18.18 Modifiable Risk Factors and Clinical Outcomes Associated with Augmented Renal Clearance in Trauma

Posted on January 31, 2019

M. B. Mulder1, M. S. Sussman1, S. A. Eidelson1, C. A. Karcutskie1, M. A. Cohen1, A. T. Vidalin1, G. A. Lama1, R. S. Iyengar1, P. M. Elias1, C. I. Schulman1, N. Namias1, K. G. Proctor1  1University Of Miami,Dewitt Daughtry Department Of Surgery: Division Of Trauma And Critical Care,Miami, FL, USA

Introduction:  

Augmented renal clearance ((ARC) defined as creatinine clearance (CLCr) > 130 mL/min) has a reported incidence from 14 to 80% in critically ill patients and is associated with therapy failures for renally-cleared drugs. While the awareness of ARC has increased, the clinical implications of this phenomenon are less defined. The objective of this study was to identify modifiable risk factors and clinical outcomes associated with ARC in severely injured trauma patients. 

Methods:  

In 199 trauma ICU patients with a Greenfield Risk Assessment Profile ≥  8, 24-hour CLCr was correlated with demographics, interventions (IV fluids, pressors, mechanical ventilation), clinical estimates of GFR (by Cockroft-Gault (CG), modification of diet in renal disease (MDRD), or chronic kidney disease epidemiology (CKD-EPI)), and clinical outcomes (infection, VTE, length of stay (LOS), and mortality). Patients with previous nephrectomy or renal transplant were excluded (n=6).  Univariate and multivariate analysis identified risk factors with significance defined at p≤0.05. Values are M±SD if parametric and median [interquartile range] if not. 

Results:

The population was 46±20 years, 68% male, BMI 28±6 kg/m2, 72% blunt mechanism of injury, and injury severity score (ISS) of 24±10. Admission SCr was 0.95 [0.78-1.2] mg/dL, CLCr was 152±74 ml/min, VTE incidence was 14%, ARC incidence was 57%, and mortality 11%. Clinical estimates of GFR by CG, MDRD, and CKD-EPI underestimated CLCr by 14%, 19%, and 18% respectively (all p≤0.001). CLCr was lower in patients receiving transfusions (123±74 v 167±67 ml/min, p≤0.001), pressors (117±71 v 162±73 ml/min, p≤0.001), with positive cultures (138±73 v 161±73 ml/min, p=0.041), and in those who expired (92±58 v 159±73 ml/min, p≤0.001).  Univariate analysis of over 15 risk factors and clinical outcomes were analyzed for ARC; values that were significant (p≤0.05) are shown in Table 1. On multivariate logistic regression male gender (OR 4.5 [1.8-11]), SCr (OR 0.17 [0.041-0.71]), age (OR 0.96 [0.94-0.99]) and transfusions (OR 0.24 [0.011-0.54) were independent predictors of ARC (all p≤0.01).  

Conclusion:
ARC occurs in half of all high-risk trauma ICU patients and is underestimated by standard clinical equations. ARC is associated with younger males, less transfusions, fewer infections, and reduced mortality. These clinical outcomes are counterintuitive.  Further investigations are warranted to delineate the implications and causality of ARC.

18.17 Superficial Surgical Infections in Operative Abdominal Trauma Patients: A TQIP Analysis

Posted on January 31, 2019

S. R. Durbin1, J. Peschman2, D. Milia2, T. Carver2, C. Dodgion2  1Medical College Of Wisconsin,Milwaukee, WI, USA 2Medical College Of Wisconsin,Department Of Surgery,Milwaukee, WI, USA

Introduction:
Surgical site infections (SSIs) have a substantial impact on economic and health indices for patients and healthcare institutions. The aim of this study was to identify risk factors for superficial surgical site infections in operative abdominal trauma patients using a national cohort.

Methods:
A retrospective analysis of all adult trauma patients treated within the Trauma Quality Improvement Database (TQIP) who underwent an abdominal operation from 2010-2015 was performed. Risk factors for infection were evaluated using χ2, Wilcoxon rank sum, and multivariate logistic regression.

Results:
There were 46,611 patients who underwent an exploratory laparotomy for trauma. The median age was 33 [24-50]; 36,337 (77.7%) were male and 24,968 (55.5%) were Caucasian. There were 20,750 (44.8%) penetrating injuries and the median injury severity score (ISS) was 21 [13-30]. 12,407 (26.5%) suffered a hollow viscus injury with one (9,457;20.2%), two (2,618;5.6%) or > three (354;0.7%) organs injured respectively.  Overall, 1,944 (4.2%) patients developed an SSI, with >15% of all colon and duodenal injuries developing an infection. On multivariate analysis, after controlling for gender, type of injury, shock on arrival, smoking status, diabetes, and steroid use or chemotherapy use, colonic injuries conferred the greatest independent risk for superficial SSIs (OR 2.91 [2.44-3.48]) followed by duodenal injuries (OR 1.92 [1.21-3.06]) small bowel injuries (OR 1.56 [1.30-1.87]) gastric injuries (OR 1.43 [1.07-1.91]), BMI >30 (1.33 [1.15-1.55]), African American race (OR 1.22 [1.03-1.47]), increasing ISS (OR 1.01 [1.01-1.02] and increasing age (OR 1.01 [1.002-1.01]. Additionally, an increasing number of hollow viscus injuries was associated with a greater risk for superficial SSI; one (OR 2.75 [2.32-3.26]), two (OR 3.81 [2.98-4.89]) or three (OR 6.85 [4.20-11.17]) organs injured respectively.

Conclusion:
The incidence of superficial SSI in operative abdominal trauma patients increases with higher ISS, increased BMI, increased age and an increasing number hollow viscus injuries. Colon and duodenal injuries, in particular, impart the greatest risk.  Consideration should be given to avoiding primary skin closure in patients with these risk factors as a way to mitigate SSIs in this patient population.

 

18.16 Blunt Esophageal Injuries: A 5-Year Review from the Trauma Quality Improvement Program (TQIP)

Posted on January 31, 2019

M. Osama1, M. Hamidi1, A. Northcutt1, E. Zakaria1, N. Kulvatunyou1, T. O’Keeffe1, A. Tang1, L. Gries1, B. Joseph1  1University Of Arizona,Trauma And Acute Care Surgery,Tucson, AZ, USA

Introduction:
Blunt esophageal injuries are rare and their management is evolving. Diagnosis and management of these injuries remains a challenge. The aim of our study was to analyze the trends of operative intervention and mortality after blunt esophageal injuries.

Methods:
A 5-year (2010-2014) analysis of all trauma patients with esophageal injuries from the Trauma Quality Improvement Program (TQIP) was performed. All adult patients (>18 years) with blunt esophageal injuries were included. Patients were identified using ICD-9 codes and the following data points were obtained: demographics, admission vitals and procedures performed. Outcome measures were trends of operative intervention and mortality after esophageal injuries. Operative intervention was defined as any surgical procedure performed for the repair of esophagus (primary, secondary, diversion, and resection).  

Results:
808,196 trauma patients were analyzed of which 153 patients had blunt esophageal injuries. Mean age was 43 ± 25 years, 75.2% (115) were male, and 79.7% (122) were white. Median [IQR] ISS was 21[9-34]. 75.2% (115) of the injuries were from motor vehicle collisions while 15.7% (24) were from falls. Overall, mortality rate was 17% (26) of which 50% were within the first 24 hours of injury and 19.1% in the following 7-days. 19.6% (30) patients underwent surgical intervention for repair of esophageal injuries. 14 patients had primary repair of laceration while 1 patient had resection of esophagus. The rate of operative intervention has increased while the mortality has decreased over the 5-years study period (Figure 1). On regression analysis, SBP (OR= 0.85; 95% CI, 0.61-091; p=0.01), pulse (OR= 1.31; 95% CI, 1.2-1.9; p<0.001), ISS (OR= 1.8; 95% CI, 1.3-2.5; p<0.001), and treatment at a non-teaching hospital (OR= 2.5; 95% CI, 1.5-6.6; p<0.001) were independently associated with mortality.

Conclusion:
The rate of operative intervention of esophageal injuries has increased while the mortality rate has decreased over the study period. Teaching status of hospital was associated with decrease in hospital mortality. Understanding why teaching hospitals have better outcomes may help improve survival from this rare injury.
 

18.15 Heparin-Sparing Anticoagulation Strategies are Viable Options for Patients on Veno-venous ECMO

Posted on January 31, 2019

K. T. Carter3, R. Panchal3, J. A. Shake1, A. Panos1, R. P. Cochran1, L. Creswell1, M. E. Kutcher2, H. Copeland1  1University Of Mississippi,Cardiothoracic Surgery,Jackson, MS, USA 2University Of Mississippi,Trauma/Critical Care,Jackson, MS, USA 3University Of Mississippi,Surgery,Jackson, MS, USA

Introduction:  Extracorporeal membrane oxygenation (ECMO) is an option for severe pulmonary dysfunction, but has traditionally been limited by anticoagulation requirements.  Recently, the absolute requirement for anticoagulation has been challenged, potentially allowing inclusion of patients with high bleeding risk or other relative contraindications to anticoagulation.  We hypothesize that among veno-venous (VV) ECMO patients, there is no difference in mortality, bleeding, or thrombotic events between a heparin-sparing and a full therapeutic anticoagulation strategy.

Methods:  We conducted a single center retrospective review of all adult patients placed on VV ECMO from 10/2011 to 5/2018.  Beginning in October 2014, a heparin-sparing protocol was implemented.  VV ECMO was initiated without heparin or with low dose continuous intravenous (IV) heparin (activated clotting time goal 140-180) unless indications developed to start continuous therapeutic heparin.  We compared ECMO patients, pre and post protocol change.  The primary endpoint was survival; secondary endpoints were bleeding, thromboembolic events, and transfusion requirements.

Results: Forty adult patients were placed on VV ECMO during the study period: 17 patients (147 circuit-days) before and 23 patients (214 circuit-days) after an institutional change to a heparin-sparing protocol.  Prior to the protocol change, a heparin-sparing strategy was used during 27% of the total ECMO days compared with 89% after.  The post-protocol group had a lower mean body mass index (29 ± 8 vs. 38 ± 13, p=0.01), were cannulated earlier (median 1 vs. 5 day after ICU admission, p=0.03), and more commonly required inotropic support (87 vs. 41%, p=0.01), but were otherwise similar.  There were no significant differences between groups in survival to decannulation (61 vs. 82%, p=0.18 ) or discharge (52 vs. 71%, p=0.33).  Incidence of bleeding (41 vs. 45%, p=1.0) and thromboembolic events (18 vs. 36%, p=0.38) as well as need for transfusion (0.9 vs. 1 unit per circuit-day, p=0.87) were also similar in the pre vs. post groups.  After the protocol change, 75 circuit-days (35%) were completely heparin-free. Nine patients (39%) in the post-group had entirely heparin-free courses of ECMO (total 33 circuit-days); survival and the incidence of bleeding and thrombotic complications did not differ in this heparin-free subset.

Conclusion: There is no difference in survival, bleeding, thrombotic complications, or transfusion requirements between a heparin-sparing strategy and a full therapeutic heparin strategy in the management of VV ECMO.  VV ECMO can be an option in patients with traditional contraindications to anticoagulation, especially in trauma patients at high risk for bleeding.

 

18.14 “Brain Injury—Does it Matter if by Stroke or by Injury?”

Posted on January 31, 2019

E. Winford1, L. Tennakoon3, L. Knowlton3, M. Martin2, K. Staudenmayer3  1Meharry Medical College,School Of Medicine,Nashville, TN, USA 2Stanford University,Department Of Medicine,Palo Alto, CA, USA 3Stanford University,Department Of Surgery,Palo Alto, CA, USA

Introduction: Cerebrovascular accident (CVA) and traumatic brain injury (TBI) are leading causes of disability and are increasingly common in older adults. While both result in some form of brain injury, it is unknown whether the impact they have on outcomes is the same. We hypothesized that short-term outcomes and 6-month healthcare utilization would be worse for stroke.

Methods: The 2014 National Readmissions Database (NRD), a longitudinal, all-payer database was used. Patients were included in the study if they were ≥65 years and were admitted with a primary diagnosis of either CVA or TBI during the first half of the year. Patients were excluded if their primary diagnosis was “transient ischemic attack” for CVA; for TBI, patients with evidence of multiple injuries were excluded (AIS≥2 in any other body region). The primary outcome was mortality during index hospitalization. Secondary outcomes included rates of tracheostomy and gastrostomy, and 6-month inpatient hospitalization days and costs. Unadjusted and adjusted analyses were performed. Weighted numbers are reported.

Results: 175,415 patients met inclusion and exclusion criteria. Mortality during the index hospitalization was lower for CVA (8.8% vs 10.3% for CVA and TBI, respectively, p<0.001). After controlling for known confounders, CVA was associated with lower inpatient mortality (OR 0.92, 95% CI 0.86-0.98). Rates of gastrostomy and tracheostomy differed for both groups. For CVA vs. TBI, 4.9% vs. 2.8%, (p<0.001) of patients had a gastrostomy, whereas 0.9% vs. 1.3% had a tracheostomy (P=0.0002). This held true in adjusted analysis, where CVA was associated with a higher likelihood of gastrostomy and lower likelihood of tracheostomy vs. TBI (gastrostomy OR 1.95, 95% CI 1.75-2.17; tracheostomy OR 0.81, 95% CI 0.68-0.95). Despite these differences, 6-month inpatient costs and hospital days were similar. For CVA vs. TBI, 6-month costs were $15,507.77 vs. $15,945.47 (p=0.16) and hospital days were 7.0 days vs. 6.5 days (p<0.001). There was no difference when controlling for known confounders.

Conclusion: Brain injury is the final common pathway of CVA and TBI. Following CVA and TBI, healthcare costs and LOS are the same. Interestingly, rates of tracheostomy and gastrostomy are different for the two populations. This may be due to different needs, but also might be due to differences in patient management by different types of physicians. This suggests an opportunity for physicians caring for CVA and TBI patients to further explore and improve care by comparing management of patients who sustain brain injury by any mechanism.
 

18.13 Inpatient Rehabilitation Reduces the Likelihood of Chronic Pain After Injury: A Multi-Center Cohort Study

Posted on January 31, 2019

J. P. Herrera-Escobar1, R. Manzano-Nunez1, S. S. Al Rafai1, A. Toppo1, K. Han2, N. Bhulani1, G. Kasotakis3, G. Velmahos2, A. Salim4, A. H. Haider1,4, D. Nehra4  1Brigham And Women’s Hospital,Center For Surgery And Public Health,Boston, MA, USA 2Massachusetts General Hospital,Division Of Trauma, Emergency Surgery & Surgical Critical Care,Boston, MA, USA 3Boston University,Boston, MA, USA 4Brigham And Women’s Hospital,Division Of Trauma, Burn, And Surgical Critical Care,Boston, MA, USA

Introduction:  Most of the 96% of patients who survive traumatic injuries require ongoing rehabilitation after discharge. However, the long-term impact of post-discharge rehabilitation care remains poorly understood. In this study, we sought to compare the likelihood of persistent pain needing medication 6 to 12 months after traumatic injury between patients discharged to an inpatient rehabilitation facility (IRF) and a skilled nursing facility (SNF).

Methods:  Moderate to severe [Injury Severity Score (ISS) ≥ 9] trauma patients admitted to 1 of 3 Level-I trauma centers were interviewed between 6 and 12 months after injury from 2015 to 2018. During the interview, the presence of daily pain and need of pain medications were evaluated using the Trauma Quality of Life (T-QoL) questionnaire. This information was linked to the index hospitalization through the trauma registry. Inverse probability of treatment weighting (IPTW)–adjusted logistic regression analysis was performed to compare the likelihood of persistent pain requiring medication among patients who were discharged to IRF vs SNF. Factors used for the propensity score calculation included age, gender, insurance, comorbidities, injury type, polytrauma, ISS, head injury, extremities injury, intensive care unit admission, ventilator use, complications, length of stay, and hospital.

Results: A total of 519 patients were included in the analysis: 389 discharged to IRFs and 130 to SNFs respectively. Unweighted and weighted baseline characteristics are described in Table 1. In unweighted analysis, rates of pain needing medication were 29.1% for IRF vs 40.8% for SNF (P: .013). After IPTW adjustment, propensity score distribution between IRF vs SNF achieved adequate balance and standardized differences were less than 10% (Table 1), which indicated that patients of both groups were subsequently comparable. IPTW-adjusted rates of pain needing medication were 28.9% for IRF vs 47.5% for SNF. In the IPTW-adjusted logistic regression analysis, IRF was associated with a significant reduction in the likelihood of chronic pain after injury (OR 0.45; 95% CI, 0.23-0.90; P = .023).

Conclusion: IRFs may be associated with a reduction in the likelihood of chronic pain after trauma. The reasons for this difference are unknown, but it is likely to be multifactorial and perhaps we should consider IRF over SNF for patients at high risk for chronic pain after traumatic injury.

 

18.12 Efficacy of Platelet Transfusion for Antiplatelet Reversal in Traumatic Intracranial Hemorrhage

Posted on January 31, 2019

E. Baughman1,2, J. G. Hein1, M. Jackson1,3, T. W. Wolff1,4, M. L. Moorman1,2, U. Pandya1,2, M. C. Spalding1,2  1OhioHealth Grant Medical Center,Division Of Trauma And Acute Care Surgery,Columbus, OH, USA 2Ohio University Heritage College of Osteopathic Medicine,Athens, OH, USA 3Northeast Ohio Medical University,Rootstown, OH, USA 4OhioHealth Doctors Hospital,Department of Surgery,Columbus, OH, USA

Introduction: Increases in vascular disease prevalence have led to as many as half of all US adults aged 45 to 75 being prescribed antiplatelet agents. Patients on antiplatelet medication with traumatic intracranial hemorrhages (tICH) have been shown to have a 3 to 15 times higher rate of mortality. Some institutions adopted the practice of giving platelet transfusions to patients with tICH on pre-injury antiplatelet therapy. Although intuitive, there is little matched cohort data to justify this practice and many studies are biased by disease burden. The aim of our study is to understand the efficacy of platelet transfusion for patients with tICH on pre-injury antiplatelet medication.

Methods:  We identified patients on pre-injury antiplatelet medication admitted with a tICH to an urban, Level 1 trauma center between January 1, 2014 and June 30, 2018. Per institutional guidelines, patients admitted prior to September 2017 were transfused platelets, and those admitted after were not. The primary outcome was mortality. Secondary outcomes were: need for neurosurgical intervention, Intensive care unit length of stay, need for increased level of care, and discharge destination. Demographics, prehospital medications, comorbidities, injury characteristics, and hospitalization events were also evaluated. Chi squared analyses and t tests were used to compare the two groups.

Results: When comparing the platelet transfusion group (449) versus no transfusion group (102), demographics, prehospital medications, comorbidities, injury characteristics, and hospitalization events were not significant, including age (73.13 vs. 75.74, p=0.062), injury severity score (16.21 vs. 15.35, p=0.339), head abbreviated injury scale (7.10 vs. 6.93, p=0.890), Glasgow coma scale (11.9 vs. 11.5, p=0.354) and length of stay (5.39 vs. 5.55, p=0.772). The primary outcome of mortality was nonsignificant (p=0.193), with a 10% and 6% mortality in the transfused and non-transfused groups respectively. Secondary outcomes of neurosurgical intervention (11.6% vs. 7.8%, p=0.300), Intensive care unit length of stay (1.33 vs. 1.46 days, p=0.698), need for increased level of care (6.68% vs. 9.80%, p=0.273), and discharge destination (p=0.662) were also nonsignificant for transfused versus non-transfused groups. Subgroup analysis of patients with subarachnoid hemorrhage (SAH), a specific type of tICH, did reveal a significant difference in mortality (6.25% vs. 0%, p=0.022) and discharge destination (p=0.035) between the transfused versus non-transfused group.

Conclusion: This early analysis indicates that platelet transfusion may have a significant effect on mortality for patients with a SAH on pre-injury antiplatelet medication. Analysis of all tICH data agrees with previous literature supporting no empiric transfusion of platelets for tICH patients on pre-injury antiplatelet medication. Further study to validate this finding and to assess the impact on hematoma expansion should be undertaken.

18.11 What Are We Looking For: Frailty Scores Lacking Uniformity in Identifying Patients

Posted on January 31, 2019

H. K. Weiss1, B. Cook2, B. W. Stocker1, N. Weingarten1, K. E. Engelhardt3, J. Posluszny2  1Feinberg School Of Medicine – Northwestern University,Chicago, IL, USA 2Northwestern University,Department Of Surgery,Chicago, IL, USA 3The Medical University Of South Carolina,Department Of Surgery,Charleston, SC, USA

Introduction: Screening patients for frailty is traditionally done at the bedside. This process includes screening for comorbidities, physical activity, emotional health, and nutrition. However, recent studies have attempted to identify frailty using non-bedside, electronic medical record (EMR)-based, and primarily comorbidity-focused frailty assessments. Our objective is to determine how the bedside Trauma and Emergency General Surgery (TEGS) frailty index (FI) compares to non-bedside frailty assessments in uniformity of detecting patients.

Methods: We retrospectively reviewed our quality improvement (QI) project database consisting of geriatric ( ≥65 year old ) TEGS patients. Patients were screened with the TEGS FI, a literature validated, 15-question assessment performed at the bedside, including comorbidities, physical activity, emotional health, and nutrition. We reviewed the EMR to calculate non-bedside frailty scores: the Enterprise Data Warehouse (EDW) Frailty Assessment score, a 6-point score from an EMR-based database, the NSQIP mFI-11, and the NSQIP mFI-5 (see Table 1). Based on 31% of the patients being frail as defined by the TEGS FI, a score ≥ 3 on the mFI-11 and ≥ 2 on the mFI-5 was considered frail. We compared overlap of frailty diagnoses between the four different frail groups. We then compared illness and disease severity among groups (Charlson Age-Comorbidity Index (CCI), ASA, SOFA, APACHE II, and P-POSSUM).

Results: 71 geriatric TEGS patients were included, of which 22 (31%) were frail on the TEGS FI, 24 (33%) on the EDW FI, 25 (35.2%) on the mFI-11, and 29 (40.8%) on the mFI-5. Of the patients identified as frail on the TEGS FI, only 13 patients (59%) were frail on the EDW FI, 13 patients (59%) on the mFI-11, and 15 patients (68%) on the mFI-5. Only 7 (32%) patients of the 22 frail patients identified by the bedside TEGS FI were frail by all 4 frailty assessments. When compared to the TEGS FI, illness severity scores did not differ amongst groups (ex. CCI: TEGS, 5.4; EDW, 5.1 (p=0.55); mFI-11, 5.8 (p=0.41); mFI-5, 5.6 (p=0.67)).

Conclusion: There was minimal overlap between the bedside TEGS FI and the non-bedside FIs, suggesting these various frailty scoring systems are identifying different cohorts of patients. There was no difference in traditional illness-severity scores between frail patients identified on the bedside and non-bedside FIs, suggesting no difference in disease or comorbidity between groups. The bedside and non-beside frailty assessments are both assessing for frailty, yet they are resulting in markedly differing patient populations. Larger sample size and further study analyzing clinical outcomes will help to demonstrate if there is a superior approach to identifying frailty.

 

18.10 The Outcomes Of Severe Traumatic Brain Injury In The Elderly: Is It Age Or Is It Frailty, Or It Is Both??

Posted on January 31, 2019

A. Azim1, K. Prabhakaran1, D. Samson1, G. Lombardo1, J. Con1, A. El-Menyar1, R. Latifi1  1Westchester Medical Center,Surgery,Valhalla, NY, USA

Introduction:  Pneumonia, Acute Respiratory Distress Syndrome(ARDS) and  Acute Kidney Injury (AKI) are common in elderly trauma patients. Presence pf these comorbidities in severe Traumatic Brain Injury (sTBI) patients makes clinical course more complicated. The aim of this study was to review the factors that influence outcomes in patients with sTBI.

Methods: All trauma patients ≥ 65 years old with sTBI (GCS ≤8) identified in the trauma registry of our Level I ACS verified trauma center during 2011-2016 were studied. Data points including Injury characteristics, demographics, comorbidities were collected and analyzed. Previously validated 11 variable modified frailty index (mFI) as utilized to calculate frality status. Outcome variables included death and complications.

Results: There were 194 patients; 55.2% age 65-79 years (Group I) and 44.8% age >80(Group II). Modified Frailty Index(mFI) scores were ≥3 in 36 (18.6%). The injury severity score(ISS) was ≥25 in 56.7%. Older patients had abbreviated injury score (AIS) head > 3 (71.3% for ≥80 and 56.1% for 65-79, p=0.036), and cirrhosis of the liver (19.5% for ≥80 versus 4.7% for 65-79, p=0.001). 64.4% of patients ≥80  and 45.8%  of patients 65-79, (p=0.01) died. Patients with higher frailty score had an increased probability of experiencing acute kidney injury(AKI) (6.3% for mFI 0-2 versus 25% for mFI ≥3, p=0.002) as well as acute respiratory distress syndrome(ARDS) (22.8% for mFI 0-2 versus 38.9% for mFI ≥3, p=0.046). In subgroup analysis, mortality was significantly more likely in the mFI 0-2/age ≥80 group (68.1%) than in the mFI 0-2/age 65-79 group (45.1%, p=0.036). Older non-frail patients were less likely to experience AKI than younger more frail patients: 5.8% for mFI 0-2/age ≥80 versus 27.8% for mFI ≥3/age 65-79 (p=0.013) (see Table 1).

Conclusions: Age is associated with increased mortality, whereas frailty is associated with increased complications (particularly AKI) in this patient population. Identification of frailty status of the patient can help in risk startification and appropriate resource allocation in patients resulting in better outcomes in severe TBI patients

 

18.09 The relationship between bonding social capital and long-term outcomes after traumatic injury

Posted on January 31, 2019

R. Manzano-Nunez1, J. P. Herrera-Escobar1, A. Toppo3, D. Blake2, N. Levy-Carrick5, T. DeRoon-Cassini4, G. Velmahos6, G. Kasotakis2, A. Salim1, A. Haider1, D. Nehra1  1Brigham And Women’s Hospital,Surgery,Boston, MA, USA 2Boston University,School Of Medicine,Boston, MA, USA 3Tufts Medical Center,Boston, MA, USA 4University Of Wisconsin,Madison, WI, USA 5Brigham And Women’s Hospital,Psychiatry,Boston, MA, USA 6Massachusetts General Hospital,Surgery,Boston, MA, USA

Introduction:  Social capital (SC) refers to the quality and quantity of social relations. Bonding SC refers to close relationships between family members or good friends. An individuals' health outcomes are known to be closely associated with individual-level SC, but the effect of SC after trauma is not known. We aim to determine the association between individual level bonding SC and long-term outcomes after injury. 

Methods: Adult trauma patients with an ISS ≥9, admitted to one of three Level I Trauma Centers between 2015-2018 were contacted by phone 6 and 12 months post-injury. Patients were asked to complete an interview to assess their Health-Related Quality of Life (HRQoL) using the 12-item Short Form Survey (SF-12) and the validated Trauma Quality of Life (TQoL) survey.  As a part of the TQoL survey all patients were asked to respond on a 5-point Likert scale to the following question: “My injuries have negatively changed my relationships with my family, friends, or intimate partner.”  We used responses to this question as a proxy of Bonding SC. Respondents who chose strongly agree/agree and strongly disagree/disagree were considered to have weak and strong bonding SC respectively. Multivariable adjusted linear regression models were used to determine independent cross-sectional associations between weak Bonding SC and HRQoL.For the subset of patients that completed the survey at 6 and 12 months post-injury longitudinal analyses were conducted to evaluate recovery trajectories.

Results: A total of 609 patients completed the phone screen at 6 months post-injury. Of these, 480 were classified as having strong bonding SC and 129 as having weak bonding SC. Patients with weak bonding SC were significantly younger [52±21.2 vs 61±19.7years; p<0.0001] and had significantly higher ISS [15.1±7 vs 13.7±6.8; p=0.03]. After multivariable linear regression, weak bonding SC was an independent predictor for both worse mental (β=-14.34, 95% CI: -16.57 to -12.11; p<0.001) and physical health (β=-6, 95% CI: -8.32 to -3.69; p<0.001) 6 months post-injury. Of the 609 patients recruited, 229 patients were successfully followed at 12 months post-injury. Of these 229 patients, 175 were classified as having strong bonding SC and 54 as having weak bonding SC. Longitudinal analyses showed that there were significant differences in recovery trajectories between individuals with strong bonding SC when compared to those with weak bonding SC with respect to mental (p=0.005) and physical health (p=0.037) (Figure).

Conclusion:  Identifying individuals with deficient close social relationships during their follow-up after trauma may help guide interventions to improve their long-term recovery.

 

18.08 The Association of ABO Blood Groups and Trauma Outcomes

Posted on January 31, 2019

M. W. Sauder1,2, T. Wolff1,3, M. C. Spalding1,2, U. B. Pandya1,2  3OhioHealth Doctors Hospital,Department Of Surgery,Columbus, OHIO, USA 1OhioHealth Grant Medical Center,Division Of Trauma And Acute Care Surgery,Columbus, OH, USA 2Ohio University,Heritage College Of Osteopathic Medicine,Dublin, OH, USA

Introduction:
Certain ABO blood types have been identified as risk factors for a variety of disease processes including acute respiratory distress syndrome, acute kidney injury, myocardial infarction, and venous thromboembolism. However, there is a relative paucity of literature regarding the implications of ABO blood type on characteristics and outcomes of traumatically injured patients. A recent study concluded that blood type O was associated with higher mortality in severely injured patients in Japan. The purpose of this study was to determine the association of ABO blood types with outcomes in traumatically injured patients in the United States.

Methods:
This retrospective study evaluated all category 1 and 2 trauma alerts at an urban, Level 1 trauma center from January 1, 2017 through December 31, 2017. Data was obtained from the institutional trauma database and electronic medical record. Patients were excluded if they were pregnant, less than 16 years old, or if blood type data was unavailable. Recorded outcomes included: ABO blood group, mortality, Injury Severity Score (ISS), race, ventilator days, transfusion requirements, massive transfusion protocol, injury type, mechanism of injury, and complications. Data analysis was performed using descriptive statistics including chi-squared and analysis of variance (ANOVA) calculations. 

Results:
A total of 3,779 patients met inclusion criteria. The proportions of ABO blood types represented by the patients in our sample data were not significantly different than published national averages. Likewise, no significant differences in age, gender, or ISS were present between blood types. Blood type AB was associated with a statistically significant increase in mortality rate in severely injured (ISS>15) Caucasian patients compared to non-AB blood types (39% vs. 16%; p=0.01). This relationship was not consistent among African-American patients (p=0.37). Neither race exhibited differences in hospital length of stay, intensive care unit length of stay, or ventilator days.

Conclusion:
Blood type AB is associated with increased mortality in severely injured Caucasian patients. This is in contrast to findings in Japanese and African American patients. Though this requires further validation, there is a potential correlation between ABO blood type, ethnicity, and trauma outcomes.

18.07 Chemoprophylaxis and Venous Thromboembolism in Traumatic Brain Injury at Different Trauma Centers

Posted on January 31, 2019

E. O. Yeates1, A. Grigorian1, S. D. Schubl1, C. M. Kuza2, V. Joe1, M. Lekawa1, B. Borazjani1, J. Nahmias1  2University Of Southern California,Anesthesia,Los Angeles, CA, USA 1University Of California – Irvine,General Surgery,Orange, CA, USA

Introduction: Patients presenting after severe traumatic brain injury (TBI) are at an increased risk of developing venous thromboembolism (VTE). Due to concerns of worsening intracranial hemorrhage, some clinicians are hesitant to start VTE chemoprophylaxis in this high-risk population. We hypothesized that American College of Surgeons (ACS) verified Level-I trauma centers are more likely to start VTE chemoprophylaxis in adults with severe TBI, compared to Level-II centers. Additionally, we hypothesized that Level-I centers would start VTE chemoprophylaxis earlier and have a lower risk of VTE.

Methods: The Trauma Quality Improvement Program (2010-2016) was queried for patients with a severe grade (>3) for abbreviated injury scale (AIS) of the head. Those that died within 24 hours or had an AIS grade of 6 were excluded. Patients were compared based on the treating hospital: Level-I versus Level-II. A multivariable logistic regression analysis was performed.

Results: From 204,895 patients with severe TBI, 143,818 (70.2%) were treated at a Level-I center and 61,077 (29.8%) at a Level-II center. Compared to severe TBI patients treated at a Level-II center, those at a Level-I center had a lower rate of midline shift >5 mm (26.2% vs. 27.6%, p=0.01), but higher rates of severe AIS grades for the spine (0.9% vs. 0.7%, p<0.001) and lower extremity (0.6% vs. 0.4%, p<0.001). There was no difference in total length of stay (LOS) or intensive care unit (ICU) LOS between the two groups (p>0.05). The Level-I cohort had a higher rate of using VTE chemoprophylaxis (43.2% vs. 23.3%, p<0.001) and shorter median time to chemoprophylaxis (61.9 vs. 85.9 hours, p<0.001), with a lower rate of deep vein thrombosis (4.9% vs. 5.8%, p<0.001) compared to Level-II centers. After controlling for covariates, Level-I centers had a higher likelihood of starting VTE chemoprophylaxis (OR=2.47, CI=2.39-2.56, p<0.001), but had no difference in the risk of VTE (p=0.41) compared to Level-II centers.

Conclusion: ACS Level-I trauma centers were found to be more than twice as likely to start VTE chemoprophylaxis and administered it nearly 24-hours sooner than Level-II centers. However, this did not translate to a decreased risk for VTE events at Level-I centers compared to Level-II centers on a multivariable analysis. Future prospective studies are warranted to evaluate the timing, safety, and efficacy of early VTE chemoprophylaxis in severe TBI patients.
 

18.06 A Novel Method of Quantifying Pain Associated with Rib Fractures

Posted on January 31, 2019

P. Farley1, R. L. Griffin2, J. O. Jansen3, P. L. Bosarge3  1University Of Alabama at Birmingham,Emergency Medicine,Birmingham, Alabama, USA 2University Of Alabama at Birmingham,Epidemiology,Birmingham, Alabama, USA 3University Of Alabama at Birmingham,Acute Care Surgery,Birmingham, Alabama, USA

Introduction:

Rib fractures are a major problem characterized by pain, which is poorly understood. Measuring total pain experience, taking into account the duration as well as intensity, could facilitate comparisons of different treatments. The aim of this study was to evaluate the feasibility of quantifying pain reported over the course of an admission and to identify factors associated with more pain in patients with rib fractures.

Methods:

Patients admitted to a Level-I academic trauma center with rib fracture or flail chest between 2015 and 2017 were included. For each patient, the maximum pain score (verbal or non-verbal) was calculated for each hospital day. Total pain reported was defined as the sum of the area under the curve (AUC) of the max pain scores plotted against time. The AUC was calculated based on the trapezoidal rule. A general linear model was used to determine demographic, injury, and clinical predictors of the pain AUC. In a post-hoc analysis, models were stratified by Injury Severity Score categories (i.e., 1-8, 9-14, 16-75) to determine whether predictors differed by injury severity.

Results:

We identified 3,713 patients. Increased pain experience was observed for those aged 40-59 years compared to those 18-39 years (B=6.1, p=0.002); ISS 9-14 (b=11.5, p<0.001) and ≥16 (B36.9, p<0.0001) compared to ISS 1-8; patients with flail chest compared to those with multiple rib fractures (B=17.1, p<0.001); and patients who underwent rib fixation (B=20.7, p=0.004). Decreased pain experience was observed for male gender (B=-3.7, p=0.032) and blunt mechanism of injury (B=-13.7, p<0.0001). Associations were broadly similar when the analysis was stratified by overall injury severity, though blunt mechanism was not associated with pain among ISS 1-8 and age ≥60 was associated with decreased pain among those with ISS 16-75 (B=-7.4, p=0.026).

Conclusion:

This study demonstrates the feasibility and usefulness of measuring patients’ total pain experience over the duration of their admission. As reported, pain—a product of pain and analgesia—is a subjective but highly relevant measure of patients’ experience. Our study identifies a number of predictive factors, some expected (such as overall injury severity, reflecting additional injuries) and some unexpected. Increased overall pain experience following fixation may be the result of severe pain prior to intervention. We are planning further work to advance the concept of total reported pain experience.
 

18.04 Novel Oral Anticoagulants vs LMWH for Thromboprophylaxis in Operative Pelvic Fractures

Posted on January 31, 2019

M. Khan1, J. Con1, F. Jehan1, R. Latifi1  1Westchester Medical Center,Surgery,Valhalla, NY, USA

Background: Pelvic fractures have been identified as a risk factor for venous thromboembolic (VTE) complications. Recent literature shows the superiority of novel oral anticoagulants (NOACs) over low molecular weight heparin (LMWH) for thromboprophylaxis in orthopedic patients. The aim of our study was to evaluate the impact of NOACs vs. LMWH on outcomes in patients with operative pelvic fractures.

Methods: We performed a 2-year (2015-16) analysis of the ACS-TQIP database. We included all adult patients with isolated blunt pelvic fractures who were managed operatively and received post-operative thromboprophylaxis with either LMWH or NOACs (Factor Xa inhibitor and direct thrombin inhibitor). Patients were stratified into two groups based on the type of thromboprophylactic agent (NOACs vs. LMWH) and were matched in a 1:2 ratio for demographics, admission vitals, injury parameters, hospital stay, facility, and timing of initiation of thromboprophylaxis. Primary outcomes were rate of DVT and/or PE. Secondary outcomes were pRBC transfusions, and intervention for hemorrhage control after initiation of thromboprophylaxis. 

Results: We analyzed 11,219 patients with pelvic fractures. A total of 3,529 patients with isolated pelvic fractures were included of which 708 patients were matched (NOACs: 236; LMWH: 472). Mean age was 61±12 and median ISS was 12[10-16]. Matched groups were similar in demographics, vitals and injury parameters, hospital stay, and timing of initiation of thromboprophylaxis. Overall 5.8% of patients had DVT, and 1.8% PE%. Patients who received NOACs were less likely to develop DVT (2.9% vs. 7.2%, p=0.01) compared to LMWH. There was no difference in PE (1.6% vs. 1.9%, p=0.28) between the two groups. Similarly there was no difference in post prophylaxis blood products transfusion, and post-prophylaxis intervention for hemorrhage control(Table 1.).

Conclusion: In patients with operative pelvic fracture, thromboprophylaxis with novel oral anticoagulant is associated with lower rate of DVT. There was no association between type of thromboprophylactic agent with PE. Further prospective clinical trials should evaluate the role of NOACs for thromboprophylaxis in high-risk trauma patients.

18.03 Repeat CT Head is Not Indicated for Patients on Novel Anticoagulants with Mild TBI

Posted on January 31, 2019

C. Cohan1, G. Beattie1, D. Dominguez1, M. Cochran1, B. Palmer1, G. P. Victorino1  1University of California San Francisco – East Bay,Department Of Surgery,Oakland, CA, USA

Introduction:  Guidelines regarding management of patients on anticoagulation suffering from mild blunt traumatic brain injury (TBI) are unclear. Repeat CT head (CTH) is commonly performed on anticoagulated patients after an initial negative CTH to assess for delayed intracranial hemorrhage (ICH-d). The reported incidence of ICH-d in anticoagulated patients is 0.51-6%. Current literature focuses on patients taking prehospital warfarin. Since approval of several novel oral anticoagulants (NOACs) including rivaroxaban, apixaban, and dabigatran, there has been a steady increase in trauma patients presenting on these medications. The rate of ICH-d in this patient population is unknown. We hypothesized that the incidence of ICH-d in patients on NOACs would be low, similar to patients on warfarin, and that routine repeat CTH after initial negative CTH in patients on NOACs is not indicated. 

Methods:  We performed a retrospective chart review of all adult patients on anticoagulation presenting to our level I trauma center for blunt trauma between February 2016 and May 2018. We excluded patients who had a positive initial CTH, who did not have a repeat CTH within 24 hours of initial scan, or who had a GCS <13 on arrival. CTH was repeated 4-12 hours after initial CTH. Clinical outcomes including ICH-d, discharge GCS, neurosurgical intervention, readmission, and death were assessed. Comparisons were made using a chi squared test or ANOVA. Data are presented as mean ± standard error of the mean.

Results: A total of 218 patients met inclusion criteria with an average initial GCS of 14.7 ± 0.03. The following groups were evaluated: warfarin only (n=133), NOAC only (n=68), and anticoagulation (warfarin or NOAC)-ASA combination regimen (n=17). The average INR for each group was 2.7 ± 0.24, 1.4 ± 0.03, and 2.2 ± 0.26, respectively (p<0.01). Average age in years for each group was 78.3 ± 1.1, 75.7 ± 0.86, and 72.9 ± 3.0, respectively (p=0.2). The overall incidence of ICH-d after initial negative CTH was 2.5%. The incidence in the NOAC only group was 1.5% (1/68) vs. 2.3% (3/133) in the warfarin only group, (p=0.71). In the warfarin group, 66% (2/3) of patients with ICH-d had a supratherapeutic INR on presentation. There were no ICH-d events in the anticoagulation-ASA combination group.

Conclusion: To our knowledge, this is the largest study of patients on novel anticoagulants to assess ICH-d in mild TBI. We found that the incidence rates of ICH-d are similar between patients taking NOACs and warfarin. In the one patient with ICH-d on novel anticoagulation, no neurosurgical intervention, decline in GCS, readmission or death occurred. Our findings suggest there is no indication for repeat imaging in patients on novel anticoagulation presenting with mild TBI. Limiting unnecessary imaging in this substantial and growing population may save time, reduce costs, and improve allocation of resources. 

18.02 Role of Extracorporeal Membrane Oxygenation (ECMO) In Trauma Patients: A Five-Year Analysis

Posted on January 31, 2019

W. Chen1, M. Zeeshan1, E. Zakaria1, N. Kulvatunyou1, T. O’Keeffe1, L. Gries1, A. Tang1, A. Northcutt1, B. Joseph1  1University Of Arizona,Trauma And Acute Care Surgery,Tucson, AZ, USA

Introduction:
Extracorporeal-membrane-oxygenation (ECMO) has been utilized in the neonatal respiratory distress syndrome. However, its role in trauma is evolving. The aim of our study was to evaluate the survival, trends of utilization and identify factors associated with mortality after ECMO in trauma patients over a 5-year period.

Methods:
We performed a 5-year (2008-2012) review of all trauma patients in the NTDB. We included all patients who underwent ECMO. Our primary outcome measures were trends of utilization and survival after ECMO in 5-year study period. Secondary outcome measures were in-hospital complications and factors associated with mortality after ECMO. Regression and trend analysis were performed.

Results:
Of the 808,211 trauma patients identified, 179 patients underwent ECMO. Mean age was 33y±15, 80% were male, and median ISS was 22[13-33]. Mechanism of injury (MOI) was blunt in 82%. Overall mortality was 34% (penetrating: 21% vs. blunt: 37%, p=0.03), and 31% (56/179) were discharged to SNF. The utilization of ECMO increased from 13.9 to 32.2 per 100,000 trauma admissions (p=0.01) while mortality rate decreased from 46.2% to 24% (p<0.001) during the study period (Figure 1). 47.5% of patients who received ECMO had ARDS, 45.7% had pneumonia, 32.1% had cardiac arrest, 22.8% developed AKI and 14.8% developed sepsis. On regression analysis, increasing age (OR: 1.12[1.03-1.25]) and ISS (OR 1.17[1.03-1.26]) were independently associated with mortality while centers that were performing multiple ECMO were associated with improved mortality (OR: 0.55[0.12-0.82]). 

Conclusion:
Two thirds of patients who undergoes ECMO survived. The use of ECMO in trauma patients has increased while the mortality rate has declined. Patients who undergo ECMO in trauma centers that frequently perform ECMO tend to do better. Further studies are required to better define the role of ECMO in trauma and identify the subset of population that may benefit from this procedure. 
 

18.01 Are We Out of the Woods Yet? The Aftermath of Resuscitative Thoracotomy

Posted on January 31, 2019

J. L. Fitch1,3,4, S. Dieffenbaugher2, M. McNutt2, C. C. Miller1, D. J. Wainwright2, J. Villarreal1, Q. Zhang1, G. Hall1, S. Gordy1, J. Ward1, C. Wilson1, J. Suliburk1, M. A. Davis1, S. R. Todd1  1Baylor College Of Medicine,General Surgery,Houston, TX, USA 2McGovern Medical School at UTHealth,General Surgery,Houston, TX, USA 3Naval Medical Center Portsmouth,General Surgery,Portsmouth, VA, USA 4Uniformed Services University Of The Health Sciences,General Surgery,Bethesda, MD, USA

Introduction:  Survival following traumatic cardiac arrest is low, but resuscitative thoracotomy (RT) is lifesaving for select patients. Data exists on those who are likely to survive RT but is limited regarding hospital course and prognosis following admission to the intensive care unit (ICU). The objective of this study was to describe the hospital course and prognosis for RT survivors admitted to the ICU.

Methods:  This was a retrospective review of all adult trauma patients who underwent RT following traumatic arrest at the only two level one trauma centers serving our metropolitan area. Data evaluated include patient demographics, injury characteristics, hospital course, and outcome.   

Results:  Over 66 months ending June 2017, there were 52,624 trauma activations for both centers. 298 (0.6%) patients underwent RT, and 96 of these (32%) survived to ICU admission. Of these initial survivors, the mean age was 35.8±14.5 years.  79 (82%) were male, 36 (38%) sustained blunt trauma, and the mean injury severity score was 32.3±13.7. 67% of deaths in the ICU occurred within the first 24 hours of admission. 90% of those alive at day 21 survived to discharge. Of those admitted to the ICU, 22% of blunt and 34% of penetrating patients survived to discharge. The mean ICU length of stay (LOS) for survivors was 24.1±17.9 days, while the mean hospital LOS was 43.9±32.1 days. Survivors averaged 1.9±1.5 complications; most commonly acute kidney injury, deep surgical site infection, and deep vein thrombosis. 24 of 28 patients surviving to discharge went home or to a rehabilitation center.

Conclusion:  Survival following RT is 9.4%, but there is an increased likelihood of survival with each day the patient remains alive. Families should be counseled to expect a long hospital course with a high likelihood of complications. The overall prognosis for survivors of RT may not be as bleak as previously assumed.

 

17.20 ROTEM as a Predictor of Mortality in Trauma Patients with an Injury Severity Score Greater than 15.

Posted on January 31, 2019

A. R. Smith1, S. Karim1, R. J. Reif1, W. C. Beck1, J. R. Taylor1, B. Davis1, A. Bhavaraju1, M. K. Kimbrough1, K. W. Sexton1  1University Of Arkansas for Medical Sciences,Department Of Surgery, Division Of Trauma And Acute Care,Little Rock, AR, USA

Introduction:  The Injury Severity Score (ISS) is an important tool for grading the severity of injury to trauma patients. Major trauma is commonly defined using an ISS threshold of 15 and has been shown to correlate with mortality likelihood, length of hospital stay, and the need for major surgery. Assessing hemostatic function in a timely manner is crucial for these patients in order to reduce the risk of mortality. Rotational thromboelastometry (ROTEM) is a whole blood viscoelastic hemostasis analyzer that allows for the detection of and differentiation between coagulopathies. The purpose of our study is to retrospectively evaluate trauma patients with an ISS greater than 15 who also received ROTEM to determine if ROTEM is a better predictor of mortality than conventional coagulation testing. 

Methods:  We performed a retrospective review of all trauma patients with ROTEM and an ISS greater than 15 admitted to the emergency department between November 2015 and August 2017. A total of 301 patients met the aforementioned criteria and were included in this study. Univariate and bivariate statistics were performed using JMP Pro (Cary, NC). Each patient was sorted into groups based on their coagulation phenotype (hypocoaguable, normal coagulation, hypercoaguable) for both ROTEM and conventional coagulation tests (Partial Thromboplastin Time [PTT], Prothrombin Time [PT], International Normalized Ratio [INR]) and compared the test results to their relation with mortality. Nominal logistic regression was performed.

Results: For the 301 patients included in the study, significant predictors of mortality included ROTEM APTEM Clotting Time (CT), ROTEM APTEM Alpha Angle, and ROTEM INTEM Clot Formation Time (CFT), with ROTEM APTEM CT being the most significant. On nominal logistic regression, APTEM CT (p=.007), APTEM Alpha Angle (p=.028), and INTEM CFT (p=.037) were the only significant predictors. PTT (p=.059), PT (p=.141), and INR (p=.634) were not significant predictors of mortality in this data set. 

Conclusion: ROTEM APTEM Clotting Time, which is the time from start of measurement until initiation of clotting in the presence of aprotinin, a fibrinolysis inhibitor, is a significant predictor of mortality in trauma patients with an ISS greater than 15. ROTEM APTEM Alpha Angle and ROTEM INTEM Clot Formation Time are also significant predictors of mortality, whereas conventional coagulation tests did not have a significant contribution to predicting mortality in this patient population.
 

17.19 Back to the Basics: Trauma Team Assessment and Decision Making is Associated with Improved Outcomes

Posted on January 31, 2019

M. A. Vella1, R. Dumas1,2, K. Chreiman1, M. Subramanian1, M. Seamon1, P. Reilly1, D. Holena1  1University Of Pennsylvania,Traumatology, Surgical Critical Care And Emergency Surgery,Philadelphia, PA, USA 2University Of Texas Southwestern Medical Center,General And Acute Care Surgery,Dallas, TX, USA

Introduction:  Teamwork and decision making are critical elements of trauma resuscitation. While assessment instruments such as the non-technical skills (NOTECHS) tool have been developed, correlation with patient outcomes is unclear. Using emergency department thoracotomy (EDT) as a model, we sought to describe the distribution of NOTECH scores during resuscitations. We hypothesized that patients undergoing EDT whose resuscitations had better scores would be more likely to have return of spontaneous circulation (ROSC).

Methods:  Continuously recording video was used to review all captured EDTs during the study period. We used a modification of the NOTECH instrument to measure 6 domains (leadership, cooperation/resource management, communication/interaction, assessment/decision making, situation awareness/coping with stress, and safety) on a 3-point scale (1 = best, 2 = average, 3 = worst).  For each resuscitation, an overall total NOTECH (6-18 points) score was calculated. The primary outcome metric was ROSC. Associations between demographic, injury, and NOTECH variables and ROSC were examined using univariate regression analysis.

Results: 61 EDTs were captured during the study period. 19 patients had ROSC (31%) and 42 (69%) did not. The median NOTECH score for all the resuscitations was 9 [IQR 8-11]. As demographic and injury data (age, gender, mechanism, signs of life) were not associated with ROSC in univariate analysis, they were not considered for inclusion in a multivariable regression model between NOTECH scores and ROSC.  The association between overall NOTECH score and ROSC did not reach statistical significance, p=0.09, but examination of the individual components of the NOTECH score (Table 1) demonstrated that compared to resuscitations that had “average” (2) or “worst” (3) scores on “Assessment and Decision Making,” resuscitations with a “best” score were 5.3x more likely to lead to ROSC, p=0.017 (OR 5.3, CI 1.2-31.9).

Conclusion: While the association between overall NOTECH scores and ROSC did not reach statistical significance, assessment and decisions making did.  In patients arriving in cardiac arrest who undergo EDT, better team performance is associated with improved rates of ROSC.  Future analysis of the timing and quality of elements of resuscitation using video review may elucidate the mechanistic underpinnings of these findings.

 

17.18 Level 1 Trauma Surgeon Staffing: Is more really better?

Posted on January 31, 2019

A. Ansari1, A. Kothari1, E. Eguia1, M. Anstadt1, R. Gonzalez1, F. Luchette1, P. Patel1  1Loyola University Chicago Stritch School Of Medicine,Department Of Surgery,Maywood, IL, USA

Introduction:
Trauma is the fourth leading cause of death in the United States. Care in level 1 trauma centers is associated with improved outcomes and the determinants of this relationship continue to be studied. The objective of this study was to determine if the number of trauma surgeons on staff at level 1 trauma centers impacted outcomes.

Methods:
This study utilized data from the American College of Surgeon’s (ACS) National Trauma Data Bank (NTDB) for years 2013-2016. Inclusion criteria was set as all patient presenting to a Level 1 trauma centers with severe traumatic injuries defined as an Injury Severity Score (ISS) of 15 or greater. The primary outcome was patient survival. A multivariable logistic regression model was constructed to estimate the adjusted effect of trauma surgeon staffing on the primary outcome.

Results:
A total of 180,999 encounters were included in this study. Injured patients that received care at a trauma center with less than 4 staff surgeons had a mortality of 16.0% vs those at a trauma center with > 4 surgeons 12.4% (P=0.01). After controlling for injury severity, age, sex, and race, the odds of mortality were 0.70 (95% CI 0.53 – 0.92) comparing high staff to low staff centers. Secondary outcomes, including length of stay, ventilator time, and ICU length of stay did not differ based on trauma center staffing.

Conclusion:
Current ACS requirements for trauma surgeon staffing at Level 1 trauma centers require that there be a minimum of one trauma surgeon per center. Based on our evaluation, there seems to be clinical improvement in outcomes when a center has 4 or greater trauma surgeons on staff. This warrants further evaluation at the requirements for trauma surgeon staffing at level 1 trauma centers.
 

« Previous Page
Next Page »