100.17 Redefining Early Tracheostomy Timing in Trauma Patients with Rib Fractures

Posted on January 31, 2019

K. C. Shue1,2, A. A. Fokin2, J. Wycech2,3, J. Lozada3, R. Stalder3,5, A. Tymchak1,2,3, I. Puente1,2,3,4  1Florida Atlantic University,College Of Medicine,Boca Raton, FL, USA 2Delray Medical Center,Trauma Services,Delray Beach, FL, USA 3Broward Health Medical Center,Trauma Services,Fort Lauderdale, FL, USA 4Florida International University,College Of Medicine,Miami, FL, USA 5Wake Forest University,Winston-Salem, NC, USA

Introduction:
Timing of tracheostomy remains a subject of debate along with whether early or late tracheostomy leads to better outcomes. The distinction of a 7-day cut off timing for early tracheostomy (ET) is not clearly substantiated. The aim of this study was to examine possible benefits of prompt tracheostomy (PT) in patients with rib fractures.

Methods:
This IRB-approved retrospective cohort study, included 124 patients with radiologically-confirmed rib fractures admitted to two Level 1 Trauma Centers, between January 2012 and December 2017, who have undergone a tracheostomy procedure. Patients were divided into 2 groups: PT group, which included 20 patients who had tracheostomy performed within 3 days of intubation and late tracheostomy (LT) group, which included 104 patients who had tracheostomy performed on day 4 or beyond from initial intubation. Analyzed variables included age, injury severity score (ISS), number of ribs fractured (RFX), presence of flail chest, bilateral rib fractures, pulmonary contusion (PC), traumatic brain injury (TBI), Glasgow Coma Scale (GCS), rate of maxillofacial co-injuries, rate of ventilation-associated pneumonia (VAP), duration of mechanical ventilation (DMV), duration of tracheostomy (DoT), intensive care unit length of stay (ICULOS) and hospital length of stay (HLOS).

Results:

Between Groups PT and LT there was no statistical difference in mean age (43.7 vs 51.6), ISS (27.0 vs 27.6), RFX (4.8 vs 5.3 ribs), incidence of flail chest (5.0% vs 8.7%), bilateral rib fractures (35.0% vs 28.4%), rate of PC (60.0% vs 56.7%), presence of TBI (50.0% vs 45.2%) and GCS (9.0 vs 9.5), all with p>0.05. Only maxillofacial co-injuries were identified more often in the PT Group than in the LT Group (80.0% vs 32.7%, p<0.001).

Within the evaluated outcome variables PT Group had significantly shorter DMV (22.9 vs 28.3 days, p=0.008) and ICULOS (15.4 vs 23.2 days, p<0.001). Analysis did not show significant differences between the two Groups’ incidence of VAP (40.0% vs 34.6%), DoT (20.8 vs 16.7) and HLOS (39.6 vs 40.5), with all p>0.3.

Conclusion:
Prompt tracheostomy in patients with rib fractures lead to shorter duration of mechanical ventilation and ICU length of stay, therefore prompt tracheostomy should be considered, particularly if maxillofacial co-injuries are present.

100.16 Iliac Vessel Injuries: An Analysis of the National Trauma Data Bank

Posted on January 31, 2019

A. D. Person1, S. Brown1, G. Mendoza1, D. Keleny1, D. Rigg1, M. Dale1, J. Dabestani1, D. K. Agrawal1, J. A. Asensio1  1Creighton University Medical Center,Department Of Trauma Surgery And Surgical Critical Care,Omaha, NE, USA

Background: Iliac vessel injjuries are uncommon, even in high volume trauma centers.  THey incur high morbidity and mortality.  The objectives of this study were to review the National Trauma Data Bank experience with iliac injuries, identify predictors of outcome, and then report the outcomes of these injuries.

Methods: The National Trauma Data Bank was queried for pre-hospital and admission data for all iliac vessel injuries.  The main outcome measured was survival.  Statistical analysis included univariate and stepwise logistic regression.

Results: 1,590 patients were identified from 1,466,887 patients from the National Trauma Data Bank.  Incidence of iliac vessel injury was 0.1%. Mean age of patients with these injuries was 35.7 +/- 18.8. Mean Revised Trauma Score (RTS) was 5.98 +/- 2.8, while the mean Injury Severity Score (ISS) was 26.1 +/- 13.4. The mechanism of injury distribution between penetrating and blunt was 804 (51%) and 777 (49%) respectively. In terms of the vessels injured, the iliac artery was injured, there were 1,174 arterial injuries, 651 venous injuries, and 84 unspecified injuries.  Span data on surgical repairs was available in 380 patients.  Surgical management is as follows: arteriorrhaphy 129 (34%), venorrhaphy 170 (45%), primary repair of artery and/or vein with vein and/or prosthetic patch 52 (14%), and interposition grafts 29 (8%). In total, there were 1,040 survivors (65.4%) and 550 non-survivors (34.6%).

Multiple factors, able to be assessed on presentation demonstrated statistical significance between survivors and non-survivors. Mean Systolic Blood Pressure on presentation was 110.36 vs. 83.45 (p<0.001) between survivors and non-survivors respectively.  Mean Revised Trauma Score was 6.80 vs. 4.41 (p<0.001). Mean Injury Severity Score was lower in survivors at 23.62 vs. 31.17 (p<0.001). Glasgow Coma Scale was 12.5 vs. 7.2 (p=0.002).

Conclusion: Iliac vessel injuries are uncommon in the landscape of American trauma as denoted by their low incidence reported in the National Trauma Data Bank.  Mortality remains high. Predictors of outcome may be identified from the initial admitting data. When stratified to mechanism of injury, this idea remains valid and also appears to correlate well with mortality.

100.14 Outcomes in Traumatic Hemorrhagic Shock in Patients with Prior Coronary Intervention

Posted on January 31, 2019

B. A. Adesibikan1, J. Hwabejire2, T. Oyetunji3, M. Williams1, S. Siram1, E. Cornwell1, W. Greene4  1Howard University College Of Medicine,Washington, DC, USA 2Massachusetts General Hospital,Boston, MA, USA 3Children’s Mercy Hospital- University Of Missouri Kansas City,Kansas City, MO, USA 4Emory University School Of Medicine,Atlanta, GA, USA

Introduction:  The ability of patients who had prior coronary intervention for coronary arterial disease to withstand major traumatic hemorrhagic shock is unknown. In this study, we examined clinical outcomes in blunt trauma patients with a history of prior coronary intervention (coronary artery bypass graft, CABG, or percutaneous coronary intervention with stenting) who developed hemorrhagic shock

Methods:  This is an analysis of the Inflammation and the Host Response to Injury Database. All patients who had a history of pre-injury coronary intervention were selected for analysis.

Results: There were 67 patients who met criteria for inclusion in the study. Their mean age was 67 years (σ=13). The majority (93%) were Whites. The mortality in this cohort was 19.4%, compared to 15.9 % in those with no prior history of coronary intervention (p=0.438). In the coronary intervention group, patients who died following shock were older (78 ±9 vs. 66 ±13 years, p=0.002), hypotensive on presentation (ER systolic BP 91±22 mmHg vs. 110 ±31 p=0.039), had higher ER lactate (5.6 ±3.2 vs. 3.6 ±2.3 mg/dL, p=0.016), had higher APACHE II score  (33.5 ±5.2 vs. 28.2 ±8.7 p=0.038), and received more blood products (2839±1883 vs. 1793±1270 mL, p=0.019). There was no difference in Injury Severity Score (ISS), incidence of abdominal compartment syndrome, rate of acute respiratory distress syndrome (ARDS), or rate of post-injury myocardial infarction between survivors and non-survivors. Survivors were more likely to have a pre-injury history of myocardial infarction (33.3% vs. 0, p=0.014) while non-survivors were more likely to have a pre-injury history of congestive heart failure (38.4% vs. 13.0%, p=0.046). Age was the only independent predictor of mortality in the multivariable analysis mortality (OR: 1.14, CI 1.03-1.26, p=0.014).

Conclusion: In patients with prior coronary intervention who suffered blunt traumatic hemorrhagic shock, age is an independent predictor of mortality. The role of pre-injury congestive heart failure as a potential risk factor for mortality in this cohort deserves further study.

 

100.13 Severly Injured Trauma Patients Can Be Successfully Managed On A General Care Ward.

Posted on January 31, 2019

M. Duverseau1,2, D. Suma1,2, R. Patel2, S. Galvin1, A. Conquest2, M. Schurr1,2  1Mountain Area Health Education Center,Asheville, NORTH CAROLINA, USA 2Mission Hospitals,Asheville, NORTH CAROLINA, USA

Introduction: Intensive care unit beds are in demand in large regional referral hospitals; therefore, non-intubated poly-trauma patients are often admitted to general care (GC) wards. We hypothesized that trauma patients with ISS≥15 and unexpected intensive care unit admission (U-ICU) after initial admission to GC had increased morbidity and mortality. We also hypothesized that those requiring U-ICU could be predicted based on admission parameters.

Methods: This was a retrospective review of patients ≥18 years old, admitted to GC with blunt trauma and ISS≥15 from April 2015-March 2017. Demographics were collected along with injury patterns and complications. Statistics included Chi square, Fisher’s exact, Mann-Whitney, and t-tests.

Results: 502/986 (50.9%) were directly admitted to GC. Prevalence of U-ICU was 9.8% (49/502 patients). The only admission predictor of U-ICU was history of myocardial infarction (8/49, 16.3%, vs 21/453, 4.6%, P<.001). Outcomes in table 1.

Conclusion: Half of all severely injured patients can be managed on GC ward; however 9.8% of patients will require U-ICU admission. Admission history of myocardial infarction predicts those at risk. Severely injured patients with U-ICU admission have significant cardiopulmonary complications, but can be rescued with no increase in overall mortality.

 

100.12 Alcohol is protective against mortality in TBI, especially in low energy blunt trauma

Posted on January 31, 2019

W. M. Brigode1, C. Cohan1, G. Beattie1, G. Victorino1  1University of California – San Francisco – East Bay,Department Of Surgery,Oakland, CA, USA

Introduction: Alcohol (EtOH) poses a diagnostic and therapeutic challenge in traumatic brain injuries (TBI) given its impact metabolically and neurologically. Studies on the interaction of EtOH and TBI have had conflicting results. Most suggest increased complications secondary to elevated EtOH, however, the correlation with mortality in TBI remains controversial. We sought to elucidate the patient factors that explain the mixed results and supply further insight into EtOH physiology in TBI. We hypothesized that trauma mechanism, brain injury severity, and blood alcohol concentration (BAC) would influence the impact of EtOH on mortality in TBI.

Methods: We performed a single-institution retrospective review from January 1, 2016 to June 30, 2018 of adult trauma patients admitted to our urban Level I Trauma Center with any measured BAC and a diagnosis of TBI. Exclusion criterion included patients in whom BAC was not assessed. The primary outcome assessed was mortality comparing EtOH positive and negative patients, and at different threshold values of BAC. Secondary outcomes included discharge Glasgow Coma Scale (GCS), Abbreviated Injury Score (AIS) of the head and neck, ventilator days, ICU and total length of stay (LOS), and mechanism subtypes.

Results: Admission EtOH was assessed in 583 patients with TBI, with 256 testing positive for EtOH and 327 testing negative. Overall, intoxicated patients had lower mortality when compared with sober patients (4.7% vs 8.9% respectively, p=0.05). The absence or presence of EtOH showed a two-fold mortality difference, while raising the threshold of BAC to 0.15% improved the benefit to three-fold (mortality 3.1% with BAC<0.15, 9.0% with BAC>0.15, p=0.01). EtOH was also associated with significantly higher discharge GCS (14.2 vs 13.4, p=0.01), reduced likelihood of mechanical ventilation (20.3% vs 30.0%, p=0.008), fewer ventilator days (1.3 vs 2.4, p=0.01), and decreased hospital LOS (7.4 vs 9.4, p=0.02). There was no effect of EtOH on mortality when patients were stratified by AIS-head/neck or admission GCS (p>0.05). When grouped into blunt (n=558) and penetrating trauma (n=25) the mortality benefit remained two-fold for blunt (3.7% intoxicated vs 7.6% sober, p=0.049) but was absent in penetrating trauma. In the blunt trauma sub-group, when trauma associated with automobiles was excluded, the mortality of positive EtOH was increased three-fold from the mortality of sober patients (2.8% intoxicated vs 9.0% sober, p=0.01).

Conclusion: Alcohol appears to be protective against mortality in brain-injured trauma patients. This appears to be most significant in blunt traumatic mechanisms that do not involve motor vehicles. There may be a dose related effect, as we found a three-fold difference in mortality when using a higher BAC threshold of 0.15%. Further research should be done to unravel ongoing controversies in the influence of EtOH on TBI.

 

100.11 Early Versus Late Pulmonary Embolism in Trauma: Not All Pulmonary Embolisms are Created Similarly

Posted on January 31, 2019

S. Gambhir1, A. Grigorian1, A. Ashbaugh1, D. Spencer1, D. Ramakrishnan1, J. Nahmias1  1University Of California – Irvine,Surgery,Orange, CA, USA

Introduction: Pulmonary embolism (PE) is an uncommon complication occurring in up to 5% of trauma patients. In small previous studies, it has been demonstrated that patients with long bone fractures are associated with a higher risk of early PE while those with severe head injuries are at higher risk for late PE. We sought to confirm risk factors for early and late PE previously reported in the literature, hypothesizing that early PE would be associated with long bone fractures and late PE would be associated with above-the-knee DVT.

Methods: This was a retrospective analysis using the trauma registry at a Level-I center from 2010-2017. Patients with early PE (≤  4 days) were compared to those with late PE (> 4 days) with bivariate and multivariable analysis.

Results:  A total of 77 patients (0.6% incidence of PE) were included with 43 (53.1%) diagnosed with late PE. Compared to those with early PE, those with late PE had similar age (p=0.97) and rate of obesity (p=0.20) but a higher rate of severe (grade>3) abbreviated injury scale (AIS) for the head (39.5% vs. 15.2%, p=0.02). Those with late PE had a higher rate of blood product transfusion (39.5% vs. 6.1%, p<0.001), craniotomy (23.3% vs. 3.0%, p=0.01), venous thromboembolism chemoprophylaxis (86.0% vs. 42.4%, p<0.001) and above-the-knee DVT (29.7% vs. 3.1%, p=0.004), compared to those with early PE. The incidence of long bone extremity fractures was similar (p>0.05). After adjusting for age≥ 65, injury severity score ≥ 25 and gender, independent risk factors for late PE included above-the-knee DVT (OR=12.01, CI=1.34-107.52, p=0.03), blood transfusion (OR=8.99, CI=1.75-46.22, p=0.009) and craniotomy (OR=8.82, CI=1.03-75.97, p=0.04) while the only independent risk factor for early PE was smoking (OR=4.56, CI 1.06-19.66, p=0.04).

Conclusion: The strongest risk factor for late PE is above-the-knee DVT. Contrary to previous reports, long bone extremity fractures are not at higher risk for early PE and severe head injuries are not at higher risk for late PE but this may be related to our study being underpowered. The only independent variable associated with higher risk of early PE is a history of smoking. A future prospective multicenter study will help confirm our findings and determine if early screening of smokers may be beneficial

100.10 Evidence-Based Improvement of the National Trauma Triage Protocol for Pediatric Trauma Patients

Posted on January 31, 2019

M. Zeeshan1, M. Hamidi1, T. O’Keeffe1, N. Kulvatunyou1, A. Tang1, E. Zakaria1, L. Gries1, A. Northcutt1, B. Joseph1  1University Of Arizona,Trauma And Acute Care Surgery,Tucson, AZ, USA

Introduction:
Systolic-blood-pressure (SBP) <90 mm-Hg is one of the criteria for trauma center need (TCN) in the national trauma triage protocol (NTTP). The aim of our study was to determine the impact of substituting SBP<90 with shock-index (SI)>1 on triage performance.

Methods:
We performed a 2-year (2014–2015) analysis of all pediatric trauma patients (age<17 y) in the pediatric TQIP. Our outcome measure was TCN (ISS≥16, need for an emergent-operation, death in the emergency department, or ICU stay>1d). AUROC and triage characteristics were compared between SBP<90 and SI>1.0. Regression analysis was performed to compare the mortality between patients triaged under the current protocol of SBP<90 and patients triaged using the newly defined protocol (SI>1.0).

Results:
We included 60,842 patients. Compared with SBP<90, SI>1.0 had a higher sensitivity (57.2% vs. 42%) but lower specificity (73.3% vs. 81.5%) for TCN. Substituting SBP<90 mm Hg with SI>1.0 resulted in a decrease in under-triage rate of 2494 patients (4.1%) but an increase in over-triage of only 912 patients (1.5%). There was no difference in aOR mortality between the new and old triaged positive system (OR: 1.01 [0.95-1.08]).

Conclusion:

Appropriate triage of pediatric trauma patients will lead to better outcomes. Utilizing SI>1instead of SBP<90 mm-Hg in the NTTP results in significant reduction in under-triage rate without causing a large increase in over triage. SI is simple to use and has a better discriminatory power when compared to SBP. 

100.09 Pelvis Fracture Protocol Effects on Utilization and Time to Hemorrhage Control

Posted on January 31, 2019

J. B. Miller1,2, R. Griffin1, K. Hendershot1  1University Of Alabama at Birmingham,Department Of Surgery,Birmingham, Alabama, USA 2Orlando Regional Medical Center,Department Of Surgical Education,Orlando, FL, USA

Introduction:  Blood loss from pelvis fractures is a significant cause of morbidity and mortality. These patients sometimes require angioembolization and/or pelvis packing for hemorrhage control. A new institutional protocol was developed to facilitate more timely and appropriate control of hemorrhage, utilizing both Interventional Radiology (IR) and pelvis packing in the operating room (OR). Our hypothesis was that our protocol would increase utilization of both IR and OR and decrease time to IR.

Methods:  A single-institution retrospective analysis was done of adult trauma patients with blunt pelvis fractures from January 1, 2015 to June 30, 2017.  The new protocol was implemented midway through the study period. Data was analyzed prior to implementation and several months after protocol initiation.  

Results: A total of 1083 patients had pelvis fractures during the study period. Of these, 85 patients underwent OR or IR procedures.  Patients were excluded for unrelated OR or IR interventions, leaving 40 patients included in the final analysis.

In the pre-protocol time period there were 12 patients: 1 underwent OR pelvis packing followed by IR embolization and 11 underwent IR embolization alone.  During the post-protocol period there were 28 patients: 2 underwent OR pelvis packing alone, 7 underwent OR pelvis packing followed by IR embolization, and 19 underwent IR embolization only.

Time to IR for pelvis hemorrhage in the pre-protocol period was 5 hours 6 minutes; post-protocol was 4 hours 30 min.

Conclusion: Our protocol resulted in increased utilization of both IR angioembolization and OR pelvis packing for patients with bleeding pelvis fractures; our time to IR decreased, although only by 36 minutes.
 

100.08 Efficacy of an Order Set for Appropriate VTE Prophylaxis Use in Trauma Patients

Posted on January 31, 2019

M. S. Stumpf1, S. O’Malley1, G. Prellwitz2, J. Sutyak1, S. Ganai1, E. Mackinney1, M. Smith1  1Southern Illinois University School Of Medicine,Surgery,Springfield, IL, USA 2Memorial Medical Center,Springfield, IL, USA

Introduction:  Inappropriate venous thromboembolism (VTE) prophylaxis may pose harm to those at high-risk of VTE including trauma patients. To improve timing and utilization of VTE prophylaxis, an admission VTE order set was implemented in a Level-I Trauma center, but rather than mandating calculation of Caprini score, it prompted the user to determine risk level and prophylaxis type based on an explanation of the score in the order set. We hypothesized that use of a parsimonious order set process would improve appropriate use of mechanical and/or chemical prophylaxis according to risk categorization.

Methods:  A retrospective cohort study was conducted to evaluate the effects of the order set implementation in June 2016 on VTE prophylaxis at a Level-I trauma center using a number-generated random sample of trauma patients pre- (Spring 2016) and post-intervention (Spring 2017). Caprini scores were calculated for each patient to determine guideline-appropriate recommendations for mechanical and/or chemical prophylaxis based on risk level. ”Appropriate use” was defined as strict adherence to guidelines, while “extra use” was defined as going beyond guideline recommendations by risk level. Comparisons by proportion were made using Fisher’s exact test.

Results: After exclusions, a total of 54 patients in 2016 and 34 patients in 2017 were analyzed. No significant difference in injury severity scores (pre 9.1±7.5; post 8.9±7.0; p=0.15) or admission Caprini score risk levels (p=0.27) were seen between cohorts. Appropriate use of DVT prophylaxis across all risk levels was 24.1% pre-intervention and 41.1% post-intervention (p=0.10). At the high-risk level, appropriate use increased significantly from 17.9% to 60.0% (p=0.005).  No significant difference was seen in appropriate use across the low to moderate risk levels, which trended down from 30.7% to 16.7% (p=0.45).  Extra use of VTE prophylaxis options increased from 72.2% to 92.9% across the low to moderate risk levels (p=0.21).

Conclusions: The data support the hypothesis that a simple order set process can improve appropriate VTE prophylaxis use according to risk categorization, but these findings were significant only in the high-risk subgroup and not across all groups. It appears that risk stratification without prompt for calculation is imprecise and users leaned towards ordering more than recommended in low to moderate risk groups. Further examination of harms and benefits of this approach will require a greater sample size to assess association with outcomes including VTE and bleeding incidence.

 

100.07 Risk Factors For Early Mortality After Tracheostomy

Posted on January 31, 2019

A. A. ARISHI1, A. Mohamed Ahmed1, W. Qu1, F. C. Brunicardi1, M. M. Nazzal1, J. Ortiz1, M. F. Osman1  1University Of Toledo Medical Center,Surgery,Toledo, OH, USA

Introduction:
Tracheostomy is a high-risk procedure that is almost always performed on patients in respiratory failure. The purpose of this study was to investigate factors associated with early mortality after tracheostomy.

Methods:
Adult patients with respiratory failure who underwent tracheostomy were selected from the National Inpatient Sample (NIS) database (2008-2012) of the Healthcare Cost and Utilization Project (HCUP). All the diagnoses and procedures were identified by ICD-9-CM codes. Patients were assigned into futile group (FG, died within 30 days after admission) and non-futile group (NFG). Demographics, hospital characteristics, and risk factors were compared between the 2 groups. The indicators of early mortality were identified by multivariable logistic regression. All statistical analyses were done with IBM SPSS statistical software ver. 24. Type I error level was set at 0.05.

Results:
A total of 86,303 patients (7,007/8.1% in FG, and 79,296/91.9% in NFG) were included in this study. The overall mortality rate was 16% with an 8.1% 30-day mortality rate. FG patients were significantly older (65±15 vs. 61±17 years), more likely to be male (57.8 vs. 56.2%), Caucasian (66.1 vs. 63.8%), and were covered by Medicare insurance (57.8 vs. 51.4%) than NFG patients (all p values <.001). In addition to age (adjusted OR=1.018, 95%CI 1.016-1.020, P<.001) and male gender (adjusted OR=1.093, 95%CI 1.038-1.151, p<.001), septicemia/sepsis, coagulopathy, liver disease, renal failure, fluid/electrolyte imbalance, and congestive heart failure were identified as among the most important predictors for early mortality after tracheostomy in multivariable analysis (Table 1).

Conclusion:
16% of the patients with respiratory failure will die after tracheostomy during their hospital stay, and about half of these deaths occur within 30 days after admission. In addition to older age and male gender, severe infections, coagulopathy, liver disease, renal failure, fluid/electrolyte imbalance, and congestive heart failure would also significantly increase the risk of early mortality. This study demonstrates that tracheostomy is a high-risk procedure associated with a significant mortality rate, which should be reviewed pre-operatively with the patient and the family.

100.06 The Complexity of Advance Directives in the SICU at a Tertiary Level Medical Center

Posted on January 31, 2019

H. A. Lyfoung1, C. T. Minshall1, M. Cripps2, S. Kartiko1  1Baystate Medical Center,Springfield, MA, USA 2University Of Texas Southwestern Medical Center,Dallas, TX, USA

Introduction:

Critical care has made tremendous advances in recent years that have improved outcomes in many disease processes. However, the ability to sustain a patient’s life often requires invasive procedures, which frequently detract from patient comfort. Understanding our patient’s expectations and goals may allow us to make better decisions regarding their care. We retrospectively reviewed changes in the code status, determination of health care proxy and disposition for patients in our surgical intensive care unit (SICU) to characterize our patient population in an attempt to identify how we can better align patient goals with medical care. 

Methods:

A retrospective chart study was performed on all patients admitted to the SICU in a tertiary level medical center from February to June 2018. We obtained information on advanced directives, Physician Order for Life-Sustaining Treatment (POLST), Health Care Proxy (HCP), gender, age, emergent/elective nature of surgery, documentations from providers, care managers, social workers, and family meetings. All patients that did not have an advanced directive at admission were assumed to be full code, until further clarification was available.

Results:

104 patients were admitted to the SICU during the study period: 40% trauma, 30% general surgery, and 30% were from other surgical subspecialties. 29% of patients had advance directives, 13% had POLST and 41% had a documented HCP. Twelve patients (12%) were determined to be DNR/DNI. This was temporarily suspended for surgical intervention in 11 (92%) prior to admission to the ICU. All but one patient had their DNR/DNI reinstated. Most patients (80%) admitted to the SICU had undergone a surgical intervention: 46% were emergent, 15.4% were urgent, and 19% were elective. 40% of patients without an advanced directive had an “on site” goal of care discussion, which resulted in a change from full code to DNR/DNI in 48% of the patients. 40% of patients, who did not have a HCP at SICU admission, had one at the time of SICU discharge. In the SICU, 20% of patients died including the 6% who expired after transitioning to comfort measures; while 4% were discharged to rehab or home. Most (76%) patients were transferred to a lower level of care in the hospital.  

Conclusion:

This study characterizes our SICU patients and begins to highlight the difficult task of aligning patient goals and providing appropriate critical care. Most patients who are admitted to the SICU do not have an advance directive or a health care proxy. However, even when they do have an advance directive, the acuity of their surgical disease cause the patient or their HCP to hold the DNR/DNI status. While a lot of goal of care conversation is conducted in the SICU, there remain a large number of patients who was discharged from the SICU and still do not have a HCP or POLST form completed. A greater focus on what patients want versus what they need is needed to provide better care.

100.05 GRUMPS: Ganey Response-Unit Matched Patient Satisfaction

Posted on January 31, 2019

J. Y. Valenzuela1, L. Szydziak1, M. Grossman1  1Southside Hospital / Northwell Health,Division Of Trauma Critical Care,Bay Shore, NY, USA

Introduction: Patient satisfaction scores may be useful to assess physician, nurse, and hospital performance. We hypothesize that cohorting trauma patients (TP) on a specific trauma/surgical unit (3T) would improve all Press Ganey (PG) metrics for TP.

Methods: 12 month retrospective analysis of PG administrative and trauma registry data in an ACS verified Level 2 trauma center. TP were compared to non-trauma patients (NTP) with respect to response rate and satisfaction scores. Baseline characteristics were compared to determine association with response. Nurse and physician satisfaction scores were compared between TP and NTP hospital-wide and on 3T. Comparisons were made using X2 or Wilcoxon test with significance set at p=0.05.

Results: 2936 returned PG surveys were matched to the trauma registry. 168 TP responded to the survey. Response rate was lower for TP compared to NTP (17% vs 22%, p=0.0021). There was no difference in age or sex between TP and NTP responders. Compared to non-responding TP, responding TP were likely to be female (p<0.001), age >65 (p=0.018), nonsubstance user (p=0.036), and have isolated hip fracture (p=0.0001). Overall, TP were dissatisfied with nursing care compared to NTP (p<0.003) with no difference in perceptions of physician care. For unit-based scores, 3T had higher scores from NTP compared to other wards (p<0.013). There was no difference in satisfaction for TP whether on 3T or other units, but on 3T, TP were less satisfied with nursing and physician care than NTP (Figure 1).

Conclusion: PG response rate is lower for TP and those who respond may not be representative of all trauma patients. Cohorting TP on 3T did not result in higher PG scores and in fact had the opposite affect. Conversely, NTP were more satisfied with care on 3T than other units. TP may receive less time and attention than other surgical NTP on 3T or more difficult complex TP may be selected for 3T admission versus other units.

 

100.04 Multidisciplinary Bedside Procedure Service

Posted on January 31, 2019

P. T. Evans1, M. C. Smith2, S. J. Schneeberger1, K. M. Prendergast1, A. B. Peetz2, E. B. Kopp2, N. E. Collins2, O. D. Guillamondegui2, B. M. Dennis2  1Vanderbilt University,School Of Medicine,Nashville, TN, USA 2Vanderbilt University Medical Center,Division Of Trauma And Surgical Critical Care,Nashville, TN, USA

Introduction: Patients with persistent respiratory failure and/or neurologic deficits often require tracheostomy and/or gastrostomy prior to ICU liberation. At our institution, a specialized multidisciplinary consult service was created to facilitate the placement of these devices. We describe the outcomes of a multidisciplinary bedside procedure service in our institution.

Methods: A multidisciplinary service consisting of a trauma surgeon, nurse practitioner, and critical care nurse was created to efficiently evaluate and place tracheostomies and gastrostomy tubes in the inpatient setting.  The service manages the device post-procedure until removal. Patients for whom bedside percutaneous tracheostomy or percutaneous endoscopic gastrostomy (PEG) were performed by our institution’s tracheostomy and PEG service were identified from an institutional database for bedside procedures (August 2013-January 2018). Demographic data, data on complications, as well as time from consult request to procedure were recorded. Descriptive statistics were calculated.

Results: 1,075 tracheostomies and 592 PEGs were identified. Median age was 53.1 for tracheostomy and 59.8 for PEG. Median BMI were 28.1 and 25.7 respectively. 99% of tracheostomies and 77% of PEGs were performed in ICUs. There were no periprocedural airway losses. 2% of patients who underwent tracheostomy had complications requiring an additional procedure. 96% of PEGs were performed successfully at bedside. 23 patients (4%) required an additional procedure after attempted PEG, 14 of which were for procedures aborted at bedside. 75.5% of tracheostomies and 61% of PEGs were performed within 1 day of consult.

Conclusion: A multidisciplinary consult service is a safe and efficient model for performing bedside percutaneous tracheostomies and PEGs in the ICU. With the increasing need for efficiency in our current healthcare economic climate, such a model can help streamline care of the ICU patient and facilitate liberation.

100.03 Failure to Perform the Trauma Tertiary Survey is Associated with Higher Acuity and Time of Year

Posted on January 31, 2019

A. L. Holmstrom1, H. K. Weiss2, K. C. Ott1, R. J. Ellis1,3, M. B. Shapiro1, A. D. Yang1,3  1Northwestern University,Department Of Surgery,Chicago, IL, USA 2Feinberg School Of Medicine – Northwestern University,Chicago, IL, USA 3Feinberg School Of Medicine – Northwestern University,Surgical Outcomes And Quality Improvement Center,Chicago, IL, USA

Introduction:  The tertiary survey is an important component of trauma patient evaluation and has been shown to reduce the rate of missed injuries, particularly in polytrauma and critically ill patients.  More complicated injury patterns and hospital stays in these patients may make formal evaluation for missed injuries challenging.  Despite their importance, no contemporary studies have evaluated rates of tertiary survey completion at trauma centers. The objectives of this study were (1) to determine the frequency of trauma tertiary survey performance at a Level 1 academic trauma center, and (2) to identify patient and encounter characteristics associated with failure to receive the trauma tertiary survey.

Methods:  Retrospective cohort study performed at a single Level 1 academic trauma center.  Trauma encounters from July 2016 through June 2017 were included in the analysis.  A random 20% sample from each month was audited for performance and documentation of the trauma tertiary survey. The primary outcome was documentation of the trauma tertiary survey, with secondary outcomes of interest being the ordering of additional studies and detection of additional injuries based on tertiary evaluation. Associations between outcomes and categorical patient/encounter characteristics were assessed using chi-square tests.

Results: Among 297 trauma encounters, a complete tertiary survey was documented in 91 (30.6%), a partial tertiary survey was documented in 114 (38.4%), and no tertiary survey was documented in 92 (31.0%). Among patients with a completed tertiary survey, 10.2% underwent additional evaluation and 3.9% were diagnosed with additional injuries. A tertiary survey was significantly less likely to be documented earlier in the academic year (56.5% in the first academic quarter vs 86.1% in the last academic quarter, p<0.001).  Documentation of a tertiary survey was also less likely to occur in patients following penetrating trauma (51.9% vs. 72.0% for blunt, p=0.004), more severe injuries (56.3% in high acuity trauma activations vs. 76.2% in low acuity trauma activations, p=0.003), and in those patients requiring intensive care (48.3% vs. 73.2%, p<0.001).  There was no difference in tertiary exam performance based on age, sex, time of day, or in operative traumas.

Conclusion: Performance and documentation of a trauma tertiary survey at a Level 1 academic trauma center was inconsistent, with higher acuity and sicker patients being less likely to receive a thorough tertiary assessment. Significant improvement in performance during the academic year suggests that some variation may be driven by the resident learning curve, and thus improvement in tertiary survey performance may be achieved through focused quality improvement efforts. Design and implementation of rigorous, standardized mechanisms to ensure tertiary survey documentation may improve tertiary survey performance and reduce the rate of missed traumatic injuries.

100.02 Is Echocardiography Needed After Negative Contrast CT to Rule Out Penetrating Cardiac Trauma?

Posted on January 31, 2019

M. Mazzei1, E. Acevedo1, M. Ethridge1, E. Dauer1, Z. Maher1, J. H. Beard1, A. S. Pathak1, T. A. Santora1, A. J. Goldberg1, L. O. Sjoholm1, L. Mason1  1Temple University,Department Of Surgery,Philadelpha, PA, USA

Introduction: The optimal radiographic evaluation of patients with penetrating thoracic trauma is unclear. Thoracic computerized tomography (CT) with IV contrast is widely available as an initial screening test, and many institutions confirm initial findings with routine transthoracic echocardiography (TTE). This is a resource-intensive and time-consuming approach, and it is not clear that TTE provides additional diagnostic accuracy over that of contrast CT alone. We hypothesize that after negative contrast CT, routine TTE does not lead to a change in the diagnosis of penetrating cardiac injuries.

Methods: A retrospective cohort study was conducted of all patients with penetrating thoracic trauma with concern for cardiac involvement at our urban level I trauma center from January 2006 to December 2015. Data was extracted from the trauma registry and the medical record. All patients who underwent TTE to rule out cardiac injury after after a negative contrast chest CT were included. Results and outcomes after imaging were evaluated.

Results: 84 patients with penetrating thoracic injury underwent screening TTE after a negative thoracic CT, including 44.7% with gunshot and 55.3% with stab wounds (avg. ISS 14.7, chest-ISS 2.4). In this population, chest CT with IV contrast showed 100% negative predictive value for cardiac injury.  There were no major discordances between TTE and CT; in 38.3 % of cases, TTE was limited by technical difficulty. Minor discordances between TTE and CT (i.e. trivial pericardial effusion) occurred in 15% of patients, but addition of TTE did not result in any changes in management.

Conclusion: In the setting of a negative thoracic CT, TTE provides little additional diagnostic information and should not be routinely used for the evaluation of the patient with penetrating thoracic trauma. Contrast thoracic CT is an acceptable screening test to rule out significant cardiac injuries.

 

100.01 State-level Estimates of Potentially Preventable Trauma Deaths in the United States

Posted on January 31, 2019

Z. G. Hashmi1, C. K. Zogg1, M. P. Jarman1, J. C. McCarty1, T. Uribe-Leitz1, E. Goralnick1, A. Salim1, A. H. Haider1  1Brigham And Women’s Hospital,Surgery,Boston, MA, USA

Introduction:  Recent estimates suggest that nationwide trauma quality improvement has the potential to save more than 200,000 lives in just over a decade. Citing this enormous burden, the National Academies of Sciences, Engineering and Medicine (NASEM) has mandated several measures aimed at achieving “Zero Preventable Deaths after Injury.” However, the geographic distribution of these potentially preventable trauma deaths, which will be critical in planning and implementing regional quality improvement efforts, remains unknown. The objective of this study is to determine the state-level estimates of potentially preventable trauma deaths in the United States.

Methods:  We included all adult (age≥16) trauma patients with blunt and/or penetrating injuries reported to the State Inpatient Database 2014 for Arizona, Florida, Kentucky and New York. Hierarchical logistic regression, adjusting for age, sex, race, mechanism of injury, intent of injury, anatomic injury severity (Trauma Mortality Prediction Model), presence of severe head injury and hospital volume, was used to benchmark hospitals into quintiles of reliability-adjusted mortality rate. Generalized linear modeling was then used to estimate the relative-risk of death for patients treated at hospitals among the best-performing quintile across the four states versus the remainder of hospitals in each state. This relative-risk of death was used to estimate the number of potentially preventable trauma deaths in each state if all patients were treated at the best-performing hospitals.

Results: A total of 176,240 trauma patients treated at 331 hospitals across the four states were analyzed. The overall adjusted mortality rate was 2.3%. A total of 67 hospitals were benchmarked into the best-performing quintile across the four states with an adjusted mortality rate of 1.6%. The remainder of the hospitals in each state had a significantly higher relative risk of mortality than the best-performing hospitals (Table). A total of 2,052 [95% confidence interval (1,843-2,228)] trauma deaths were deemed potentially preventable across the four states with the most preventable deaths occurring in New York [923 (851-987)] and the highest proportion occurring in Kentucky (18.2 per 1000 trauma admissions).

Conclusion: Large geographic variation exists in the number and proportion of potentially preventable trauma deaths. This study establishes state-specific targets for regional trauma quality improvement initiatives aiming to achieve “Zero Preventable Deaths After Injury.”

« Previous Page