53.14 Trauma/Acute Care Surgeries Performed in the Emergency Room Impact on the Timeliness.

Posted on January 18, 2018

K. Ito1, K. Nakazawa1, T. Nagao1, H. Chiba1, T. Fujita1  1Teikyo University Hospital Trauma And Resuscitation Center,Department Of Emergency Medicine, Division Of Acute Care Surgery, Teikyo University School Of Medicine,Tokyo, , Japan

Introduction:  Timely surgical interventions are important for patients with trauma or acute general surgical diseases. At our institution, we have emergency rooms (ERs) with the operating room (OR) set-up which allow surgeons to perform thoracotomy and/or laparotomy without transferring patients to the OR. This practice pattern is not standard in the United States or other western countries. We conducted this study to assess the overall outcomes of this practice, as well as, to test the hypothesis that the ERs with OR set-up improve the timeliness of surgery for trauma / acutely ill general surgical patients.

Methods:  Patients who underwent emergent surgeries by our acute care surgery group (4/2013 – 6/2017) were reviewed. Patients’ demographics, diagnoses, the location of operation (ER vs regular OR),  type of operations, time from admission to operation, preoperative interventions, postoperative outcomes, and in-hospital mortality were analyzed. These data were compared with patients who underwent surgery in the ER (ER group) and patients who underwent surgery in the OR (OR group). Parametric data were analyzed by Chi-square test. Non-Parametric data were analyzed by Mann-Whitney U test.

Results: There were 322 consecutive patients (105 traumas [33%] and 217 emergent general surgeries [67%]) who met inclusion criteria. Among them, there were 68 patients who underwent surgery in the ER (21%, ER group) and 254 patients in the OR (79%, OR group). Compared to the OR group, The ER group had more trauma patients (74% vs 24%, p<0.001). The time from admission to operation was shorter in the ER group than the OR group (median 57 minutes [range 4 – 1069] vs 170 minutes [range 25 – 1320], p<0.001). For trauma patients, the Injury Severity Score was higher in the ER group than the OR group (median 34 [range 1 – 59] vs 9 [4 – 45], p<0.001). The in-hospital mortality rate was higher in the ER group than the OR group (42% vs 13%, p<0.001). Details of surgeries in the ER were shown on the Table 1.

Conclusion: The ERs with the OR set-up can allow surgeons to start surgery quicker. These surgeries tended to be performed in sicker patients and likely associated with higher. Further study is warranted for selecting patients who benefit best from this approach.

 

53.11 Elderly Pedestrians Struck by Vehicles Are More Likely to Be Admitted in the Morning

Posted on January 18, 2018

D. C. Patel1, T. Li1, N. K. Dhillon1, N. T. Linaval1, L. Kirillova1, D. R. Margulies1, E. J. Ley1, G. Barmparas1  1Cedars-Sinai Medical Center,Los Angeles, CA, USA

Introduction:  Elderly patients are at high risk for mortality following traumatic injury and prevention is of paramount importance. This study’s aim was to identify specific times of the day during which the elderly are more likely to be injured.

Methods:  The Los Angeles County TEMIS database was retrospectively reviewed for all pedestrians ≥ 18 years who were struck by vehicles over a 16-year period (2000-2015). Elderly (≥ 65 years) patients were compared to non-elderly (18-64 years) with respect to the time of admission and mortality.

Results: Of 36,358 patients, 4,871 (13%) were elderly. The proportion of elderly pedestrian injuries was highest between 6-11am (average 21%), peaking between 10-11 am (23%). This proportion started to abruptly drop at noon, reaching the lowest levels between 2-3 am (2%). Compared to their non-elderly counterparts, the elderly were more likely to have a Glasgow Coma Scale ≤ 8 (11% vs. 7%, p < 0.01), a systolic blood pressure < 90 mmHg (6% vs. 3%, p < 0.01), and a higher Injury Severity Score (median 9 vs. 5, p < 0.01). Overall mortality was 6%, significantly higher in the elderly (14% vs. 4%, p<0.01). Elderly pedestrians had almost a 7-fold higher adjusted odds for death (AOR: 6.8, p<0.01).

Conclusion: Elderly pedestrians struck by vehicles are more likely to be admitted during the morning hours and their mortality risk is high. Preventative strategies with lower speed limits or high surveillance during morning hours in highly populated areas may result in a decreased incidence of these injuries.

53.12 The Depth of Sternal Fracture Displacement is Not Associated with Blunt Cardiac Injury

Posted on January 18, 2018

L. Heidelberg1, R. Uhlich1, P. Bosarge1, J. Kerby1, P. Hu1  1University Of Alabama at Birmingham,Acute Care Surgery,Birmingham, Alabama, USA

Introduction:
Despite little evidence to support an association of sternal fractures with blunt cardiac injury (BCI), displaced sternal fractures are viewed as more severe and more likely to result in BCI. Common recommendations include inpatient evaluation for observation and additional diagnostic screening. Little information exists regarding the depth and severity of sternal fracture displacement and it remains unclear if there is increased risk of BCI with increasing sternal fracture depth. The purpose of this study was to quantify fracture severity by the degree of displacement and evaluate the association of fracture severity with BCI.

Methods:
A retrospective review was performed at an American College of Surgeons verified level 1 trauma center from 2011-2014. All adult patients admitted to the trauma surgery service were eligible for inclusion, with patients excluded for pregnancy, age <18 years old, or lack of imaging. Patients with sternal fracture were identified from the trauma registry using ICD-10 codes. Sternal fracture displacement was measured by posterior displacement in the axial plane of computerized tomography of the chest. Fracture displacement was defined as mild (>0 mm, <5 mm), moderate (≥5 mm, <10 mm), or severe (≥10 mm). BCI was diagnosed using electrocardiogram or echocardiograph and graded according to standard AAST grading.  Analysis was performed using χ2 and Student's t-test or one-way ANOVA for categorical and continuous variables respectively. Multivariate regression analysis was subsequently performed to assess the association of sternal fracture displacement with BCI. The primary outcome of interest was the association of BCI with severity of sternal fracture displacement. 

Results:

235 patients with sternal fractures were identified and eligible for inclusion during the study period. 45% of patients suffered a displaced fracture, with the majority of these being mild (65.1%) or moderate (24.5%) in severity. Only 10.4% of patients had severely displaced fractures. Overall, 42.6% of patients were diagnosed with BCI. Of patients with BCI, there was no difference in mean fracture displacement when compared to patients without BCI (2.4 vs 1.6 mm, p=0.07)[Office1] . There was no significant increase in BCI with sternal fracture displacement when compared to patients with non-displaced fractures (44.3% vs 41.1%, p=0.62)[Office2] . While among patients with displaced fractures, the incidence of BCI increased with increasing severity (39.1% vs 50.0% vs 63.6%, p=0.25), this was not significant. Further, neither fracture displacement (OR 1.10, CI 95% 0.65-1.88) nor severe displacement (OR 2.34, CI 95% 0.64-8.54) were associated with significantly increased risk of BCI on multivariate analysis.

Conclusion:
There is no significant association between the depth of sternal fracture displacement and BCI. Further evaluation and management for BCI should be reserved in the absence of additional symptoms or findings.

53.09 A Review of Hydroxocobalamin Use in Patients with Inhalation Injury at a Regional Burn Center

Posted on January 18, 2018

J. S. Vazquez1,2, L. S. Johnson1,2, T. E. Travis1,2, L. T. Moffatt2, J. W. Shupp1,2  1MedStar Washington Hospital Center,The Burn Center, Department Of Surgery,Washington, DC, USA 2Firefighters’ Burn And Surgical Research Laboratory,Washington, DC, USA

Introduction:
Hydrogen cyanide is produced during combustion of several different types of household materials such as synthetic polymers. This small lipid soluble molecule inhibits oxidative phosphorylation and produces a severe lactic acidosis in affected patients. Early identification of cyanide (CN) toxicity allows for treatment with Hydroxocobalamin, a CN binding agent that forms cyanocobalamin, a non-toxic substance excreted by the kidneys. However, even though the medication has a mild safety profile, its high cost should guide judicious use. We evaluated the appropriateness of a single burn center’s administration of Hydroxocobalamin in patients with inhalation injury suspected of having CN intoxication.

Methods:
In this single center retrospective study, pharmacy records where queried for patients that received Hydroxocobalamin between January 2014 through June 2017. Twenty-seven patients received treatment based on the suspicion of inhalation injury and CN toxicity and where included in this study. The electronic medical record of those patients was queried to collect data regarding survival status, clinical parameters, and details surrounding Hydroxocobalamin administration. Blood CN levels were drawn either prior to or immediately after Hydroxocobalamin administration in 20 patients. Clinical criteria for out of hospital administration of Hydroxocobalamin by EMS was decreased GCS in the setting of inhalation injury. As for patients arriving to the trauma bay clinical criteria for physician administration of Hydroxocobalamin was acidosis out of proportion to CO2 retention and an elevated CO level in the setting of inhalation injury.

Results:
A total of 27 patients (mean age 48.9 years; 17 men) were treated with Hydroxocobalamin based on the suspicion of CN poisoning in the setting of inhalation injury. House fires (74%) were the most common cause injury overall. Sixteen patients had GCS <8. A concomitant burn injury was present in 78% of patients (%TBSA 0-95). After correcting for smoking status, 70% of patients had CN levels above normal range (Max 2.79mg/L). Presence of an elevated CN level was used as a surrogate for accuracy of administration. All patients arrived at the burn center within the cyanide half-life. Mortality was 37% in this patient cohort.

Conclusion:

A high index of suspicion should be maintained for the presence of CN toxicity in patient with concomitant burn and inhalation injury. Treatment for CN poisoning should be initiated in any patients with inhalation injury, unexplainable lactic acidosis, or impaired consciousness that cannot be explained only by CO poisoning. In this study, we demonstrated that the medication was administered correctly in 70% of patients. Given the broad therapeutic index of Hydroxocobalamin and the mortality associated with cyanide toxicity, this medication should be used when the correct clinical scenario is present, although the high cost might be a barrier for some centers around the world.

53.10 The Bigger They Are, the Harder They Fall: Obesity and Severity of Proximal Humerus Fractures

Posted on January 18, 2018

R. Belayneh1,2, J. Haglin1, A. Lott1, S. Konda1, K. A. Egol1  1New York University School Of Medicine,New York, NY, USA 2Howard University College Of Medicine,Washington, DC, USA

Introduction: The prevalence of obesity in adults has increased significantly in the United States and worldwide. It has been extensively reported in the literature to cause not only medical problems, but musculoskeletal issues as well. In addition to being associated with high rates of osteoarthritis as compared to normal weight populations, obese patients have a 48% increased risk of trauma, including minor injuries and fractures.  Obesity is also associated with an increased risk of injury to the upper limbs resulting from falls from an individual’s own height, including proximal humerus fractures, which account for 4-6% of all fractures. The purpose of this study is to evaluate if there is a relationship between obesity and proximal humerus fracture characteristics.

Methods:   : Proximal humerus fractures at one academic medical center were prospectively followed. Fractures were classified according to the international AO-Müller/Orthopedic Trauma Association (AO/OTA) classification in order to determine their severity. All Type OTA 11-A proximal humerus fractures were categorized as less severe and type OTA 11-B or 11-C were categorized as more severe. Patients’ Body Mass Indexes (BMI) were calculated and used to identify two groups, BMI ≥ 30 kg/cm (obese) and < 30 kg/cm (non-obese). Variables such as age, gender, height, weight, Charlson Comorbity Index (CCI), AO classification; number of complications, latest follow-up shoulder range of motion (ROM), latest follow-up Disabilities of the Arm, Shoulder, and Hand (DASH) survey scores were also recorded. Independent t-tests were used for statistical analysis of continuous variables and χ2 tests for categorical variables. Regression analysis was performed to determine if BMI was a predictor of severity of fractures as determined by the AO classification. Statistical significance was considered as  p<0.05.

Results: Overall, 190 patients who sustained proximal humerus fractures were available for analysis where the average age at time of injury was 59.6±13.9 years. There were 56 OTA 11-A, 67 OTA 11-B, and 67 OTA 11-C fracture types. 58 patients (30.5%) were obese, while 132 patients (69.5%) were non-obese. No significant differences were seen between groups in regards to age, gender, height, CCI, complication rates, or functional and clinical outcomes as determined by follow-up DASH scores and shoulder ROM, respectively. Statistical analysis also demonstrated that obese patients had greater fracture severity per the AO classification (P=0.025).

Conclusion: Based on the results of this study, obesity is associated with more severe fractures of the proximal humerus as determined by the AO/OTA classification. However, there are no differences outcomes or complication rates between obese patients and non-obese patients. With increasing rates of obesity, this relationship may have important epidemiological implications in the future, including predicting proximal humerus fracture burden and severity in society. Additionally, orthopaedic surgeons should be reassured that performing proximal humerus fixation in obese patients yields similar outcomes and complication rates to non-obese patients.
 

53.07 Impact of Red Blood Cell Transfusion in Severe Pediatric Thermal Injury

Posted on January 18, 2018

A. Nordin1, N. Shah2, R. Devine1, R. Fabia1, R. K. Thakkar1  1Nationwide Children’s Hospital,Department Of Pediatric Surgery,Columbus, OH, USA 2Ohio State University College Of Medicine,Columbus, OH, USA

Introduction:
Blood transfusions in adult trauma patients increase the risk of nosocomial infections, and similar results have been found in adult burn patients. However, the literature regarding transfusion practices and their consequences in pediatric burns is limited. We therefore sought to determine the impact of packed red blood cell (pRBC) transfusions in pediatric patients with severe thermal injury, and hypothesized that pRBC administration would be associated with increased risk of infection and greater length of stay.

Methods:
We utilized our institutional trauma registry to identify all patients 0-18 years old admitted for burns 10% total body surface area (TBSA) or greater between 2007and 2015. Data points collected included demographics, mechanism of injury, TBSA and clinical outcomes including overall and intensive care unit (ICU) length of stay (LOS), number of operative procedures, and the number of nosocomial infections. Nosocomial infections were defined as positive cultures treated with a complete course of antibiotics. The volume of pRBC administered was also analyzed, and patients who received transfusions were compared against those who did not. Variables were analyzed using either student’s t-test or chi square analysis as appropriate.

Results:
142 patients were identified during the study period, and 15 were excluded for incomplete records. Of the remaining 127, 39 (30.7%) received pRBC transfusions and 88 (69.3%) did not require transfusion. The two groups of patients were not significantly different in terms of age, sex or weight. Mean TBSA in the transfused group was 28.7%, as compared to 14.9% among those not transfused (p < 0.001). The volume of blood transfused increased with TBSA burned. Transfused patients had significantly longer overall LOS (45.5 v 11.7 days; p < 0.001), ICU LOS (18.2 v 1.1 days; p < 0.001), and increased ventilator days (8.2 v 0.6 days; p < 0.001).The average number of operations (4.5 v 0.6; p < 0.001) and infections (3.5 v 0.2; p < 0.001) also increased relative to non-transfused patients. There were no mortalities in either group.

Conclusion:
Transfusions in pediatric burn patients are associated with adverse outcomes, including prolonged hospital and ICU LOS and increased infections. However, pRBC transfusion increased with increasing TBSA, suggesting a possible confounding factor. Our results add to existing literature on the prevalence and impact of blood transfusions in severe pediatric burns, and may have implications for the resuscitation and management of children with severe thermal injuries.
 

53.08 Use of Serum Amylase Levels to Diagnose Traumatic Pancreatic Injury

Posted on January 18, 2018

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

Introduction:
Pancreatic injury is a rare, although potentially devastating consequence of trauma. Diagnosis of and appropriate grading of these injuries remains challenging, with variable sensitivity reported for CT. Other methods, such as magnetic resonance cholangiopancreatography (MRCP) or endoscopic retrograde cholangiopancreatography (ERCP) may improve diagnostic accuracy, however are limited to specialized centers and take longer to perform. Serum amylase has been suggested for use as an adjunctive test to help identify patients with pancreatic injury, but remains controversial. We sought to evaluate its role in the diagnosis and management of pancreatic trauma.

Methods:
A retrospective case control study was performed at an American College of Surgeons verified level 1 trauma center from 2011-2017. All adult patients admitted to the trauma surgery service were eligible for inclusion, while patients with pregnancy or age <18 years old were excluded. Patients with pancreatic injury were identified from the trauma registry using ICD-10 codes. Corresponding controls with thoracic trauma, but without pancreatic or hollow viscus injury, were identified and matched using injury severity score (ISS), age, then gender in a 1:1 fashion. Pancreatic injuries were graded according to the AAST guidelines (Grades 1-5), with major injury identified as ≥ grade 3 (pancreatic ductal injury). Serum amylase levels were recorded from admission and throughout hospitalization. Hyperamylasemia was defined as a serum amylase >103 U/L, according to institutional standard. Analysis was performed using χ2 or Student's t-test for categorical and continuous variables respectively. The primary outcome of interest was admission serum amylase level. Secondary outcomes included serum amylase levels in isolated pancreatic injury and grade of pancreatic injury with elevated serum amylase.

Results:
51 patients with pancreatic injury and 51 corresponding controls were identified. Admission hyperamylasemia was identified in 18 patients with pancreatic injury and 2 controls (p<0.001). Average admission serum amylase levels were significantly increased in patients with pancreatic injury compared to without (122.24±136.37 vs 53.90±60.59, p=0.002), which persisted when controlling for hollow viscus injury (145.93±171.13 vs 48.93±32.35, p=0.006). Among patients with pancreatic injury, 36% (18/51) had hyperamylasemia on admission. Hyperamylasemia did not predict major pancreatic injury (50.0% vs 40.6%, p=0.57), need for operative intervention (100% vs 84.4%, p=0.15), or pancreatic resection (38.9% vs 37.5%, p=0.94) when compared to patients with normal admission amylase. 

Conclusion:
Pancreatic injury results in higher mean levels of admission serum amylase. However, measurement of serum amylase is of questionable clinical value as hyperamylasemia fails to differentiate major and minor pancreatic trauma or need for operative intervention. 
 

53.06 What’s in a name? Provider perception of injured John Doe patients

Posted on January 18, 2018

C. F. Janowak1, S. K. Agarwal2, B. L. Zarzaur3  1University Of Cincinnati,Trauma And Surgical Critical Care,Cincinnati, OH, USA 2University Of Wisconsin,Department Of Surgery,Madison, WI, USA 3Indiana University School Of Medicine,Department Of Surgery,Indianapolis, IN, USA

Introduction:
We previously demonstrated that unidentified, aliased patients, John Doe’s (DOE) are one of the highest risk and medically fragile populations of injured patients.  Aliasing can result in misplaced information and confusion that must be overcome by healthcare professionals.  DOE alias use is institutionally dependent and not uniform.  We sought to determine if healthcare practitioners experience confusion and deliver compromised care by caring for injured DOE patients.

Methods:
After obtaining institutional review board (IRB) approval we surveyed critical care nurses, nurse practitioners, resident physicians and surgeons who care for DOE patients at two academic level I trauma centers with separate DOE alias practices.  Surveys asked whether caring for DOE patients created possible or actual confusion, and possible or actual patient care errors.  In one institution (System 1) only unidentified patients were given an alias that was reconciled when information became available.  In the other (System 2) all trauma patients were admitted with an alias that was reconciled within 24 hours.  Respondents were invited to complete an anonymous questionnaire regarding the care for DOE patients.  Results were analyzed with Wilcoxon rank-sum tests and significance was assessed at a level of 0.05.

Results:
Out of 176 total respondents, 120 (68.2%) reported from System 1, and 56 (31.8%) from System 2.  Overall 53.1% reported that DOE alias use can cause serious confusion possibly resulting in errors affecting patient care.  Specifically, 31.3% reported experiencing actual confusion, although only 4% reported actual errors.  Nurses had significantly higher perceived risk of confusion in the system of all DOE versus selective DOE assignment (35.4% vs. 8.2%, p < 0.01).  Resident physicians reported significantly more frequent actual mistakes within the System 2 versus System 1 (61.9% vs. 17.5%, p < 0.01) despite finding no significant difference in resident perception of confusion (54.5% vs. 37.5% respectively, p = 0.2).

Conclusion:
Our study sheds light on clinical consequences of EMR use and aliases for end users.  We show that nurses perceive there are greater potential complications associated with DOE aliases use, and this varies dependent upon the system used for managing unidentified patients.  Provider confusion, risk for error, and patient safety should be considered in DOE alias use.
 

53.04 Treatment of ICU Delirium with QTc Prolonging Medications Does Not Lead to Cardiac Arrhythmias

Posted on January 18, 2018

J. Zakko1, A. Francis2, C. V. Murphy2, D. A. Eiferman1  1Ohio State University Wexner Medical Center,Department Of Surgery,Columbus, OH, USA 2Ohio State University Wexner Medical Center,Department Of Pharmacy,Columbus, OH, USA

Introduction:  ICU delirium is common in the SICU population and many patients are treated with medications that can cause QTc prolongation, which is a risk factor for the development of Torsades de Pointes (TdP). Serial ECGs are often ordered in this population to assess for QTc prolongation, and effective medications are routinely discontinued due to an increase in QTc for fear of progression to TdP. There is limited data available to determine the risk of TdP in the surgical intensive care unit (SICU) patient population as well as any morbidity from QTc prolongation. This study aims to determine if QTc prolongation is associated with development of TdP.

Methods: A single-center retrospective cohort study was conducted to evaluate QTc prolongation and development of TdP amongst non-cardiac SICU patients being treated for ICU delirium at a large academic medical center. Delirium treatment included at least one of the following medications: haloperidol, risperidone, quetiapine, or olanzapine. QTc prolongation was defined as QTc > 500 milliseconds or >20% increase from baseline. Exclusion criteria included ventricular pacing, bundle branch blocks, incarceration, pregnancy, patients on select antipsychotics prior to admission, and congenital long QT syndrome. The primary outcome was to determine prevalence of QTc prolongation and TdP. Secondary outcomes included SICU mortality and risk factors for QTc prolongation. Univariate and multivariate logistic regression models were constructed for assessment of the outcomes.

Results: Eighty patients were eligible for evaluation. Eight (10%) patients had QTc prolongation. There were no cases of TdP. Assessing patient demographics, concomitant antiarrhythmic and antidepressant use, history of heart disease, diuretic use, hepatic dysfunction, SICU length of stay, length of delirium treatment, and SICU mortality in univariate and multivariate analysis yielded no statistically significant association with development of QTc prolongation, morbidity, or mortality.

Conclusion: Among SICU patients receiving treatment for ICU delirium, the frequency of QTc prolongation was only 10% in our cohort with no cases of TdP. Furthermore, univariate and multivariate analysis did not demonstrate a significant correlation between QTc prolongation and morbidity or mortality. These results suggest that QTc prolongation due to medications used to treat delirium does not lead to the development of cardiac arrhythmias. Furthermore, we propose that serial ECG monitoring may be overused and the subsequent discontinuation of delirium-treating medications may not be necessary in this patient population.

53.05 The Epidemiology of Injuries and Related Surgical Intervention in Aftermath of Tornados in America

Posted on January 18, 2018

M. Rajaei1, R. Griffin2, P. Hu1, T. Swain2, J. Kerby1  1UAB,Division Of Acute Care Surgery, Department Of Surgery, School Of Medicine,Birmingham, ALABAMA, USA 2UAB,Department Of Epidemiology, School Of Public Health,Birmingham, ALABAMA, USA

Introduction: Since 1900, natural disasters in the United States have affected more than 26.5 million people and caused almost $737 billion in damage. Tornadoes are common in North America, where the majority of the 1,000 annual tornadoes occur. These disasters result in hundreds of deaths and further injuries every year. Studying the most common injuries and associated operative interventions in the aftermath of the disasters contributes to more efficient mass casualty management. The purpose of this study was to describe the most prevalent injuries and operative procedures perform on victims of tornadoes in the United States.

Methods: A cross-sectional study was performed using National Trauma Data Bank. Tornado-related injuries were identified using ICD-9 codes. Patient information was collected on demographic, injury location, procedures, and clinical characteristics. Injury severity was measured by Abbreviated Injury Scale (AIS), Injury Severity Score (ISS), and Glasgow Coma Scale (GCS).

Results: From 2003-2013, 1,059 individual-injuries were included. Age was 44±39 years with 53% female. The length of hospital and ICU stay were 10±23 and 8±16 days respectively. The average ISS and GCS were 17±22 and 13±5, with 24.6% having a GCS < 8. Most common injury locations were thorax (53.3%), head (52.2%), lower (51.7%) and upper (53.6%) extremities. Severe injuries happened most commonly in the thorax (40.0%), head (15.2%), and lower extremities (15.2%). Head (85.6%), thorax (79.9%), and spinal (64.3%) injuries were the most common injuries among fatal cases. Severe thoracic injuries occurred nearly twice as common as severe head injuries in patient who didn’t survive the incidents (66.3% vs. 35.9%). Patients were equally admitted to the ICU (35.6%) or regular floor (31.5%) with 20.1% requiring operative intervention prior to admission. The most common procedures were performed on skin/ soft tissue (17.9%); fractures/ dislocations (16.9%); injured vessels (11.0%); chest (7.9%); and muscle/ tendons (5.6%).

Conclusion: ~~Head and thoracic injuries were most common in tornadoes, with the latter being the most prevalent severe injury associated with mortality. While more study is needed, these data could potentially inform disaster preparedness and injury prevention approaches.

53.03 Prehospital Blunt Traumatic Cardiac Arrest: Is It Worth the Resuscitative Investment?

Posted on January 18, 2018

N. R. Manley1, J. Holley1,2, J. Martin2, T. Stavely1, M. Croce1, P. E. Fischer1  1University Of Tennessee Health Science Center,Surgery,Memphis, TENNESSEE, USA 2Memphis Fire Department,Memphis, TENNESSEE, USA

Introduction:  Prehospital resuscitation of patients with blunt traumatic cardiac arrest (BTCA) is known to have extremely poor outcomes when success is considered hospital discharge. However, the results are unknown when survival is defined as hospital discharge or organ donation. Prehospital protocols must be adjusted to decrease futile resource utilization while not sacrificing potential survivors or donors, as one donor can potentially provide 7 organs. We sought to identify factors associated with survival (discharge or organ donation) of patients with BTCA.

Methods:  All adult patients with BTCA transported by our city EMS for 2013-2015 were included.  Data on demographics, prehospital resuscitation, prehospital injury assessment, hospital course and outcomes were collected and analyzed.  Survival was considered either hospital discharge or consideration for organ donation.  

Results:  There were 61 patients with BTCA and overall survival was 5% (1 discharge, 2 organ donation). Two kidneys and one liver were successfully procured allowing 3 transplants.  There was no difference in survivors regarding mean prehospital CPR time (25 vs 24 min) or initial cardiac rhythm.  On prehospital assessment, 36 patients (59%) had documented head trauma and 2 survived (8%).  22 patients (36%) had torso trauma and none survived.  All patients required extensive prehospital resources with an average ambulance service time of 103 minutes. 

Conclusion:  Prehospital resuscitation of BTCA patients with evidence of torso trauma appears futile and these patients should be declared in the field.  However, BTCA patients with no apparent trauma or injuries isolated to the head/face have a small chance of survival to discharge or organ donation and resuscitation should be attempted until further study can elucidate more prehospital predictors of death.

 

53.02 Impact of Aggressive Treatments in Trauma: Using the Emergency Department Thoracotomy to Death Ratio

Posted on January 18, 2018

D. C. Patel1, N. K. Dhillon1, A. Ko1, C. Colovos1, N. Melo1, D. R. Margulies1, E. J. Ley1, G. Barmparas1  1Cedars-Sinai Medical Center,Los Angeles, CA, USA

Introduction:  More than 50 years after its introduction, Emergency Department Thoracotomy (EDT) remains a heavily debated procedure due to the absence of high level evidence in its support. We have previously shown that there is significant variation among trauma centers in performing EDT for trauma patients in extremis, with liberal such centers performing additional, unnecessary procedures. We sought to explore the correlation between trauma center practices in regards to the performance of EDT and survival of patients admitted to these centers. We hypothesized that centers that are liberal in performing EDT would not necessarily have increased survival.

Methods:  Level I and II trauma centers contributing data to the National Trauma Data Bank between 2007 to 2011 were included. Centers with < 200 subjects reported and ≤ 25 ED deaths (EDD) during the study period were excluded. The data was aggregated and the counts for EDT and EDD were calculated for each center. All centers were then divided into quartiles based on the ratio of EDT:EDD. A multivariate logistic regression model was then utilized to calculate the adjusted odds ratio (AOR) for mortality for patients admitted to each quartile. Patients admitted to centers with the lowest quartile (Q1) were used as the reference group. The primary outcome was overall mortality and mortality among subgroup of patients, including those with penetrating trauma and those with critical injuries (ISS≥25).

Results: A total of 174 trauma centers admitting 1,432,811 subjects were included. The median EDT:EDD ratio ranged from 0 for Q1 to 17.6% for centers in the highest quartile (Q4). Q4 centers were more likely to be Level I (Q4: 53% vs. Q1: 44%, p=0.03) and an academic center (Q4: 67% vs. Q1: 42%, p=0.05). Compared to patients admitted to Q1 centers, those admitted to Q4 centers had a significantly higher adjusted mortality (AOR: 1.06, p<0.01). This difference applied also to the subgroup of patients with a penetrating injury (AOR: 1.21, p<0.01). There was no difference in mortality for patients with critical injuries (AOR: 1.02, p=0.52).

Conclusion: Trauma centers where emergency department thoracotomy is liberally performed for trauma patients in extremis had higher adjusted mortality compared to less liberal centers. This paradoxical finding might be explained in part by triaging patients at extremely high risk for mortality to these centers that are liberal in performing those procedures. Further investigation of this phenomenon is required to identify areas for potential improvement and standardization of the management of the trauma patient in extremis, avoiding unnecessary interventions.
 

52.16 Facilitating Factors In Same-day Discharge After Laparoscopic Appendectomy

Posted on January 18, 2018

O. Cheng1, L. Cheng2, S. Burjonrappa2  1Stony Brook University Medical Center,Stony Brook, NY, USA 2Montefiore Medical Center,Bronx, NY, USA

Introduction:  Appendicitis has been cited to be the most common abdominal disorder that requires acute care surgery in the pediatric population. Enhanced Recovery After Surgery (ERAS) protocols are multimodal perioperative care pathways designed to achieve early recovery after surgical procedures by maintaining pre-operative organ function and reducing the profound stress response following surgery. Such protocols have been found to enhance quality of care for surgical patients, as well as improve recovery and shorten hospital stays. ERAS protocols have been proven to help colorectal surgeries but there are no protocols in place yet for appendectomies in the pediatric population. The purpose of this study is to determine the key factors that facilitate same-day discharge and early return to normal activities after laparoscopic appendectomies in children.

Methods: This is a single-center retrospective chart review of pediatric patients (<18 years old) who underwent appendectomies for acute appendicitis from January 2015 to April 2017. The patient population was divided into two groups: those with same-day discharge and those who were discharged one or more days after surgery. Same day discharge (SDD) was defined as discharge less than 24 hours of surgical admission. Patient factors, including pre-hospital, pre-operative, peri-operative, and post-operative factors, were compared and analyzed between the two groups and statistically evaluated using Fisher two-test for categorical data and student t-test for continuous variables.

Results: 248 patients were found under ICD-9 and ICD-10 codes for acute appendicitis. Of these, 63 were excluded due to perforated appendicitis, non-operative management, interval appendectomies, or misdiagnosis. The remaining 185 had laparoscopic appendectomies; 59.5% (n=110) were SDDs and 40.5% (n=75) stayed more than one day. No significant difference was found for time between ER arrival and surgical admission (5.27 vs 5.38 hours; p=0.8) but SDD patients had a significantly shorter time between surgical admission and operation (5.8 vs 11.4 hours; p<0.001). SDD patients also had fewer intra/post-operative complications (1.8% vs 13%; p<0.01) and patients with complications were more likely to stay. There was no significant difference in readmission rates between the two groups (2.73% vs 2.63%; p=1). Total hospital costs were significantly less for SDD ($29,200 vs $33,700; p<0.001). See table for more values.

Conclusion: Surgical leadership can be effective in facilitating same-day discharge without increasing readmission rates or complications, and helps reduce hospital costs, decreases chances of nosocomial infection, and increases patient and family satisfaction.

 

52.14 Outcomes of Circumcision in Children with Single Ventricle Physiology

Posted on January 18, 2018

J. A. Sujka1, R. Sola1, A. Lay1, S. St. Peter1  1Children’s Mercy Hospital- University Of Missouri Kansas City,Kansas City, MO, USA

Introduction:
Children with single ventricle physiology (SVP), especially hypoplastic left heart syndrome, have been shown to have a high morbidity and mortality after major non-cardiac surgical procedures. Elective circumcision, a cosmetic procedure, is one of the most common operations for infants and children with a very low morbidity with <1% of patients requiring readmission or reoperation. The purpose of our study was to review our institutional experience with SVP children undergoing circumcisions to determine the peri-operative course and outcomes.

Methods:
With IRB approval, we performed a retrospective review of children with SVP who underwent an elective circumcision from 2000 to 2017. Children with SVP include hypoplastic left heart syndrome, double outlet right ventricle, and tricuspid atresia. Children with non-single ventricle physiology congenital heart disease and those children undergoing circumcision in combination with another case were excluded from the study. Patient’s demographics, surgical characteristics and outcomes were analyzed. All means reported ± standard deviation.

Results:

There were 15 males who underwent elective circumcision from 2000 to 2017.  The mean gestational age at birth was 37.7 ± 2.65 weeks.  Their mean age at the time of their surgery was 1.39 ± 0.82 years old. Their mean weight was 9.7 ± 1.6 kg. Fourteen were Caucasian and one was African American.  Eleven of fifteen (73%) children were diagnosed with hypoplastic left heart syndrome, all had undergone their first stage procedure with 10 of 15 (66%) undergoing a Norwood Reconstruction as their first stage operation. 

All children underwent a circumcision due to uncomplicated phimosis. Thirteen (87%) of the children underwent a freehand circumcision.  Eighty four percent underwent their circumcision after their 2nd stage cardiac operation with only two patients having their circumcision after their 3rdstage cardiac surgery. 

The mean operative time was 20 ± 7 minutes and there was a mean total length of stay of 247 ± 98 minutes. None of the children were admitted after their circumcision. There were no intraoperative complications.  Post-operative complications included two (16.7%) hematomas with one requiring surgical intervention.  Both complications occurred in patients undergoing a freehand circumcision. There were no deaths within the first 30 days after surgery. There were no unplanned readmissions in one year after surgery.  

Conclusion:

Children with single ventricle physiology who undergo elective circumcision may have higher risk of complications, especially bleeding, compared to the general population.  Further investigations with more patients to better define the risk and allow for definitive recommendations are needed.

52.12 Symptom Resolution and Volumetric Reduction of Abdominal Lymphatic Malformations with Sclerotherapy

Posted on January 18, 2018

H. J. Madsen1,2,5, A. Annam2,3,5, R. Harned2,3,5, T. A. Nakano2,4,5, L. O. Larroque1,2, A. M. Kulungowski1,2,5  1Children’s Hospital Colorado,Divsion Of Pediatric Surgery,Aurora, CO, USA 2Children’s Hospital Colorado,Vascular Anomalies Center,Aurora, CO, USA 3Children’s Hospital Colorado,Interventional Radiology,Aurora, CO, USA 4Children’s Hospital Colorado,Center For Cancer And Blood Disorders,Aurora, CO, USA 5University Of Colorado,School Of Medicine,Aurao, CO, USA

Introduction:   Lymphatic malformations are congenital lesions that arise from errors in vascular embryogenesis.  Cystic lymphatic malformations are categorized based on the size of the lymphatic channels as microcystic, macrocystic, or combined.  Abdominal lymphatic malformations are rare.  Surgical resection of abdominal lymphatic malformations has been the mainstay of therapy but recurrence is high.  We sought to determine the effectiveness of sclerotherapy for the treatment of abdominal lymphatic malformation with regards to symptom resolution and volume reduction.  

Methods:   A single-center, retrospective review from 2014-2017 was conducted evaluating patients with abdominal lymphatic malformations.

Results:  Eight patients were included; 7 patients were male.  Macrocystic lymphatic malformation was the predominant type (n=7); one patient had microcystic disease.    The average age at time of first treatment was 6.8 years (range, 0-17 years). The most common presenting symptoms were distention (n = 7), abdominal pain (n = 6), infection (n = 2) and anemia (n = 1).  Preprocedural cross-sectional imaging was performed for all patients with an average pretreatment volume of 2983.7cm3 (± 4228.6 cm3).   Lymphatic malformations were accessed using ultrasonographic guidance followed by injection of opacified doxycycline.  Patients were treated with a mean of 7.75 cycles (range, 2-16 cycles) of doxycycline sclerotherapy.  Catheters were left in place for a median of 3 treatments (range, 1-7 treatments).  Complications included: intraperitoneal extravasation of doxycycline (n = 1) and infection of the abdominal lymphatic malformation (n=1).  The extravasation was managed conservatively and remained asymptomatic.  The infection was treated successfully with intravenous antibiotics and drainage.  One patient went on to surgical resection of the lymphatic malformation due to inability gain access to the lymphatic malformation.  Postprocedural imaging was available for 87.5% (7/8) patients; one patient is awaiting posttreatment imaging.  Magnetic resonance imaging was obtained after sclerotherapy for 6 patients with 83.3% (n = 5) showing resolution of the lymphatic malformation.  Lymphatic malformations’ volumes decreased by 97.1% after sclerotherapy.  The average remaining volume was 85.7 ± 226.7 cm3 (p = 0.07).  The patient undergoing surgical resection had follow-up abdominal ultrasonography that showed no recurrence.  All patients had resolution of presenting symptoms.  Follow-up duration was 11.8 months (range, 6-24 months).   

Conclusion: Initial results demonstrate that sclerotherapy is a safe and effective treatment for abdominal lymphatic malformations providing symptom resolution and volumetric reduction.

 

52.09 Surgeon Accuracy in Identifying Children with Simple Appendicitis

Posted on January 18, 2018

Y. R. Yu1,2, E. H. Rosenfeld1,2, S. Dadjoo1,2, M. E. Lopez1,2, S. R. Shah1,2, B. J. Naik-Mathuria1,2  1Texas Children’s Hospital,Pediatric Surgery,Houston, TX, USA 2Baylor College Of Medicine,Surgery,Houston, TX, USA

Introduction:
Non-operative management (NOM) of simple appendicitis is a proposed alternative to traditional appendectomy. This study assessed the accuracy of surgeons’ prediction of appendicitis severity.

Methods:
From February to August 2016, pediatric surgeons prospectively predicted whether patients had simple or complex appendicitis before the operation based on clinical data, imaging, and general assessment at a single tertiary care pediatric hospital. Surgeon confidence in their prediction using a 5-point Likert scale was documented. Results were analyzed using receiver operating characteristic (ROC) curves to determine area under the curve (AUC) and optimal cut-off points of clinical findings for diagnosing simple appendicitis. Sensitivity, specificity, positive predictive value (PPV), and negative predictive values (NPV) were also calculated. Predictions were compared to findings using χ2 . A p-value < 0.05 was considered statistically significant.

Results:

Of 125 cases, 73 (58%) were male and the median age was 9 years (range 1-18 years). Simple appendicitis was predicted in 77 (62%) and complex appendicitis in 48 (38%). Surgeons were generally confident of their predictions (simple: 87% certainty, complex: 88% certainty). Predictions were accurate in 59 (77%) of simple cases and 45 (94%) of complex cases. Although surgeon prediction was more accurate than individual imaging or clinical findings and was highly sensitive (95%) for diagnosing simple appendicitis, specificity was only 71% (Table).

Eighteen cases (14%) were inaccurately predicted as simple when they were actually complex.  Of these, 6 (33%) were gangrenous, 17 (94%) had focal/no abdominal tenderness, 15 (83%) were well-appearing, 11 (61%) had ultrasound reported as simple appendicitis, 11 (61%) had ≤ 2 days of symptoms, and 8 (44%) were afebrile (<100.4 oF).

Lower WBC (<15.5×103/uL, AUC 0.61, p=0.05), afebrile (<100.4 oF, AUC 0.86, p<0.01), and shorter symptom duration (≤ 1.5 days, AUC 0.71, p<0.001) were associated with simple appendicitis. 

Conclusion:
Successful NOM for appendicitis works best in patients with simple appendicitis. While surgeon prediction of simple appendicitis is more accurate than ultrasound or clinical data alone, a significant error rate still exists. 

52.10 Gastrointestinal Outcomes in Congenital Diaphragmatic Hernia

Posted on January 18, 2018

M. A. Verla1,2, T. C. Lee1,2, C. C. Style1,2, P. E. Lau2, A. R. Mehollin-Ray1,3, C. J. Fernandes1,4, S. C. Fallon2, C. A. Ikedionwu2, S. G. Keswani1,2, O. O. Olutoye1,2  1Texas Children’s Hospital,Fetal Center,Houston, TX, USA 2Baylor College Of Medicine,Michael E. DeBakey Department Of Surgery,Houston, TX, USA 3Texas Children’s Hospital,Department Of Radiology,Houston, TX, USA 4Texas Children’s Hospital,Department Of Pediatrics – Newborn Section,Houston, TX, USA

Introduction:  Congenital diaphragmatic hernia (CDH) is a major anomaly with high mortality and significant long-term comorbidities. Survivors are at risk for developing gastrointestinal (GI) disorders, some requiring corrective abdominal surgical procedures. The purpose of this study was to examine the GI outcomes in CDH.

Methods:  This is a retrospective review of all infants evaluated for CDH at a single institution tertiary fetal center from April 2004 to March 2017. Data analyzed included prenatal imaging data and postnatal GI outcomes. Imaging features included liver position, stomach position, lung to head ratio, total fetal lung volume (TFLV), observed to expected total fetal lung volume (O/E TFLV) and laterality of defect. GI outcomes were defined as gastro-esophageal reflux disease (GERD), gastroparesis, and abdominal surgeries including fundoplication, gastrostomy tube (G-tube) placement, or exploratory laparotomy for bowel obstruction. Data were analyzed using chi-square, ANOVA, Mann-Whitney U test and student’s t-test as appropriate; a p-value of <0.05 was considered significant. Results are listed as percentages or median (interquartile range) unless otherwise stated.

Results: A total of 250 infants were evaluated over a thirteen year period. Of these, 55.2% were male (n=138). Seventy seven percent (n=192) had left sided CDH, 64% had an intrathoracic stomach, and 70% had an intrathoracic liver morphology prenatally. Postnatally, 78 neonates (31%) required ECMO (Extracorporeal Membrane Oxygenation). The cohort survival was 74% with a median age at repair of 3 days of life [3 – 6 days]. Of the 184 CDH survivors, 99% received total parental nutrition (TPN) for a median of 18 [14 – 28] days. After repair, enteral feeds were initiated at a median of 7 [5 – 10] days and full enteral feeds attained at a median of 11 [7 – 190] days. Of the survivors, 57% (n=105) had GERD requiring medication and 12% (n=21) had gastroparesis. Of those with GERD, 68% (n=71) had an intrathoracic stomach morphology prenatally and 27% (n=28) failed medical management requiring a fundoplication. Overall, 48% (n=89) of survivors had some form of abdominal surgery separate from the CDH repair. Neonates with unfavorable prenatal indices were more likely to require ECMO. There is a significant correlation between ECMO use and having a G-tube, GERD, gastroparesis, fundoplication, or need for other abdominal surgical intervention (p <0.05, Table 1).

Conclusion: A significant number of infants with CDH will have some form of GI morbidity postnatally. Those with severe CDH requiring ECMO are at an increased risk for needing a G-tube or other GI surgery. The results of our study may aid in pre- and postnatal counseling.

52.11 Assessment of Outcomes and Costs for Common Surgical Procedures at Children’s and Non-Children’s Hospitals

Posted on January 18, 2018

C. Tom1, R. P. Won1, E. Saab2, A. D. Lee2, S. Friedlander3, S. L. Lee1,2,3  1Harbor-University Of California Los Angeles Medical Center,Department Of Surgery,Torrance, CA, USA 2University Of California Los Angeles,Department Of Surgery,Los Angeles, CA, USA 3Los Angeles Biomedical Research Institute,Torrance, CA, USA

Introduction:  The benefits of managing children after pediatric surgical procedures or injury at children’s hospitals (CH) are well established. However, little is known about the outcomes and costs of managing common pediatric surgical procedures at non-children’s hospitals (NCH). The purpose of this study is to compare the outcomes and costs of appendectomy and cholecystectomy between CH and NCH.

Methods:  Using the Kids’ Inpatient Database (KID), we compared costs and outcomes for pediatric patients (< 18 years old) who underwent appendectomy and cholecystectomy in 2003, 2006, 2009, and 2012 at CH and NCH. We analyzed the patient demographics, complexity of disease, surgical management (rate of laparoscopy), length of stay (LOS), morbidity, and costs between CH and NCH. 

Results: The majority of appendectomies (NCH = 223,924 vs CH = 83,048) and cholecystectomies (NCH = 18,406 vs CH = 8,774) were performed at NCH. Overall, CH cared for younger children, had higher costs, and increased LOS compared to NCH. Results of the univariate analysis are summarized in the table. On multivariate analysis for appendectomies, CH treated younger children with higher rates of perforated appendicitis, and were associated with increased use of laparoscopy, longer LOS, lower complication rates, and higher costs. Conversely, multivariate analysis for cholecystectomy demonstrated that CH were associated with decreased use of laparoscopy, longer LOS, and higher costs with no difference in the complication rate compared to NCH. 

Conclusion: Variations in surgical management, outcomes, and costs related to common surgical procedures in children exist between CH and NCH. Appendectomies at CH are associated with improved outcomes and higher utilization of laparoscopy despite treating more advanced disease, but have longer LOS and higher costs.  Cholecystectomies at CH are associated with no difference in outcomes, but are less likely to utilize laparoscopy, have longer LOS, and higher costs. Opportunities exist at both CH and NCH to improve the quality of care and lower expenses for common surgical diseases in children.  

 

52.08 Pediatric Single-Site Laparoscopic Appendectomy: Predictors of Conversion to Multiport Appendectomy

Posted on January 18, 2018

N. L. Gates1, R. D. Rampp1, S. D. Bhattacharya1  1University Of Tennessee College Of Medicine,Department Of Surgery,Chattanooga, TN, USA

Introduction:  Single-incision laparoscopic appendectomy (SILA) in the pediatric population has been well described in small case series and met analyses.  Our children’s hospital has adopted this modality for nearly all appendectomies since 2012.  In a review of 1000 consecutive cases from 2012-2017, we hoped to identify factors that portend conversion from SILA to multiport appendectomy.  We compared our cohort of conventional three port laparoscopic appendectomy (CLA) for outcomes including operative time, post-op length of stay, complications and readmission.

Methods:  A retrospective chart review of 1000 patients who underwent appendectomy from March 2012 to February 2017 at a single Children’s Hospital was performed. The type of appendectomy performed (single incision, conventional multiport, open), if the case was converted from single incision to multiport or open, and perforation status were recorded. Demographic data identified included age, sex, and BMI. Outcomes analyzed were operative time, length of stay, and postoperative complication/readmission rate. 

Results: Of 1001 appendectomies during the study period, 959 (95.9%) were initiated with plan for SILA,    35 (3.5%) were initiated CLA, and 6 (0.6%) were initiated via open approach.  Of those initiated SILA, 884/959 (92.2%) were able to be completed without additional port placement.   Cases initiated SILA for perforated appendicitis had a higher rate of conversion to multiport (22/169, 13%) than cases initiated SILA for non-perforated appendicitis (45/782, 5%; p<0.01).   Cases which were not able to be completed SILA were statistically significantly more likely to be older, male patients, or have increased BMI.  When compared to cases which were initiated CLA for perforated appendicitis, SILA- regardless of conversion- remained statistically similar for operative times and length of stay but had higher return to emergency department.  We found no statistically significant risk factors among any of the subgroups identified for increased morbidity such as small bowel obstruction, UTI, readmission, or abscess/reinterventions. 

Conclusion: Single incision laparoscopic appendectomy appears to be a safe and easily adopted modality for the treatment of appendicitis in pediatric populations with no increased morbidity.   Parents of children that are obese, males or present with perforation should be counselled regarding the possibility of additional port placement or considered for initiating conventional laparoscopic appendectomy.

 

52.04 Inter-rater Reliability of a Grading System for Congenital Diaphragmatic Hernia Defect Size is Fair

Posted on January 18, 2018

C. E. Hunter1, Z. M. Saenz1, D. Nunez1, L. Timsina2, B. W. Gray1  1Indiana University School Of Medicine,Division Of Pediatric Surgery, Department Of Surgery,Indianapolis, IN, USA 2Indiana University School Of Medicine,Center For Outcomes Research In Surgery, Department Of Surgery,Indianapolis, IN, USA

Introduction:
The Congenital Diaphragmatic Hernia Study Group (CDHSG) registry is a vital multi-institutional tool to help track outcomes of CDH patients to improve prognosis and patient care. The CDHSG asks surgeons to categorize each patient’s diaphragmatic defect size as grade A, B, C, or D based on published guidelines. A reliable grading system of these defects is important for individual patient prognosis and clinical research. The reported size of the defect has been correlated with patient outcomes, such as survival. However, the inter-rater reliability of this system has not been evaluated. The goal of this study was to evaluate the inter-rater reliability of the CDHSG grading system.

Methods:
The operative notes from patients that underwent surgical repair of a unilateral CDH at a single institution between 2010 and 2016 were collected. Forty-six operative notes were cropped to include only the information necessary to grade the hernia defect A-D based on the CDHSG grading system guidelines. The defects were graded by 9 pediatric surgeons of differing experience levels, and the inter-rater reliability was determined by calculating a Cohen’s kappa (κ). The following cutoffs were used to interpret κ: ≤ 0 – no agreement, 0.01-0.20 – none to slight agreement, 0.21-0.40 – fair agreement, 0.41-0.60 – moderate agreement, 0.61-0.80 – good agreement, 0.81-1.00 – very good to perfect agreement.  Data was also collected on intraoperative findings (liver up vs. down, ECMO status, need for patch repair) and patient outcomes (length of stay, survival).

Results:

Overall, there was 57.49% agreement across all raters, corresponding to a fair level of agreement (κ=0.395, p<0.001). Between any two raters, agreement ranged from no agreement (21.74% agreement, κ= -0.027) to good agreement (82.61% agreement, κ= 0.7543). All 9 of the surgeons agreed in only 2 of the 46 patients, both of which were assigned an “A” grade. Four patients received 3 different grades: 3 received grades A, B, and C, and 1 received grades B, C, and D. No patients were given all four grades. Overall, there was 87% survival (n=40). Inter-rater agreement was similar despite different operative findings and outcomes (p > .05): survival yes/no (κ=0.3690, κ=0.3518), need for ECMO yes/no (κ=0.3323, κ=0.3362), patch repair yes/no (κ=0.2050, κ=0.1916), and liver up/down (κ=0.2941, κ=0.4404).

Conclusion:
This single institution study shows that while the CDHSG grading system is not random, it produces only a fair amount of agreement between pediatric surgeons when grading the severity of hernia defects. Patient outcomes and intraoperative findings do not affect levels of agreement. Future research will examine intra-rater reliability of this system and will work to provide a more reliable system for grading the severity of CDH defects. 

« Previous Page
Next Page »