintubation within 24 hours. Prior to self-extubation, reintubated patients had on average longer hospital length of stay (7.79 days vs 4.65 days, p = 0.0217), longer ICU length of stay (4.86 days vs 3.19 days, p = 0.0295), and more time on the ventilator (4.86 days vs 3.23 days, p = 0.0454). This group required higher levels of FIO2 (49.6% vs 42.2%, p = 0.0116) and breathed at a higher respiratory rate (18.5 vs 16.5, p = 0.0370) prior to self-extubation. No statistically significant differences in other ventilatory parameters were noted. Weaning parameters were obtained in 20 patients (24%), of whom six required reintubation (30%, p = 0.685).  There were no statistically significant differences in weaning parameters between the two groups.

 

Conclusion: It is generally safe to observe patients after a self-extubation episode as most do not require reintubation. Prior to self-extubation, reintubated patients had increased hospital and ICU lengths of stay, and prolonged ventilatory dependence. The reintubation group required higher levels of FIO2 and demonstrated increased respiratory rates prior to self-extubation. There was no association between weaning parameters and later reintubation. Additionally, only 24% of patients were on weaning trials although most patients did not require reintubation. Further study is needed to identify patients who may be appropriate for extubation despite failure of a spontaneous breathing trial or to expand criteria for weaning protocols.

 

 

55.14 Clinical Characteristics of Trauma Centers in the United States

Posted on January 18, 2018

G. Eckenrode1, E. J. Kaufman1, D. N. Holena2, C. C. Branas3, M. Narayan1, R. Winchell1  1Weill Cornell Medical College,Surgery,New York, NY, USA 2University Of Pennsylvania,Department Of Surgery,Philadelphia, PA, USA 3Columbia University,Mailman School Of Public Health,New York, NY, USA

Introduction:
Trauma center (TC) care has been shown to reduce mortality after injury by approximately 25%. The American College of Surgeons Committee on Trauma establishes criteria for TC designation, but little is known about differences among TCs. We hypothesized that the injured population would differ among centers, yielding fields of emphasis, and perhaps expertise, among U.S. TCs.

Methods:
We used 2014 data from the National Trauma Data Bank to analyze level I and II TCs. TCs were excluded if they had < 100 adult patients. For each TC, we calculated the proportion of adult patients falling into each of 10 categories of mechanism (penetrating injury, motor vehicle crash, pedestrian and bicyclist crash), injury type (splenic injury, blunt multisystem injury, fractures requiring fixation, traumatic brain injury, severe injury), and population (geriatric trauma and geriatric hip fractures (age ≥ 75)). These categories were drawn from the areas of interest determined by the Trauma Quality Improvement Program. We included an 11th variable for total number of traumas. TCs were classified as “high” in an area of interest if the proportion of that center’s patients falling into the area in question was in the top quartile of all TCs. We performed a cluster analysis to identify the relationship among these features. We computed a matrix of Jaccard measures of similarity between these variables. An average linkage function then clustered these variables and yielded a dendrogram of the results. The length of the bars shown on the dendrogram x-axis represents the degree of association among characteristics, with lower values indicating greater similarity among variables.

Results:
The 653 included TCs varied widely in the proportion of their patients falling into each area of interest. Median total traumas were 1,033 (interquartile range  535-1606). Median percent severe trauma (ISS ≥ 25) was 4.4% (2.4-6.6%). Median percent penetrating was 4.9% (3.1-7.6%). Median percent geriatric was 24.3% (16.2-32.5%). We observed several clusters of trauma center characteristics, as shown in the Figure. Centers with high proportions of geriatric trauma also tended to fall in the top quartile for geriatric hip fractures. Centers high in motor vehicle trauma tended to have high proportions of severe injuries, splenic injuries, and blunt multisystem injuries. Centers with high proportions of penetrating trauma also had high proportions of bicycle and pedestrian injuries.

Conclusion:
We identified clustering patterns amongst trauma center characteristics that suggest there are several distinct types of trauma centers in the United States. Further investigation is needed to identify variation in outcomes amongst trauma center types.
 

55.10 Feasibility of Fluoroscopy-Free Endovascular Navigation in Trauma Patients of Different Ages

Posted on January 18, 2018

B. Marmie1, C. Sanderfer1, J. Fuchs1, A. Kamenskiy1, P. Aylward1, M. Tommeraasen1, J. MacTaggart1  1University Of Nebraska Medical Center,Surgery,Omaha, NE, USA

Introduction: Fluoroscopy-free endovascular device navigation is receiving increased attention due to its potential to control noncompressible hemorrhage in pre-hospital settings. Many strategies have been proposed to calculate catheter and wire lengths to reach specific aortic zones without the aid of fluoroscopy. It remains unclear whether certain anatomical characteristics that are particularly prevalent in older subjects, such as wide bifurcation angles or vessel tortuosity, will prevent safe fluoroscopy-free device navigation. Our goal was to test the ability to blindly navigate guidewires to aortic occlusion zones through simulated patient-specific aortic anatomies.

Methods: A total of n=86 trauma patient CTAs (5-93 years old, average age 53±2 years) were used to build 3D models of the aorta and its branches using Mimics software. The models were exported into a Mentice VIST G5 simulator using the Case-IT capability. The physical guidewire was represented in the simulator as either a 0.035” J-curve guidewire [JCW] or a 0.035” 35° angle tip hydrophilic guidewire [ATHW]. An electric automatic wire-feeding mechanism was used to advance a physical guidewire at a constant rate into the simulator for a total of six trials performed in each anatomy with each guidewire for a total of 1442 guidewire passages. Final locations of the guidewire tip were recorded and percentage of unsuccessful attempts to advance the guidewire from the femoral access site to the aortic target zone was calculated.

Results: Overall 88% of simulations ended with the guidewire in the target aortic zone. Overall frequency of misplacement increased with age for both wires (p=0.04 for JCW and p=0.04 for ATHW respectively, see Figure). In subjects <50 years there were no differences in misplacements between the wires (p=0.21), but in subjects >50 years, ATHW had statistically more misplacements than the JCW (p=0.04). The most common misplacement locations were ipsilateral internal iliac and contralateral common iliac arteries. No misplacements were observed in 20-29-year-old anatomies for either of the guidewires.

Discussion: Aging is associated with increased guidewire misplacement during simulated fluoroscopy-free endovascular navigation. With anatomies older than age 50, misplacements become more common, particularly when using the ATHW wire, likely due to increased vessel tortuosity and widening of the aortoiliac bifurcation with age. Though results need to be confirmed in vivo, fluoroscopy-free endovascular navigation may be feasible in subjects younger than 50 years (<5% misplacements), but the risk of aberrant vessel catheterization more than doubles in older subjects. This risk might be reduced through device design and proper device selection.

55.11 Limitations in Aggressive Management of Gunshot Wounds to the Brain

Posted on January 18, 2018

L. A. Robinson1, P. Arnold3, T. J. McDonald2, S. Berry1, A. Bennett1, J. Howard1, J. L. Green1, R. D. Winfield1  1University Of Kansas Medical Center,Department Of Surgery,Kansas City, KS, USA 2University Of Kansas Health System,Trauma Services,Kansas City, KS, USA 3University Of Kansas Medical Center,Department Of Neurosurgery,Kansas City, KS, USA

Introduction:  Gunshot wounds to the brain (GSWB) carry high lethality and uncertain recovery, but aggressive resuscitation has been associated with increased survival and organ donation. Evidence-based recommendations guide resuscitative thoracotomy in trauma victims; a similar approach is needed in GSWB to optimize resource utilization. An immediate question is whether patients with GSWB undergoing cardiopulmonary resuscitation (CPR) during trauma resuscitation recover or become organ donors. We hypothesized that patients with GSWB undergoing CPR during trauma resuscitation would not survive to discharge or organ donation.

Methods:  We performed a retrospective review of patients with traumatic brain injury (TBI) from June 1, 2011 to May 31, 2016 at our level 1 trauma center. Analysis focused on patients with isolated GSWB who received CPR in the field, at a referring hospital, or the ED, with the primary endpoint of survival and secondary endpoint of organ donation.

Results: There were 74 patients with GSWB. Thirteen patients received CPR during initial resuscitation, with 100% mortality. Four arrived to the ED with signs of life, two of these patients and two with no signs of life had return of spontaneous circulation and survived to admission. There were no organ donors among these patients. The 100% mortality and 0% donation rate among  GSWB patients undergoing CPR during initial resuscitation was significantly worse than patients with GSWB not undergoing CPR (23% mortality, 45% donation, both p<0.001) or patients with isolated blunt TBI undergoing CPR during initial resuscitation (n=13, 13% mortality, p<0.001, 0% donation, NS) during the same time period.

Conclusion: Patients with GSWB who require CPR during initial resuscitation do not survive to discharge or organ donation. Aggressive resuscitation of these patients is futile and wastes scarce resources. Based on our data and existing evidence, we propose a novel algorithm to guide the management of patients presenting with GSWB. (Figure 1).
 

55.08 Validation Of The Surgical Apgar Score In Trauma Patient Undergoing Emergent Exploratory Laparotomy

Posted on January 18, 2018

A. Masi1, M. Choudhary1, K. Chao1, K. Barrera1, L. Dresner1, M. Muthukumar1, R. Gruessner1, V. Roudnitsky1  1Kings County Hospital Center,Surgery/Trauma,Brooklyn, NY, USA

Introduction:

The Surgical Apgar Score (SAS) has been shown to correlate with postoperative morbidity and mortality in patients undergoing elective general surgery, but has never been validated in a Trauma patient population.

Methods:

We retrospectively collected data on demographics, medical history, type of surgery, and post-operative outcomes for any patient undergoing emergent laparotomy due to trauma during the period Jan 2014-Aug 2016 at our Level 1 trauma center. We categorized patients in 3 groups according to their SAS, a 10-point scoring system calculated using limited intra-operative data (blood losses, lowest mean arterial pressure, lowest heart rate). Differences between SAS groups were evaluated with Pearson’s χ2 and ANOVA as appropriate. The study primary end-points were overall morbidity ( post-operative complication according to Clavien’s classification) and 30-day mortality, the secondary end-points were ICU Length of Stay (LOS) and Hospital LOS.

Results:

During the study period 177 patients underwent emergent laparotomies due to trauma. After exclusion of patients with intra-operative mortality and/or missing variables, 160 were available for analysis (SAS 0-3: n=27; SAS 4-6: n=91; SAS 7-9: n=42). The three groups were similar in regards to demographic, functional status and underlying co-morbidities (including coronary artery disease, COPD, diabetes, chronic renal failure). Patients in the lower SAS groups had higher ASA score (p <0.002) and higher Injury Severity Score ( p <0.001). Low SAS scores were associated with significant post-operative morbidity( 77.8 %vs 57.1% vs 45.2 %, p < 0.028 ) and 30-day mortality (29.6% vs 4.4 %vs 2.4%, p <0.001), a prolonged ICU length of stay ( p= 0.001) and hospital LOS (p <0.001).

Conclusion:

The SAS is easily calculated from three routinely available intra-operative measurements. The SAS correlates with fixed pre-operative risk (acute conditions, pre-existing comorbidities, ISS, and operative complexity) and It allows real time assessment of patient morbidity and mortality and may help for better triage of patient to ICU vs PACU setting.

55.09 Rapid Ground Transport Of Trauma Patients A Moderate Distance From Trauma Center Improves Survival

Posted on January 18, 2018

B. N. Taylor1, N. Rasnake1, K. McNutt1, B. J. Daley1  1University Of Tennessee Medical Center,Surgical Critical Care,Knoxville, TENNESSEE, USA

Introduction: There is debate within the EMS community over the value of calling a helicopter for trauma patients within a moderate distance/< 45 minutes, from a trauma center. Helicopter EMS (HEMS) generally have a wider scope and more advanced training than the ground EMS (GEMS). GEMS, on the other hand, have the benefit of being able to immediately initiate rapid ground transport to the center without having to wait for the HEMS to fly to the scene, land, and assume patient care. 

Methods:  We retrospectively analyzed patients brought to a level I trauma center that were admitted with blunt traumatic injuries between 2010 and 2015 in the Trauma Quality Improvement Program (TQIP) database. Two analyses were performed, one in which the patient’s reported initial scene vitals met criteria for step one of the CDC’s 2011 National Field Triage Guidelines (NFTG), and the other in which the patient’s initial scene vitals met those same guidelines and/or had a pulse greater than 110 beats per minute. Patients were categorized on scene to ED transport mode, either HEMS or GEMS. Inclusion criteria were a HEMS response time to the scene that was between 15 and 45 minutes with a transport time from the scene to the ED that was between 10 and 35 minutes, or a GEMS transport time from the scene to the ED that was between 15 and 45 minutes. Statistical significance (p< 0.05) was established through logit regression. Mortality rates were then calculated within each transport mode based population.

Results: 400 subjects were included in the analysis of patients meeting the first step of the NFTG, with 212 HEMS patients and 188 in the GEMS group. HEMS had a higher mortality rate at 0.184 and GEMS at 0.101 which was statistically significant (p=0.019).  When 606 subjects meeting the first step of the NFTG or with a pulse greater than 110 beats per minute were analyzed, the results were statistically significant again (p=0.000), with the HEMS category having a higher mortality rate at 0.154 and the GEMS category having a lower mortality at 0.056.

Conclusion: This data demonstrates that scene to ED time is paramount and rapid ground transport should be used in blunt trauma patients when the scene is a up to a moderate ground distance away from the trauma center and there is a moderate to prolonged HEMS response time. In both analyses, hemodynamically unstable trauma patients had a lower rate of mortality following ground transport. We recognize that there may be a subset of patients at these distances who could benefit from HEMS response, particularly if the flight crew can offer more advanced and specialized techniques, however every effort should be made to minimize the scene to ED time, and HEMS response, scene, and transport time must be considered. This study only analyzed the patients within a moderate distance of the trauma center, and at longer distances or in different environments; HEMS transport may indeed minimize the scene to ED time.

 

55.06 Variability of Fluid Administration During Exploratory Laparotomy for Abdominal Trauma

Posted on January 18, 2018

J. E. Baker1, G. Katsaros1, G. E. Martin1, C. Wakefield1, A. T. Makley1, M. D. Goodman1  1University Of Cincinnati,Surgery,Cincinnati, OH, USA

Introduction: Perioperative fluid management during surgery for trauma remains a debated issue as the exact quantity and type of fluid replacement is not well established. Several studies have observed a decrease in postoperative complications following major abdominal surgery when a limited perioperative fluid regimen is utilized; however, few studies have investigated outcomes resulting from the type of crystalloid used during trauma laparotomy. In this analysis, we sought to identify 1) which crystalloids were utilized most commonly during trauma laparotomy, 2) whether transfusion was associated with the amount and of crystalloid used, and 3) whether crystalloid use affected markers of resuscitation.

Methods: A retrospective review of 504 patients who underwent laparotomy for abdominal trauma between 2014-2016 was performed. Medical records were reviewed to establish estimated blood loss (EBL); postoperative laboratory values; anesthesia provider present; and volume and type of crystalloid and blood products infused during trauma laparotomy. Correlation analysis was performed to investigate relationships between the amount of crystalloid infused and EBL as well as between the amount of crystalloid received and the number of blood products transfused.  The relationship between the amount and type of each crystalloid received and post-operative markers of resuscitation was investigated. 

Results: Lactated Ringer’s was the most commonly used crystalloid, followed by normal saline and normosol (50.7% vs. 34.7% vs. 14.6% of all crystalloid infused, respectively). During all laparotomies, 1.9% of patients received no crystalloid, 38.5% received one type of crystalloid, 47.9% received 2 different types of crystalloid, and 11.6% received 3 different types of crystalloid. A total of 50.3% of patients received one or more units of packed red blood cells (pRBCs) during laparotomy. A moderate correlation was identified between the amount of crystalloid received and time spent in the OR, and was similar amongst anesthesia providers (resident, r=0.65; CRNA, r=0.62). No correlation was observed between the amount of crystalloid received and EBL (r=0.09). There was no correlation between the amount of crystalloid and the number of pRBCs (r=0.02); this remained true for those patients who underwent massive transfusion. For those who received more than 5 units of pRBCs, there was a moderate correlation between the number of different crystalloids received and the volume of crystalloid received (r=0.695). There was no correlation between any type of crystalloid received and postoperative sodium, lactate, or pH.

Conclusion: A wide variation was observed in the amounts and types of crystalloids administered intraoperatively for patients undergoing laparotomy for trauma. Interestingly, the amount of crystalloid infused only correlated with time in the OR rather than EBL. Furthermore, the type of crystalloid received did not influence postoperative markers of resuscitation. 

 

55.07 Seniority is not a Privilege: Analysis of Triage & Access to Trauma Centers for patients ≥65 years

Posted on January 18, 2018

T. Uribe Leitz1, D. J. Sturgeon1, A. F. Harlow1, M. P. Jarman1, S. Lipsitz1, Z. Cooper1, A. Salim1, A. H. Haider1  1Brigham And Women’s Hospital,Center For Surgery And Public Health,Boston, MA, USA

Introduction: Trauma is the fastest growing cause of death and disability among individuals ≥65 years in the United States (US) leading to nearly 1 million hospitalizations and 54,000 deaths in 2013. Regional studies suggest older patients may not be fully benefitting from the US’s advanced trauma care system and Trauma Centers (TC) as many older trauma patients are under-triaged to Non-Trauma Centers (NTC). Our objective was to determine if older patients are under-triaged to NTCs at a national level and identify any regions where this is more likely to occur. 

Methods: We identified trauma patients that were transported by ambulance from a residential address, skilled nurse facility, or accident site to a hospital in Medicare 2008-2014 data (patients aged ≥65 years). We merged this with data from the American Trauma Society and the American Hospital Association to identify level of TC that these patients were transported to. TC care was defined as care received at a Level I or Level II TC and the rest were classified as NTC. Under-triage was defined as patients with an injury severity score (ISS) ≥16 treated at a NTC, per the ACS Committee on Trauma recommendations. We performed multivariable logistic regression with TC status as the outcome variable adjusting for sex, race, age, ISS, functional status and miles traveled.

Results:A total of 1,162,960 patients ≥65 years of age were transported and admitted to a hospital during our study period. Thirty four percent (399,933/1,162,960) were treated at a TC. Mean age was 83.3 (SD 7.9), more than 69% were female, and the majority were white (>90%). Patients traveled a median of 5 miles to receive care. Seventy two percent of the injuries were due to falls. Of the patients that received care at a TC 31.1% (124,419/399,933) were from the Midwest. Of the patients that received care at a NTC 44.0% (335,504/763,027) were form the South. Multivariable analysis showed that patients in the Midwest, West, and South had 1.1 (95% CI, 1.0- 1.1), 1.6 (95% CI, 1.5 – 1.7), and 2.2 (95% CI ,2.1- 2.2) higher odds of being under-triaged, respectively, when compared to the Northeast after adjusting for potential confounders. Fifty-nine percent (55,778/93,878) of patients with an ISS of ≥16 were treated at a TC and 41% (38,100/93,878) at a non-TC.

Conclusion: At at a national level, nearly 40% of severely injured trauma patients ≥65 are under triaged to a NTC. This problem appears to be particularly worse in the Southern US, despite the fact that most field triage guidelines currently call for expedited transport of older trauma patients to TC. Tools to improve field triage and enhance trauma center access for older patients are urgently needed.

 

55.05 Rotational Thromboelastometry (ROTEM) Thresholds for Blood Component Therapy in Injured Patients

Posted on January 18, 2018

G. R. Stettler1, H. B. Moore1, G. R. Nunns1, J. Chandler2, A. Arsen Ghasabyan2, E. Peltz1, M. J. Cohen2, C. C. Silliman1, A. Banerjee1, A. Sauaia1, E. E. Moore2  1University Of Colorado,Surgery,Aurora, CO, USA 2Denver Health Medical Center,Surgery,Denver, CO, USA

Introduction: Goal directed hemostatic resuscitation based on thrombelastography (TEG) has a survival benefit compared to conventional coagulation assays. While TEG transfusion thresholds for patients at risk for massive transfusion (MT) have been defined, similar cutoffs do not exist for the other commonly used viscoelastic assay; ie, rotational thrombelastometry (ROTEM).  The purpose of this study was to develop ROTEM blood product thresholds in patients at risk for MT.

Methods: ROTEM was assessed in trauma activation patients admitted from 2010 to 2016 (n=222).  Receiver operating characteristics curve (ROC) analysis was performed to test the predictive performance of ROTEM measurements for blood products in patients requiring MT defined as >10 units of RBCs or death in the first 6 hours. The Youden Index defined optimal thresholds for ROTEM-based resuscitation for each aspect of clot formation and breakdown.  Transfusions were guided by TEG and physicians were blinded to ROTEM results.

Results:As anticipated, patients who received a MT (n=37) were sicker with more abnormal physiologic and laboratory values (Table).  Prolonged clot initiation is an indication for plasma and reflected by EXTEM CT.  EXTEM CT was longer in patients with MT compared to those without (87 vs 64 seconds, p<0.0001).  Abnormal dynamics of clot formation is an indication for fibrinogen products and reflected by EXTEM Angle.  EXTEM Angle was shallower in MT patients compared to those that did not (54 vs 69 degrees, p<0.0001).  Low clot strength is an indication for platelets and reflected by EXTEM CA10.  CA10 was less in MT compared to non-MT patients (30.5 vs 50 mm, p<0.0001).  Increased fibrinolysis is an indication for antifibrinolytics and reflected by EXTEM CLI60.  CLI60 was lower in patients that had MT than those that did not (47 vs 94 percent, p=0.0006).  EXTEM CT yielded an area under the ROC curve (AUROC) = 0.7116, and a cut point of >78.5 sec.  EXTEM angle had an AUROC = 0.865, and a cut point of <64.5 degrees.  EXTEM CA10 had an AUROC = 0.858, with a cut point of <40.5 mm.  CLI60 had an AUROC = 0.6788 with a cut point at < 74%.

Conclusion:We have identified ROTEM thresholds to guide hemostatic resuscitation during MT in trauma, and propose plasma transfusion for EXTEM CT >78.5 seconds, fibrinogen products for angle <64.5 degrees, platelet transfusion for CA10 <40.5 mm, and antifibrinolytics for CLI60 <74%.

 

55.03 Considerations for Rib Plating versus Conservative Treatment of Rib Fractures

Posted on January 18, 2018

I. Puente1,2,4, E. Picard1, J. Wycech2, G. DiPasquale2,3, R. Weisz1,2, A. Fokin2  4Broward Health Medical Center,Trauma,Fort Lauderdale, FL, USA 1Florida Atlantic University,College Of Medicine,Boca Raton, FL, USA 2Delray Medical Center,Trauma,Delray Beach, FL, USA 3Health Care District Palm Beach County,Trauma,Palm Springs, FL, USA

Introduction: Surgical fixation of rib fractures (SSRF) has expanded in use as a rival to conservative treatment.  Disagreement remains regarding plating in patients with pulmonary contusion or without flail chest. The goal was to determine guidelines for surgical treatment.

Methods: In this IRB approved retrospective-cohort study, 48 patients underwent SSRF between 1/1/2011 and 6/1/2017 at a Level 1 trauma center. This group was matched to 50 nonoperative (NO) controls based on age, mechanism of injury, number of ribs fractured (RFX) and occurrence of flail chest. Incidence of pulmonary contusion and timing to rib plating were recorded. The hospital length of stay (HLOS), ICU LOS, and duration of mechanical ventilation (DMV) were then compared.

Results: SSRF (n=48) and NO (n=50) groups were similar in age (mean 58 y.o. in both), ISS (SSRF:20.0 vs NO:23.5, p=0.2), and RFX (6.7 in both, p=0.9). SSRF was performed 1-18 days of admission. SSFR within 18 days had significantly longer HLOS (SSRF:14.5 vs NO:9.2,p=0.007) and ICU LOS (SSRF:8.7 vs NO:4.2,p=0.001). If plating was done within 3 days, there was no longer a significant increase in HLOS (SSRF:12.4 vs NO:10.9,p=0.7) or ICU LOS (SSRF:7.2 vs NO:3.9,p=0.07). Patients without flail chest plated within 18 days (n=27) had increased HLOS (SSRF:14.7 vs NO:6.5,p<0.001) and increased ICU LOS (SSRF:8.9 vs NO:2.8,p=0.001) compared to NO controls (n=32). Within 3 days to plating, both HLOS (SSRF:12.3 vs NO:6.4,p=0.02) and ICU LOS (SSRF:7.4 vs NO:2.3,p=0.05) remained longer for SSRF patients (n=13) without flail chest compared to their controls (n=14). Patients with flail chest plated within 18 days (n=21) had similar HLOS (SSRF:14.2 vs NO:13.9,p=0.9) and ICU LOS (SSRF:8.3 vs NO:6.1,p=0.2) compared to their controls (n=18). Within 3 days to rib plating, HLOS and ICU LOS remained consistent for SSRF patients (n=10) compared to their controls (n=9). Patients without pulmonary contusion plated within 18 days (n=21), had similar HLOS (SSRF:13.6 vs NO:8.9,p=0.13), but significantly longer ICU LOS (SSRF:8.9 vs NO:2.7,p<0.001) compared to their NO controls (n=34). Within 3 days to plating, HLOS remained similar and ICU LOS longer for SSRF patients (n=10) compared to their controls (n=15). Patients with pulmonary contusion plated within 18 days (n=27) had significantly longer HLOS (SSRF:15.2 vs NO:9.8,p=0.03) but similar ICU LOS (SSRF:8.5 vs NO:6.8,p=0.50) compared to their NO controls (n=16). Within 3 days to plating, HLOS became similar (SSRF:12.2 vs NO:12.7,p=0.5) and ICU LOS remained similar for SSRF patients (n=13) compared to their controls (n=8). Differences in DMV were insignificant in all analyzed cases.

Conclusion: Plated patients had increased HLOS and ICU LOS. These differences were nullified when patients had flail chest, or had pulmonary contusion (if surgery was done within 3 days). Possible advantages of rib plating, including pain reduction and return to work, may be more obvious in long term follow-up.

 

55.04 Using Artificial Intelligence to Predict Prolonged Mechanical Ventilation and Tracheostomy Placement

Posted on January 18, 2018

J. Parreco1, R. Kozol1, R. Rattan1  1University Of Miami,Miami, FL, USA

Introduction:

Early identification of critically ill patients who will require prolonged mechanical ventilation has proven to be difficult and there are no established guidelines. The purpose of this study was to use artificial intelligence and machine learning techniques to identify patients at risk for prolonged mechanical ventilation (PMV) and tracheostomy placement.

 

Methods:

The Multiparameter Intelligent Monitoring in Intensive Care III database was queried for all intensive care unit (ICU) stays with mechanical ventilation and surviving hospitalization. PMV was defined as mechanical ventilation for more than 7 days. Machine learning classifiers with a gradient boosted decision trees algorithm were created for the outcomes of PMV and tracheostomy placement. The classifiers were trained using 10-fold cross validation. The variables used by the classifiers were six different severity of illness scores calculated on the first day of ICU admission including their components as well as thirty comorbidities. Mean receiver operating characteristic (ROC) curves were calculated for the outcomes and variable importance was quantified.

 

Results:

There were 20,262 ICU stays identified and PMV was required in 13.6% and tracheostomy was performed in 6.6% of all patients. The figure shows the mean ROC curves for the outcomes with shaded portion representing the range of cross validation folds. The classifier for predicting PMV was able to achieve a mean area under the curve (AUC) of 0.82 with an accuracy of 83% and specificity of 88%. The mean AUC for tracheostomy was 0.83 with an accuracy of 91% and specificity of 96%. The variable with the highest importance for predicting PMV was the Sequential Organ Failure Assessment (SOFA) Score (13%) and the most important comorbidity in predicting tracheostomy was cardiac arrhythmia (12%).

 

Conclusion:

Machine learning classifiers can be easily incorporated into existing electronic medical record systems. This study demonstrates their utility for the early identification of patients at risk for prolonged mechanical ventilation and tracheostomy. Application of these identification techniques could lead to improved outcomes by allowing for early intervention.

 

 

55.01 Hearing Impairment Impedes Physical Function Recovery after Falls in the Elderly

Posted on January 18, 2018

J. R. Oliver1, C. J. DiMaggio2,3, P. R. Ayoung-Chee2  1New York University School Of Medicine,New York, NY, USA 2New York University School Of Medicine,Department Of Surgery, Division Of Trauma And Acute Care Surgery,New York, NY, USA 3New York University School Of Medicine,Department Of Population Health,New York, NY, USA

Introduction:  Falls are the leading cause of both fatal and nonfatal injuries in adults ≥ 55 years of age. In older adults hearing impairment is associated with an increased risk of falling. This can be due to reduced spatial awareness and coexistent vestibular pathology leading to impaired balance. It has yet to be examined if and how hearing impairment impacts physical function recovery following falls. This study examines the association between hearing impairment and physical function recovery over 6 months following falls in the elderly. 

Methods:  Data were prospectively collected from patients ≥ 55 years of age admitted to a Level 1 trauma center after falling from standing height. Patients were eligible for the study whether or not they had hearing impairment or sustained an injury (Injury Severity Score (ISS) ≥ 1) from their fall. Hearing ability was self-reported and analyzed dichotomously as impaired or not impaired. A self-reported physical function measure (PROMIS-PF) was administered to patients at four time points (upon admission, immediately post-discharge, 3 and 6 months post-fall) to evaluate physical function recovery. PROMIS-PF scores range from 20 to 100 with a minimally important difference of 4.

Results: There were 144 patients in the cohort; 11 patients died. Follow-up rates post discharge, 3 months post-fall and 6 months post-fall were 47.9%, 43.8%, and 44.4% respectively. 31.3% of patients had impaired hearing. 67.4% sustained an injury from their fall, with a mean ISS of 5.8 (sd = 5.4). Immediately after discharge a similar proportion of hearing impaired (61.5%) and non-impaired (60.0%) patients had significant declines in PF scores (p = 1.00). The average PF decline was similar between both groups (hearing impaired 7.3, sd = 13.4) (non-impaired 9.7, sd = 14.4). Six months post-fall a substantially larger proportion of hearing impaired patients continued to have physical function impairment (62.5%) compared to normal hearing individuals (31.5%, p = 0.05). Multivariable regression demonstrated a significant association between hearing impairment and failure to recovery to pre-injury physical function after adjusting for ISS, Charlson Comorbidity Score, and age (OR = 4.77, 95% CI = 1.33 – 19.91, p = 0.02).

Conclusion: Hearing impairment has been shown to be a potentially modifiable risk factor for falls in the elderly. This study demonstrates hearing impairment is also significantly associated with worse recovery following falls. We should focus on improving hearing in elderly individuals as both an injury prevention and injury recovery strategy.

55.02 Sex-Based Differences in Transfusion Need After Severe Injury: Findings of the PROPPR Study

Posted on January 18, 2018

M. McCrum2, B. Leroux1, T. Fang1, E. Bulger1, S. Arbabi1, C. E. Wade3, E. Fox3, J. B. Holcomb3, B. Robinson1  1University Of Washington,Surgery / General Surgery,Seattle, WA, USA 2University Of Utah,Surgery / General Surgery,Salt Lake City, UT, USA 3University Of Texas Health Science Center At Houston,Surgery / General Surgery,Houston, TX, USA

Introduction: Previous studies examining sex-based differences in outcomes after traumatic injury have revealed inconsistent findings, potentially due to heterogeneous study populations and variation in management strategies. We sought to evaluate if sex-based differences in outcomes exist for severely injured patients with hemorrhage who underwent protocolized damage control resuscitation. We hypothesized that female sex is protective against mortality in this setting, and that females would achieve hemostasis more rapidly than males.

Methods: We performed a secondary analysis of severely injured patients predicted to receive a massive transfusion who were randomized to 1:1:1 vs 1:1:2 plasma-platelet-red blood cells (RBCs) ratios at 12 level I North American trauma centers in the Pragmatic Randomized Optimal Platelet and Plasma Ratios (PROPPR) trial. We used Cox regression to evaluate the sex-based differences in time to hemostasis and mortality. Poisson regression with robust standard errors was used to evaluate differences in volume of blood products transfused in each phase of the study. Pre-randomization and randomization periods occurred prior to hemorrhage control, whereas the post-randomization period occurred after anatomic hemostasis was achieved. All analyses were adjusted for clinical site, age, race, height, base deficit, INR, mechanism of injury, AIS scores, and transfusion arm (1:1:1 vs 1:1:2).  

Results:We identified 134 females and 546 males in the study population. Female patients were older (age 40.5 [IQR 26-60] vs 33 [24-47], p<0.001), with higher ISS (29 [21-41] vs 26 [16-38], p=0.04), and with a higher proportion of blunt (76 vs 47%, p<0.001) and severe head injury (AIS ≥3, 33 vs 20%, p0.01); BMI did not differ by sex (p=0.50). There were no differences between the sexes in mortality over the first 24-hours (HR for female=1.08, 95% CI 0.66, 1.74; p=0.77), 30-days (HR=1.00, 95% CI 0.69, 1.45; p=0.99), or time to hemostasis (HR=1.02, 95% CI 0.82, 1.26; p=0.88). Differences by sex were not found in the volume of plasma, platelets or RBCs received in the pre-randomization or randomization periods. However, in the post-randomization period, females received 38% less plasma (95% CI 11%, 57%; p=0.01), 48% less platelets (95% CI 22%, 65%; p=0.001) and 48% less RBC (95% CI 20%, 67%; p=0.01) compared to males (Figure 1).

Conclusion:Mortality rates and time to hemostasis of severely injured trauma patients with hemorrhage did not differ by sex. Female sex appears to be protective against the need for “catch up” blood required after initial hemostasis. Further work is needed to elucidate sex-based differences in coagulopathy and hemostasis during and following damage control resuscitation.

54.17 Risk Factors for Prolonged Need for PEG Tube In Adult Trauma Patients: Experience of a Level I Trauma Center

Posted on January 18, 2018

B. K. Yorkgitis1, O. A. Olufajo2, L. Gurien1, E. Kelly2, A. Salim2, R. Askari2  1University Of Florida-Jacksonville,Acute Care Surgery,JAcksonville, FL, USA 2Brigham And Women’s Hospital,Trauma, Burns And Surgical Critical Care,Boston, MA, USA

Introduction: PEG tubes are a means of providing nutrition to patient that are unable to consume their caloric intake.  PEG tubes placed in trauma patients after the acute trauma differ from commonly studies malignancy patients.  Often, patients and/or care givers inquires about the length of time alternative means of nutrition will be required.  Our aim was to identify risk factors for prolong need for PEG tube (>90 days).

Methods: The trauma database of a Level I, urban trauma center was used to identify all trauma patients who had PEG tubes placed between January 1, 2009 to December 31, 2014.  Exclusions included age <18 years, pregnancy, patients without follow-up evaluations, and those with pre-existing PEG tubes. Variables examined included: age, sex, injury severity score (ISS), initial GCS, admission albumin, ventilator days, ICU days, hospital LOS, and Charlson Comorbidity Index (CCI). Multivariable logistic regression models were used to identify risk factors for having PEG tubes >90 days.

Results:

During the study period 9772 charts were reviewed for PEG tube placement.  Of these patients, 285 (2.9%) underwent attempted PEG tube, of which 281 (98.9%) were successful. Five attempts were aborted due inadequate visualization; one was able to be subsequently placed with an additional attempt. There were 9 (3.2%) in-hospital deaths in patients successfully receiving PEG tube placement during their index trauma admission; none were related to the procedure. There were 18 (6.2%) PEG tube related complications. Upon review of the living patients at discharge, 195 had adequate clinical documentation of PEG tube presence or decannulation status to allow for analysis.

Of these patients, 149 (76.4%) were male, the average age was 57.5 years with 80 (41.0%) being over 65 years.  The average admitting albumin was 3.7 g/dL and CCI was 1.1. First recorded mean GCS was 10.7 and ISS 23.2. The mean hospital LOS was 23.8 days, ICU LOS 16.5 days and ventilator days while in the hospital was 11.5 days. The mean CCI was 1.1.

The findings are summarized in Table 1. Of those requiring a PEG tube during their index admission, 77 (41.4%) required their PEG tube >90 days. Statistically significant characteristics identified upon univariate analysis include increasing age, age>65 years, higher CCI and longer in-hospital ventilator days.  Upon logistic regression, CCI>1 (OR1.27, CI 95% 1.03-1.56, p=0.02) and longer in-hospital ventilator days (OR 1.05 CI 95% 1.02-1.09, p<0.01) were predictive of need for prolonged PEG tube.

Conclusion:

Patients and their surrogates often inquire about the length of time a PEG tube may be needed after trauma.  This retrospective chart review identified CCI>1 and prolonged in-hospital ventilator days as risk factors for necessity of PEG tube >90 days. This may assist clinicians and patients or their surrogates in decision making when faced for the need for alternative routes for nutrition. 
 

54.19 A New Paradigm for the Acute Care Surgeon: The Active Cancer Patient Population

Posted on January 18, 2018

C. R. Horwood1, S. Byrd1, E. Schneider1, K. Woodling1, J. Wisler1, A. P. Rushing1  1Ohio State University,General Surgery,Columbus, OH, USA

Introduction:
With advances in cancer treatment, patients with active cancer are living longer and developing common conditions that fall outside the realm of oncologic treatment. There is little data describing the clinical decision-making and subsequent outcomes that occur in patients with cancer that present with an emergency general surgical diagnosis. The purpose of this study is to evaluate factors that determine operative management (OM) versus non-operative (NOM) management and subsequent outcomes of colonic emergencies within a cancer population.

Methods:
A single institution registry at a comprehensive cancer center was queried to identify patients with an active cancer diagnosis that had a surgical consult placed for one of three acute colonic surgical diagnoses between 2011 and 2016. The diagnoses included colitis (with or without peritonitis), bleeding diverticulosis, and diverticulitis. Records were retrospectively reviewed for patient demographics, baseline clinical characteristics, OM vs NOM interventions, and outcomes. Primary outcomes examined include hospital length of stay (LOS), 30-day mortality, disposition at discharge, and post-operative complications.

Results:
A total of 87 patients were evaluated of which 38 (43.7%) underwent OM vs 49 (56.3%) who underwent NOM. Differences in initial lab values included median white blood cell count (WBC) and median serum lactate. Median WBC was 8.1 in the OM group vs 4.6 in the NOM group, (p<0.001), and median serum lactate was 2.32 in the OM group vs 0.95 in the NOM group, (p<0.001). Thirteen patients had peritonitis at the time of surgical consult in the OM group vs zero in the NOM group, (p<0.001). With regard to post-operative outcomes, there was no difference in hospital LOS, but there was a difference in 30-day mortality and discharge to home between the groups. Median LOS in the OM group was 13.5 days vs 9 days in the NOM group (difference of 4.5 days, p=0.16). However, 30-day mortality was 32% in the OM group vs 6% in the NOM group (p=0.003) and discharge to home occurred in 35% of patients in the OM group compared to 80% in the NOM group (p<0.001). Within the OM group, 70% of patients had at least one post-operative complication.

Conclusion:
Not surprisingly, patients with higher WBCs, lactic acidosis, and peritonitis were more likely to undergo surgical intervention. Operative management was offered to sicker patients, and as a result, this group had a higher 30-day mortality and were less likely to be discharged home. While more work is needed to evaluate risk stratification for post-operative morbidity, particularly as it compares to non-cancer patients undergoing similar surgical procedures, these preliminary outcomes should be discussed with the active cancer patient prior to pursuing surgical intervention.
 

54.15 Surgeon-Placed Endoscopic Stent As Rescue Therapy For Failed Esophageal Repair After Trauma

Posted on January 18, 2018

K. Chow1, H. N. Mashbari1, M. Hemdi1, E. Smith-Singares1  1University Of Illinois At Chicago,Division Of Surgical Critical Care,Chicago, IL, USA

Introduction:
Esophageal trauma represents an uncommon but potentially catastrophic injury with a reported overall mortality of up to 20%. The management of iatrogenic and spontaneous perforations have been previously described with well-established guidelines which have been mirrored in the trauma setting. Esophageal leaks are the most feared complication after primary surgical management and present a challenge to salvage. There has been increasing reports in the literature supporting the use of removable covered metal stents to heal esophageal perforations and leaks in the non-trauma setting. We present the largest case series of four patients presenting with external trauma induced esophageal injuries, with failure of initial primary surgical repair with subsequent development of a leak successfully managed with esophageal stenting.

Methods:
A retrospective review of four patients who underwent endoscopically placed indwelling covered metal stents after leakage of primary esophageal repair was performed. Demographic information, hospital stay, additional interventions, complications, imaging studies, ISS scores, and outcomes were collected.

Results:
Our cohort consisted of 4 patients with penetrating injuries to the chest and neck with esophageal injuries (3 thoracic and 1 cervical esophageal injuries) managed with esophageal stenting after leaks were diagnosed following primary repair. Leaks were diagnosed on average post-operative day 9 with stents placed within 3 days after diagnosis. There was complete resolution of their esophageal fistulas with all patients resuming oral intake (averaging 72 days after stent placement). Three patients (75%) required further endoscopic interventions to adjust the stent due to migration or for dilations due to strictures. Mortality was 0%, all patients survived to be discharged from the hospital with average ICU length of stay of 30 days.

Conclusion:
The use of esophageal stenting has progressed over the last few years, with successful management of both post-operative upper gastrointestinal leaks as well as benign, spontaneous, or iatrogenic esophageal perforations. While the mainstay of traumatic esophageal injuries remains surgical exploration, debridement, and repair with perivisceral drainage; our case series illustrates that the use of esophageal stents is an attractive adjunct that can be effective in the management of post-operative leaks after attempted repair of external penetrating traumatic injuries to the esophagus.
 

« Previous Page
Next Page »