56.01 Screening For Anal Dysplasia With Anal Pap Smears: Clinical Follow-up And Correlation

J. Son1, E. Lawson1, S. Selvaggi2, B. Harms1, E. Carchman1, R. Striker3, C. I. Voils1, C. B. Geltzeiler1, C. B. Geltzeiler1  1University Of Wisconsin,Colorectal Surgery,Madison, WI, USA 2University Of Wisconsin,Pathology,Madison, WI, USA 3University Of Wisconsin,Infectious Disease,Madison, WI, USA

Introduction:

Anal dysplasia screening and surveillance guidelines are poorly defined and based on little data. Although recommendations on type of surveillance are controversial, most practitioners recommend follow-up clinical examination for patients with an abnormal anal pap smear. Our objective was to determine how often an anal pap was followed by clinical exam at our institution and how often histology correlated with pap cytology.

Methods:  

All anal pap results at a single tertiary academic center from 2008 to 2018 were collected. Retrospective chart review was performed on all patients with cytology results demonstrating dysplasia (high-grade squamous intraepithelial lesions (HSIL) or low-grade squamous intraepithelial lesions (LSIL)). We examined patient risk factors as well as their clinical follow-up within 1 year. Clinical exam was defined as digital rectal examination, anoscopy, or high resolution anoscopy (HRA). We also examined if cytology accurately predicted histologic dysplasia. 

 

Results:

A total of 327 anal pap smears demonstrated dysplasia (25% HSIL and 75% LSIL) in 182 patients. 92% of patients were male, 97% HIV positive, and 73% had documented anal receptive intercourse. 75% of dysplastic anal paps were followed by clinical exam within 1 year and 50% were biopsied. Of the 45 HSIL anal paps that were followed by biopsy, only 38% confirmed high-grade disease on histology, 24% demonstrated low-grade disease, and the remaining were negative. In contrast, of the 119 LSIL anal paps that were followed by biopsy, 44% confirmed low-grade disease on histology, 22% were upgraded to high-grade disease. 3% had invasive squamous cell carcinoma on biopsy after LSIL pap.

Conclusion:

This single center study demonstrates that only 75% of abnormal anal paps were followed up with clinical exam within 1 year and only 50% were biopsied for histologic confirmation. When biopsied, only 66% of dysplastic paps demonstrated dysplasia or invasive disease on histologic examination. There is room for improvement in our institution to consistently follow-up with clinical exam after abnormal pap. Our data suggests this is especially important considering anal pap cytology is an imperfect predictor of histologic dysplasia and invasive disease. Clinical exam may be more important as the initial screening test.

55.20 Imaging Characteristics of Mucosal Appendicitis

K. Gee1, R. Jones1, S. Preston1, A. Beres1,2  1University Of Texas Southwestern Medical Center,Surgery,Dallas, TX, USA 2Children’s Medical Center,Surgery,Dallas, Tx, USA

Introduction:  Mucosal appendicitis is a controversial entity that is histologically separate from transmural appendicitis. Therefore, the ability to distinguish mucosal appendicitis from transmural appendicitis preoperatively is of importance. We hypothesize that patients with mucosal appendicitis can be distinguished based upon specific preoperative imaging findings.

Methods:  After IRB approval, charts were evaluated from all patients who underwent laparoscopic appendectomy at our institution during 2015. Patients with mucosal appendicitis were identified and matched 2:1 to a random cohort of nonperforated transmural appendicitis cases. Demographic and clinical data were collected, including imaging and pathology findings. Preoperative predictors of mucosal appendicitis were modeled using binomial logistic regression analysis. Predictive factors tested included sex, age, use of multiple imaging studies, and the following imaging features: whether the appendix was visualized, compressibility, hyperemia, thickened appendiceal wall, dilation of appendix, presence of abdominal free fluid, fluid collection, surrounding echogenicity, dilated bowel, lymphadenopathy, and appendicolith.

Results: Mucosal appendicitis was identified in 103 patients. Female patients were 1.8 more likely to have mucosal appendicitis (95% CI 1.1-3.3). Compressibility on ultrasound was associated with 2.9 times more likelihood of mucosal disease (95% CI 1.3-6.5). Non-dilated appendix correlated with 3.8 times more likelihood of mucosal infection (95% CI 1.4-9.8), and lack of free fluid led to 2.1 times more likelihood of mucosal disease (95% C I1.1-3.7). Most significantly, lack of echogenic changes in surrounding tissue was predictive of mucosal appendicitis, conferring 4.8 times the risk (95% CI 2.3-10.0). Finally, patients who underwent multiple imaging tests were 3.1 times more likely to have mucosal appendicitis (CI 1.2-8.0). Together, these variables can successfully predict presence of mucosal appendicitis on final pathology report at a rate of 74.5%, and explain 29% of the variance in diagnosis of mucosal versus transmural appendicitis (p <0.0001 ).

Conclusion: Mucosal appendicitis is a controversial entity. When compared to transmural appendicitis, US findings of mucosal appendicitis are more likely to include non-dilated, compressible appendix, lack of free fluid. Additional studies examining the presenting symptoms and post-operative complications of mucosal versus transmural appendicitis may further delineate the clinical significance of mucosal appendicitis.

 

55.19 A Systematic Review of Malrotation Presenting Between 1 and 19 Years of Age

K. T. Gemayel1, G. Romero-Velez2, S. Burjonrappa1  1University Of South Florida College Of Medicine,Department Of Pediatric Surgery,Tampa, FL, USA 2The Children’s Hospital at Montefiore,Department Of Pediatric Surgery,Bronx, NY, USA

Introduction:
Midgut malrotation is a congenital anomaly of intestinal rotation and fixation where in an incomplete rotation or a complete failure of rotation of the primitive intestinal loops around the superior mesenteric artery axis occur during the fetal period. We have personally evaluated two adolescent cases of malrotation, and it is the intent of this review to further investigate the intraoperative presentation and clinical symptoms of adolescent malrotaion in addition to our findings. Adolescent and adult age malrotation are difficult to diagnose clinically and the focus of this case series and systematic review is malrotation after the age of 1 year and inclusive of the adolescent population.

Methods:
To further evaluate intraoperative findings and clinical symptoms associated with adolescent malrotation, systematic analysis was done using Cochrane, Medline and Pubmed databases. Database search was guided using keywords “malrotation”, “volvulus”, “Ladd’s procedure”, “pediatrics”, and “adolescent”. Inclusion criteria were selected based on case series in the English language, with patients in-between 1-19 years of life, with clear cut description of intraoperative findings with perioperative evaluation. Exclusion criteria was non-English literature, age under 1 year and older than 19 years without a clear cut description of findings. Operative findings were categorized based on Stringer’s classification of malrotation, with type II and III essentially representing pre-axial and post-axial malrotation.

Results:
Systematic review yielded 75 cases in which were distributed by Stringer's classification into norotation (n= 14), duodendal malrotation (n= 37), duodenal and cecal malrotation (n= 24). Intraoperative findings were analyzed for presence of volvulus, narrow mesenteric base and presence of Ladd Bands with varying clinical symptoms (as shown in figure).

Conclusion:
In conclusion, in the case of intermittent intestinal obstruction in young adults, physicians should keep in mind that early and accurate diagnosis of malrotation with an appropriate surgical treatment may save patients from unexpected complications. Due to the presence of atypical symptoms, the diagnosis of malrotation requires a high index of suspicion, appropriate diagnostic studies, and aggressive definitive surgical treatment.
 

55.18 CT scan improves pelvic fracture detection in pediatric blunt trauma without effect on management

E. G. Englert1,2, C. Buonpane1,2, G. Ares1,2,3, B. Benyamen2, C. J. Hunter1,2  1Feinberg School Of Medicine – Northwestern University,Surgery,Chicago, IL, USA 2Ann & Robert H. Lurie Children’s Hospital of Chicago,Pediatric Surgery,Chicago, IL, USA 3University Of Illinois At Chicago,General Surgery,Chicago, IL, USA

Introduction:  Pelvic fractures are uncommon yet potentially serious injuries in pediatric trauma patients. The role of computed tomography (CT) scans in addition to standard pelvic radiographs (XR) in identifying and altering management of pelvic fractures is controversial. Unnecessary use of CT scan is especially undesirable in the context of minimizing pediatric radiation exposure. We hypothesized that CT scan for pelvic fractures would not significantly change patient diagnosis or management.  In order to minimize utilization of CT scan, this study sought to use history and physical examination findings to identify high risk patients who would benefit from CT scan.

Methods:  This study is a single institution retrospective review of patients ≤18 years old who were seen by the trauma service for blunt trauma and underwent pelvic radiograph from 2014 to 2016.  Patients with pelvic XR alone were compared to those who received pelvic XR and CT scan.  Primary outcome was pelvic fracture.  Secondary outcomes included management, length of stay and survival. 

Results: 257 patients were identified, of which 156 (60.7%) had pelvic XR alone while 47 (18.3%) received pelvic XR and CT scan. Patients that were chosen for CT scan were more likely to have pelvic tenderness (OR 2.7, 95%CI 1.03-6.94, p=0.04), injury of the spine (OR 4.9, 95%CI 1.19-20.29, p=0.03) and/or abdominal tenderness (OR 5.6, 95%CI 1.87-16.98, p=0.002).

Eighteen patients (7%) were diagnosed with pelvic fracture.  Patients with pelvic fracture were 23 times more likely to have pelvic tenderness on physical examination than patients without fracture (OR 23.2, 95%CI 6.86-78.5, p<0.001).  The sensitivity, specificity, positive predictive value and negative predictive value of diagnosis of pelvic fractures on pelvic XR alone was 83.3%, 100%, 100% and 98.77%, respectively.  Eight of the 18 patients with fractures received both X-rays and CT. Pelvic injury findings were identical in 6 of those patients.  However, in two patients, XR failed to diagnose fractures that were later seen on CT scan (fractures of the acetabulum, ilium, sacrum and sacroiliac joint disruption).  The additional findings, however, did not change patient management (observation and physical therapy). 

Conclusion: CT scan is more sensitive than pelvic XR in the diagnosis of a pelvic fracture; however, in our series this did not result in a change in patient management. In patients with XR and a high clinical suspicion of more extensive injuries based on pelvic tenderness and other physical examination findings, we recommend a CT scan.

 

55.17 Intubation Outside a Pediatric Trauma Center Associated with Worse Outcomes in a Non-Urban Setting

R. B. Hawkins1, S. L. Raymond1, H. C. Hamann1, M. M. Mustafa1, J. A. Taylor1, S. Islam1, S. D. Larson1  1University Of Florida,Department Of Surgery, Division Of Pediatric Surgery,Gainesville, FL, USA

Introduction: Trauma is the leading cause of death in pediatric patients over 1 year of age. Controversy exists regarding prehospital airway management for these patients. Studies suggest that bag-valve mask ventilation is preferred in an urban setting. The purpose of this study is to evaluate differential outcomes in pediatric trauma patients who underwent endotracheal intubation at the scene of injury, referring hospital, or pediatric trauma center in a predominantly rural/suburban setting.

Methods:  A retrospective review was performed evaluating trauma patients age 18 or younger at a single institution over 10 years (2004-2014). Patients were selected who underwent endotracheal intubation and were classified based on location of intubation (scene, referring hospital, or trauma center). Fischer’s exact test and t-tests were performed to compare outcomes between groups, and multivariate regression modeling was performed to evaluate for significant predictors of mortality.

Results: 288 patients were identified who underwent endotracheal intubation related to their trauma care. 155 (53.8%) were intubated at the scene of injury, 55 (19.1%) at a referring hospital, and 72 (25%) at the trauma center. Overall mortality was 21.9%, which was highest in the scene intubation group (29.7%) compared to the referring hospital (20%) and trauma center (5.6%) groups (p<0.01). Patients intubated at the scene had higher Injury Severity Scores and lower Glasgow Coma Scale scores (p<0.01). Duration of intubation was lowest in the trauma center group (p<0.01). Complication rate was highest in the referring hospital group (p<0.05). A multivariate regression model identified location of intubation, ISS, and GCS as significant independent predictors of mortality with an area-under-the-curve of 0.9193.

Conclusion: Mortality and duration of intubation were lowest in trauma patients intubated at a pediatric trauma center. When possible, intubation can be delayed as indicated until expert care can be provided at a pediatric trauma center in a rural/suburban setting.

 

55.16 Do negative pressure incisional wound VACs decrease surgical site infections in pediatric surgery?

M. R. Phillips2, S. L. English1, E. Teeple1, A. E. Martin1, K. Reichard1, C. D. Vinocur1, L. Berman1  1Nemours/Alfred I DuPont Hospital for Children,Sidney Kimmel School Of Medicine At Thomas Jefferson University,Wilmington, DE, USA 2University Of North Carolina At Chapel Hill,Surgery,Chapel Hill, NC, USA

Introduction: Surgical site infection (SSI) rates are an important quality metric in surgery. Adult studies have demonstrated a decrease in rate of SSI with the use of negative pressure incisional wound vac device (NPIWV) dressings. No studies have examined the effect of NPIWV dressings on SSI rates in pediatric patients.

Methods:   We performed a retrospective review of patients who underwent surgery using NPIWV at our institution between February 2016 and February 2018. NPIWV dressings were applied by approximating the skin edges with buried, absorbable sutures, protecting the skin with adhesive barrier, covering the wound with black sponge, and applying negative pressure for up to 4 days. We identified a group of patients with the same CPT codes from our National Surgical Quality Improvement Program-Pediatric (NSQIP-P) data between January 2014 and January 2016, in order to compare SSI rates in the NPIWV population to historical controls.  The cohorts were compared using either a Mann-Whitney test or chi-square analysis (p<0.05).

Results: Thirty-five patients underwent surgery using NPIWV for wound closure. Sixty-four patients with similar CPT codes were identified from our institutional NSQIP-P data who did not have NPIWV dressings (Table 1). The groups were comparable.  In the NPIWV group there was 1 SSI in 35 cases (2.9%), and there were 7 SSIs in the 64 historical control patients (10.9%). The difference in SSI rates did not reach significance (p=0.25, OR: 4.18 CI [0.49-35.43]). There were no complications associated with the use of NPIWV dressings

Conclusion: NPIWV dressings can be used safely for the management of contaminated wounds in pediatric patients undergoing surgery, including neonatal patients. Our study shows a trend toward decreased SSI rates with NPIWV. However, our sample size was likely too small to reach statistical significance. These findings from a small retrospective study need to be confirmed in a larger, prospective trial.

55.15 Predictors of In-Hospital Mortality in Newborn Conjoined Twins

B. Willobee1, M. Mulder1, E. A. Perez1, A. R. Hogan1, A. Brady1, H. Neville1, J. E. Sola1, C. M. Thorson1  1University of Miami,Pediatric Surgery,Miami, FL, USA

Introduction: Conjoined twins are rare developmental anomalies. There is a paucity of literature other than case reports and small case series. The aim of this study was to examine national outcomes and identify predictors of mortality in newborn conjoined twins.

 

Methods:  Patients born with a diagnosis of conjoined twins were identified in the Kids' Inpatient Database (1997-2012). Patient demographics, associated anomalies, operative procedures and outcomes were identified and analyzed using standard statistical methods.

Results: A total of 248 patients were identified for a nationally weighted incidence of 1 per 100,000 live births. The majority were female (n=197, 80%). The most common associated anomalies were cardiac (n=90, 36%), gastrointestinal (n=43, 17%), abdominal wall defects (n=32, 13%), and genitourinary (n=28, 12%). Prematurity was seen in 59% (n=147) and 45% (n=112) were classified as low (27%), very low (8%) or extremely low (11%) birth weight. Fifty-eight patients (24%) underwent operative procedures during the primary hospitalization, including 27 (11%) who had neonatal separation surgery. 

The overall mortality was 62% with most deaths occurring just after birth (104/153, 68%) or within the first 48 hours of life (134/152, 88%). Mortality was significantly higher in infants who were female vs. male (67% vs. 36%, p=<0.001), premature vs. full term (72% vs. 46%, p<0.001), low birth weight (73% vs. 58%, p=0.031) and extremely low birth weight (96% vs. 58%, p<0.001). Congenital diaphragmatic hernias were seen in 15 patients (6%) and were uniformly fatal (100% vs. 59%, p=0.002). Mortality was highest in hospitals not designated as children’s hospital (71%) vs. hospitals with a children’s unit or free-standing children’s hospital (42%), p=0.01. Greater than 50% of all deaths were in hospitals not designated as a children’s hospitals. Multiple logistic regression revealed prematurity (OR 2.3 [1.06-4.84], p=0.034) and non-children’s hospital designation (OR 2.3 [1.09-5.02], p=0.029] to be independent predictors of mortality.

 

Conclusion: Conjoined twins are rare defects and are often associated with multiple anomalies. They experience an extremely high perinatal mortality, especially females, those who are premature, low birth weight, or with diaphragmatic hernias. This data supports caring for these complex patients at hospitals designated for children.

55.14 High Alvarado Scores Predict Appendicitis-Positive MR in the Diagnosis of Pediatric Appendicitis

J. Sincavage1,2, C. Buonpane1,2, B. Benyamen2, E. Benya2, T. Lautz1,2, C. J. Hunter1,2  1Feinberg School Of Medicine – Northwestern University,Surgery,Chicago, IL, USA 2Ann & Robert H. Lurie Children’s Hospital of Chicago,Pediatric Surgery,Chicago, IL, USA

Introduction:  The clinical presentation of appendicitis is variable, and radiologic imaging is often used to help make the diagnosis. The accuracy of magnetic resonance imaging (MR) for the diagnosis of appendicitis has been validated and has replaced computed tomography (CT) in some centers, particularly in the setting of a non-diagnostic ultrasound (US). Nonetheless, MR is not widely used in this setting due to high cost, variability of MR capacity among institutions, time required for the study, and high rates of negative scans.  Clinical scoring systems such as the Alvarado Score (AS) have been used to aid in the diagnosis of appendicitis but have limited efficacy when used in isolation. We hypothesized that the AS could predict patients in which an MR would be diagnostic of appendicitis.

Methods:  This study is a single institution, retrospective review of patients ≤18 years of age who received US in the workup of suspected appendicitis in 2017 at a tertiary care children’s hospital with 24-hour MR capabilities. Non-diagnostic US studies were defined as non-visualization of the appendix or inability to rule out appendicitis. Imaging results were reviewed and AS were calculated from data obtained in patient EMR notes.  Primary outcome was a confirmed pathologic diagnosis of appendicitis.

Results: 408 patients met inclusion criteria, of whom 60.5% had a non-diagnostic US.  48% percent of patients then received MR for further evaluation, of which 72% were negative for appendicitis.  Of the patients who had a non-diagnostic US, MR, and data available for AS calculation, 20 were low score (0-3), 51 moderate score (4-6) and 30 high score (7-10).  High AS patients were five times more likely to undergo appendectomy than low/moderate AS patients (OR 5.3, 95% CI 2.1-13.5, p=0.0004).  Patients with low/moderate AS had a higher rate of negative MR after US than patients with high AS (82.5% versus 41.4%, p = 0.0001). The number of MRs needed to identify one case of appendicitis in the high AS group was 1.75, 5.5 in the moderate AS group, and 6 in the low AS group.

Conclusion: MR after non-diagnostic US in the workup of suspected appendicitis has increased diagnostic value in patients with high AS.  Based on our results, we believe that the AS can assist health care providers in determining the additional utility of MR in the workup of appendicitis.  In order to optimize usage of MR at institutions with access to this imaging modality, we do not recommend MR in the setting of low/moderate AS.

55.13 Differences in Outcomes Based on Sex for Pediatric Patients Undergoing Pyloromyotomy

R. L. Massoumi1, C. Tom2, E. Howell2, C. P. Childers1, R. Sakai-Bizmark3,4, S. L. Lee1,5  1University Of California – Los Angeles,Surgery,Los Angeles, CA, USA 2Harbor-UCLA,Surgery,Torrance, CA, USA 3Los Angeles Biomedical Research Institute,Torrance, CA, USA 4Harbor-UCLA,Pediatrics,Torrance, CA, USA 5UCLA Mattel Children’s Hospital,Pediatric Surgery,Los Angeles, CA, USA

Introduction:  

While male and female patients are known to have varied responses to medical interventions, how sex influences surgical outcomes has not yet been clearly elucidated. Data are particularly lacking in children. Our study aimed to determine the effect of sex on surgical outcomes after pyloromyotomy.

Methods:  

Using the Kids’ Inpatient Database (KID) for the years 2003, 2006, 2009, and 2012, we performed a retrospective review of all patients under 1 year of age who underwent pyloromyotomy for hypertrophic pyloric stenosis. Endpoints included patient mortality, any complications during their hospitalization (derived from ICD-9 codes and including during procedure, gastrointestinal, respiratory, or cardiovascular), hospital cost (calculated using cost to charge ratios), and total length of hospital stay (LOS). The primary comparison was outcomes for males versus females. Regression models were adjusted by race, age group and complications with region and year fixed effects. Hierarchical logistic regression was used for complication and mortality rates, hierarchical negative binomial regression was used to assess LOS, and hierarchical multivariable linear regression was performed for cost analysis. 

Results:
Of 51,298 weighted operations reviewed, 41,331 (80.6%) were in males and 9,967 (19.4%) were in females. There was no statistically significant difference in distribution of races between the two genders.  Female patients were more likely to be greater than 29 days old on the day of operation (83.4% of females versus 79.5% of males P < 0.001). Multivariable analysis comparing males and females is summarized in Table 1. Females had a higher risk of patient mortality, complications, LOS, and cost.

Conclusion:
Our study demonstrated that females had worse outcomes following pyloromyotomy compared to males.  Females had increased mortality, complications, LOS, and cost. These findings are striking and are important to consider when treating either sex to help set physician and family expectations perioperatively. Further studies are needed to determine why such differences exist to develop targeted treatment strategies for both females and males with pyloric stenosis.

55.12 The Natural History of Uncorrected Biliary Atresia

I. N. Lobeck1, I. P. Lim2, R. Karns3, J. Bezerra3, G. Tiao2  1Wayne State University,General Surgery,Detroit, MI, USA 2Cincinnati Children’s Hospital Medical Center,General And Thoracic Pediatric Surgery,Cincinnati, OH, USA 3Cincinnati Children’s Hospital Medical Center,Division Of Gastroenterology, Hepatology, And Nutrition,Cincinnati, OH, USA

Introduction: Biliary atresia (BA) is a cholangiopathy of infancy which, without intervention, progresses to death in the first two years of life. We report our institutional experience of the pathophysiology and natural history of BA without portoenterostomy.

Methods: After Institutional Review Board approval, a retrospective chart review of patients who underwent primary liver transplant for BA without portoenterostomy during 2003-2015 was performed. Data collected included demographics, laboratory and clinical history at diagnosis, listing and transplantation. Statistical analysis was performed utilizing pairwise correlation analyses, time-series regression modeling and association testing between clinical variables and surgical outcomes.

Results:Sixteen patients were identified (63% male; mean age 165±68 days at diagnosis and 362±223 days at transplant). At presentation, symptoms were predominantly of cholestasis and mild synthetic dysfunction, with normal platelet counts (mean direct bilirubin 8.1mg/dL, INR 1.3, platelets 256,000). At transplantation, direct bilirubin rose to 34 fold and INR 2 fold over normal; mean platelet count decreased to 128,000. 45% (n=5) endured variceal bleeding.  Patients with bleeding episodes before transplant tended to have more postoperative complications (p=0.08, R2=0.45). The strongest predictors of postoperative course were platelets, change in INR between diagnosis and transplant, and bilirubin. Each increase of direct bilirubin by 1 unit resulted in one day increase of ICU stay (p=0.002). INR also predicted EBL, with each increase of 1 unit resulting in additional 838ml of blood loss (p=0.022). Platelet count showed a borderline association with postoperative outcomes, with each increase of 50,000 units in platelets linked to decreasing the risk of infection one year postoperatively by half (p=0.058).

Conclusion: At presentation, patients with uncorrected BA have cholestasis and mild synthetic dysfunction. Disease progression entails increasing cholestasis, portal hypertension, and synthetic dysfunction. The greatest predictors of post-transplant outcomes include bilirubin, platelets and INR. Early liver transplantation is warranted once the opportunity for portoenterostomy is missed.

 

55.11 Diverting Outcomes? Anti-TNFα Therapy in Pediatric Crohn’s Disease

I. P. Lim1, A. Chernoguz2, R. A. Falcone1, D. Von Allmen1, B. A. Rymeski1, J. S. Frischer1  1Cincinnati Children’s Hospital Medical Center,Colorectal Center, Division Of Pediatric General & Thoracic Surgery,Cincinnati, OH, USA 2Tufts Medical Center,Floating Hospital For Children,Boston, MA, USA

Introduction:   Colorectal and perianal Crohn’s disease in the pediatric population poses a formidable therapeutic challenge.  In an effort to preserve bowel length while controlling symptoms, fecal diversion is an attractive option.  However, fecal diversion has been associated with an increased risk of a permanent stoma in the adult population.  We hypothesize that fecal diversion is more likely to be temporary in children when used in combination with anti-tumor necrosis factor (anti-TNFα) agents.

Methods: Records of patients with perianal and colonic Crohn’s disease who underwent fecal diversion (July 2006 – July 2018) at our institution were reviewed. Perioperative outcomes were analyzed using Fisher’s exact test and indepent t-test.

Results: Fecal diversion was performed in 36 patients with colorectal and perianal Crohn’s disease at a mean age of 14 years (range 2.1 to 20.7). Segmental resections were performed in 25 patients (69.4%) and proctocolectomies in 5 (16.7%). Intestinal continuity was restored in 22 (61%) patients within an average of 10.2 months (range 2 to 29), but 6 (27.2%) required re-diversion for recurrent, severe colitis or perianal disease. Re-diversion occurred at a median of 228 days after initial restoration of intestinal continuity (range 20 to 893 days).  At the conclusion of the follow-up period, 16 (44.4%) of the patients retained intestinal continuity.

     When comparing patients who were diverted for colonic versus perianal Crohn’s, there was no difference in age at time of stoma creation nor was there a difference in the duration of fecal diversion (Table).  Both groups had similar rates of initial intestinal continuity and anti-TNFα therapy after diversion.  Moreover, both groups had similar rates of re-diversion. 

     There was no difference in age at the time of fecal diversion between patients whose stomas were reversed (174.9 months, standard deviation 40.2) and those who remained diverted (157.3, standard deviation 73.0; p = 0.36).  Twenty-two received anti-TNFα therapy prior to diversion.  Of the 31 patients who received anti-TNFα therapy after diversion, 20 (55.6%) had their stomas reversed and 11 (30.6%) did not (p = 1.0).  All 6 patients who required re-diversion received anti-TNFα therapy after their initial stoma creation.

Conclusions: Despite widespread use of anti-TNFα agents, fecal diversion in pediatric colorectal or perianal Crohn’s patients is associated with low rates of sustainable restoration of intestinal continuity.

55.10 Tracheostomy to Support Traumatic Brain Injured Children and Adolescents

V. Young1, P. Evans1, H. Phelps1, M. Raees1, S. Zhao1, C. Lovvorn1, A. Greeno1, B. Brake1, C. Shannon1, H. Lovvorn1  1Vanderbilt University Medical Center,Nashville, TN, USA

Introduction:
The Brain Trauma Foundation recommends early tracheostomy in adult patients sustaining profound traumatic brain injury (TBI) and requiring prolonged mechanical ventilation (MV). This practice has been recommended for the management of pediatric patients, but its applicability and efficacy have not been reported. The purpose of this study was to evaluate practice biases and patient variables between adult and pediatric providers regarding tracheostomy in children sustaining TBI.

Methods:
The comprehensive trauma registry of a single center was queried to identify patients ≤ 18 years who presented between 2013-2017 with profound TBI and required MV. Demographic characteristics, clinical parameters, and outcomes were compared between patients treated at independent adult and pediatric hospitals and those who did or did not receive tracheostomy. Trauma care providers at both the adult and children’s hospitals were surveyed regarding their practice tendencies and inclinations regarding the role of tracheostomy in these patients.

Results:
In this cohort (n=197), 43 (21.8%) patients received tracheostomy. Demographic characteristics were not significantly different between children who did and did not receive tracheostomy. Adult (n=29, 39.7%) compared to pediatric (n=14, 11.3%) providers placed more tracheostomies (p<0.001). Tracheostomy occurred earlier in the adult than the pediatric hospital (day 6.10 ± 2.93 versus 15.14 ± 8.55, p=0.002). Injury severity score (ISS; OR 1.12, CI 95 1.06-1.18; p=0.0438), intracranial pressure (ICP) monitor duration (OR 1.11, CI 95 1.07-1.17; p=0.0162), and facial fractures (OR 2.22, CI 95 1.57-3.15; p=0.0284) were associated with tracheostomy placement. Among survivors, Kaplan-Meier analysis showed a median ventilation of 10 days with tracheostomy versus 3 days without (p<0.001). A multivariate analysis of age, gender, and ISS also found tracheostomy to correlate with longer MV, ICU duration, and hospital days (p<0.0001). Detailed survey of adult and pediatric trauma providers suggested biases for earlier and more liberal tracheostomy in TBI patients treated at the adult hospital.

Conclusion:
Attitudes and practice patterns differed between adult and pediatric providers regarding tracheostomy to support pediatric TBI patients. Potential predictors for early tracheostomy in young patients sustaining profound TBI included facial fractures, duration of ICP monitoring, and ISS. In multivariate analysis of age, gender, and ISS, tracheostomy correlated with longer duration of total MV, increased ICU days and longer hospital stays. Further work is needed to evaluate if tracheostomy is indicated earlier in these patients.
 

55.09 A Prospective Comparison of Pediatric Ovarian Mature Cystic Teratoma and Functional Pathology

I. Vannix1, A. Munoz1, R. Hazboun1, V. Pepper1, D. Moores1, E. Tagge1, J. Baerg1  1Loma Linda University Children’s Hospital,Division Of Pediatric Surgery,Loma Linda, CA, USA

Aim: To prospectively identify the unique features, radiologic-pathologic correlation and ovary preservation for pediatric ovarian mature cystic teratoma (MCT) compared to a control group. 

Methods:  Between January 2013 and December 2017, 68 girls were identified, ten infants were excluded, 58 between 2-18 years were included. MCT’s were compared to controls undergoing operation for functional adnexal pathology.  Tumor markers were obtained for all and were negative.  Chi-square compared categorical and t-tests compared continuous variables.  Radiologic-pathologic correlation was documented for ultrasound (US), computed tomography (CT) and magnetic resonance imaging (MRI).  Sensitivity, specificity, and accuracy were examined for identification of torsion with US and CT as sample sizes were similar. Ovary preservation was recorded. 

Results: Of 58 girls, 28 had 30 MCT’s, of which two had synchronous, bilateral tumors.  Four girls with unilateral MCT’s torsed (13.3%).  Of 30 controls, 8 had solitary cysts without torsion (26.7%), 22 torsed (73.3%), 4 with cysts (18.0%) and 18 without (82.0%).  

The mean age for MCT’s vs. controls was similar 12(3.0) vs. 11(4.0) years (p=0.28).  Each had similar ultrasound-measured volumes (p=0.58) and similar proportions were post-menarchal (p=0.79).

Girls with MCT’s had significantly lower body mass indexes (p=0.01), pain >1 month (p=0.02), more palpable masses (p=0.01), less torsion (p=<0.01), fewer laparoscopic procedures (p=0.02) and more oophorectomies (p=0.01).  All four ovaries were preserved in the two girls with bilateral, synchronous MCT’s (13.3%).  For unilateral MCT’s, two-thirds (66.7%) resulted in oophorectomy.

Conclusion: When compared to functional adnexal pathology, girls with MCT’s have a lower BMI, longer pain history, more palpable masses and less torsion.  Ultrasound for first-line imaging is recommended as it has similar radiologic/pathologic correlation for all lesions, but better sensitivity for torsion.  MRI is a feasible second imaging study to avoid CT radiation, and improve diagnostic capability. Promotion of ovary preservation for all MCT’s is warranted, as bilateral synchronous lesions resulted in complete ovary preservation.

55.08 Early Bolus Feeding after Fundoplication With or Without Gastrostomy

J. A. Sobrino1, H. Alemayehu1, J. A. Sujka1, L. A. Benedict1, T. A. Oyetunji1, S. D. St.Peter1, J. D. Fraser1  1Children’s Mercy Hospital- University Of Missouri Kansas City,Department Of Surgery,Kansas City, MO, USA

Introduction:
Failure of fundoplication has been linked to technical and patient-related matters. One clinical concern that persists is the early initiation of bolus feeding leading to retching and wrap disruption. Here, we evaluate this relationship in fundoplication performed with minimal dissection of the hiatus.

Methods:
A retrospective review of patients undergoing laparoscopic fundoplication with or without gastrostomy was performed from 2014 to 2016. Demographics, details of the operation, details of post-operative feeding regimen, and one-month follow up were obtained. All values are reported as medians with interquartile ranges (IQR).

Results:
Over 2 years, 58 patients were included. The median age was 200 days (IQR 118, 538). The most common preoperative symptom was retching in 58% of the patients (n=33).The median operative time was 73 minutes (IQR 57, 115). Majority of the patients (67%, n=39) had a gastrostomy tube placed during their fundoplication and 8 patients (14%) already had a gastrostomy tube.
Feeds were initiated on the day of surgery in 97% of patients, with 28 patients (51%) receiving bolus feeds and 26 (48%) on continuous feeds. Three patients (10%) on continuous feeds were started at their goal rate. Median time to conversion from continuous feeds to bolus was 2 dys (IQR 1, 6 dys) for those that were on bolus feeds pre-operatively.
Median follow up time was 1.87 mo (.83, 2.17 mo). Twenty-three patients (39.6%) reported retching, with no between those on continuous vs bolus feeds (34.4% vs 44.8%, p=0.59). 18 patients (37%) had ongoing reflux symptoms prompting further work up. One patient from the bolus feeding group was diagnosed with a wrap disruption – a non-significant difference between the groups (3% vs 0%, p=1). This patient’s symptoms were controlled and she was allowed to gain weight with a successful repair 18 months later.

Conclusion:
Bolus feeding after laparoscopic fundoplication does not lead to increased rates of wrap disruption.
 

55.07 Effect of Inaccurate Pediatric Total Body Surface Area Estimate on Fluid Resuscitation

R. D. Shelby1, A. B. Nordin1, C. McCulloh1, J. Shi2,3, R. Fabia1,2, R. K. Thakkar1,2,3  1Nationwide Children’s Hospital,Pediatric Surgery,Columbus, OH, USA 2The Research Institute at Nationwide Children’s Hospital,Center For Pediatric Trauma Research,Columbus, OH, USA 3The Research Institute at Nationwide Children’s Hospital,Center For Injury Research And Policy,Columbus, OH, USA

Introduction:  Fluid resuscitation remains the cornerstone of acute burn management. There is a delicate line between providing enough fluid volume and over resuscitation – one that can lead to significant complications. Total body surface area (TBSA) is the primary component in calculation of required fluid volume in burn injury resuscitation. We have previously shown that TBSA estimations for pediatric patients performed by prehospital emergency medical service providers were significantly higher than those reported by burn specialists at an American Burn Association (ABA)-verified pediatric burn center. Still, the relationship between TBSA overestimation and over resuscitation remains unclear. We sought to evaluate whether over estimation of TBSA prior to arrival at an ABA-verified pediatric burn center led to significantly higher fluid volume administration, resulting in over resuscitation.

Methods:  A retrospective chart review was performed of our trauma registry of patients admitted to our burn center. Inclusion criteria were children presenting with large burns, defined as TBSA burns ≥15%, from 2007 to 2015. The TBSA estimate prior to arrival to the burn center, TBSA determined by the ED physician, TBSA determined by a burn specialist, total fluid volume given before arrival to the burn center, and demographics were reviewed. The experimental standard TBSA was set as the TBSA determined by pediatric burn specialists at our ABA-verified center. This was compared with prehospital emergency medical service providers, outside hospital physicians, and our burn center ED physicians to determine presence of overestimation of TBSA. TBSA’s ≥10% or ≤10% from the standard was designated as inaccurate. Expected fluid volume was then calculated using the standard TBSA and our burn resuscitation formula. Over resuscitation was defined as receipt of fluid volume ≥50% of the expected volume.Statistical significance was determined using a paired t-test with P < 0.05.

Results: A total of 72 patients ≤18 years old met inclusion criteria. Average TBSA was 19.5 ± 12.6%, and age of 2.5 ± 0.53 years. The most common mechanism was flame (50%) followed by scald burns (47.2%). 44.4% (n=32) of the patients received +50% of their expected resuscitative intravenous fluids based on the burn specialist determined TBSA, while 55.6% of patients were either appropriately resuscitated (n=10, 13.9%) or under resuscitated (n=30, 41.7%) (P=0.408).

Conclusion: Inaccuracies in TBSA calculation can lead to potentially life-threatening and disabling complications in pediatric patients with thermal injury. However, we were unable to demonstrate an association between over resuscitation and inaccurate TBSA in this cohort. Though less than majority of our cohort was over resuscitated, these patients received an average of 150-200% of excess fluid- a statistic that remains substantial, and requires further investigation.
 

55.06 Simulated Street Crossing Intervention is Associated with Lower Rate of Pedestrian Injury in Children

C. M. McLaughlin1, W. Barry1,2, E. Barin1, M. Mert3, C. Lowery4, J. Upperman1,2, H. Arbogast4, A. M. Jensen1,2  1Children’s Hospital Los Angeles,Pediatric Surgery,Los Angeles, CA, USA 2University Of Southern California,Department Of Surgery,Los Angeles, CA, USA 3Southern California Clinical and Translational Science Institute,Los Angeles, CALIFORNIA, USA 4Children’s Hospital Los Angeles,Injury Prevention,Los Angeles, CA, USA

Introduction:  Unintentional injury is the leading cause of death in children. Elementary school children are at high risk for street crossing-related injuries. We hypothesized that an interactive pedestrian safety educational intervention is associated with lower incidence of pedestrian-related injuries in elementary school-aged children.

Methods:  An interactive pedestrian safety educational intervention was implemented at target Los Angeles County elementary schools beginning in 2009. The intervention was designed to be administered over one day and included formal didactic education followed by simulated street crossings using a life-size set (Figure). A retrospective cohort study was conducted of 10 schools from 2012-2017. A statewide traffic records database was queried for reported pedestrian-related injuries in elementary school-aged (4-11 y) children at intervention school districts. Pedestrian-related injury incidence was compared one year before and after the intervention. A standardized rate ratio was calculated by comparing injury incidence at intervention school districts to the entire city of Los Angeles. A negative binomial model was used and p<0.05 was considered statistically significant.

Results: The total number of injuries reported in elementary school-aged children in intervention school districts was n=6 before and n=2 after the intervention, resulting in a significantly lower incidence rate (1.22 vs 0.40 per 10,000 children/year; p<0.001). The standardized rate ratio still reflected a significantly lower incidence of pedestrian-related injury after the intervention (rate ratio 0.28; 95% CI 0.11-0.73).

Conclusion:
Interactive street safety education at Los Angeles elementary schools was associated with a lower incidence of pediatric pedestrian-related injury. The observed difference was still significant after adjusting for injury incidence in the entire city of Los Angeles. These data suggest formal pedestrian safety education should be included into injury prevention efforts in similar urban communities.
 

55.05 H is just a Number? Using the Hirsch Index to Describe Academic Productivity Amongst Pediatric Surgeons

K. N. Lucey1, K. T. Anderson1, M. A. Bartz-Kurycki1, M. C. Henry1  1University of Arizona Medical Center,General Surgery,Tucson, AZ, USA

Introduction: The Hirsch Index (h-index) is a measure of academic output based on the number of published works and the subsequent citations of work. It is frequently used as a factor in academic promotions. The utility of the h-index in pediatric surgery has not been established.

Methods: Faculty members of pediatric surgery fellowship programs as listed by the American Pediatric Surgical Association (APSA) were identified. Faculty rank and gender as noted on department and institutional websites were captured. Faculty of programs that did not provide a clinical appointment were noted to be instructors. Published articles, citations with and without self-citations, book citations, and h-index with and without self-citations or book citations were collected through Scopus. Publication years were defined by 2018 minus the year publications were noted by Scopus. Student’s t-test and linear regression were used for analysis.

Results: From 58 pediatric surgery fellowship programs, 456 faculty were identified who had a record in Scopus, and whose demographics were available from their institutional websites. The cohort for analysis included 42 (9.2%) department chairs, 84 (18.4%) professors, 145 (31.8%) assistant professors, 120 (26.3%) associate professors, 63 (13.8%) instructors and 2 (0.4%) professor emeritus. Most faculty were male (n=331, 72.6%). During an average of 20.5±9.6 publication years, the mean h-index was 16.8±.12.6 with a range from 1 to 111. H-index did not change significantly with the removal of self-citation (16.4±12.2, p=0.65) nor with the removal of book citations (15.8±11.5, p=0.20). H-index increased with publication years and faculty rank but the trajectory of the index differed by gender (figure). Adjusting for publication years and faculty rank, female faculty had a lower overall h-index (Male faculty mean h-index: 18.7±13.6 vs Female faculty 11.9±7.8, p<0.01).

Conclusions: According to the Hirsh index, male faculty out produce female faculty, even after accounting for length of publishing time and faculty rank in pediatric surgery. Further research is needed to evaluate causes of the discrepancy in academic productivity.

55.04 Natural History of Gastrojejunostomy Tubes in Children

R. E. Wilson1, P. K. Rao1, A. J. Cunningham1, S. Krishnaswami1, E. N. Dewey1, M. C. Boulos1, N. A. Hamilton1  1Oregon Health And Science University,Pediatric Surgery,Portland, OR, USA

Introduction:
Gastrojejunostomy (GJ) tubes are frequently used to provide enteral nutrition in patients who do not tolerate gastric feeds. GJ tubes are reported to have a high rate of minor complications, most commonly thought to be migration into stomach, requiring unplanned urgent interventions. However, there is currently insufficient literature on the lifespan of GJ tubes, reasons for failure, and recommendations for optimal techniques and timing of replacement. We aimed to evaluate the natural history of GJ tubes in pediatric patients to guide clinical management.       

Methods:
We reviewed all pediatric patients who underwent GJ tube placement or exchange at our institution from January 2012 to July 2018. Demographic data was collected, as was time and indication for replacement or removal of GJ tubes. End points for the study include permanent removal of GJ tube or mortality. Current feeding status of each patient was also recorded.

Results:

Seventy-nine patients underwent 205 GJ tube procedures. Four had prior fundoplication. The tubes lasted a median of 98 days (interquartile range = 54-166) and patients had a median of 2 GJ tubes. The most common indication for tube change was a structural/mechanical problem with the tube, occurring 56 times (43.1%). These included broken balloons (27, 20.8%), loose connector rings (17, 13.1%) and tube plugging (9, 6.9%). Other indications for tube replacement included dislodgement of tube from tract (45, 34.6%), migration of tube into stomach or esophagus (11, 8.5%), routine change (9, 6.9%), or other (9, 6.9%). Thirty-four percent of tubes replaced were able to be performed without general anesthesia or sedation (Table 1).

 

Twelve patients (15.2%) died from their primary disease during the study period. Thirty-two patients (40.5%) ultimately tolerated gastric feeds. Nine (11.4%) of these patients required subsequent fundoplication. The remaining 23 patients (29.1%) progressed to gastric feeds without subsequent fundoplication. Conversion to gastric feeding without subsequent fundoplication occurred at a median time of 212 days.

Conclusion:
Gastrojejunostomy tubes offer a safe and effective feeding option in patients who do not tolerate gastric feeds.  Most tubes fail due to intrinsic structural/mechanical issues and not secondary to migration into the stomach. No additional operative therapy is needed in 44% of patients, as many ultimately tolerate gastric feeds or suffer early mortality from their primary disease. Finally, exchange of GJ tubes without anesthesia is a viable option in many cases and can increase feasibility of long-term GJ use.

55.03 Pediatric Breast Abscess: National Epidemiology and Management at Children’s Hospitals

J. A. Sobrino1, J. A. Sujka1, L. A. Benedict1, J. D. Fraser1, T. B. Lautz2, T. A. Oyetunji1  1Children’s Mercy Hospital – University Of Missouri Kansas City,Department Of Surgery,Kansas City, MO, USA 2Ann & Robert H. Lurie Children’s Hospital of Chicago,Department Of Surgery,Chicago, IL, USA

Introduction:
Pediatric breast abscesses are rare. Data is lacking on their epidemiology and management. Adult literature has demonstrated that aspiration and antibiotics are a safe and effective first line treatment with incision and drainage (I&D) reserved for complicated collections or failure of aspiration. I&D is of particular concern in the pediatric population given the unique concern of injury to the pre-pubertal breast bud leading to a cosmetic defect. The purpose of our study was to evaluate the current national management of pediatric breast abscesses.

Methods:
A retrospective review was conducted utilizing the Pediatric Health Information System (PHIS) database including patients less than 18 years old with appropriate ICD 9 and 10 diagnostic codes for breast abscesses and corresponding procedure codes for I&D or aspiration. Patients with length of stay greater than 5 days were excluded. Details were collected on demographics, imaging, procedures, comorbidities, length of stay, and 30-day re-admissions. Medians are reported with interquartile ranges.

Results:
A total of 3,381 patients were included. Only female patients were analyzed, and this included 2,796 patients (83%). Median age at presentation was 13 years (7, 15). Seventeen percent (481) occurred in patients less than a year old. Toddler and early childhood ages accounted for 4% of patients each before increasing in middle childhood (9%) and early adolescence (60%). The racial composition was 46% white and 36% black.
Fifty-seven percent of patients had no imaging performed and 43% (1,199) underwent an ultrasound. Four hundred and eighty patients (17%) underwent I&D, 192 (7%) underwent aspiration, and 60 (2%) patients were treated with interventional radiology guided drainage. I&D rates varied from 16% to 28% by age group and 0 to 50% by hospital. There were no differences in rates of aspiration (p = 0.536) or I&D (p = 0.056) between age groups.
Mean length of stay was 1.3 days. Length of stay was longest in infants (1.8 days, p = < 0.005). The length of stay was shorter for I&D compared to aspiration (1.6 vs 1.9 days, p = 0.002). The overall 30 day readmission rate was 17% (482). There was no difference in readmission rate between those treated with I&D compared to aspiration (16% vs 14%, p = 0.64).

Conclusion:
Despite concerns for damage to a developing breast bud, rates of I&D are high throughout the nation. Overall the management of pediatric breast abscesses is variable and needs standardization with long term follow up of the two most common methods of drainage. Further study is needed.
 

55.02 Recurrence of Central Venous Catheter Infections after Treatment in Pediatric Surgical Patients

B. D. Hosfield1, A. K. Bagwell2, R. J. Vandewalle1, A. P. Ladd1  1Indiana University School Of Medicine,Pediatric Surgery,Indianapolis, IN, USA 2Indiana University School Of Medicine,Indianapolis, IN, USA

Introduction:   Children requiring long-term central venous access are at risk for catheter associated blood stream infections (CLABSI).  Treatment of CLABSI requires either removal of the catheter or antimicrobial treatment to sterilize the catheter and bloodstream, which is again at risk for CLABSI development.  The purpose of this study is to identify factors that predispose the development of a second CLABSI.  

Methods:   A single-center institutional database was queried from 1/1/2010 to 12/31/2016 for patients aged 0-18 years old with a tunneled central venous catheter (tCVC) and diagnosis of CLABSI.  The studied cohort were patients who developed a second CLABSI from the same tCVC.  Patients undergoing treatment for malignances or solid organ/hematologic transplants were excluded.  Factors associated with the treatment of the index CLABSI and the recurrence were analyzed.  

Results:  Thirty-one patients treated for CLABSI met study criteria of developing a second CLABSI.  The median age for original tCVC placement was 15 months (IQR 4.5-33 months).  Twenty-six patients in the cohort (83.9%) had short bowel syndrome (SBS).  Three additional patients with SBS were identified who could not treat their CLABSI with antibiotics (clearance rate of 89.7%).  The median time from tCVC placement to first CLABSI was 67 days (IQR 32.5-125.5 days).  Multivariate analysis revealed a significantly shorter length of time from tCVC placement to first CLABSI for patients with SBS versus other indications (59 vs. 161 days, respectively; p=0.012) when controlling for age, gender, and culture data.  The median time between initial and recurrent CLABSI in the study group was 49 days (IQR 24.5-91 days).  There was no difference in median days from first CLABSI to second CLABSI for patients with SBS versus other diagnosis (48.5 vs. 53 days, respectively; p=0.747).  Patients with isolates of Staphylococcus Aureus at the first CLABSI had a significantly longer length of time from first to second CLABSI when compared to those with other bacteria isolates on initial CLABSI cultures (104 vs. 43 days; p=0.001).

Conclusion:  Initial CLABSI appears to occur earlier in pediatric patients with SBS compared to other indications for tCVC placement.  Additionally, CLABSI due to isolated Staphylococcal Aureus infections confer a longer time before CLABSI recurrence among those lines salvaged with intravenous antibiotics.