33.09 Pre-injury ACS-TQIP Funtional Status Predicts Decline after Traumatic Injury in Older Adults

Posted on January 25, 2017

J. R. Cherry-Bukowiec1, N. Werner1, L. Min2, J. Ha2, P. K. Park1  1University Of Michigan,Surgery,Ann, MI, USA 2University Of Michigan,Geriatrics,Ann Arbor, MI, USA

Introduction:
The ACS-TQIP geriatric trauma management guidelines incorporate a 4 question, 5-item functional status (FS) screen to measure functional outcomes in older patients. We examined the performance of ACS-TQIP screen in injured and non-injured patients over time to deterine performance of this abbreviated scale was comparable to a full 12-scale ADL/IADL ability score over time.

Methods:
Functional status collected every 2 years by the nationally-representative Health and Retirement Study (HRS) was linked with CMS Inpatient Claims data to identify older (age >=65) patients with a functional status evaluation prior to hospitalization for injury and a subsequent evaluation following injury within two years. For comparison to natural functional decline, interview intervals that did not include an acute injury hospitalization served as controls. Changes in functional status by the ACS-TQIP FS were evaluated as a function of time from injury to follow-up interview and compared to the 12-scale ADL/IADL ability score.

Results:
Between 1998-2010, 20,562 acute hospitalizations occurred between two consecutive HRS interviews; 1056 were in patients with severe traumatic injury. In patients with no preinjury functional disability (N=715), mean ACS-TQIP FS and 12-scale ADL/IADL ability scores were both lower than in control patients (in excess of 1 ADL) during the post-injury period, with persistent decreases seen up to 2 years after injury. Similar, sustained decreases were seen in patients with mild preinjury disability(N=205). In patients with severe preinjury functional disability (N=136), overall performance remained limited.

Conclusions:
In a national cohort of older trauma patients with no or mild prior history of disability, the ACS-TQIP FS screen identified sustained new functional impairment up to 2 years after the initial injury in a similar pattern as the full 12-scale ADL/IADL ability score.

33.08 Using Machine Learning Techniques to Develop a Mobile Application to Predict Unplanned Reintubations

Posted on January 25, 2017

A. N. Kothari1, J. Attisha2, S. A. Brownlee1, A. Cobb1, C. Fairman1, K. Halvorsen1, W. Hopkinson1, H. H. Ton-That1, P. C. Kuo1  1Loyola University Chicago Stritch School Of Medicine,Surgery,Maywood, IL, USA 2DePaul University,College Of Computing And Digital Media,Chicago, IL, USA

Introduction:
Unplanned intubations are associated with significant patient morbidity and mortality. Given the important impact of unplanned intubations on both ventilator days and mortality, reductions in the number of unplanned intubations has far-reaching quality improvement implications. The overarching goal of this project was to create a mobile application for use during bedside rounds to identify patients at high daily risk for unplanned intubation. 

Methods:
A single-center, retrospective review of patients who underwent a surgical procedure that met National Surgical Quality Improvement Program inclusion criteria and subsequently received care in the Surgical Intensive Care Unit (SICU) was conducted. The primary predicted outcome was an unplanned intubation. For each machine learning algorithm tested, a grid search was performed to identify the best hyper-parameter values. Each model was trained and tested 20 times on random balanced subsets of the dataset. The following performance metrics were used to choose the final model: accuracy, sensitivity, and specificity. Mobile application usability was measured using the System Usability Scale (SUS). 

Results:
A total of 24,198 surgical encounters meeting NSQIP inclusion criteria were identified from January, 2012 – July, 2015. Only patients that received care in the SICU were included in the final analytic cohort (n=4,487). The overall unplanned intubation rate was 6.6%. A total of 8 different machine learning algorithms were tested: Decision Trees, Naïve Bayes, K-Nearest Neighbors, Linear Discriminant Analysis, Support Vector Machines, Random Forests, Stacked Ensemble Method, and Bucket Ensemble Method. Model performance of each is shown in Table 1. The Stacked Ensemble algorithm was integrated into a mobile application for bedside use and received a mean SUS score of 74 (SD 11.2, usability above average) by a pilot group of surgical faculty, residents, and intensive care nurses.  

Conclusion:
The use of machine learning algorithms can create a high-performing predictive tool. Accurate prediction of high-risk individuals in the SICU using a mobile application will allow for the implementation of targeted interventions to better assist those patients most vulnerable to having unplanned intubation occurrences.  
 

33.07 Platelet Transfusion Does Not Improve Outcomes in Brain Injured Patients on Antiplatelet Therapy

Posted on January 25, 2017

P. Maluso1, J. Holzmacher1, C. L. Reynolds3, M. Patel4, S. Holland5, N. Gamsky1, H. Moore2, E. Acquista6, M. Carrick7, R. L. Amdur1, H. Hancock5, J. Dunn9, B. Sarani1  1George Washington University School Of Medicine And Health Sciences,Surgery,Washington, DC, USA 2Carle Foundation Hospital,Surgery,Urbana, IL, USA 3East Carolina University Brody School Of Medicine,Surgery,Greenville, NC, USA 4Vanderbilt University Medical Center,Surgery,Nashville, TN, USA 5San Antonio Military Medical Center,Surgery,Fort Sam Houston, TX, USA 6University Of North Carolina At Chapel Hill,Surgery,Chapel Hill, NC, USA 7Medical Center At Plano,Surgery,Plano, TX, USA 8Columbia University College Of Physicians And Surgeons,Surgery,New York, NY, USA 9Medical Center Of The Rockies,Surgery,Loveland, CO, USA

Introduction: Brain injury is the most common cause of death following trauma. Platelet dysfunction is associated with worsening hemorrhage following brain injury (TBI). Use of antiplatelet medications (APM) is common, especially in elderly patients but the efficacy of platelet transfusion remains unknown. Thrombelastography platelet mapping (TEG-PM) assesses platelet function. We hypothesize that platelet transfusion can reverse the effects of APM but does not improve clinical outcomes in TBI patients on APM.

Methods: A 2 year prospective, observational study at 6 US trauma centers was performed. Patients over 17 years old on APM with CT evident TBI after blunt injury were enrolled. Patients underwent TEG-PM and brain CT on arrival and repeat imaging within 24 hours. Platelets were transfused and repeat TEG-PM was ordered at physician discretion. Demographics, platelet transfusion, brain CT and TEG-PM results, length of stay (LOS), and injury severity score (AIS) were abstracted. Groups were compared using student t-test. 

Results: 66 patients were enrolled (89% aspirin, 34% clopidogrel, 2% ticagrelor). 23 patients underwent platelet transfusion (table). The transfused group had significantly higher AIS and CT brain injury scores. TEG variables were not significantly associated with the decision to transfuse platelets(table 2). MA(AA) increased and %inhibition(AA) decreased significantly following transfusion but other parameters were unchanged. CT brain injury scores did not change after transfusion (mean change 0.10 ± 0.41, p=0.1). Among non-transfused patients with repeat CT scores available (n=29), the change in mean Marshall CT score was 0. Among transfused patients (n=23), mean CT score change before/after transfusion was 0.22 ± 0.60 (p=0.1). Transfusion was associated with longer LOS (7.8 v 3.5 days, p<0.01), but this was not significant after controlling for AIS. There was no difference in mortality. 

Conclusion:Platelet transfusion significantly decreases degree of platelet inhibition in the arachidonic acid pathway in TBI patients but is not associated with change in CT brain injury scores or clinical outcomes.

 

33.06 Late Relative Lymphopenia is Associated with an Increased Infection Rate in Pediatric Burn Patients

Posted on January 25, 2017

Z. Diltz3,4, R. A. Devine4, K. Wheeler4, J. Shi4, H. Xiang4, R. Fabia1,5, M. W. Hall2,4,6, R. K. Thakkar1,4,5  1Nationwide Children’s Hospital,Department Of Pediatric Surgery,Columbus, OH, USA 2Nationwide Children’s Hospital,Department Of Critical Care Medicine,Columbus, OH, USA 3Ohio State University,College Of Medicine,Columbus, OH, USA 4Nationwide Children’s Hospital,The Research Institute,Columbus, OH, USA 5Ohio State University,Department Of Surgery,Columbus, OH, USA 6Ohio State University,Division Of Pulmonary, Critical Care And Sleep Medicine,Columbus, OH, USA

Introduction:

Burn injury is estimated to be the fourth leading cause of death in children in the United States, according to the World Health Organization, and each year roughly 745,000 children under age 17 require medical attention for burn injuries.  These patients are at high risk for adverse outcomes including infectious complications which remain a leading cause of morbidity for burn patients.  Both increased (leukocytosis, neutrophilia) and decreased (lymphopenia) white blood cell (WBC) counts have been reported in this setting.

We designed a retrospective study to test the hypothesis that abnormalities in WBC counts that are present beyond the first two days of burn injury will be associated with increased nosocomial infection risk.  

Methods:
We used our institution’s trauma registry to identify patients aged 0-18 years old with burns of at least 10% total body surface area (TBSA) from 2005 to 2015.  Demographic data, mechanism of injury, and clinical outcomes including infections were collected and verified through chart review.  Complete blood counts with differentials were recorded through the first week of hospitalization following injury.  Abnormal WBC data were defined as high total WBC count (leukocytosis), high percentage of neutrophils (relative neutrophilia), or low percentage of lymphocytes (relative lymphopenia) according to age-based laboratory norms.  Late abnormalities were defined as those noted on post-burn days 3 – 7. Nosocomial infection was defined as a positive culture plus receipt of a full course of antibiotics. 

Results:

140 burn patients TBSA≥ 10% were identified during the study period. A higher percentage of patients had late relative lymphopenia (67.2%) than late leukocytosis (10.6%) or late neutrophilia (32.0%).  There were no significant differences in age or burn TBSA between subjects with and without late relative lymphopenia.  The group of patients with late relative lymphopenia had a significantly higher nosocomial infection rate (71.8%) than those with normal lymphocyte percentages on or after day 3 (42.1%) (p=0.0287).  This was not true for patients with late leukocytosis (p=0.34) or late relative neutrophilia (p=1.0).

After controlling for age, gender, mechanism of injury, and TBSA with multivariable logistic regression analysis, the adjusted odds of nosocomial infection were significantly lower in subjects without late relative lymphopenia (AOR = 0.18; 95% CI = 0.04-0.81).  Patients with late relative lymphopenia had longer mean hospital and ICU lengths of stay, but the differences were not statistically significant.

Conclusion:

Late relative lymphopenia following severe pediatric thermal injury is associated with the subsequent development of nosocomial infection even when controlling for burn size and other factors. This should be the subject of a future prospective study in a larger sample size. 

33.05 Abnormal Platelet Function Contributes to Hypercoagulability After Trauma

Posted on January 25, 2017

S. A. Eidelson1, C. A. Karcutskie1, C. I. Schulman1, N. Namias1, K. G. Proctor1  1University Of Miami,Department Of Surgery,Miami, FL, USA

Introduction:

Thomboelastography (TEG) has had an increasingly useful role in critical care.  It has been postulated that platelets may contribute to hypercoagulability after trauma.  There is not yet definitive data on when these patients become hypercoagulable, or how long this persists.  We hypothesize that patients will gradually develop abnormal platelet function after trauma.

Methods:
Prospective study of patients sustaining blunt or penetrating trauma admitted to the intensive care unit from 8/2011-4/2015.  Patients were all deemed high risk for venous thromboembolism with a Greenfield Risk Assessment Profile Score of 10 or greater.  All patients received bilateral lower extremity duplex ultrasounds and TEGs on admission and then weekly afterwards.  Additionally, all patients received mechanical and chemical thromboprophylaxis.

Results:
One hundred patients received TEGs on both admission and at one week.  The cohort was aged 47±20 years, with 76% blunt trauma and 24% penetrating trauma.  The deep venous thrombosis rate was 22%.  The maximum amplitude (MA) TEG value, which indicates platelet function, was in the hypercoagulable range in 7 patients (7.0%) on admission.  At week 1, MA was in the hypercoagulable range in 62 patients (62%), significantly higher compared to admission (p<0.001).  Fifty-five of these patients also received TEGs at week 2.  In this subgroup, hypercoagulable MA values were seen in 5 (9.1%), 36 (65.5%), and 37 (67.3%) patients at admission, week 1, and week 2, respectively.  Again, there were significantly more hypercoagulable patients at week 1 than admission (p<0.001).  There was no difference in hypercoagulability at weeks 1 and 2.

Conclusion:
Despite the use of chemical thromboprophylaxis, trauma patients appear to exhibit abnormal platelet function as early as week 1.  This hypercoagulability seems to persist to at least week 2.  Antiplatelet therapy may play a significant role in thromboprophylaxis after trauma.
 

33.04 Platelet ADP Receptor Inhibition Predicts Need for Platelet Transfusion but not Massive Transfusion

Posted on January 25, 2017

G. R. Stettler1, H. B. Moore1, G. R. Nunns1, A. Sauaia1, A. Ghasabyan2, C. C. Silliman1,3,4, A. Banerjee1, E. E. Moore1,2  1University Of Colorado Denver,Department Of Surgery,Aurora, CO, USA 2Denver Health Medical Center,Department Of Surgery,Aurora, CO, USA 3Bonfils Blood Center,Denver, CO, USA 4Children’s Hospital Colorado,Aurora, CO, USA

Introduction: Platelet dysfunction has been documented early after trauma, but the impact on outcome has not been determined. A recent randomized trial of blood component therapy did not document a survival advantage of early platelet transfusion. Our experimental work suggests metabolites generated during shock inhibit platelet function. Our previous clinical work suggested inhibition to adenosine diphosphate (ADP) was a more sensitive marker of trauma induced platelet dysfunction than arachidonic acid. We hypothesize that platelet function assessment via the ADP response will be associated with platelet transfusion but platelet count will be a superior predictor of massive transfusion.

Methods: Subjects without coagulation-related comorbidities or medications were enrolled from 2014-16. Thromboelastography (TEG) Platelet Mapping was assessed in trauma activation patients (TAP, n=270, field or ED arrival blood samples) and healthy volunteers (n=89). Values of MA-ADP (mm) and ADP receptor inhibition (%ADP-INH) presented as median, IQR or n,%. The results of the platelet assessment was not known to the physicians managing the patient, who based their decision for platelet transfusion based on rTEG MA.

Results:Compared to controls, TAP patients showed early low MA-ADP [59, 53-65, vs. 36, 17-50] and higher %ADP-INH [40, 23-69, vs. 4, 0-14] (p<0.0001). MA-ADP and %ADP-INH were both significantly (all p<0.05) correlated with NISS (Spearman Rho: -0.37, +0.26), temperature (Rho=+0.18, -0.14), GCS (Rho=+0.24, -0.20), and platelet count (Rho=+0.21, -0.16). %ADP-INH was higher in blunt trauma compared to penetrating (47, 23-81 vs. 36, 22-60; p=0.03). Adjusted for platelet count, MA-ADP predicted massive transfusion (MT=>10PRBC or death/6hrs, p=0.02), while %ADP-INH did not (p=0.16). Adjusted for platelet count, MA-ADP and %ADP-INH predicted platelet transfusion requirements. Platelet count remained a powerful predictor of MT (p<0.0001).  Platelet count is a powerful predictor of MT or death within 6 hours (AUC 0.84), ventilator free days <3 (AUC 0.69), ICU free days <6 (AUC 0.63), death within 24 hours (AUC 0.82), and requirements for platelet transfusions (AUC 0.75). Platelet count was a better predictor of need for MT or death in the first 6 hours compared to %ADP-INH (AUC 0.66; p=0.037) (Fig).

Conclusion:%ADP-INH predicts the need for platelet transfusion, but not for massive transfusion. Compared to platelet function, platelet count is consistently a better predictor of post injury outcomes. The predictive value of dysfunction at the ADP receptor is improved by adjustment for total platelet count.

33.03 Safety of Early Venous Thromboembolism Prophylaxis for Isolated Blunt Splenic Injury: A TQIP Study

Posted on January 25, 2017

B. Lin1, K. Matsushima1, L. De Leon1, G. Recinos1, A. Piccinini1, E. Benjamin1, K. Inaba1, D. Demetriades1  1University Of Southern California,Acute Care Surgery,Los Angeles, CALIFORNIA, USA

Introduction:

Non-operative management (NOM) has become the standard of care in hemodynamically stable patients with blunt splenic injury. Due to the potential risk of bleeding, there are no widely accepted guidelines for an optimal and safe timeframe for the initiation of venous thromboembolism (VTE) prophylaxis in patients undergoing NOM. The purpose of this study was to explore the association between the timing of VTE prophylaxis initiation and NOM failure rate in isolated blunt splenic injury. 

Methods:

After approval by the institutional review board, we utilized the American College of Surgeons (ACS) Trauma Quality Improvement Program (TQIP) database (2013-2014) to identify adult patients (≥18 years) who underwent NOM for isolated blunt splenic injuries (Grade III/IV/V). Patients were excluded if they expired within 24 hours of admission or required surgical management of splenic injury within 12 hours after admission. Failure of NOM was defined as any splenic surgeries after 12 hours of admission. The incidence of overall NOM failure was compared between two groups: 1) VTE prophylaxis <48 hours after admission (early prophylaxis group), and 2) VTE prophylaxis ≥48 hours (late prophylaxis group). Similarly, we compared the incidence of NOM failure after the initiation of VTE prophylaxis between the early and late prophylaxis group. Multiple logistic regression analysis was performed for NOM failure adjusting for clinically important covariates including the timing of VTE prophylaxis initiation. 

Results:

A total of 816 patients met the inclusion criteria; median age: 34 years (IQR 23-52), 67% male gender, median ISS: 13 (IQR 10-17), 679 patients (83.2%) with severe splenic injury (Grade IV/V). Of the patients who met the inclusion criteria, VTE prophylaxis was not administered in 525 patients (64.3%), whereas VTE prophylaxis was given < 48 hours and ≥48 hours after admission in 144 and 147 patients, respectively. Among patients who received VTE prophylaxis, angioembolization of the spleen was performed in 30 patients (10.3%). Overall NOM failure rate was 13.4% (39/291). While overall NOM failure rate was significantly lower in the early group compared to late prophylaxis group (4.9% vs. 21.8%, p<0.001), there was no significant difference in the NOM failure rate after the initiation of VTE prophylaxis between two groups (3.5% vs. 3.4%, p=1.00). In the multiple logistic regression analysis, early initiation of VTE prophylaxis was not significantly associated with NOM failure (OR: 1.19, 95% CI 0.31-4.51, p=0.80).

Conclusion:

Our results suggest that early initiation of VTE prophylaxis (<48 hours) does not increase the risk of NOM failure in patients with isolated splenic injury. Further prospective study to validate the safety of early VTE prophylaxis is warranted.   
 

33.02 An Analysis of Trauma Intubations Performed by Emergency Physicians at a Level 1 Trauma Center

Posted on January 25, 2017

K. N. Williams1, P. Rhee1, T. O’Keeffe1, B. Joseph1, M. Singer1, A. Tang1, G. Vercruysse1, N. Kulvatunyou1, J. Sakles1  1University Of Arizona,Division Of Trauma,Tucson, AZ, USA

Introduction:  There is variability amongst trauma centers as to who primarily manages the airway of trauma patients, with some utilizing anesthesia personnel and others emergency department physicians.  It is not known if this impacts the success rates for intubation in trauma patients. The aim of this study was to examine success rates and complications from a prospectively collected database. Our hypothesis was that ED physicians can effectively intubate trauma patients with few complications.

Methods:  An analysis of a prospectively collected database of all adult (≥18 years old) trauma patients requiring intubation at a level I trauma center over a 6 year period (2009-2015) was performed. The database was a quality improvement database collected in the trauma bays and was matched to the institutional trauma registry for additional data.  All initial intubation attempts were performed by an emergency physician.  After intubation, the physician that performed the intubation completed a structured data collection form that included: demographics, complications, and the presence of difficult airway characteristics (DACs). Our primary outcome was first pass success of intubations. Secondary outcomes were number of attempts, success in patients with DACs and immediate complications. 

Results: 972 patients met analysis criteria. The successful intubation rate by emergency medicine physicians was 98%. An Anesthesiologist was called for 7 patients (0.7%) and a surgical airway was required in only 12 (1.2%) patients. Five were for failed intubations, four for cardiac arrest and three for primary surgical airway. First attempt success rate was 80% and the second attempt success rate was 95%. The first emergency physician was successful in intubating the patient 93% of the time and only 7% needed rescue by a senior resident or attending. PGY1 residents (n=126) had a first attempt success rate of 69% and complication rate of 26.2%, compared to PGY3 residents (n=393) with an 82% (p=.019) first attempt success rate and an 18.8% (p=.082) complication rate. The overall complication rate was 21%. Complications included, desaturation (14%), esophageal intubation (2%), cardiac arrest (0.9%), aspiration (0.8%), hypotension (0.5%), dysrhythmia (0.5%), dental/airway trauma (0.3%) and laryngospasm (0.1%). When comparing the first three years of the study period to the latter three years, the first pass success rate was 74% during the first half of the study period, compared to 86% (p=.001) during the second half (when GlideScope Videolaryngoscopy (GVL) was increasingly utilized).  There was no difference in overall complication rates between the first and second half of the study period (p=0.84).

Conclusion: The overall rate of successful intubation in trauma patients is high and the need for emergency rescue is very low; Emergency Department physicians can safely intubate trauma patients with high success rates comparable to those reported in the anesthesiology literature.

 

 

33.01 Fractured Care and Outcomes After Injury: National Estimates of 30-day Hospital Readmissions and Mortality

Posted on January 25, 2017

B. P. Smith1, C. E. Fick2, D. N. Holena1  1University Of Pennsylvania,Surgery,Philadelphia, PA, USA 2Georgetown University Medical Center,Washington, DC, USA

Introduction:  Hospital readmissions data is a critical aspect of patient care as it is directly related to patient outcomes and healthcare expenditures. Data suggest some patients experience improved outcomes with readmission to their index hospital versus another hospital. Much of the information known about re-admission for injured patients is based on single center or state level data sets, which report unplanned 30 day readmission rates between 2 and 7%. The purpose of this study was to define national estimates for trauma readmissions and compare outcomes of patients re-admitted to index hospitals compared to non-index hospitals.

Methods:  We performed a retrospective cohort study using the 2013 National Readmission Database.  Inclusion criteria were primary ICD-9CM diagnosis codes indicating injury (800.00-959.9, excluding 905-909, 910-924, 930-939), age≥18years, maximum Abbreviated Injury Score≥3, and non-elective admission.  The index hospital was defined as the center of first injury admission, and readmission was defined as any non-elective re-admission within 30 days of discharge.  The proportion of discharges readmitted to the index vs. non-index hospital were tabulated, and discharge weights and hospital strata were used to generate national estimates.

Results: After weighting, there were 350,102 trauma admissions meeting inclusion criteria (60% male, mean age 58 (SD 22) years, median Injury Severity Score 13 (IQR9-17)).  Median index length of stay was 4 (IQR2-9) days.  Of these, 31,558 (9%) had ≥ 1 30-day readmission of which only 22,372 (71%) were ever readmitted to the index hospital. Complications related to intracranial injuries, septicemia, and procedural outcomes accounted for nearly 25% of re-admission diagnoses. After adjusting for age, index length of stay, mechanism of injury, and injury severity score, diagnoses most strongly associated with re-admission to the index hospital included complications of surgery (OR 3.3, 95% CI 2.6-4.1), complications of devices (1.9, 1.5-2.4) and acute cerebrovascular accident (1.7, 1.4-2.1). Overall mortality for patients readmitted in 30 days was 4.83% and did not vary between those readmitted to index vs. other centers (4.81% vs. 4.89%, p =0.84). The most common admitting diagnosis resulting mortality in readmitted patients was sepsis, accounting for 36% of 30 day-readmission deaths.

Conclusion: Using a nationally representative dataset, we show the unplanned 30 day re-admission rate for injured patients is 9.0%, which far exceeds most single center and state reports. Most striking is that nearly 1/3 of injured patients are re-admitted to hospitals that differ from the index hospital. Although we are unable to demonstrate a mortality difference between re-admission locations, we do add crucial data to patient centered outcomes such as recovery from brain injury, and outcomes related to procedural and operative intervention for injury.

 

32.10 A Survey of Seat Belt and Helmet Use Immediately Post Collision in the Injured Patient

Posted on January 25, 2017

M. Meyer1, M. C. Spalding1, M. S. O’Mara1  1Grant Medical Center/Ohio University Heritage College Of Medicine,Trauma Surgery,Columbus, OH, USA

Introduction:
Seatbelt and helmet use have been well established as effective primary prevention measures.   A collision increases post-collision use of prevention, but knowing why those in a crash were not using the prevention in the first place could lead to better intervention strategies.  We hypothesized that patients admitted to a level one trauma center after a motor vehicle or motorcycle crash will have noncompliance with primary prevention measures (seatbelts and helmets), and the reasons for noncompliance will not align with the benefits of prevention.

Methods:
208 consecutive patients over a two month period at a level one trauma center. All patients had been involved in a motor vehicle or motorcycle collision.  Each patient gave consent and answered a 17 question survey on the circumstances of their collision, focusing on primary prevention measures use.  Demographic and injury information were also collected from the patient medical record.  43 patients were excluded, due to inability to communicate, early discharge, or being less than 18 years of age.

Results:
Seatbelt users (92/129, 71.3%) were older (46 vs. 39 years, p = 0.038), had more people in the vehicle (2.0 vs. 1.5, p = 0.004), and admitted to speeding more often (11.5% vs. 0%, p = 0.04).  Helmet wearers (11/36, 30.6%) varied only in gender, with women more likely to wear their helmets (OR 9.6, p = 0.009).  Stated reasons for seatbelt use were “habit” (43), safety (26), and required by law (24).  Most patients who were not wearing their seatbelt could not state a reason or felt it had been a bad choice (21), were not in the habit of wearing one (4), or were a backseat passenger (4).  Reasons for wearing a helmet were primarily safety (7) and habit (4).  Patients not wearing helmets could not state a reason (10) or did not like wearing helmets (7).  Two patients stated that helmets were unsafe to wear and three stated they would wear them if a law was in place to do so (p < 0.0001). Veracity was assessed by comparing patient report of alcohol use to blood alcohol screening.  11% reported alcohol use before operating their vehicle, while 22% had a positive alcohol level (p = 0.007).  12% denied their proven alcohol consumption.  32% not wearing their helmets denied proven consumption.

Conclusion:
Consistent use of primary prevention devices relies upon establishing them as a habit.  Education and mandated laws are good ways of establishing this habit.  We saw this consistent pattern across our patients who had just been injured in a crash.  Laws can overcome individual dislike of a device, eventually establishing a habit.  Intervention at the time of injury may be useful in improving compliance, as many patients could state that not using prevention was a bad choice.  More important is the group of patients that still maintained their choice to not put on a helmet was the right one.  For them, education might save their life.
 

32.09 Computed Tomography In Trauma: Effect Of Provider Experience And Training Level On Patterns Of Use

Posted on January 25, 2017

K. Habeeb1, T. R. Wojda1, A. Z. Hasani2, J. D. Nuschke2, Z. K. Zhang2, B. A. Hoey1, W. S. Hoff1, P. G. Thomas1, S. P. Stawicki1  1St. Luke’s University Health Network,Department Of Surgery,Bethlehem, PENNSYLVANIA, USA 2Temple University,St. Luke’s University Hospital Campus,Bethlehem, PA, USA

Introduction: Our group’s previous work suggested that greater "tonnage" of computed tomography (CT) for trauma may be associated with lower mortality. However, questions remain regarding the association between CT scan utilization and traumatologist level of experience. Based on empirical observations, we hypothesized that increasing provider experience may be associated with lower reliance on CT scanning, and that trauma fellows utilize CT imaging more than attendings.

Methods: Institutional registry consisting of 32,026 records (Jan 1998 – Dec 2015) at our Regional Level I Trauma Center was reviewed, excluding 4,346 patients who underwent emergency surgery or died before CT imaging was performed. The resulting sample was analyzed for: mortality, trauma provider level of training/experience, and CT scan “tonnage” per provider. We also collected demographic and injury information (gender, age, injury severity score [ISS], revised trauma score [RTS], mechanism). We then compared CT utilization and mortality between attendings and fellows during trauma resuscitation events (TRE). Data analyses were carried out using Chi-squared testing, Mann-Whitney U-testing or Analysis-of-Covariance (ANCOVA, correcting for injury mechanism and demographics), with statistical significance set at α=0.05.

Results: A total of 27,372 patient records were analyzed (60.3% male, median age 45 yrs, 95% blunt trauma, median ISS 5.00, median RTS 7.84, median hospitalization of 2 days). Seventy-nine ATLS-certified traumatologists (12 attendings, 67 fellows) were examined. Median mortality per traumatologist was 2.3%, with median number of 2.2 CT scans per TRE. There was no difference in average utilization of CT scans among attendings (2.1±0.1 per TRE) and fellows (2.2±0.1 per TRE). Patient mortality did not differ when the trauma team was led by an attending (3.7±0.2%) versus a fellow (3.3±0.4%). The number of CT’s per provider decreased with provider experience, with the median number of scans per TRE declining from 2.1 during the first decade of clinical experience to 1.9 during the subsequent decade in practice (p<0.05). The median number of CT scans for first year attendings (1.8 per TRE) was significantly lower than for first year fellows (2.2 per TRE, p<0.04). While the number of CTs per TRE increased to 2.3 among second year fellows (p<0.05), the same was not true for second year attendings.

Conclusion: We found important correlations between traumatologist level of experience and CT scan utilization. Despite lower utilization of CT scans among attendings, there was no associated mortality difference. Based on the overall number of CT scans performed during the entire study period, potential cost savings associated with fellows utilizing advanced imaging in-line with attending levels would amount to nearly $13 million, highlighting the need for clinical education in this important area.
 

32.08 Trends in Firearm Related Injuries in Children and Young Adults Admitted to US Hospitals

Posted on January 25, 2017

M. Nuno1, M. K. Srour1, A. V. Lewis1, R. F. Alban1  1Cedars-Sinai Medical Center,Trauma And Critical Care Surgery,Los Angeles, CA, USA

Introduction:
Firearm violence in the USA results in the injury and death of thousands of individuals annually. In an effort to curtail this public health concern, firearm prevention strategies such as the Brady handgun violence prevention act – the Brady law have been proposed. Given the gaps inherent is some of these laws, these prevention strategies have resulted in limited success.  The objective of this study was to evaluate the role of state-level gun laws, age and race on firearm related injuries and mortality among children and young adults admitted to US hospitals.

Methods:
A total of 27,566 children and young adults were identified using the Kids’ Inpatient Sample (KID) database (2000, 2003, 2006, and 2009). Data was obtained from the Healthcare Cost and Utilization Project (HCUP) of the Agency for Healthcare Research and Quality (AHRQ) and all statistical analyses were conducted in SAS 9.2. Trends of injuries were explored in terms of state-level gun laws, age, and race. Admitting hospitals were stratified into 5 categories (A, B, C, D and F, with A representing states with the most strict and F states with the least strict laws) based on the Brady Campaign to prevent Gun Violence that assigns scorecards for every state. Descriptive statistics were provided and multivariate logistic regression was applied to evaluate factors associated with in-hospital mortality.

Results:
A total of 27,566 children and young adults were analyzed in this study. Most patients were young adults of age 15-19 years (87.3%), male (89.7%), black (53.7%), and admitted as emergent/urgent cases (85.0%). Most patients were discharged from teaching (81.9%) hospitals with large bedsize (70.9%), and located in southern states (34.2%). States with weaker gun laws had an increased rate in accidents while more assaults were documented in states with stronger gun laws. Accidents were significantly more common in children age 0-4 while assaults were prevalent in younger adults. Whites experienced more firearm related accidents while Black and Hispanics were victims of more assaults. Overall mortality was 6.4%; after adjusting for multiple factors we found that race (p=0.009), age (p<0.0001), and the type of firearm related injury (p=0.0011) were associated with mortality. Hispanics compared to Whites (OR 1.36, 95% CI: 1.03-1.78), children age 5-9 (OR 2.03, 95% CI: 1.30-3.17) compared to young adults (15-19), and suicides (OR 15.6, 95% CI: 11.6-20.9) in comparison to accidents had an increased risk of in-hospital mortality.

Conclusion:
Firearm related injury type was strongly correlated with state-level gun laws, age and race of victim. Accidents were most prevalent in states with weak gun laws, young children and Whites while assaults prevailed in states with stricter gun laws, young adults, and Black and Hispanics. Further disparities in mortality were found by race, age, and type of injury.
 

32.07 The Impact Of Gcs-age Prognosis (Gap) Score On Geriatric Tbi Outcomes

Posted on January 25, 2017

M. Khan1, A. Azim1, T. O’Keeffe1, L. Gries1, K. Ibraheem1, A. Tang1, G. Vercruysse1, R. Friese1, B. Joseph1  1University Of Arizona,Trauma And Surgical Critical Care/Department Of Surgery,Tucson, AZ, USA

Introduction:
As the population ages, increasing number of elderly patients sustain traumatic brain injury (TBI). Communication of accurate prognostic information plays a crucial role in informed decision making for these patients. The aim of our study was to develop a simple and clinically applicable tool that accurately predicts the prognosis in geriatric TBI patients

Methods:
One-year (2011) retrospective analysis of geriatric TBI patients (h-AIS≥3 and age≥65) in the National Trauma Data Bank was performed and patients dead on arrival were excluded. We defined and calculated a GCS and Age Prognosis (GAP) score (Age/GCS score) for all patients. Our outcome measures were mortality and discharge disposition (Home versus Rehab/SNiF). ROC analysis was performed to determine the discriminatory power of GAP score.

Results:
A total of 8,750 geriatric patients with TBI were included. Mean age was 77.8± 7.1 years, median [IQR] GCS was 15 [14-15], and median [IQR] head-AIS was 4[3-4]. Overall mortality rate was 14.1% and 42.7% patients were discharged home. As the GAP score increased, mortality rate increased and discharge to home decreased. ROC analysis revealed excellent an discriminatory power for mortality (AUC: 0.826). Above a GAP score of 12, mortality rate was greater than 60%, more than 35% patients were discharged to Rehab/SNif and less than 5% of patients were discharged home.

Conclusion:
For geriatric patients with TBI, a simple GAP score reliably predicts outcomes. A score above 12 results in drastic increase in mortality and adverse discharge disposition. This simple tool may help clinicians provide accurate prognostic information to patient families.
 

32.06 Standard Enoxaparin Dosing Provides Inadequate Thromboprophylaxis in Orthopaedic Trauma

Posted on January 25, 2017

D. L. Jones1, A. Prazak2, K. I. Fleming3, T. Higgins1, C. J. Pannucci3  1University Of Utah,Orthopaedic Surgery,Salt Lake City, UT, USA 2University Of Utah,Salt Lake City, UT, USA 3University Of Utah,Plastic Surgery,Salt Lake City, UT, USA

Introduction:
Fixed doses of enoxaparin are routinely used in orthopaedic trauma surgery to lower the risk of perioperative venous thromboembolism (VTE). Despite prophylaxis, however, breakthrough VTE events remain high, particularly those with immobilizing lower extremity or pelvic fractures. Based on anti-Factor Xa levels (aFXa), a growing body of literature demonstrates that this “one size fits all” approach to standard enoxaparin dosing leaves a significant number of patients inadequately prophylaxed and vulnerable to VTE events. We explored enoxaparin metabolism in orthopaedic trauma patients on twice per day enoxaparin by monitoring peak and trough anti-Factor Xa (aFXa) levels as well as their association with gross weight.

Methods:
We prospectively enrolled post-operative orthopaedic trauma patients undergoing acute fracture or non-union surgery. All patients received enoxaparin prophylaxis at 30mg twice per day, initiated within 36 hours after surgery. Steady-state peak and trough aFXa levels, which measure enoxaparin effectiveness and safety, were drawn four and twelve hours after the third dose, respectively. Goal peak aFXa levels were 0.2-0.4IU/mL and goal trough levels > 0.1IU/mL. Patients with out of range peak aFXa levels had real time enoxaparin dose adjustment based on a written protocol, followed by repeat aFXa levels. Stratified analyses examined variation in peak aFXa by patient weight.

Results:
To date, 60 orthopaedic patients on 30mg twice-daily enoxaparin have been enrolled. Initial peak aFXa levels were out of range in 46.7% of patients, with 10% having undetectable levels. Trough aFXa levels were undetectable in 81.3% of patients. Dose adjustment resulted in 50% more patients reaching in-range levels. Gross weight was strongly associated with peak steady state aFXa level (Figure 1; grey box represents appropriate, in range aFXa levels). Patients with gross weight over 75 kg were significantly more likely to have inadequate aFXa levels when compared to patients ≤75 kg (63.2% vs. 13.6%, p=0.0003). 

Conclusion:
Enoxaparin 30mg twice daily provided inadequately prophylaxis in nearly half of all of orthopaedic trauma surgery patients. A gross weight > 75 kilograms resulted in a significantly higher likelihood to have inadequate peak aFXa levels. Given the number of patients under-prophylaxed, inadequate enoxaparin dosing may explain some breakthrough VTE events seen in orthopaedic trauma. A weight-based enoxaparin dosing protocol may provide a more satisfactory strategy to VTE prophylaxis. Future directions aim to correlate VTE and bleeding events with peak and trough aFXa levels after orthopaedic trauma surgery and explore the effect of injury severity on predicting enoxaparin metabolism.
 

32.05 Asymptomatic Screening in Trauma Patients Reduces Risk for Pulmonary Embolism

Posted on January 25, 2017

D. Koganti1, A. Johnson1, S. Stake1, A. Wallace1, S. Cowan1, J. Marks1, M. Cohen1  1Thomas Jefferson University,Surgery,Philadelphia, PA, USA

Introduction:
Now that deep vein thrombosis (DVT) is linked to reimbursement and publicly reported metrics, hospitals are pressuring trauma programs to discourage lower extremity (LE) venous duplex ultrasounds (VDUS) in asymptomatic patients. Current evidence is ambiguous and controversial. We aimed to evaluate LE VDUS screening practices at our institution for risk reduction for pulmonary embolism (PE).

Methods:
Patients admitted to an urban level-1 trauma center between 2005 and 2015 were retrospectively reviewed, excluding patients with a length of stay (LOS) <4 days. We performed propensity-matching of screened to unscreened patients based on gender, transfer, spinal procedure, spinal cord injury, or spinous, femur, pelvis, tibia and upper extremity fracture. In our matched samples, we performed a chi-squared analysis to determine association of screening with PE, absolute risk reduction and number needed to treat.

Results:
Of the 11,280 trauma patients admitted, 5,611 met LOS criteria. Of these patients, 2,687 (48%) underwent asymptomatic LE VDUS screening. Propensity matching identified 1,915 unscreened patients with a similar risk profile. The rate of PE was significantly higher in our matched unscreened sample [1.72% (n=33) vs 0.45% (n=12), p<0.001, Figure]. The absolute risk reduction was 1.28%, suggesting that the number needed to screen to prevent one PE is 78 high-risk patients.

Conclusion:
The data demonstrate significant risk reduction for pulmonary embolism in propensity-matched patients at our institution over a 10-year period. The screened patients still have a higher risk factor profile than the matched cohort suggesting that the actual risk reduction might even be greater than 1.28%. This data can help define the best population for routine screening and determine the cost-effectiveness of screening programs.
 

32.04 Environmental and Community Determinants of Injury Mortality

Posted on January 25, 2017

M. P. Jarman1, R. C. Castillo1  1Johns Hopkins Bloomberg School Of Public Health,Department Of Health Policy And Management,Baltimore, MD, USA

Introduction: Disparities in access to trauma care and injury mortality persist for rural, low income, and minority populations despite nearly 50 years of effort to regionalize and standardize trauma care in the US. Little is known about the contribution of injury incident location to these disparities. This study sought to examine the role of environmental and community-level factors in predicting injury mortality.

Methods: Injury incident locations (n = 11,070) in the 2015 Maryland eMEDS Patient Care Reporting system were geocoded and linked with individual and hospital characteristics drawn from the Maryland Adult Trauma Registry, as well as environmental factors present at each injury scene using data from the Maryland Department of Planning and the Maryland State Highway Administration, and community-level factors at the census tract level from the United States Census Bureau. Multivariate logistic regression models were used to estimate odds of death associated with environmental factors present at the scene of the injury incident and community-level factors at the census tract level, while controlling for total pre-hospital time, injury severity, comorbidities, age, sex, and race.

Results: Relative to patients who travel less than 25 miles from the injury incident scene to a trauma center, patients who traveled 50-75 miles were 3.88 times more likely to die (p = 0.003), and patients who traveled 75-100 miles were 7.15 times more likely to die (p < 0.001). Odds of death for patients traveling 25-50 miles did not differ from those traveling less than 25 miles. Compared to commercial land use, patients who were injured at locations with residential land use were 53% more likely to die (p = 0.038), and those injured at locations with transportation land use were 2.00 times more likely to die (p = 0.079). Odds of death increased by 5.90% for every 5% increase in the proportion of residents using private vehicles to commute to/from work (p = 0.030), and by 8.67% for every 5% increase in the proportion of residents with commutes longer than 25 minutes (p = 0.004)

Conclusions: Distance from the injury scene to a trauma center appears to be a significant determinant of injury mortality, independent of pre-hospital time. Residential and transportation land and community characteristics related to transportation also appear to increase odds of injury mortality. These factors may contribute to persistent disparities in trauma mortality. The findings of this study can inform policy and practice decisions regarding organization of trauma systems, delivery of pre-hospital care, and injury prevention in geographic areas at high risk for fatal injuries.

32.02 Effectiveness of ATOMAC Guideline for Blunt Pediatric Injury: A 3-Year 10-Center Prospective Study

Posted on January 25, 2017

D. M. Notrica1,11,12, C. S. Langlais1, M. E. Linnaus1,11, K. A. Lawson2, J. W. Eubanks6, A. C. Alder3, N. M. Garcia2, R. W. Letton5, D. W. Tuggle2, T. Ponsky8, D. Ostlie1, A. Bhatia10, S. D. St. Peter9, C. Leys7, R. T. Maxson4, D. M. Notrica1,11,12  1Phoenix Children’s Hospital,Phoenix, AZ, USA 2Dell Children’s Medical Center,Austin, TX, USA 3Children’s Medical Center Dallas, Part Of Children’s Health,Dallas, TX, USA 4Arkansas Children’s Hospital,Little Rock, AR, USA 5The Children’s Hospital At OU Medical Center,Oklahoma City, OK, USA 6LeBonheur Children’s Hospital,Memphis, TN, USA 7American Family Children’s Hospital,Madison, WI, USA 8Akron Children’s Hospital,Akron, OH, USA 9Children’s Mercy Hospital,Kansas City, MO, USA 10Children’s Healthcare Of Atlanta,Atlanta, GA, USA 11Mayo Clinic,Phoenix, AZ, USA 12University Of Arizona College Of Medicine – Phoenix,Phoenix, AZ, USA

Introduction: Prior guidelines had required bedrest equal to the grade of injury +1 day. The ATOMAC guideline is an evidence-based published guideline for management of pediatric blunt liver and spleen injury (BLSI). The guideline allows for an abbreviated period of bedrest, and provides a detailed algorithm for management. The purpose of this study was to prospectively evaluate the effectiveness of the algorithm to safely guide care and confirm the safety of the abbreviated bedrest included in the algorithm.

Methods: After IRB approval, data was prospectively collected on patients ≤18 years of age admitted with a BLSI identified by Computed Tomography. Data collected included injury details, hospital details, and clinical outcomes. The algorithm was amended during the study to make early recurrent hypotension a failure point. Descriptive statistics are reported. Length of stay (LOS) was compared to a LOS equal to grade + 1 day.

Results: A total of 1008 children were included; 499 liver injuries (50%), 410 spleen injuries (41%), and 99 with both (10%).  Median age was 10.3 years [IQR 5.9, 14.2]. At initial presentation, 286 (28%) had recent or ongoing bleeding and were assigned to the bleeding pathway; 242 (24%) were tachycardic and 129 (13%) were hypotensive. Concomitant traumatic brain injury was present in 189 (19%).  There were 23 in-hospital deaths (2.5%), 2 due to bleeding. Of the 717 patients clinically assessed and started on the stable pathway, 10 (1.5%) crossed over to the algorithm’s unstable pathway. While minor deviations were common, only 1 patient (0.1%) was at risk of a negative outcome if they followed the original algorithm, resulting in the algorithm amendment. In patients with isolated injuries, median [IQR] lengths of stay by grade of injury (in days) were 0.94 [0.75, 2.17], 1.21 [0.83, 1.89], 1.65 [1.17, 2.08], 2.00 [1.46, 3.29], and 3.23 [2.35, 4.88] for isolated injuries grade 1-5, respectively, totaling 678 days, compared to an expected LOS of 1,211days.

Conclusion:The original ATOMAC guideline was safely applied to 99.9% of 1008 children with BLSI.  With the modification for recurrent hypotension in the guideline published last year, the guideline could have safely guided care for 100% of the children with BLSI. Ninety-one (9%) patients reached the algorithm endpoint where continued NOM could no longer be recommended; 22 (24%) of these were still managed nonoperatively at the surgeon’s discretion. Ten patients (1.5%) crossed over from the stable to the unstable pathway. The algorithm saved 533 hospital days over the prior guideline. In the largest prospective study ever conducted of pediatric BLSI, the ATOMAC guideline performed well in guiding non-operative management of patients with BLSI.

 

32.01 Empowering Bystanders to Intervene: The Health Belief Model and Chicago’s South Side

Posted on January 25, 2017

L. C. Tatebe1, S. Speedy1, S. Regan3, A. Boone1, F. Cosey-Gay2, L. Stone3, M. Shapiro1, M. Swaroop1  3University Of Illinois At Chicago,CeaseFire Chicago,Chicago, IL, USA 1Northwestern University,Trauma/Critical Care,Chicago, IL, USA 2University Of Chicago,Chicago, IL, USA

Introduction:  The paucity of trauma centers in the south side of Chicago leads to prolonged transport times and increased morbidity and mortality for those affected by penetrating trauma. An evidence-based community-driven Trauma First Responders Course (TFRC) could potentially mitigate this effect, but does not currently exist. Bystanders are present at 60-97% of traumas and are more likely to assist with prior training. However, the bystander effect remains a major barrier. We hypothesize that by utilizing the Health Belief Model as a framework, we can characterize the factors in our community that lead to bystander non-interference. Through this, neighborhood focus groups will facilitate effective course development and thus improved patient outcomes and community empowerment.

Methods:  The Health Belief Model determines the likelihood of an action by examining individual perceptions of susceptibility, self-efficacy, and severity of a health issue, then applying the modifying factors of knowledge and cues to action. The resulting perceived threat is then modulated by perceived benefits and barriers of performing the action. Written surveys and focus group questions were developed to specifically address each of these facets. Focus groups were conducted by 2 guides over an hour with 8-10 community members, including youth at highest risk for violence and key community leaders. Data were collected via surveys and recordings and analyzed quantitatively as well as qualitatively.

Results: The focus groups demonstrated consistency across many of the factors examined, see Figure. Participants felt the perceived susceptibility for witnessing an injury was high. Half stated they worry that they or someone they know will get hurt "a lot" or "all of the time" and that they are "unsure," "not confident," or "not at all confident" in their ability to render aid, demonstrating a sense of low self-efficacy. Only 39% of responders stated they had any form of first aid training, and less than 10% stated they had advanced training. A majority said they did not know when to intervene, which often stemmed from a concern that interference would lead to increased harm. There was significant fear of social or legal retaliation for helping a victim of violence, while 92% would want a stranger to help if they or someone they know was injured. Overall, the likelihood of bystander intervention was deemed to be low.

Conclusion: Critical factors have been identified through the structure of the Health Belief Model that contribute to bystander non-intervention in our community. These will need to be addressed during the development and implementation of an effective TFRC to empower community members to overcome the bystander effect.

31.10 Market Competition Influences Access to Renal Transplantation

Posted on January 25, 2017

J. T. Adler1, D. C. Chang1, H. Yeh1  1Massachusetts General Hospital,Boston, MA, USA

Background: In kidney transplantation, practice patterns and utilization vary across the 58 Donor Service Areas (DSAs) in the United States.  However, most patients would assume that his or her treatment is based on unique patient factors, and not on any economic or healthcare market factors.  Nevertheless, market competition has been shown to contribute significantly to variability in both patient and graft outcomes after kidney transplantation. Relatively little is known about how competition affects access to transplantation (ATT) after the initiation of dialysis. Understanding how competition relates to ATT may improve outcomes and access in kidney transplantation.

Methods:  The United States Renal Data System, which contains all new registrations for dialysis, was queried for adult incident dialysis registrations in 2008; final follow-up was through June 2013. Cox proportional hazards models were used to estimate the time from onset of dialysis to ATT, which was defined as enrolling on the kidney waitlist or receiving a living donor kidney transplant. To measure market competition, the Herfindahl-Hirschman Index (HHI) was calculated for each DSA and categorized as no (HHI 1), low (HHI 0.52-0.97), medium (HHI 0.33-0.51), or high (HHI 0.09-0.32) competition.  

Results: 104,789 adults began dialysis in 2008. After adjusting for age and comorbidities (Table), increasing competition was strongly associated with increased ATT (HR 1.57, P < 0.001, high vs. no competition). Additional factors associated with increased ATT were male sex (HR 1.16, P < 0.001) and being under the care of a nephrologist at time of dialysis initiation (HR 1.70, P < 0.001). Compared to Whites, Asian patients had increased ATT (HR 1.38, P <0.001), while Black patients had decreased ATT (HR 0.77, P < 0.001).

Conclusion: Even after adjusting for comorbidities and other known patient factors, market competition plays a strong role in ATT. Understanding the interplay between competition and the conduct of kidney transplantation may improve ATT, and it deserves further investigation.

 

31.09 Nationwide Assessment of Readability of Liver Transplant Education

Posted on January 25, 2017

Y. J. Bababekov1, F. C. Njoku1,2,3, B. Cao1,2,4, J. T. Adler1, J. J. Pomposelli5, D. C. Chang1, H. Yeh2  1Massachusetts General Hospital,Codman Center For Clinical Effectiveness In Surgery/ Surgery/Harvard Medical School,Boston, MA, USA 2Massachusetts General Hospital,Transplantation/Surgery/Harvard Medical School,Boston, MA, USA 3University Of California – Irvine,School Of Medicine,Irvine, CA, USA 4University Of Rochester,School Of Medicine,Rochester, NY, USA 5University Of Colorado Denver,Transplant Surgery/Surgery/School Of Medicine,Aurora, CO, USA

Introduction:  Patient education during the transplant process is known to increase compliance and access to transplantation which predict better patient outcomes. Health literacy is associated with compliance; however, patients and caregivers may have suboptimal liver transplant health literacy. In response to the Plain Writing Act of 2010, the Centers for Disease Control and Prevention developed the Clear Communication Index (CCI) to create and assess public communication materials with a focus on readability. This is of increasing importance given the organ donor shortage and unacceptably high rate of rejection from non-adherence to treatment protocols. However, transplant center educational content varies and the readability of educational materials is unknown. We therefore assessed the existence and quality of education materials available on line at United Network for Organ Sharing (UNOS) and liver transplant centers across the United States.

Methods:  The Scientific Registry of Transplant Recipients was searched for adult liver transplant centers (n=115). Two independent reviewers assessed patient education information on the websites of UNOS and all liver transplant centers for readability via the Clear Communication Index (CCI), developed by the Center for Disease Control and Prevention. The CCI is a validated tool of 20 scored items that yield a percentage score; a score ≥ 90% indicates the communication material is easy to understand. 

Results: All liver transplant centers had websites, but only 39% (45/115) had printable PDFs; UNOS had both a website and PDF material. The median CCI score for websites and PDFs was 72.6% (IQR 67.5%, 75.0%) and 70% (IQR 64.3%, 75.0%), respectively. No patient information website or PDF had a CCI score ≥ 90% (Figure 1). State, organ procurement organization (OPO), number of centers per OPO, or UNOS region did not predict CCI scores for websites or PDFs. There was no difference in CCI scores by reviewer.

 

Conclusion: Patient education materials for liver transplantation may be presented in a way that is not easy to understand for patients and caregivers, and would benefit from increased quality of readability. Access to care and patient compliance may be improved with revision of patient education materials. 

« Previous Page
Next Page »