17.04 National evaluation of adherence to quality measures in esophageal cancer

Posted on January 18, 2018

A. Adhia1, J. Feinglass1, K. Engelhardt1, M. DeCamp1, D. Odell1  1Northwestern University,Chicago, IL, USA

Introduction: Esophageal cancer is the leading cause of death among GI malignancies and the incidence of the disease is rising faster than any other solid organ tumor. Patients frequently present with locally advanced disease (stage III), contributing to challenges in treatment decision making.  Our objective was to assess adherence to four novel quality measures in patients with stage III esophageal cancer.

Methods:  18,555 patients diagnosed with stage III esophageal cancer were identified from the National Cancer Database (NCDB) between 2004 and 2014.  Four quality measures were defined from NCCN guidelines: administration of induction therapy, >15 lymph nodes sampled at resection, surgery within 120 days of neoadjuvant treatment, and R0 resection.  The association of patient demographic and treatment variables (age, sex, location of lesion, histology, income, education, race and ethnicity and year of diagnosis) with measure adherence was assessed using logistic regression. Risk of all-cause mortality was assessed comparing adherent and non-adherent cases using Cox modeling.  Kaplan-Meier survival estimates of groups that adhered to none, one of four, two of four etc. quality measures were performed.

Results: Adherence was high for three of the quality measures: neoadjuvant treatment (92.7%), timing of surgery (82.5%) and completeness of resection (91.5%).  However, nodal evaluation was adequate in only a minority of patients (20.0%). Advanced age, Medicaid insurance status, lower level of education and Black or Hispanic ethnicity were all associated with statistically significant increased odds of non-adherence for all measures.  Adherence improved in the more recent time period, with cases after 2008 having improved adherence in administration of induction therapy (OR = 2.58 in 2012-2014 period) and adequate nodal staging (OR = 2.49 in 2012-2014).  Achieving adherence was associated with a statistically significant decrease in all-cause mortality for administration of induction therapy (HR = 0.70 [0.62, 0.78]), nodal staging (HR = 0.67 [0.63, 0.70]), and R0 resection (HR = 0.48 [0.43, 0.53]), but not for timing of surgery (HR = 0.93 [0.85, 1.02]).  Survival improved as the number of quality measures an individual patient adhered to increased (Figure).

Conclusion: Adherence to quality measures in the care of patients with stage III esophageal cancer is associated with improved survival.  Understanding variability in measure adherence may identify potential targets for cancer quality improvement initiatives.

17.01 Effect of preoperative liquid diet on liver volume and MRI estimated proton density fat fraction

Posted on January 18, 2018

T. SUZUKI1, R. B. Luo1, J. C. Hooker2, Y. Covarrubias2, T. Wolfson2, A. Schlein2, S. Liu1, J. B. Schwimmer3, L. M. Funk5, J. A. Greenberg5, G. M. Campos6, B. J. Sandler1, S. Horgan1, S. B. Reeder4, C. B. Sirlin2, G. R. Jacobsen1  1University Of California – San Diego,Division Of Minimally Invasive Surgery, Department Of Surgery,San Diego, CA, USA 2University Of California – San Diego,Liver Imaging Group, Department Of Radiology,San Diego, CA, USA 3University Of California – San Diego,Division Of Gastroenterology, Hepatology, And Nutrition, Department Of Pediatrics,San Diego, CA, USA 4University Of Wisconsin,Departments Of Radiology, Medical Physics, Biomedical Engineering, Medicine And Emergency Medicine,Madison, WI, USA 5University Of Wisconsin,Department Of Surgery,Madison, WI, USA 6Virginia Commonwealth University,Division Of Bariatric And GI Surgery,Richmond, VA, USA

Introduction: Liver volume (LV) and fat content are important considerations during bariatric procedures as increased liver volume not only increases the difficulty of intra-operative visualization but also elevates the risk of bleeding complications. The aim of this study was to evaluate the impact of a preoperative liquid diet (PLD) on LV and magnetic resonance imaging (MRI) estimated proton density fat fraction (PDFF) as a measure of liver fat content, in morbidly obese patients undergoing bariatric surgery (BS). 

Methods: This prospective multi-institutional study was approved by an institutional review board (IRB) and was Health Insurance Portability and Accountability Act (HIPAA) compliant. After providing informed consent, patients meeting National Institutes of Health (NIH) criteria for BS underwent MRI at baseline and post PLD. LV and PDFF were estimated from 3D chemical shift encoded MRI (CSE-MRI) anatomical images and PDFF maps, using the OsiriX (Pixmeo SARL, Bernex, Switzerland) imaging software. Primary outcomes were patient weight, body mass index (BMI), LV and PDFF. Secondary outcomes were relationships between the changes in BMI, LV and PDFF. Data were analyzed with paired t-test and Wilcoxon-Mann-Whitney tests. Pearson correlation was used to assess the relationships between measures. Relative reduction rate of BMI was defined as: (baseline BMI – post BMI) / baseline BMI ×100 (%). Relative reduction rate of LV was defined as: (baseline LV – post LV) / baseline LV ×100 (%). The absolute reduction rate of PDFF was defined as: baseline PDFF ?post PDFF  (%).

Results:One-hundred-twenty-four patients scheduled for BS were recruited to be part of the study between October 2010 and June 2015. 102 patients (87 females, 85.3%, mean age 48.0 ± 12.8 years) underwent MRI at baseline and post PLD. The mean liquid diet duration was 17.1 ± 8.8 days. Post PLD, mean weight decreased from 119.6 ± 19.1 kg/m2 to 114.8 ± 18.7 kg/m2 (p<0.0001). BMI decreased  from 43.6 ± 6.4 kg/m2 to 41.9 ± 6.3 kg/m2 (p<0.0001) with a mean relative reduction of 4.1 ± 2.2 %. LV decreased from 2277.2 ± 578.0 cm3 to 1985.0 ± 510.6 cm3 (p<0.0001) with a mean relative reduction of 12.3 ± 10.1 %. PDFF decreased from 13.6 ± 9.4 % to 10.4 ± 7.8 % (p<0.0001) with a mean absolute reduction of 3.2 ± 4.3 %. Pearson correlations analyses revealed statistically significant relationships between the relative reductions in LV and BMI (r=0.5253, p≤0.0001), between the absolute reduction in PDFF and relative reduction in BMI (r=0.2451, p=0.0140), and between the absolute reduction in PDFF and relative reduction in LV (r=0.3861, p=0.0001).

Conclusion:PLD significantly reduced LV and PDFF. This highlights the importance of PLD in the improvement of LV and PDFF in morbidly obese patients and underscores the reason why PLD is routinely performed at our institutions.

16.19 Outcomes and Hospital Resource Utilization in Older Adult Patients After Motor Vehicle Crashes

Posted on January 18, 2018

P. P. Patel1, L. Gryder1, C. McNicoll1, C. Katona1, P. McGrew1, P. Chestovich1, J. Fildes1, D. Kuhls1  1University Of Nevada,Trauma & Critical Care,Las Vegas, NEVADA, USA

Introduction: As the average life expectancy increases, more older adults continue to drive or travel as passengers in motor vehicles. Crashes involving the elderly have worse outcomes compared to younger patients. The purpose of this study is to describe the injury burden, hospital resource utilization (HRU), hospital charges, and disposition incurred by older adult patients after a motor vehicle crash (MVC).

Methods: The Statewide Vehicular Trauma Database was queried for all individuals age ≥65 injured in a MVC from 2005-2014. Patients were stratified by age: 65-74, 75-84, and ≥85. Relevant data include demographics, crash factors, and injury severity score (ISS). Primary outcome was hospital mortality, with secondary outcomes of hospital and ICU length of stay (LOS), hospital charges, and disposition. Analysis was by Chi-squared and Kruskal-Wallis tests, with p<0.05 considered significant.

Results: A total of 2,029 individuals met inclusion criteria. The average age was 75, majority were Caucasian, restrained, and seated in the driver position. Gender distribution was equal. Injury and HRU was significant for a higher average ISS and an increased mean number of hospital and ICU days in the 75-84 age group. Upon nonparametric analysis, the ≥85 group showed significantly increased ISS, hospital and ICU LOS, and hospital charges.  Patients ≥85 were also more likely to die or require disposition to a rehab facility or a nursing home after discharge.

Conclusion: This study demonstrates that although the older adult population is considered a high-risk group, there are significant differences in injury burden, outcomes and HRU within this cohort. Older adults had greater injury severity requiring a higher resource utilization while achieving less desirable outcomes. As the number of older adult trauma patients grows, special attention should be placed on those over age 85 to enhance their recovery after a MVC.

 

16.20 The Influence of Pancreatic Division Technique on Pancreatic Leak Rates Following Traumatic Injury

Posted on January 18, 2018

P. Hu1, R. Uhlich1, J. Kerby1, P. Bosarge1  1University Of Alabama at Birmingham,Acute Care Surgery,Birmingham, Alabama, USA

Introduction:
Pancreatic injury is a rare, potentially devastating consequence of abdominal trauma. While low grade injuries may be successfully managed conservatively, injuries to the pancreatic duct or head typically require operative intervention. Complications following pancreatic resection are historically associated with high rates of morbidity and mortality. We sought to evaluate the influence of intra-operative techniques on postoperative complications.

Methods:
A retrospective case control study was performed at an American College of Surgeons verified level 1 trauma center from 2011-2017. All adult patients admitted to the trauma surgery service were eligible for inclusion, while patients with pregnancy or age <18 years old were excluded. Patients with pancreatic injury were identified from the trauma registry using ICD-10 codes. Pancreatic injuries were graded according to the AAST guidelines (Grades 1-5), with major injury identified as ≥ grade 3 (pancreatic ductal injury). Patients were stratified into cohorts according to the method used for pancreatic division and resection, including stapled, cut and oversewn, stapled and oversewn, or cautery. Pancreatic leak was defined as a drain amylase level three times greater than normal serum amylase (103 U/L), according to institutional standard. Analysis was performed using χ2 and Student's t-test or one-way ANOVA for categorical and continuous variables respectively. The primary outcome of interest was the rate of pancreatic leak following resection. 

Results:

52 patients were identified with pancreatic injury[PLB1] . The majority of patients (90.4%) underwent operative management. Pancreatic resection was required in 40.4% (21/52), with a majority undergoing stapled resection (52.3%). The remaining resections were performed by cut and oversew (14.3%), stapled and oversewn (23.8%), and cautery (9.5%). Pancreatic leak was identified postoperatively in 76.2% (16/21) of patients, with no significant difference between the different types of resection (p=0.27).

 

Conclusion:

Pancreatic injury requiring resection results in significant rates of postoperative leak, regardless of intraoperative technique. Drain placement should be strongly considered at the time of initial operation

16.17 Clinical Outcomes In Patients Requiring Dialysis After Trauma: A National Trauma Database Analysis

Posted on January 18, 2018

A. E. Siletz1, J. Grotts2, C. E. Lewis1, A. Tillou1, H. Cryer1, A. Cheaito1  1University Of California – Los Angeles,Department Of Surgery, David Geffen School Of Medicine At UCLA,Los Angeles, CA, USA 2University Of California – Los Angeles,UCLA Department Of Medicine Statistics Core, David Geffen School Of Medicine At UCLA,Los Angeles, CA, USA

Introduction: AKI requiring renal replacement therapy (RRT) represents the most severe form of post-traumatic AKI, and has been associated with poor outcomes.  Incidence and clinical impact vary by study due to variations in study population and definitions. The objective of this study was to determine the correlation between initiating dialysis and clinical outcomes in trauma patients using a national dataset.  

Methods:  All patients in the National Trauma Database from 2013-2014 with a diagnosis of AKI based on ICD9 code during admission for trauma were reviewed. Patients were excluded if they had no signs of life on arrival, were under age 18, or had pre-existing end-stage renal disease. A propensity score based on ISS, penetrating injury, age, and gender was used to match patients with AKI requiring dialysis with those with AKI who did not need dialysis.  A multivariate logistic regression model using dialysis, ISS, injury type, age, and gender as covariates was also constructed. 

Results: Among adult patients surviving to admission for trauma without pre-existing end-stage renal disease, the incidence of AKI was 1.07%  and the incidence of AKI requiring dialysis was 0.02%.  17668 trauma patients with AKI, of which 282 received dialysis, were compared.  Older age, male gender, black/African American race and Medicare and Medicaid coverage were significantly associated with dialysis (p<0.02). Penetrating injury was more likely to be associated with dialysis than blunt injury (OR 3, 95% CI 2.3-4, p<0.001) and dialysis patients had higher median ISS scores (26.5, IQR 18-35.2 vs. 17, IQR 9-29, p <0.001).  When patients were matched using a propensity score based on ISS, penetrating injury, and age, dialysis patients were found to have higher complication rates including unplanned intubation (OR 3, 95% CI 1.6-5.6, p <0.001), unplanned return to the operating room (OR 7.3, 95% CI 3.8-14, p <0.001), acute lung injury/acute respiratory distress syndrome (OR 4.7, 95% CI 3-7.3, p <0.001), pulmonary embolism (OR 3.1, 95% CI 1.3-7.2, p =0.013), severe sepsis (OR 11.3, 95% CI 6.4-19.9, p <0.001), myocardial infarction (OR 4, 95% CI 1.5-10.7, p =0.009), and death (OR 3.8, 95% CI 2.7-5.2, p <0.001).  Median hospital stay (27 vs. 8 days, p <0.001), ICU stay (19 v. 5 days, p < 0.001), and number of ventilator days (16 vs 5 days (p < 0.001) were significantly higher for dialysis patients.  In a multivariate logistic regression model, initiating dialysis was significantly associated with development of acute respiratory distress syndrome (OR 4.8, 95% CI 3.1-7.6, p < 0.001), severe sepsis (OR 12.2, 95% CI 7.0-22.2, p < 0.001), and mortality (OR 4.0, 95% CI 2.9-5.6, p<0.001).

Conclusion: 

AKI requiring dialysis after trauma is rare. Risk factors include high ISS and penetrating injury.  The need for dialysis after AKI during admission for trauma is associated with increased complications, length of hospital stay, and mortality. 

 

16.18 Current Nutritional Practices and Associated Outcomes in Critically-Ill Trauma Patients

Posted on January 18, 2018

B. E. Haac1, R. Van Besien1, R. Jenkins1, A. Geyer2, J. Diaz1, D. Stein1  1University Of Maryland,R Adams Cowley Shock Trauma Cener,Baltimore, MD, USA 2Air Force Institute Of Technology (AFIT/ENC),Wright-Patterson AFB, OHIO, USA

Introduction: Nutrition is an important component of support for critically-ill trauma patients who often present in a state of catabolic stress but there is limited recent research on this topic specific to trauma patients. We sought to examine nutritional practices in a critically-ill trauma population and to identify baseline factors and outcomes associated with timing, content and route of nutrition.

Methods:  We conducted a retrospective review of adult critically-ill trauma patients admitted to the intensive care unit (ICU) for >72 hours. A multivariable regression model was built for each nutritional variable and outcome variable. Outcomes evaluated include number of operative trips, hospital and ICU length of stay (LOS), ventilator days, mortality, discharge destination and hospital-acquired infections.

Results: 683 patients (mean ISS 24.4) were included. 461 patients received tube feeds within the first 7 days of admission. Two-thirds (n=297, 64%) of these were initiated on early enteral tube feeding within 48 hours. Injury pattern was associated with timing of nutrition initiation, time to goal tube feed rate and percent of goal calories and protein received. Specifically, severe head injury (brain AIS=5) was independently associated with reaching a goal tube feed rate (aOR 3.1, P<0.01) and receiving a greater percent of goal calories (aOR 2.1, p=0.01) in the first 48 hours of admission whereas patients without head injury (brain AIS=0) were less likely to initiate nutrition within 24 hours of admission (aOR 0.3, p<0.01). Higher admission GCS was also associated with decreased odds of reaching goal tube feeds within 24 hours (aOR 0.6, p<0.01). Later initiation of enteral nutrition after 48 hours was associated with increased ICU LOS (aOR 1.4, p<0.01) and more ventilator days (aOR 1.6, p<0.01) in all patients and increased pneumonia rates in patients who required gastrointestional (GI) surgery for their injury (aOR 15.7, p=0.02). Increased percent of goal nutrition received in the first 48 hours was associated with more ventilator days (aOR 2.8, p<0.01) and longer ICU LOS (aOR 1.7, p<0.01). Increased percent of goal nutrition received in the first 7 days was associated with development of urinary tract infection (UTI) (aOR 5.4, p<0.01). Gastric tube feeds were associated with lower bacteremia incidence (aOR=0.4, p=0.01) and longer ICU LOS (aOR 1.2, p<0.01). There was no association of nutrition variables with mortality.

Conclusion: Injury pattern, especially presence of brain injury, may be predictive of ability to initiate early enteral nutrition, time to goal feeds and percent of goal nutrition received. Benefits of early initiation may include decreased LOS and ventilator days and reduced pneumonia rates in patients requiring GI surgery. Trophic feeds may be less likely to result in UTI, and gastric feeds may have a protective role in prevention of bacteremia.

 

16.14 Thyroid Trauma − Incidence, Mortality, and Concomitant Injury

Posted on January 18, 2018

D. Spencer1, A. Grigorian1, S. Schubl1, K. Awad1, D. Elfenbein1, T. Dogar1, J. Nahmias1  1University Of California – Irvine,Division Of Trauma, Burns & Surgical Critical Care,Orange, CA, USA

Introduction:  Traumatic injury to the thyroid is rare with no large studies previously published. Although the thyroid is not considered an immediately vital structure, it is in close proximity to several critical structures such as the carotid arteries, trachea, esophagus, and cervical spine. We sought to describe the national incidence of traumatic thyroid injury as well as mortality rate, rate of operative intervention, and frequencies of concomitant injury to surrounding structures. We hypothesized isolated thyroid injury would have a lower mortality compared to thyroid with concomitant carotid artery, trachea, esophagus or cervical spine injury.

Methods:  National Trauma Data Bank data from 2007-2015 was used to identify patients with thyroid injury. Demographics, associated injuries, operative repair, complications, and outcomes were analyzed. Analysis was performed by odds ratio utilizing a logistic regression model.

Results: 1,395 patients with thyroid injury were identified from over 6.7 million trauma patients. Yearly incidence was 0.02%. The majority of patients were female (79.6%), had a penetrating mechanism of injury (79.7%), and had isolated thyroid injury (59.7%). The most common concomitant injuries were to the trachea (25.9%), carotid artery (9.5%), and cervical spine (7.9%). Operative interventions most frequently performed were direct thyroid repair (13.9%), thyroid blood vessel repair (3.4%), and thyroidectomy (3.2%). No patients experienced postsurgical hypothyroidism. All-cause mortality was 15.6%. After controlling for age  ≥  65, ISS > 25, and gender, non-isolated thyroid injury was shown to be an independent risk factor for mortality when compared to isolated thyroid injury (Odds Ratio 1.67, 95% Confidence Interval 1.17 – 2.34; p<0.05).

Conclusion: Thyroid injury in trauma patients is extremely rare. Interestingly, thyroid trauma is seen more often in females than males. Isolated thyroid trauma presents less of a clinical challenge with a lower risk of mortality than those with concomitant injuries even after controlling for significant covariates. When operative intervention is required, direct thyroid repair is greater than four times more common than thyroidectomy. Regardless of injury type and operation, postsurgical hypothyroidism was not seen.

16.15 Using Injury Severity Score to Determine Venous Thromboembolism Risk in Trauma Patients

Posted on January 18, 2018

T. E. Hereford1, S. Ray1, R. D. Robertson1, M. K. Kimbrough1  1University Of Arkansas For Medical Sciences,Little Rock, AR, USA

Introduction:
Venous thromboembolisms (VTEs) continue to be a leading cause of death among trauma patients. Predicting which patients will develop a VTE can be difficult. This study investigated whether the Injury Severity Score (ISS) could be used in conjunction with the Abbreviated Injury Score (AIS) to assess a trauma patient’s risk for subsequent VTE development. 

Methods:
Participants were found by querying a trauma center registry. There were 2,213 patients included for evaluation. The patients were categorized based on their ISS and the anatomical region with the greatest injury (determined by the AIS). Odds ratios for developing VTEs were calculated for each ISS category. 

Results:
The results showed that in most categories VTE risk increased as ISS increased. Patients with trauma to their head/neck, chest, or extremities with ISS values of 21 or greater were at significantly increased risk for VTE development. Patients in these categories with an ISS less than 21 seemed to have little or only moderately increased odds of developing a VTE, although these values were not statistically significant. Patients with abdominal trauma were at increased risk even with ISS values of 11-21. 

Conclusion:
Trauma to the head/neck region, chest, and extremities (including pelvis) with Injury Severity Scores of 21 or higher had significantly increased odds of developing a VTE. Patients with abdominal trauma of any severity appeared to have increased odds of developing a VTE. Physicians should be aware of patients that fall into these categories and consider whether the risks of developing a VTE outweigh the risk of prophylactic treatment. 
 

16.16 Diminished Physiologic Reserve Predicts Mortality in the Underweight Following Hemorrhagic Shock

Posted on January 18, 2018

J. O. Hwabejire1, B. Adesibikan1, T. A. Oyetunji2, O. Omole1, C. E. Nembhard1, M. Williams1, E. E. Cornwell III1, W. R. Greene3  1Howard University College Of Medicine,Surgery,Washington, DC, USA 2Children’s Mercy Hospital- University Of Missouri Kansas City,Surgery,Kansas City, MO, USA 3Emory University School Of Medicine,Atlanta, GA, USA

Introduction:  We have previously demonstrated that extremes of body mass index (BMI) are associated with poor outcomes following blunt traumatic hemorrhagic shock. In this study, we examined the risk factors for mortality in underweight patients following blunt trauma.

Methods:  The Glue Grant database was retrospectively analyzed. Patients with BMI <18.5 kg/m2 who met criteria for hemorrhagic shock after blunt trauma were included. Survivors were compared to non-survivors using univariate analysis. Multivariable analysis was used to determine predictors of mortality.

Results: There were 102 patients who met criteria for inclusion in the study. Their mean age was 46 years (SD=20), with 62% being males, 89% Whites and 5% black. Mortality in this cohort was 52.9%, compared to 16.0 % in all comers and 14.3 % in patients with a normal BMI. Amongst the underweight, there was no differences in age, multiple organ dysfunction score, or emergency room (ER) shock index or pre-injury comorbidities between survivors and non-survivors. Compared to survivors, non-survivors were hypotensive in the ER (systolic BP: 110 ±27 vs. 87±38 mmHg, p=0.001), had higher ER lactate (7.1 ±4.1 vs. 4.1 ±2.5 mg/dL, p<0.001), were more coagulopathic (ER INR: 1.92 ±1.91 vs. 1.24±0.30, p=0.026 ), had higher APACHE II score (35±6 vs. 28±7, p<0.001), higher injury severity score, ISS (35±13 vs. 27±11, p=0.002), received more crystalloids (12696±6550 vs. 9796±4964 mL, p=0.014), and more blood (6070±4912 vs. 2240±3658 mL, p<0.001) within 12 hours of presentation.  When only patients with ISS >25 were compared, non-survivors were still more likely to be hypotensive (ER SBP: 112 ±28 vs. 87±36 mmHg, p=0.004), acidotic (ER lactate: 7.4 ±4.4 vs. 4.4 ±3.0 mg/dL, p=0.006), received more blood 6174±4926 vs. 3024±4612 mL, p=0.011) and had a higher APACHE II score (35±6 vs. 29±5, p<0.001). In the multivariate analysis, after adjusting for ISS, the only independent predictor of mortality was the APACHE II score (OR: 1.35, CI 1.08-1.1.69, p=0.009). 

Conclusion: The Acute Physiologic and Chronic Health Evaluation (APACHE) II score independently predicts mortality in the underweight after blunt traumatic hemorrhagic shock. Underweight patients appear to lack the physiologic reserve to tolerate severe trauma.

 

16.11 When We Take the Time to Look: Completion Angiography After Major Vascular Injury Repair

Posted on January 18, 2018

S. A. Moore1, J. P. Hazelton3, Z. Maher2, B. L. Frank4, J. W. Cannon1, D. N. Holena1, N. D. Martin1, A. Goldenberg-Sandau3, M. J. Seamon1  4Geisinger Health System,General Surgery,Scranton, PA, USA 1University Of Pennsylvania,Division Of Traumatology, Surgical Critical Care And Emergency Surgery,Philadelphia, PA, USA 2Temple University,Division Of Trauma & Surgical Critical Care,Philadelphia, PA, USA 3Cooper University Hospital,Division Of Trauma Surgery,Camden, NJ, USA

Introduction: Despite the operative vascular trauma advances achieved over the past several decades, these challenging injuries still result in significant morbidity and mortality.  Completion angiography (CA) immediately following repair of major vascular injury (MVI) has been advocated to limit adverse outcomes, but adequate data supporting or refuting this practice is lacking.  We hypothesized that CA after operative MVI repair identifies unsatisfactory repairs requiring intraoperative revision.  

Methods: A multi-center, retrospective cohort study of consecutive patients with operative MVI was conducted at 3 urban, Level-I centers (2005-2013).  Patients (≥15 years) with MVI of the neck, torso, or extremities (proximal to elbows/knees) requiring operative management were included.  Demographics, clinical variables and revision risk factors were analyzed with respect to our primary study endpoint, intraoperative revision following CA.  Secondary endpoints included outcomes after MVI repair.

Results: Of the 435 patients identified in the study, the majority were young (mean = 31 years) male (89%) patients with penetrating (84%) trauma.  Patients who underwent CA after repair (n= 128) were compared to patients who did not (n=303).  Although patients sustaining blunt injuries with associated fractures were both more likely to undergo CA (p<0.01), no differences with respect to age, gender, Injury Severity Score (ISS), initial systolic blood pressure, transfusion requirement and operating surgeon subspecialty were detected between study groups (all p<0.05).  Completion angiography study group patients were then more likely to undergo immediate intraoperative revision than those who did not undergo CA (CA, 21/128 [16.4%] vs. no CA, 4/303 [1.3%]; p<0.01, Figure 1).  Importantly, there were no differences in fasciotomy, delayed revision, arterial patency at discharge, or limb salvage rates between study comparison groups.

Conclusion: CA after operative repair led to intraoperative revision in 16% of MVI patients.  These data suggest that all patients undergoing operative MVI repair should undergo CA, as this additional diagnostic adjunct may prevent later adverse outcomes caused by unsatisfactory repairs.

 

16.13 Could Retained Bullet Fragments Be a Significant Source of Blood Lead Levels in Trauma Patients?

Posted on January 18, 2018

S. A. Eidelson1, C. A. Karcutskie1, A. B. Padiadpu1, M. B. Mulder1, S. K. Madiraju1, G. D. Garcia1, G. D. Pust1, N. Namias1, C. I. Schulman1, K. G. Proctor1  1University Of Miami,Miami, FL, USA

Introduction:
On Feb 17, 2017, the CDC reported that retained bullet fragments (RBF) may be a source of elevated blood lead levels (BLL) in those with no other known exposure.  This conclusion was based on voluntary reports of BLL>10 µg/dl to the CDC’s National Institute for Occupational Safety and Health. Roughly 75,000 non-fatal firearm injuries occur annually in the United States and routine screening for BLL is rarely performed. Thus, the incidence and magnitude of BLLs from RBF are unknown, but the CDC reports that any measurable BLL is unsafe.  We test the hypothesis that BLLs are elevated in trauma patients with RBF.

Methods:
BLL were measured in 23 consecutive adult patients with imaging-proven RBF admitted to an American College of Surgeon’s verified level 1 trauma center from 2/15/17-7/16/17. BLL is considered elevated at >5 μg/dL.  Data are expressed as mean±standard deviation if parametric and median if nonparametric.   Differences are assessed at p<0.05.

Results:
The study population is 95.7% male, 33±15 yrs, 25±4 kg/m2, and 70% African American. Of twenty-three patients with RBF, 35.0% (n=8) were found to have elevated blood lead levels and 74.0% (n=17) were found to have measureable lead levels. 

Conclusion:

These preliminary data provide basic proof of concept that measurable BLL occur in over half of trauma patients with RBF, regardless of days exposed.  Potential deleterious effects include impaired renal function with BLL <5 μg/dL, an increased risk for hypertension and essential tremor with BLL between 5-10 μg/dL, and neurocognitive deficits and adverse reproductive outcomes (including spontaneous abortion and reduced birthweight) with BLL ≥10 μg/dL.  Thus, patients with RBF may benefit from precautionary counseling on lead poisoning and the importance of baseline and periodic monitoring. Moving forward, there may also be a potential benefit of surgical retrieval.   

 

16.10 Provider Beliefs and Practice for the Use of Long-Acting Pain Medication in the Adult Burn Patient

Posted on January 18, 2018

K. Sloan1, J. Cartwright2, J. Liao1, Y. M. Liu1, K. S. Romanowski1  1University Of Iowa Hospitals And Clinics,Department Of Surgery,Iowa City, IA, USA 2University Of Michigan,Ann Arbor, MI, USA

Introduction: Achieving adequate pain control is a vital yet challenging component of burn management. Pharmacological interventions with opioids and adjuvants have long been the cornerstone of burn pain management. Since background pain is innate to burn injuries, long-acting pain medication (LAPM) are particularly advantageous to attain stable analgesia. However, literature is lacking surrounding their utilization and efficacy. The purpose of this survey was to assess burn providers’ beliefs and practices surrounding LAPMs in burn analgesic management.

Methods: Following approval by the Institutional Review Board and the American Burn Association (ABA) Survey Advisory Panel, a 31-item survey concerning LAPM was distributed electronically through Google Forms and REDCap to all physician, physician assistant, and nurse practitioner members of the ABA. Descriptive statistics and analysis of variance were conducted as indicated.

Results: Of 194 respondents (36.7% response rate), 101 (52%) identified as prescribers of pain medications with 93% utilizing LAPM. A majority of prescribers (73.4%) reported being likely or extremely likely to prescribe LAPM to burn patients. The most common trigger for initiation was “patient’s complaints of pain” (82%). Practitioners were evenly divided on whether burn size would influence their use of LAPM (46% no, 43% yes). Almost half of the respondents (47.25%) would utilize LAPM in burns as small as < 10% TBSA. Patient age was cited as consideration in the use of LAPM by 44% of practitioners with 13.5% of practitioners not using LAPM in patients aged 70 or older.  In adults, methadone was the most common first line therapy (44%), but was closely followed by extended-release morphine (31%). There was no consistent starting dose or regimen identified among practitioners. Only 21 (22.3%) practitioners cited that their institution had a protocol for the administration of LAPM. Clinical response was the principal reason for altering initial medication choice (18%), with excessive sedation being the chief variable stimulating reduction or with-holding of doses (90%). Analysis of provider perceptions of the effectiveness of LAPM revealed over 97% agreed/strongly agreed LAPM diminish background pain. While only 52% agreed/strongly agreed LAPM reduce the overall adjuvant pain medication requirement, over 90% agreed/strongly agreed that the usage of LAPM was associated with reduction in amount and/or dose of short-acting opioids.

Conclusions: While LAPM use was common among survey respondents and their attitudes towards it were largely positive, there was variance in individual practice and a lack of institutional protocols for use. More research into the most effective administration, dosing and weaning protocols must occur in light of the worsening opioid addiction crisis. 

16.08 Walking Under the Influence: Pedestrians Struck by Vehicles Are Commonly Intoxicated

Posted on January 18, 2018

M. Srour1, T. Li1, N. K. Dhillon1, K. Patel1, E. Gillettee1, D. R. Margulies1, E. J. Ley1, G. Barmparas1  1Cedars-Sinai Medical Center,Los Angeles, CA, USA

Introduction:  The aim of this study was to characterize alcohol intoxication among pedestrians struck by vehicles and examine whether these patients are more likely to be admitted after hours.

Methods:  The Los Angeles County TEMIS database was reviewed for all pedestrians struck by a vehicle over a 16-year period starting in 01/2000. Subjects ≥ 18 years old with available time and day of admission were selected. Patients with available blood alcohol concentration (BAC) were further analyzed and those with positive (+) BAC (> 0 gr %) were compared to those with negative (-) BAC. The primary outcome was mortality.

Results: Of 36,358 patients meeting inclusion criteria, 11,963 (33%) were tested for BAC and of those 3,846 (32%) were (+).  The proportion of (+) BAC pedestrians was low from the early morning until early evening (between 4% at 6-7 am to 8% at 5-6 pm). However, after 6 pm there was an abrupt increase in this proportion, peaking at 2-3 am (27% among all admitted pedestrians and 69% among pedestrians tested for BAC). This pattern was observed for all days of the week, but was more profound on Saturdays with the respective proportions reaching 30% and 75%.  When (+) BAC patients were compared to their (-) counterparts, they were more likely to be admitted with a Glasgow Coma Scale score ≤ 8 (10% vs. 5%, p < 0.01), with a systolic blood pressure (SBP) < 90 mmHg (3% vs. 2%, p < 0.01). Injury severity scores were similar (median 5 vs 5, p=0.66). The overall mortality was 4%, divided equally between the two cohorts. After adjusting for differences, (+) BAL pedestrians were significantly more likely to survive their injuries (AOR: 0.50, p<0.01). 

Conclusion: Pedestrians who are struck by vehicles during late hours are commonly intoxicated. These findings could have implications in developing preventative strategies to separate pedestrians from vehicles or to lower vehicle speed limit after hours in high risk areas.

16.09 How timing of surgical airway impacts in-hospital mortality in medical patients in US hospitals

Posted on January 18, 2018

I. Yi1, G. Ortega3, M. F. Nunez3, E. E. Cornwell2, M. Williams2  1Howard University College Of Medicine,Washington, DC, USA 2Howard University College Of Medicine,Department Of Surgery,Washington, DC, USA 3Howard University College Of Medicine,Clive O. Callender, MD Howard-Harvard Outcomes Research Center,Washington, DC, USA

Introduction:

Optimal timing of surgical airways in admitted patients requiring ventilator support remains elusive. Previous studies have classified tracheostomies as “early” and “late” using cut-off dates ranging from 5 to 10 days on ventilator to assess mortality. Our study aims to investigate mortality rates based on the day of the procedure and the number of days on a ventilator using a national database.

Methods:

We performed a retrospective analysis of the National Inpatient Sample (NIS) 2005–2014. We included non-trauma adult patients who underwent a surgical airway (ICD-9 31.1) procedure within 28 days of admission. We excluded patients who underwent elective and permanent tracheostomies (ICD 31.2), transfers from another facility, and patients requiring a surgical airway for the management of another localized disease (e.g. cancer or disease of the oropharynx and upper airway). We analyzed the day(s) from admission and/or day(s) from endotracheal intubation to the day the surgical airway was performed. Descriptive statistics were obtained for patient demographics, co-morbidities, length of stay, and mortality. Unadjusted and adjusted analyses were performed where appropriate to assess mortality adjusting for age, race/ethnicity, insurance, median household income, hospital type, and co-morbidities.

Results:

A total of 88,890 patients underwent a surgical airway. Most patients were White (60.5%), male (53.1%), had a mean age of 62.3 years. Most patients presented with respiratory failure (83.1%), followed by heart diseases (56.3%), sepsis (49.0%) and pneumonia (48.8%). Over the 28-days period, the average surgical airway was performed on day 13, and patients were intubated after 10 days. Most surgical airways were performed at teaching (57.7%) and urban facilities (95.9%), with an 18.9% overall mortality rate. The mortality rate was 15.7% on day 0 and 27.8% on day 28 for day of procedure, with the lowest rate at 12.6% on day 2. The mortality rate was 20.4% and 24.8% for 0 and 28 days on ventilator respectively, the lowest rate being 12.4% after 2 days on ventilator. On adjusted analysis, we found an increase by 1.6% and 1.1% in mortality rate for each day preceding the surgical airway and for each day spent on a ventilator, respectively.

Conclusion:

As time before surgical airway and number of days on ventilator increase, so does in-hospital mortality. Earlier timing of surgical airways appears to be independently associated with a modest increase in in-hospital survival compared to later surgical airways.

16.06 Traumatic Atlanto-Occipital Dissociation: No Longer a Death Sentence

Posted on January 18, 2018

D. M. Filiberto1, J. P. Sharpe1, M. A. Croce1, T. C. Fabian1, L. J. Magnotti1  1University Of Tennessee Health Science Center,Surgery,Memphis, TN, USA

Introduction: Although rare, traumatic atlanto-occipital dissociation (AOD) injuries are considered highly unstable and are associated with a high mortality rate.  In fact, these injuries were once believed to be uniformly fatal in adults. However, with recent advances in pre-hospital care coupled with early diagnosis and stabilization, these injuries are now potentially survivable. The purpose of this study was to evaluate the effect of rapid diagnosis and treatment (stabilization) of traumatic AOD following blunt injury in one of the largest single institutional series reported in the literature.

Methods:  Patients with traumatic AOD following blunt injury treated over a 17-year period were identified from the trauma registry of a level I trauma center and stratified by age, gender, injury severity (as measured by Injury Severity Score [ISS] and admission Glasgow Coma Scale [GCS] score) and severity of shock (as measured by admission base excess [BE] and 24-hour transfusions). Time to diagnosis, time to and method of stabilization, and mortality were recorded and compared. Multivariable logistic regression (MLR) was performed to determine which risk factors were independent predictors of death following AOD.

Results: 52 patients were identified: 35 men (67%) and 17 women (33%) with a mean age, admission GCS and ISS of 44, 8 and 34, respectively. Mean admission BE was -7.7 with patients requiring on average 3.7 units of packed red blood cells over the first 24 hours. 30 patients (58%) underwent stabilization: 16/30 underwent fusion, 5/30 were fitted with an external orthosis and 9/30 had a combination of both. Overall mortality was 32.7%. 16 of the deaths (94%) were secondary to severe traumatic brain injury. 3 of the deaths (17.6%) were among those patients who had undergone stabilization. Of the survivors, 34 patients (97%) were discharged neurologically intact: 15 patients went home, 15 to a rehabilitation center and 4 to a skilled nursing facility. Only one patient was discharged with neurological deficits to a rehabilitation center. There were no missed or delayed diagnosis related to AOD over the study. MLR identified admission GCS (OR 0.7; 95%CI 0.552-0.877) as the only independent predictor of death in patients with AOD after adjusting for severity of shock, injury severity, and time to stabilization.

Conclusion: Traumatic AOD remains a relatively rare injury following blunt trauma. Prompt diagnosis is crucial in promoting rapid stabilization and contributing to increasing survivability. Traumatic AOD should no longer be considered a uniformly fatal injury in adults.

 

16.07 Weight-Adjusted Enoxaparin Decreases Venous Thromboembolism? Rates in Trauma Patients

Posted on January 18, 2018

P. Martinez Quinones1, A. Talukder1, R. Latremouille3, T. Robinson2, A. Newsome2, C. White1  1Medical College Of Georgia,Surgery,Augusta, GA, USA 2Medical College Of Georgia,Pharmacy,Augusta, GA, USA 3Medical College Of Georgia,Augusta, GA, USA

Introduction: ? Traumatic injury is a major risk factor for the development of venous thromboembolic events (VTE), and is associated with increased bleeding risk. VTEs increase length of stay, cost and are associated with elevated morbidity and mortality in trauma patients. Optimal VTE pharmacologic prophylaxis in trauma patients remains unknown. Recent studies suggest that standard dosing of enoxaparin (30mg SQ Q12hrs) leads to sub-therapeutic levels of anti-Factor Xa, which are associated with increased risk for the development of VTE. We aimed to determine the efficiency of standard-dose versus weight-adjusted dose of enoxaparin.?

Methods:  As a quality improvement measure for medication use evaluation we conducted a retrospective registry review and data analysis. Patients with an initial trauma admission from January-December 2016 who received standard-dose (STD) or weight-adjusted (WT) enoxaparin were included. Patients <18 years, prior/current anticoagulation, prior VTE, known malignancy, autoimmune disorder and/or severe traumatic brain injury (TBI) were excluded. The primary outcome was incidence of VTE, including pulmonary embolism (PE) and deep venous thrombosis (DVT). Secondary outcomes included bleeding complications and length of stay.?

Results: We identified 142 patients who met inclusion criteria. Both groups (STD dose and WT dose) had comparable baseline characteristics for age, gender, race/ethnicity, mean weight and tobacco use. Mild-to-moderate TBI patients were similarly distributed, STD n=17, WT n=12  (p=0.81). VTE incidence was significantly different, STD n=17 and WT n=3 (p=0.03). No difference noted in length of stay (p=0.35) or time to onset of prophylaxis (p=0.93). No bleeding complications reported.?

Conclusion: Weight-adjusted enoxaparin dose for VTE prophylaxis decreased the risk of VTE in our trauma population sample without an increase in bleeding complications or expansion of intracranial hemorrhage in traumatic brain injury patients. Our ultimate goal is to develop a protocol for VTE prophylaxis that is safe and beneficial in patients with traumatic injuries.?

 

16.04 REBOA: How Many Patients Are We Missing? Assessing the Need in a Large Urban Trauma Center

Posted on January 18, 2018

R. P. Dumas1, D. N. Holena1, B. P. Smith1, M. J. Seamon1, P. M. Reilly1, Z. Qasim2, J. W. Cannon1  1University Of Pennsylvania,Division Of Traumatology, Surgical Critical Care And Emergency Surgery,Philadelphia, PA, USA 2University Of Pennsylvania,Division Of Emergency Medicine,Philadelphia, PA, USA

Introduction:

Resuscitative Endovascular Balloon Occlusion of the Aorta (REBOA) has shown benefit as a less invasive bridge to hemorrhage control in patients with torso trauma. The number of patients who might benefit from this procedure and the need for this intervention in an urban trauma center however, remains unclear. We sought to develop a generalizable methodology to identify the number of patients with injuries and presenting physiology amenable to REBOA, with the intention of characterizing the accuracy of our algorithm compared to traditional chart review.

Methods:

We queried the database of our Level I trauma center for all patients presenting from 2014-2015. Potential REBOA patients were included based on anatomic injuries and physiology. ICD-9 codes were used to identify REBOA-amenable injury patterns (abdominal solid organ; traumatic lower extremity amputation; major abdominal or lower extremity vascular injury; pelvic fractures) and physiology (presenting systolic pressure of ≤90 mmHg or transfusion requirement during initial trauma resuscitation). We excluded patients with injuries contraindicating REBOA (major vascular neck, thoracic, and axillary injuries). Chart review was used to confirm that our algorithm correctly identified these patients. Two reviewers experienced in REBOA then performed chart review to adjudicate algorithm-identified cases.

Results:

4818 patients were admitted from 2014-2015. 666 patients were included based on injury pattern. 186 patients received blood transfusions. 149 patients had an initial systolic blood pressure ≤90 mmHg. 309 patients had contraindications to REBOA (FIGURE 1). 64 patients (79.7% male, 67.2% African-American, 53% penetrating mechanism of injury, median ISS 18.5 [IQR 14-28]) had an injury pattern and physiology amenable to REBOA with no injury contraindications. Chart review confirmed that our algorithm correctly identified 54 (86%) of patients that had anatomic injuries amenable to REBOA with no contraindications. Review by two independent REBOA-experienced physicians revealed 29 patients (46% of those identified by algorithm) that may have benefited from REBOA. The inter-rater reliability was excellent (kappa 0.94, p<0.001). In the total cohort, 0.6% of patients may have benefited from REBOA.

Conclusion:

Our REBOA algorithm identified patients who may have benefited from early femoral arterial access but over-estimated the number of true REBOA candidates. Centers seeking to establish a REBOA program should combine an algorithm to identify potential patients with a detailed chart review to determine their center-specific REBOA candidate population. Future work should focus on revision and refinement of this algorithm for application at other institutions.
 

16.05 Psoas Muscle Area Index May Not Predict Outcomes in Trauma Patients

Posted on January 18, 2018

A. Santoro1, E. Otoo1, A. Salami1, R. Smith2, A. Joshi1  1Albert Einstein Medical Center,Surgery,Philadelphia, PENNSYLVANIA, USA 2Albert Einstein Medical Center,Radiology,Philadelphia, PENNSYLVANIA, USA

Introduction:  Sarcopenia is the age-related loss of skeletal muscle mass. Studies have described a correlation between sarcopenia, frailty, and poor outcomes in heterogeneous populations. We sought to determine the utility of psoas muscle area index (PAI), a new tool for the assessment of sarcopenia, in predicting poor outcomes amongst trauma patients. 

Methods:  This retrospective observational study was conducted using data from the trauma database at Albert Einstein Medical Center, a level 1 trauma center in Philadelphia, PA. All level 1 & 2 trauma activations for blunt and penetrating trauma between September 2014 and December 2015 were included. Patients without abdominal CT scans at the time of initial presentation were excluded. CT scans of the abdomen were used in estimating PAI at the level of the 3rd lumbar vertebral body using the formula [psoas muscle area(mm2)/height(m2)]. PAI was dichotomized with values < the 25th percentile representing the sarcopenic group, while ICU and hospital length of stay (LoS) were dichotomized with values > the 75th percentile representing prolonged stay. Outcomes of interest were in-hospital mortality, prolonged ICU, and Hospital LoS. Multivariable logistic regression was used in elucidating associations. 

Results: A total of 254 patients were included in this study, 73 (28.7%) were aged>65 years and 111 (43.7%) were female.  In-hospital mortality occurred in 13 patients (5.2%). Prolonged ICU and hospital LoS were observed in 22.8% and 24% of the study population, respectively. PAI was not associated with in-hospital mortality (OR: 0.5, CI: 0.12-2.49; p=0.426), prolonged ICU (OR: 0.95, CI: 0.48-1.89; p=0.894), or hospital LoS (OR: 1.3, CI: 0.68-2.46; p=0.425) on univariate analyses. This trend persisted following multivariable adjustment (p>0.05 for all). Similarly, PAI was not found to be predictive of outcomes following stratification by age (above or below 65 years) and gender. 

Conclusion: PAI does not seem to have any correlation with poor outcomes in trauma patients, irrespective of age or gender. Prospective studies with larger populations are needed to determine if PAI has any prognostic utility in the risk stratification of trauma patients. 

 

16.02 Risk of Recurrent Pneumothorax and Does it Really Matter?

Posted on January 18, 2018

A. F. Elegbede1, B. W. Carr1, B. L. Zarzaur1, S. A. Savage1  1Indiana University School Of Medicine,Indianapolis, IN, USA

Introduction:

Thoracic trauma is a significant source of mortality in multiply injured patients.  The vast majority will be managed solely with tube thoracostomy, however, and chest tube insertion remains one of the most commonly performed procedures.  A common complication related to chest tube removal is recurrent pneumothorax (R-Ptx).  R-Ptx has been linked to the need for additional procedures and may increase hospital length of stay (LOS).  Though multiple studies have focused on tube removal techniques to best reduce R-Ptx, there have been few reports regarding other factors that may predispose to R-Ptx. 

Methods:
 A retrospective chart review was performed for all patients requiring chest tubes from January to July 2016 at a single Level 1 trauma center.  Data collected included patient demographics, injury characteristics, information regarding chest tube placement and any complications related to the intervention.  Groups were compared using either χ2 or Kruskal Wallis tests as appropriate.  Multivariable logistic regression was used to compare outcomes.

Results:
106 patients were included in this study. 75% of the patients were male, 83% suffered a blunt mechanism, the mean age was 48 years (SD 20.7) and mean chest AIS was 3.1 (SD 0.89). There were no differences between groups in terms of gender, overall ISS, or chest AIS.  Patients in the R-Ptx group were significantly younger (40 years (SD 17) vs 54 years (SD 21), p=0.0004), had a significantly lower body mass index (BMI) (25.3 (SD 5.7) vs 29.4 (SD 7), p=0.0031), and a higher rate of penetrating trauma (28% vs 9.5%, p=0.0179).  With multivariable logistic regression, both age and BMI remained significantly associated with R-Ptx (Table 1).   When excluding patients with trace or small R-Ptx (<1 cm),  BMI remained associated with recurrence (Table 1).   ICU LOS was no different between groups (R-Ptx 4 days (IQR0, 16); no R-Ptx 6 days (IQR 2, 13), p=0.2807).  However, R-Ptx patients had a significantly shorter hospital LOS (10 days (IQR 7, 14)) compared to no R-Ptx patients (12 days (IQR 7, 20), p=0.0088).

Conclusion:
Recurrent pneumothorax is a complication of thoracostomy tube placement that may delay recovery and discharge.  In our study, patients with a higher BMI and older age were protected from recurrence compared to our younger patients suffering penetrating trauma.  This may in part be due increased subcutaneous tissue preventing entrainment of air during tube removal.  Counterintuitively, despite the R-Ptx, these patients actually had a significantly shorter hospital LOS.  The very factors which predispose patients to recurrence may allow faster recovery and earlier discharge.  R-Ptx may not have a significant negative impact on duration of hospital stay.
 

16.03 Aspirin Versus Low-Molecular Weight Heparin for VTE Prophylaxis After Lower Extremity Trauma

Posted on January 18, 2018

M. Mazzei1, M. Uohara2, S. Pasch2, E. Dauer1, Z. Maher1, L. L. Mason1, A. Pathak1, T. Santora1, L. O. Sjoholm1, A. J. Goldberg1, J. M. Lopez1  1Temple University Hospital,Department Of Surgery,Philadelphia, PA, USA 2Temple University,Lewis Katz School Of Medicine,Philadelphia, PA, USA

Introduction: Low-molecular-weight heparin (LMWH) is an accepted standard of care for extended venous thromboembolism (VTE) prophylaxis after lower extremity trauma based on data extrapolated from elective orthopedic procedures, but can be limited due to cost and compliance. The aim of this study was to investigate the potential noninferiority of daily aspirin compared to LMWH in preventing symptomatic VTE after lower extremity trauma.

Methods: A six-year, single-institution, retrospective analysis of patients requiring surgical intervention for lower extremity trauma at a level one urban trauma center was performed. Subjects were included if they had no prior history of VTE, did not sustain lower extremity vascular injury, did not require therapeutic anticoagulation for any reason, and followed with trauma or orthopedic care providers for at least thirty days post-discharge. Subjects receiving LMWH as outpatient VTE prophylaxis were compared with those who received aspirin due to an inability to afford LMWH. The primary endpoint was the development of symptomatic VTE diagnosed by venous duplex or CT-angiography within the first thirty days from discharge.

Results: Of the 1099 patients initially identified, 712 met inclusion criteria and comprised the study population. Patients receiving aspirin (n=187) rather than LMWH (n = 525) were younger (38.99 vs. 42.88 years old, p = 0.012), more likely to be male (78.61% vs. 63.62%, p < 0.0001), and spent less time in the hospital (5.8 days vs. 7.6 days, p = 0.001). Otherwise, groups were comparable in terms of most comorbidities, mechanism of injury, presenting vitals, lower extremity injury severity score, and rates of in-hospital complications. 1.59% (3) of the patients receiving aspirin developed a symptomatic VTE during the first thirty days following discharge, compared with 1.52% (8) who developed symptomatic VTE (p = 0.93) while receiving LMWH.

Conclusion: In this retrospective review, aspirin was found to be noninferior to LMWH for preventing symptomatic VTE in patients requiring surgical intervention for lower extremity trauma. Future research should be conducted to evaluate aspirin’s potential role as a safe, effective, and affordable alterative to LMWH for extended VTE prophylaxis.

 

« Previous Page
Next Page »