67.14 Outcomes from Cholecystectomy in Children: Does the Type of Stone Matter?

Posted on December 22, 2014

T. Wang1, D. W. Kays1, S. D. Larson1, J. A. Taylor1, S. Islam1  1University Of Florida,Surgery,Gainesville, FL, USA

Introduction: The number of pediatric patients undergoing cholecystectomy has increased over the past decade, including cholesterol and pigment stones. No clear data exists as to whether pediatric patients with cholesterol gallstones compared to pigment gallstones have different outcomes and should be managed differently. The purpose of this study is to report a cohort of patients with cholelithiasis due to cholesterol or hematologic disease and compare treatment and outcomes.

 

Methods: A retrospective review was performed on all pediatric cholecystectomies (defined as age 18 and younger) completed at a single institution between 2001 and 2014. Clinical data points including patient demographics, indication for operation, complications, and outcome were collected on all of the patients and analyzed. The cohort was separated into cholesterol, pigment stones, and biliary dyskinesia. Students t test and Fischer’s exact test were used as appropriate and a p value of less than 0.05 was considered significant.

 

Results:

Calculous disease accounted for 311 cases in the cohort of 382 patients. Of these a majority were cholesterol stones (68.9%). Patients with pigment stones were younger and more likely to be male gender compared to cholesterol. The table shows the differences between the groups. We found no difference in the use of ultrasound or HIDA as preoperative imaging, duration of symptoms, post operative ERCP, conversions to open procedure, or in the need for urgent procedure. Complication rates, including gallstone pancreatitis, were higher with cholesterol stones, despite comorbidities in the pigment group. 

 

Conclusion: We noted significant differences between children undergoing cholecystectomy for cholesterol vs. pigment stones in presentation, preoperative procedures, complications, and symptom resolution. Pigment stone cases were less likely to have pain or nausea as symptoms and had higher percentage of symptom resolution after surgery. Prophylactic cholecystectomy for pigment stones may not be indicated based on this cohort. A prospective study would help to better understand these data. 

67.15 30-day Outcomes of Laparoscopic and Open Intestinal Resection for Pediatric Crohn’s Disease

Posted on December 22, 2014

J. B. Mahida1,2, L. Asti1, P. C. Minneci1,2, K. J. Deans1,2, B. C. Nwomeh2  1Nationwide Children’s Hospital,Center For Surgical Outcomes Research,Columbus, OH, USA 2Nationwide Children’s Hospital,Division Of Pediatric Surgery,Columbus, OH, USA

Introduction: To compare 30-day outcomes between laparoscopic and open intestinal resection in pediatric patients with Crohn’s disease using a validated national database.

Methods: We identified all small intestine resections and partial colectomies with or without ileal resection for Crohn’s disease in the 2012 National Surgical Quality Improvement Program Pediatric (NSQIP Pediatric) database and compared demographic, clinical, and 30-day outcome characteristics between patients who underwent an open or laparoscopic resection.

Results: Of the 89 patients with Crohn’s disease who underwent an intestinal resection, 30 (34%) were performed open and 59 (66%) laparoscopically. In comparison to patients undergoing laparoscopic resection, patients undergoing open resection were more likely to be classified as American Society of Anesthesia class 3 or greater (50% vs. 25%, p=0.020).  There were no significant differences in the proportion of patients undergoing operations longer than 3 hours (open vs. laparoscopic, p-value) (30% vs. 36%, p=0.598), in postoperative length of stay (median 5 vs. 5 days, p=0.472), and in all reported NSQIP-Pediatric 30-day outcomes.

Conclusion: In children undergoing intestinal resection for Crohn’s disease, laparoscopic resections are more common with open procedures more likely to be performed in patients who are more severely ill. Overall, there were no significant differences in 30-day outcomes between laparoscopic or open resections. Additional accrual of cases within NSQIP-Pediatric will allow for future risk-adjusted analyses of outcomes.

 

67.16 Does the sidedness of congenital diaphragmatic hernia affect patient outcomes?

Posted on December 22, 2014

A. C. Akinkuotu1,2, S. Cruz1,2, D. Cass1,2,4,5, T. Lee1,2, C. Cassady1,3, A. Mehollin-Ray1,3, J. Williams1,3, R. Ruano1,4, S. Welty1,5, O. Olutoye1,2,4,5  1Texas Children’s Hospital,Texas Children’s Fetal Center,Houston, TX, USA 2Baylor College Of Medicine,Michael E. DeBakey Department Of Surgery,Houston, TX, USA 3Baylor College Of Medicine,Department Of Radiology,Houston, TX, USA 4Baylor College Of Medicine,Department Of Obstetrics And Gynecology,Houston, TX, USA 5Baylor College Of Medicine,Deparment Of Pediatrics- Newborn Section,Houston, TX, USA

Introduction:
Studies comparing outcomes of right-sided (R-CDH) and left-sided (L-CDH) congenital diaphragmatic hernia are conflicting, though most suggest R-CDH is associated with higher morbidity and mortality than those with left-sided lesions.  The purpose of this study was to examine the effect of right-sidedness on outcomes, particularly pulmonary morbidity, in neonates with CDH.  We hypothesized that neonates with R-CDH would have worse short-term pulmonary morbidity compared to those with L-CDH.

Methods:
A retrospective review of all neonates with CDH diagnosed at a tertiary children’s hospital between January 2004 and January 2014 was performed. Patient characteristics and outcomes were compared between R-CDH and L-CDH. In cases of prenatally diagnosed CDH, lung volumes were measured by fetal MRI-based calculations of the observed-to-expected total fetal lung volume (O/E- TFLV). A subgroup analysis was performed for patients with isolated CDH. Statistical analyses were performed using the Mann -Whitney test for continuous variables and Fischer’s Exact test for categorical variables.

Results:
Of 169 CDH patients; 37 (21.9 %) were R-CDH and 132 (78.1%) were L-CDH.  Fetuses with R-CDH were diagnosed at a later gestational age (29.5±8.5vs. 23.5±5.6 weeks; p<0.001) compared to those with L-CDH.  Of fetuses evaluated with fetal MRI (13 R-CDH, 91 L-CDH), there was no difference in O/E-TFLV between  those with R-CDH and L-CDH (30.9±11.2% vs. 33.9±14.5%; p=0.471). There was no difference in the median duration of intubation (10 vs. 12 days; p=0.697), hospital stay (53 vs. 46 days; p=0.927), or need for oxygen at 30 days of life (51.7% vs. 46.1%;p=0.679) between groups.  Furthermore, there was no statistical difference in the use of ECMO (40.5% vs. 30.5%; p=0.321) or 6-month mortality (27.0%vs.19.7%; p=0.366) between groups.  When we examined outcomes in those with isolated CDH alone (n=128), we found a lower incidence of prenatal diagnosis and inborn status in those with R-CDH (41% v 77%, p=0.001 and 59% v. 78%, p=0.081, respectively).  For those with isolated R-CDH, ECMO was used nearly twice as often compared to those with L-CDH (Table). 

Conclusion:
Fetuses with R-CDH are less likely to be diagnosed prenatally and thus have a higher need for ECMO and higher mortality. Otherwise, the sidedness of the hernia defect appears to have no effect on short-term pulmonary morbidity in a large, contemporary single-institution experience of neonates with intrapleural CDH. 
 

67.17 Hypertrophic Pyloric Stenosis – is Postoperative Care on a Non-Surgical Service Beneficial?

Posted on December 22, 2014

S. N. Acker1, M. Hodges2, T. M. Crombleholme1, S. Somme1, A. M. Kulungowski1, D. A. Partrick1  1Children’s Hospital Colorado,Pediatric Surgery,Aurora, CO, USA 2University Of Colorado,Surgery,Aurora, CO, USA

Introduction:   Recent progress has been made in the care of Infants with hypertrophic pyloric stenosis (HPS), including earlier operative intervention and shorter hospital length of stay.  This improvement is somewhat attributable to standardized postoperative feeding protocols, which lead to shorter hospital stay. However, these protocols tend to be developed and implemented by surgeons.  We hypothesized that patients with HPS admitted to a nonsurgical service postoperatively have longer length of stay (LOS) than those on the floor. 

Methods:   We reviewed the medical records of infants who underwent pyloromyotomy for HPS at a single institution from 4/09-7/13.

Results:  259 patients underwent pyloromyotomy (35 female; 13.5%), 205 (79%) were admitted to the floor; 47 had a planned NICU admission (18%), and 7 (3%) had an unplanned NICU admission. The groups did not differ in terms of sex, age, electrolytes at presentation, or time between surgeon evaluation and operating room (Table 1).  Floor patients had longer preoperative symptom duration (8.5 vs 5.3 days, p=0.02). Operative time was longer in NICU patients (27.2 vs 32 minutes, p<0.01).  After excluding those 7 who had an unplanned NICU admission, there was no difference in postoperative apnea or hypoxic episodes; 1 floor and no NICU patients had a bradycardic episode (p=NS).  NICU patients had a longer time to full feeds (1.4 vs 2.0 days; p=0001) and longer postoperative LOS (1.6 vs 2.2 days; p=0.0012).

Conclusion:  Patients with HPS admitted to the NICU postoperatively had a longer time to full feeds and hospital LOS.  This may improve with the implementation of a hospital wide postoperative care protocol to ensure care does not differ between hospital units. 

 

67.18 Iatrogenic Esophageal Perforation in Neonates

Posted on December 22, 2014

A. J. Hesketh1,2, C. A. Behr1,3, S. Z. Soffer1,3, A. R. Hong1,3, R. D. Glick1,3  1Cohen Children’s Medical Center,Division Of Pediatric Surgery,New Hyde Park, NY, USA 2Elmezzi Graduate School Of Molecular Medicine,Manhasset, NY, USA 3Hofstra North Shore-LIJ School Of Medicine,Hempstead, NY, USA

Introduction: Esophageal perforation is a rare complication of enteric instrumentation in children, most commonly following endoscopy for stricture dilation or foreign body extraction. In preterm and low birth weight infants, enteric tube placement for suction or feeding poses a particular hazard to the delicate esophagus. Esophageal perforation in neonates due to oro/nasogastric tube insertion may be misdiagnosed as esophageal atresia or may go undiagnosed altogether. Historically, management of this life-threatening iatrogenic disease was operative, until case reports published over the last several decades described successful non-operative management. In the present study, we review neonatal esophageal perforations at our own institution in an effort to evaluate management techniques, risk factors and outcomes.

 

Methods: Our institution’s clinical database was queried for the ICD-9 code for esophageal perforation. Seventeen patients were identified, and 9 were excluded because they were older than one year. The charts of the remaining 8 patients were retrospectively reviewed for the following information: age and sex, demographics, comorbidities, cause and type of perforation, diagnostic modalities, management decisions, complications and outcomes. All study components were conducted in accordance with institutional review board policies.

 

Results: All 8 patients had esophageal perforations resulting from traumatic esophageal intubations. Six of the 8 patients were preterm, with a mean gestational age of 27.2 ± 4 weeks and birth weight of 862 ± 690 grams. The average age at diagnosis was 7 ± 6.6 days with 5 patients diagnosed within the first week of life. In all 8 patients, chest x-ray was the initial radiologic modality utilized and a non-operative approach was instituted as the initial management plan. Six patients ultimately required surgical intervention, 5 for pneumothoraces (tube thoracostomy) and one for a large thoraco-abdominal air leak (peritoneal drain insertion). Three patients died due to sequelae of prematurity (grade IV IVH and subsequent withdrawal of care, fungemia from NEC, and sepsis from chorioamnionitis). Treatment strategies included removal of the offending tube, non per os and antibiotics. One patient was diagnosed as having esophageal atresia prior to transfer to our institution; esophagoscopy in the operating room prior to operative repair of the presumed esophageal atresia established the correct diagnosis of an esophageal perforation.

 

Conclusion: Preterm, extremely low birth weight neonates are at significant risk for traumatic esophageal perforation during nasogastric or orogastric intubation. Non-operative management may be a safe initial management of esophageal perforation in the neonatal setting, but surgical interventions are often eventually indicated. Although recent early recognition and awareness of this iatrogenic disease has improved, missed and incorrect diagnoses still occur.

67.19 Criteria for Thoracoscopic Esophageal Atresia Repair and Feasibility of Prospective Randomized Trial

Posted on December 22, 2014

M. Rojnica1, C. S. Lyttle2, G. Z. Mak1, M. B. Slidell1  1University Of Chicago,Pediatric Surgery,Chicago, IL, USA 2University Of Chicago,Center For Health And The Social Sciences,Chicago, IL, USA

Introduction:
Thoracoscopic repair (TR) of esophageal atresia (EA) has gained acceptance as an alternative to the gold-standard of open repair (OR) via thoracotomy. To date, studies comparing these approaches have been retrospective in nature and typically conclude with a call for prospective randomized trials. Currently, no consensus exists on selection criteria for TR versus OR. Using published results from EA repair as well as national outcomes data, we propose selection criteria for considering TR, but also hypothesized that the sample size required for a prospective randomized trial comparing the two approaches is prohibitively large, making such a trial infeasible.

Methods:
We identified 1109 neonates with EA in the 2009 Kids' Inpatient Database (KID).  Potential selection criteria which may preclude a thoracoscopic approach were developed based on patient weight and the absence of major cardiac anomalies. Two objective endpoints for power calculations we chosen: anastomotic leak and mortality. To compare TR versus OR we estimated a mean leak rate of 12.2% vs. 10.4% and mortality rates of 0.9% vs. 1.1% respectively based upon contemporary published data. 

Results:
The mean birth weight of our cohort was 2476 grams. Proposed inclusion criteria were compared. Cohort number one consisted of weight > 2kg and no major cardiac anomalies, and cohort number two was patients with a weight > 3kg and no major cardiac anomalies. This yielded cohorts where 72% of the infants were >2kg and 27% were >3kg. When additional cardiac exclusion criteria were applied 58% of >2kg infants and 21% of >3kg infants met our criteria for TR of esophageal atresia. Even with the liberal inclusion criteria of patients >2kg and no major cardiac anomalies, our power calculations show that a prospective randomized trial would require 8,148 patients per study arm for a total of 16,296 patients in order to detect a difference in leak rates (p >0.05) and 130,636 patients to detect a difference in mortality between the two techniques.

Conclusion:
With approximately 1100 EA births annually in the USA, a prospective randomized trial using liberal inclusion criteria requires a prohibitively large sample size. Use of conservative selection criteria of weight >3kg and no major cardiac anomalies would make this study even less feasible.
 

67.20 Management of Traumatic Duodenal Hematomas in Children

Posted on December 22, 2014

M. L. Peterson1,2, P. I. Abbas1,2, S. C. Fallon1,2, B. J. Naik-Mathuria1,2, J. R. Rodriguez1,2  1Baylor College Of Medicine,Michael E. DeBakey Department Of Surgery,Houston, TX, USA 2Texas Children’s Hospital,Divison Of Pediatric Surgery,Houston, TX, USA

Introduction: Duodenal hematomas due to blunt abdominal trauma are uncommon in children. The natural history of this injury has not been well-characterized. Treatment strategies range from non-operative management to surgical decompression. The purpose of this study was to review the experience with this injury at a large volume children’s hospital.

Methods: After IRB approval, a retrospective case series was assembled containing patients with duodenal injuries secondary to blunt abdominal trauma from 2003-2014. Patients were identified through the trauma registry and validated through ICD-9 codes. Children with duodenal perforations or deserosalization injuries not associated with a hematoma were excluded. Data collection included demographics, clinical and radiographic characteristics, and hospital course. Patients with Grade I injuries based on the AAST duodenal organ injury scale were compared to those with Grade II injuries. Statistical analysis included student’s T-test and chi square.

Results: Twenty-seven patients were identified with blunt abdominal trauma resulting in duodenal injury; 19 patients (32% female) met inclusion criteria at a mean age of 9.1 ± 4.5 years. Mechanisms of injury included direct abdominal blow/handle bar injury (n=9), non-accidental trauma (n=5), falls (n=3) and motor vehicle accident (n=2). Grade I hematomas were present in 53%, and Grade II in 47% (Table 1). Hematomas were most frequently seen in the second portion of the duodenum (n=9). CT scan was the diagnostic test of choice in all patients; 10 had associated injuries (pancreas, liver, and/or other small bowel) with five (26%) of these patients undergoing a laparotomy for concerns for hollow viscous injury. No patients underwent operative drainage of the hematoma; however, one patient was treated with percutaneous drainage. Twelve patients received parenteral nutrition (PN) for a median duration of 9.3 days (range 5-14 days). Complications included readmission for concern of obstruction requiring further bowel rest (n=1). No patients developed infection of the hematoma.

Conclusion: Duodenal hematomas as a consequence of blunt abdominal trauma in children can be successfully managed non-operatively. There is acceptable morbidity, particularly with regards to the duration of parenteral nutrition, using this strategy. The AAST grade of duodenal hematoma is associated with longer duration of PN therapy and consequently longer hospital stays. These data can assist in care management planning and parental counseling.
 

68.01 Impact of Employing Damage Control Laparotomy on Pulmonary Complications and Timing of Femur Repair

Posted on December 22, 2014

J. N. Steward1, B. A. Cotton1, J. B. Holcomb1, J. A. Harvin1  1University Of Texas Health Science Center At Houston,Houston, TX, USA

Introduction:  Damage control laparotomy (DCL) is a technique initially described in the 1980’s to address multi-trauma patients with penetrating torso injury and worsening coagulopathy. However, it application increased dramatically over the next three decades and only recently has its use been called into question with data demonstrating potential morbidity and overutilization. DCL has also been shown to delay definitive repair of orthopedic injuries (especially femur fixation), with an increase in pulmonary complications. The purpose of this study was to compare pulmonary complications and delays in definitive femur fixation among patients undergoing emergent laparotomy managed by either definitive laparotomy (DEF) of DCL. 

Methods:  Following IRB approval, our trauma registry was queried for (1) all adult patients >17 years of age, (2) undergoing emergent laparotomy (directly to the operating room within 1-hour of arrival), and sustaining femur fracture. Patients were then divided into DEF group: patients with primary fascial closure at the end of emergent laparotomy and DCL: those whose fascia was left open at the end of the initial case. The co-primary outcomes of interest were pulmonary complications and time to femur fixation (both initial and/or definitive, internal). Univariate analysis was followed by purposeful linear and logistic regression. 

Results: 106 patients met study criteria; 65 in the DCL cohort and 41 in teh DEF group. There were no differences in demographics or pre-hospital vitals between the groups, with teh exception of lower systolic blood pressurein the DCL patients (median 96 vs. 119; p=0.001). The DCL group had higher Injury severity scores (median 34 vs. 26; p<0.001), more hypotension (median 87 vs. 120; p<0.001) and shock (median -8 vs. -4; p=0.002) on arrival. While initial labs and vitals demonstarted worse physiology and coagulaopthy, the temperature (96.8 vs. 96.1), systolic (124 vs, 133), base value (-4 vs. -6) and prothrombin time (16.3 vs. 16.2) were similar. Time to definitive internal fixation was longer in the DCL group (71 vs. 32 hours; p=0.010). Multivariate modeling confirmed increased risk of pulmonary (odds ratio 1.90; p=0.039) and septic complications (odds ratio 1.86, p=0.048). 

Conclusion: Employment of DCL in teh setting of emergent laparotomy delays definitive femur fixation and is associated with increased likelihood of pulmonary and septic complications. 

 

68.02 Transfer Time and Distance Do Not Impact TBI Outcomes in a Mature Rural Regional Trauma System

Posted on December 22, 2014

S. C. Gale1,2, J. Peters1, P. Detwiler1, V. Y. Dombrovskiy2  1East Texas Medical Center,Trauma Surgery,Tyler, TX, USA 2Robert Wood Johnson Medical School,Rutgers University,New Brunswick, NJ, USA

Introduction:
After injury in rural settings, evaluation at local hospitals followed by transportation to regional trauma centers may cause delays to definitive care. We sought to determine if such delays impact outcomes in patients with traumatic brain injury (TBI) within a mature rural regional trauma system.

Methods:
The East Texas Medical Center Regional Level 1 Trauma registry was queried for all patients from 2008 to 2013. Blunt TBI patients, aged ≥18 and admitted ≤24 hours from injury, were stratified as “transfer” versus “direct” admission. Demographics, transfer distance, time from injury to ETMC and outcomes (mortality, complications, length of stay (LOS)) were compared for all study patients, for patients with Injury Severity Score (ISS) ≥15, and for patients requiring neurosurgical intervention using Chi-square; logistic regression was used to identify contributors to mortality.

Results:

For the 6-year study period, 7823 patients were admitted; 1845 met inclusion criteria: 947 direct admissions and 898 transfer patients from 50 different hospitals. For transfer patients, mean travel distance was 59.3±31.7 miles; mean time to Level 1 care after injury was 4.6±2.4 hours. Transfer patients were significantly older (55 vs 49 yrs p<0.01) and had more comorbidities, but also had lower mean ISS (15.9 vs 18.5 p<0.01) and lower mortality (7.0 vs 10.0% p<0.03), complications, and LOS. Neurosurgical intervention was equivalent between groups (p=0.88). For the most injured patients, those with ISS ≥15, mortality was similar (12.4 vs 14.8% p=0.28) between groups. After logistic regression analysis of all study patients, and of those with ISS ≥15, only age and ISS, not time or distance to definitive care, significantly predicted mortality. 

Conclusion:
Neither transfer distance, nor transfer time, independently contributed to mortality after TBI in a rural setting utilizing staged care. An established and mature regional trauma system, with initial stabilization using ATLS principles at small, rural hospitals, is effective in reducing negative outcomes for injured patients in rural settings.

 

68.03 Lower Extremity DVT Screening is Not Associated with Improved Outcomes in Trauma Patients

Posted on December 22, 2014

Z. C. Dietch1, B. Edwards1, M. Thames2, P. Shah1, M. Williams1, R. Sawyer1  1University Of Virginia,Department Of Surgery,Charlottesville, VA, USA 2University Of Virginia,School Of Medicine,Charlottesville, VA, USA

Introduction: Institutions may perform lower extremity ultrasound (LUS) screening for deep venous thrombosis (DVT) in trauma patients because therapeutic intervention is thought to reduce the incidence of pulmonary embolism (PE) in patients with DVT. However, disparate screening practices reflect a lack of consensus regarding clinical indications for screening and whether screening improves clinical outcomes. We hypothesized that LUS screening for DVT is not associated with reduced incidence of PE.

Methods: The 2012 American College of Surgeons National Trauma Data Bank Research Data Set was queried to identify 442,108 patients who were treated at institutions that reported performing at least one LUS and at least one DVT. Institutions that performed LUS on more than 2% of the admitted population were designated as screening (SC) facilities and remaining institutions were designated as non-screening (NSC) facilities. Patient characteristics and risk factors were used to develop a logistic regression model to assess the independent associations between LUS and DVT, and between LUS and PE.

Results: Overall, DVT and PE were reported in 0.94% and 0.37% of the study population, respectively. DVT and PE were more commonly reported in SC than NSC (DVT: 1.12% vs. 0.72%, p<0.0001; PE: 0.40% vs. 0.33%, p=0.0004). Patients treated at SC facilities were more severely injured (ISS>9) (39.1% vs. 34.9%, p<0.001) and significantly more likely to have at least one of 11 injuries or treatment variables commonly associated with DVT. Multivariable logistic regression demonstrated that LUS was independently associated with DVT (OR=1.44, CI 1.34-1.53) but not PE (OR=1.01, CI 0.92-1.12) (c-statistic 0.86 and 0.85, respectively). Sensitivity analyses performed at various rates for designating SC facilities, or limiting analyses to patients with length of hospital stay ≥3 days, did not alter the significance of these relationships (Table).

Conclusion: The performance of LUS in trauma patients is more likely to identify DVT but is not associated with a change in the incidence of pulmonary embolism. These findings suggest that LUS DVT screening protocols will detect many clinically insignificant DVTs for which subsequent therapeutic intervention may be unnecessary, and the use of these protocols should be questioned. Furthermore, in the absence of evidence that LUS decreases the rate of PE in trauma patients, consideration of DVT as a quality and performance measure should be abandoned until prospective, randomized trials further define the role for high-risk screening protocols.

 

68.04 Validation of comorbidity-polypharmacy score as predictor of outcomes in older trauma patients

Posted on December 22, 2014

R. N. Mubang1,3, J. C. Stoltzfus6, B. A. Hoey3, C. D. Stehly2,3, D. C. Evans4, C. Jones4, T. J. Papadimos5, M. S. Cohen1, J. Grell1, W. S. Hoff1,3, P. Thomas1,3, J. Cipolla1,3, S. P. Stawicki2  1St. Luke’s University Health Network,Department Of Surgery,Bethlehem, PA, USA 2St. Luke’s University Health Network,Department Of Research & Innovation,Bethlehem, PA, USA 3St. Luke’s University Health Network,Regional Level I Trauma Center,Bethlehem, PA, USA 4The Ohio State University College Of Medicine,Department Of Surgery,Columbus, OH, USA 5The Ohio State University College Of Medicine,Department Of Anesthesiology,Columbus, OH, USA 6St. Luke’s University Health Network,The Research Institute,Bethlehem, PA, USA

Introduction: Traditional injury severity assessment is insufficient in estimating the morbidity and mortality risk for older (≥45 years) trauma patients. Commonly used tools involve complex calculations or tables, do not consider comorbidities, and often rely upon data that is not available early in the trauma patient’s hospitalization. The Comorbidity-Polypharmacy Score (CPS), a simple sum of all pre-injury medications and comorbidities, has been found to independently predict morbidity and mortality in previously published reports. These studies, however, have been limited by relatively small sample sizes. We hypothesized that CPS is indeed associated with outcomes in older trauma patients and sought to validate this association using a large administrative dataset.

Methods: After IRB approval, a retrospective study of patients aged 45 years and older was performed using the administrative database from a Level I Trauma Center. The study period was Jan 1, 2008 to Dec 31, 2013. Abstracted data included patient demographics, injury characteristics and severity, Glasgow Coma Scale (GCS), hospital (HLOS) and intensive care unit lengths of stay (ILOS), morbidity, post-discharge destination, and in-hospital mortality. Univariate analyses were conducted with mortality, all-cause morbidity, and discharge destination as primary end-points. Variables trending toward statistical significance (p<0.20) were subsequently included in multivariate logistic regression. Data are presented as adjusted odds ratios (AOR), with p<0.05 denoting statistical significance.

Results: A total of 5,839 complete patient records were analyzed. Average patient age was 68.5±15.3 years (52% male, 89% blunt mechanism), with mean GCS 14.3. Mean HLOS increased from 4.27 days for patients with CPS≤15 to 5.55 days for those with CPS≥16 (p<0.01). Likewise, ILOS increased from a mean of 0.67 days for patients with CPS≤7 group to 1.33 days for those with CPS≥22 (p<0.01). Independent predictors of mortality included age (AOR 1.05, p<0.01), CPS (per-unit AOR 1.08, p<0.02), GCS ≤8 (AOR 48.16, p<0.01), and ISS (per-unit AOR 1.08, p<0.01). Independent predictors of all-cause morbidity included age (AOR 1.02, p<0.01), GCS ≤8 (AOR 5.4, p<0.01), ISS (per-unit AOR 1.09, p<0.01), and CPS (per-unit AOR 1.04, p<0.01). CPS did not independently predict need for discharge to a facility.

Conclusion: This study confirms that CPS is an independent predictor of morbidity and mortality in older trauma patients. However, CPS was not found to be independently associated with need for discharge to a facility in the current dataset. Prospective multicenter studies are warranted to evaluate the use of CPS as a predictive and interventional tool, with special focus on correlations between specific pre-existing conditions, potential pharmacologic interactions, and morbidity/mortality patterns.
 

68.05 CTA Grading Predicts Safe Nonoperative Management in Above-Knee Blunt Lower Extremity Vascular Injury

Posted on December 22, 2014

M. R. Noorbakhsh1, M. J. Bradley1, B. Zahoor1, S. Kyere1, K. Shanmuganathan1, D. Stein1, T. M. Scalea1  1University Of Maryland,R A Cowley Shock Trauma Center,Baltimore, MD, USA

Introduction:  The use of CT angiography for detection of extremity vascular injury has become widespread.  We sought to further characterize the vascular injuries noted on CT angiogram of the lower extremity for blunt injury with the use of a novel grading system, and to use CTA grade of vascular injury to predict safe nonoperative management.

Methods: A prospectively collected institutional database was queried for all CT angiograms for blunt lower extremity injury between March 2006 and December 2013.  Admission notes, CT angiograms, traditional angiograms, operative reports, discharge summaries, and post-discharge trauma and orthopedic clinic notes were reviewed.  Data extracted included pulse exam, presence of threatened limb, extremity vascular injuries noted on CTA and in the operating room, and any delayed claudication.  Those patients whose CT examinations revealed an injury to the common femoral, superficial femoral, profunda femoral, and/or popliteal arteries were submitted to an attending radiologist for further grading.

Results: 440 CTA examinations of the lower extremity were performed on 398 patients during the study period.  140 of these examinations revealed a vascular injury.  The sensitivity of abnormal pulse exam in predicting a vascular injury noted on CTA was 69.6%, with specificity of 53.2%.  79 above-knee vascular injuries were identified and graded on 57 CTA examinations.  52 of these injuries were immediately addressed operatively.   Results of operative and nonoperative management, stratified by injury grade, are summarized in table 1.  40/41 (98%) high grade (grades 3 and 4) injuries prompted immediate operative intervention.  32 isolated low-grade (grades 1 and 2) above-knee vascular injuries were identified, of which 25 (78%) were managed nonoperatively, with no adverse sequelae (delayed amputations or claudication).

 

Conclusion: CT angiography is an excellent test to assess the vasculature in blunt lower extremity injury.  Grading of injuries, when combined with physical exam, can predict which above-knee vascular injuries are managed nonoperatively.  Nonoperative management of low-grade injuries appears to be safe, even in the setting of a documented pulse deficit but not an acutely threathened limb, with no evidence of adverse sequelae in the 25 injuries managed nonoperatively.  Grade 3b and grade 4 injuries were generally managed operatively, with a higher percentage of acutely threatened limbs requiring immediate vascular intervention.
 

67.04 Pediatric Scald Burns: Do Cooking-related Injuries Have A Higher Injury Burden?

Posted on December 22, 2014

M. Bachier1, S. E. Hammond2, T. Jones3, R. Williams1, T. Jancelewicz1, A. Feliz1  1University Of Tennessee, Health Science Center,Division Of Pediatric Surgery,Memphis, TN, USA 2University Of Tennessee, Health Science Center,Department Of General Surgery,Memphis, TN, USA 3University Of Tennessee, Health Science Center,Departments Of Pediatrics And Preventive Medicine,Memphis, TN, USA

Introduction:
Pediatric accidental scald injuries result in frequent Emergency Room visits and hospitalizations. We investigated whether burns due to food preparations produce more significant injuries leading to more extensive care and greater morbidity.

Methods:
We performed a six-year(2007-2012) retrospective chart review at a self-standing children’s hospital. All children less than 18 years old admitted for accidental scald burns(E Codes 924, 924.2; ICD-9 codes 940–949) were selected. Demographics, injury pattern, treatment and outcome data were analyzed to compare cooking versus non-cooking scald burns. The Mann-Whitney U test, a Chi-square test, and the negative binomial were used to compare continuous, categorical, and count data between groups. A Bonferroni correction was used for multiple testing.  A p < 0.05 was considered to be statistically significant.

Results:
We reviewed 308 records of children with scald burns, 85% cooking related and 15% non-cooking related (mostly bathing injuries). For all groups the majority of patients were male, African American, had public insurance, and median age was two years. Cooking burns were preferentially distributed over the head, neck and upper body, while non-cooking burns were distributed over the lower body(p values <0.02). Despite differences in burn distribution, median total body surface area for second and third degree burns was equal for all study groups(p values >0.11). Extent of care, measured by need for consults or feeding tubes (data not shown), readmissions, number of surgical debridements, or number of clinic visits, was not significantly different between groups(Table 1). Length of stay was significantly longer for non-cooking burns when compared to all cooking-related burns(Table 1). On subgroup analyses, patients burned with semisolid substances had the shortest length of stay(p <0.02); however, these patients suffered more wound contractures and limited mobility due to scarring. Likewise, patients burned with grease had a higher rate of wound contractures and limited mobility(Table 1).

Conclusion:
The majority of accidental scald burns in our population were related to food preparation and occurred in young children. Cooking-related scald burns with semisolid and grease produce more significant morbidity than non-cooking scald burns, demonstrated by more complications and greater long term disability. These patients may benefit from more aggressive therapies to prevent long term complications. Burn prevention strategies should target safety and education during food preparation.
 

67.05 Differences in Antibiotic Utilization Based on Primary Treatment for Pediatric Empyema

Posted on December 22, 2014

K. W. Gonzalez1, B. G. Dalton1, A. L. Myers1, J. G. Newland1, S. D. St. Peter1  1Children’s Mercy Hospital- University Of Missouri Kansas City,Pediatric Surgery,Kansas City, MO, USA

Introduction:

Chemical fibrinolysis has been shown in prospective randomized trials to be as effective as surgical debridement for the treatment of empyema in children, and both therapies are widely used today.  However, no studies have effectively evaluated antibiotic treatment.  We wanted to evaluate how different treatment strategies for pediatric empyema impacted antibiotic utilization. 

Methods:

A retrospective chart review of patients with empyema who underwent chemical and/or mechanical fibrinolysis at a single, dedicated children’s hospital from 2005-2013.  Data points included duration of antibiotic therapy, culture results, presence of necrosis or abscess, and adverse reactions associated with antibiotic therapy.  Immunocompromised patients and those with additional foci of infection were excluded.  Comparative analysis was performed utilizing 2-tailed Student t-tests and Pearson correlation.  Tukey HSD was used to analyze differences between chemical and mechanical fibrinolysis.

Results:

There were 169 patients identified of which 27 underwent primary video assisted thoracoscopic debridement (VATS), 123 patients had chemical fibrinolysis via tube thoracostomy with tissue plasminogen activator (tPA), and 19 had tPA followed by VATS.  The mean (±STD) duration of total antibiotic was 25.7 ± 6.5 days (d), and following a 24 hour afebrile period was 19.4 ± 6.3d.  Patients who underwent chemical fibrinolysis had a significantly shorter mean duration of parenteral antibiotic therapy when compared to primary VATS (9.2 ± 3.6d versus 11.6 ± 5.5d, p 0.04) and VATS following tPA (9.2 ± 3.6d versus 14.3 ± 8.1d, p <0.01).   There was no correlation between the length of antibiotic therapy following discharge and last recorded white blood cell count (Figure 1).  The presence of positive blood or pleural cultures also did not correlate with duration of antibiotic therapy.  Patients with necrosis or abscess (n=26) had an increased total mean duration of antibiotics (29.3 ± 5.7d versus 25.1 ± 6.4d, p <0.01).  Seventy patients (41%) had an adverse reaction related to antibiotic use (rash, nausea/vomiting, diarrhea, antifungal, other).  There was no correlation between duration of antibiotics and complication. There were no recurrences of empyema after discontinuing antibiotics.

Conclusion:

Patients diagnosed with empyema are currently placed on a protracted and variable time course of antibiotic therapy, which seems to be influenced by primary treatment and the presence of necrosis or abscess.  Since nearly half of our population experienced side effects from antimicrobial therapy, a standardized protocol with truncated duration of treatment should be considered.

 

67.06 Utility of Intraoperative Ultrasound in Resection of Pediatric Primary Liver Tumors

Posted on December 22, 2014

A. E. Felsted1, Y. Shi1, P. M. Masand2,4, J. Goss1,5,6, J. G. Nuchtern1,3, S. A. Vasudevan1,3  1Baylor College Of Medicine,Michael E. DeBakey Department Of Surgery,Houston, TX, USA 2Baylor College Of Medicine,Department Of Radiology,Houston, TX, USA 3Texas Children’s Hospital,Pediatric Surgery,Houston, TX, USA 4Texas Children’s Hospital,Pediatric Radiology,Houston, TX, USA 5Texas Children’s Hospital,Transplantation Program,Houston, TX, USA 6Michael E. DeBakey Veterans Affairs Medical Center,Liver Transplantation,Houston, TX, USA

Introduction:
Primary hepatic tumors are rare in children. While approximately two thirds of these tumors are malignant, they account for only 1.1% of childhood malignancies annually. Surgical resection or liver transplantation are the primary curative treatments for primary malignant pediatric liver tumors. When sufficiently symptomatic, benign liver lesions may also necessitate resection. Pre-operative imaging (POI) including multidetector CT and liver MRI are critical in assessing disease extent and surgical planning. The use of intraoperative ultrasound (IOUS) as an adjunct to POI in guiding surgical management of adult liver tumors is well-documented, but has not been well-studied pediatric liver tumors. We report here a single center’s experience with POI and IOUS in resection of pediatric primary liver tumors.  

Methods:
The medical records of 22 pediatric patients diagnosed with primary liver tumors and scheduled for resection between 2003 and 2014 at one institution were reviewed to determine the utility of IOUS in pediatric liver resection. IOUS finding were compared with the patients’ pre-operative MRI and/or MDCT studies. Changes in surgical management based on IOUS findings were also noted.

Results:
Of the 22 patients reviewed, 16 were evaluated using IOUS. Of those 16 evaluated with IOUS, five patients had discordant findings. In two of these cases, MRI overestimated the tumor involvement in segment IV. IOUS showed no extension into segment IV in one case and very slight involvement of segment IVa in the other. In both instances the surgeon then was able to perform a left lateral hepatectomy rather than a left hepatectomy. In the third patient, IOUS revealed tumor extension into segment IVb which was not seen on MRI. Adenomatosis was discovered in one patient believed to have a single adenoma. An additional lesion was then biopsied, confirmed an adenoma, and resection continued as planned. In the last patient MRI showed a displaced but uninvolved right hepatic vein, but on IOUS the tumor clearly surrounded the vein necessitating a formal right hepatectomy. In four other cases IOUS was necessary to identify unusual vascular anatomy or determine resection margins. Two cases were aborted after IOUS confirmed the extent of disease and proximity to critical vessels. Vascular margins on POI and IOUS were also compared for tumors less than 2 cm from major vessels. In the 12 evaluated margins nine measurements differed less than 1 mm and three differed between 1 and 4 mm.

Conclusion:
Use of IOUS in surgical management of primary pediatric liver tumors has not been universally adopted to date. We found IOUS and POI to be equivalent when assessing tumor-to-vessel margins. However, IOUS findings altered surgical management in multiple cases – particularly when evaluating tumor involvement of segment IV. Thus, IOUS should be considered a useful adjunct to POI for surgical management of pediatric primary liver tumor.
 

67.07 When the Symptoms Come Back: Gastroparesis and Biliary Dyskinesia in Teenage Girls

Posted on December 22, 2014

S. Islam1, C. D. Jolley1, S. Islam1  1University Of Florida,Surgery,Gainesville, FL, USA

Introduction:  Gastroparesis (GP) and Biliary Dyskinesia (BD) are both poorly understood problems of motility that cause significant symptoms in patients. These conditions have not been considered to occur in the same patient. The purpose of this study is to report a cohort of patients who had both conditions, and discuss management and outcomes in that group. 

Methods:  A retrospective review was performed on all children and adolescents who were treated for either GP or BD initially and developed the second disorder within 6 months.  Clinical data was collected on all the patients and compiled and analyzed. Long-term follow up was obtained by phone interviews if necessary. 

Results:  

A total of 20 patients were included in this cohort. All were female, and the mean age was 15.7 years.  Every patient had a HIDA scan (mean EF 24.8%, all had reproduction of pain with CCK injection) and gastric emptying scan (mean t1/2 =220 minutes) performed. In 12 patients, the diagnosis of BD was made first and a cholecystectomy was performed. Three cases were found to have concurrent GP, while in the remaining 9 cases the diagnosis of GP was made when the symptoms recurred, or persisted. Four of these eight cases underwent temporary GES, and two responded and had permanent GES, while the remaining two required chronic J tube feeds.

GP was the initial diagnosis in 8 girls, and they underwent temporary followed by permanent gastric electrical stimulation (GES) therapy via implantation of the device and leads. All these patients had good response to GES with symptom improvement. They then developed right upper abdominal pain, and a HIDA scan was obtained which confirmed BD. All of these patients underwent cholecystectomy. Long term follow up of the entire cohort (mean 26 months, median 20 months) reveals that there was complete relief from the BD symptoms in ALL cases. GP symptoms were managed medically in 8 cases, with 2 patients requiring feeding tube placement. Those who responded to GES (n=10) did much better with good resolution of symptoms. 

Conclusion: BD and GP may occur in a synchronous or metachronous fashion, and occurred exclusively in Caucasian teenage girls in this cohort. Excellent long-term relief of symptoms can be obtained in a majority of these cases by treating both conditions. In teenage girls undergoing treatment for either BD or GP, a high index of suspicion should be maintained for the other condition and should be investigated if symptoms persist or recur.
 

67.08 5, 4, 3, 2, 1: Embryological Variants of Pentalogy of Cantrell

Posted on December 22, 2014

B. Kaul1,2, S. Cruz1, F. Sheikh1,2, I. J. Zamora1,2, A. Mehollin-Ray2,3, C. I. Cassady2,3, T. C. Lee1,2, D. L. Cass1,2, O. O. Olutoye1,2  1Baylor College Of Medicine,Michael E. DeBakey Department Of Surgery,Houston, TX, USA 2Texas Chlidren’s Fetal Center,Houston, TX, USA 3Baylor College Of Medicine,Department Of Radiology,Houston, TX, USA

Introduction: The Pentalogy of Cantrell is a rare congenital syndrome characterized by intracardiac anomalies and defects of the abdominal wall, lower sternum, anterior diaphragm and pericardium. A varied number of the components of the Pentalogy of Cantrell may manifest as a result of varying degrees of perturbation in early embryogenesis. The purpose of this study was to evaluate our experience with this rare disease and the various embryological variants. 

Methods: Patient charts and diagnostic imaging studies of all fetuses evaluated at Texas Children's Fetal Center for Pentalogy of Cantrell between April 2004 and June 2014 were reviewed retrospectively. All patients referred to the fetal center underwent comprehensive ultrasounds, magnetic resonance imaging (MRI) and/or echocardiography.  Data collected from patient charts  fetal and postnatal imaging findings, operative treatment, pathologic evaluation and outcomes.

Results:
 There were ten patients who presented with embryological variants of Pentalogy of Cantrell over a six-year period.  Two cases displayed the full range of embryological defects observed and eight cases exhibited variants of the classic pentalogy (Table 1).  The most common embryologic defects observed were cardiac anomalies, omphalocele and congenital diaphragmatic hernia which were present in the majority of patients.  Sternal and pericardial defects were each present in 40% of patients.  Additional anomalies present included pulmonary hypoplasia, pulmonary artery stenosis and chromosomal abnormalities.  In the Pentalogy of Cantrell, diaphragmatic and pericardial defects often occur together; however, we observed four patients with diaphragmatic defects but no defect in the pericardium and one patient with a defective pericardium but no associated diaphragmatic defect, suggesting highly specific losses of somatic mesoderm during embryological development.  One patient was lost to follow up and a second patient underwent termination of pregnancy.  Five out of the remaining eight patients survived, one of which who had the full range of embryologic defects and now attends preschool but requires speech and occupational therapy.  The remaining surviving patients have developed without serious sequalae.

Conclusion: This contemporary series demonstrates that a fetus with Pentalogy of Cantrell can survive but at the expense of significant morbidity. Those with partial components of the Pentalogy have a more normal course and excellent outcome. This report highlights the very specific embryological losses that occur during early fetal development and give rise to the spectrum of anomalies observed in the Pentalogy of Cantrell.     
 

67.09 Thyroglossal duct cyst: does surgical specialty impact complication rates?

Posted on December 22, 2014

M. S. Arda1, G. Ortega2, V. F. Pinard2, E. Jelin1, F. Qureshi1  1Children’s National Medical Center,General Pediatric Surgery,Washington, DC, USA 2Howard University College Of Medicine,Department Of Surgery,Washington, DC, USA

Introduction:  Thyroglossal Duct Cyst (TGDC) is the most common congenital neck in children.  Surgical management by sistrunk procedure is the mainstay of treatment.  Pediatric general surgeons (GS) and otolaryngologists (ENT) perform this procedure. Recurrence occurs in 6-10% of children. Studies have suggested preoperative infection, age, abscess and multicystic histology as possible underlying reasons for recurrence.  The impact of surgical subspecialty has not been studied as ENT surgeons often core out the foramen cecum at the base of the tongue.  The aim of our study is to evaluate surgical outcomes of children undergoing sistrunk procedures for TGDC based on surgeon type at an academic pediatric institution. 

Methods:  A retrospective chart review of children diagnosed with TGDC from 2004-2014 was conducted. Inclusion criteria were children who underwent the sistrunk procedure and surgeon type was either GS or ENT. Basic demographic, preoperative clinical signs, age at surgery, surgeon type, operative characteristics, postoperative complication and recurrence rates were extracted. The primary outcome was recurrence rate of TGDC. Statistical analysis was conducted via descriptive statistics, t-test, and Pearson chi-square with p < 0.05 deemed statistically significant.

Results: A total of 165 patients met our inclusion criteria. 53.3% were male with swelling being the most frequent preoperative sign (92.7%), followed by infection (17.4%) and fistula (15.7%). The mean age at surgery was 5.4 years. The mean volume of TGDC excised was 4.7cm3 and most were single cyst (90.5%) The most common complications were swelling (82.0%), followed by seroma (9.0%) and wound infection (7.3%). ENT surgeons performed 106 procedures (64.2%), the mean age at surgery for GS and ENT were 5.6 and 5.3 years, respectively (p=0.67). The mean volume of TGDC for GPS and ENT were 2.5 and 5.9 cm3,  (p<0.01). The overall recurrence rate was 8.4%. The complications by surgeon type were swelling (GPS 18.6%, ENT 17.9%, p=0.90), seroma (GPS 8.5%, ENT 9.4%, p=0.83), and wound infection (GPS 6.8%, ENT 8.5%, p=0.69). The recurrence rates were 11.7% and 6.6% for GPS and ENT, respectively (p=0.25).  

Conclusion: Our study demonstrates that there is no difference in recurrence rates for TGDC excision between GS and ENT specialties. There was no added benefit of foramen cecum resection. Furthermore, no preoperative predictive factors for recurrence were identified.  This study is limited by its retrospective nature and a larger prospective study may be able to identify differences between GS and ENT procedures. 

 

67.10 Injury Patterns Associated with Pediatric Bicycle Accidents: Experience Of A Level 1 Trauma Center

Posted on December 22, 2014

C. J. Allen2, J. Tashiro1, J. P. Meizoso2, J. J. Ray2, C. I. Schulman2, E. A. Perez1, D. Lasko1, H. L. Neville1, K. G. Proctor2, J. E. Sola1  1University Of Miami,Pediatric Surgery,Miami, FL, USA 2University Of Miami,Trauma And Critical Care,Miami, FL, USA

Introduction:  Traffic accidents represent a leading cause of severe injury in children in the USA.  Bicycles are connected to more pediatric injuries than any other consumer product, other than automobiles. Whereas patterns of injury in motor vehicle accidents have been well characterized and have led to major safety initiatives and treatment guidelines, information related to pediatric bicyclist injuries is lacking.   With this recognition, our purpose is to identify major injury patterns associated with bicycle accidents in children admitted over a decade at a large pediatric level 1 trauma center. 

Methods:  From January 2000 to December 2012, consecutive pediatric admissions (≤17y) at a Level I trauma center were retrospectively reviewed for mechanism of injury, demographics, initial laboratory values and vital signs, injury patterns, Injury Severity Score (ISS), operative intervention, length of stay (LOS), and survival.  Analysis was performed to recognize injury patterns and outcomes significantly associated with bicycle related accidents. Parametric data presented as mean±standard deviation and nonparametric data presented as median(interquartile range).

Results: A total of 80 pediatric patients were admitted following bicycle related trauma (4% of all pediatric trauma admissions). The cohort had an age of 11±4y, ISS of 11±10, was 48% black, and 81% male. Injury patterns included 21% isolated head, 21% isolated abdominal, 13% isolated extremity, and 35% multiple injuries.  15% required immediate operative intervention (6.3% abdominal, 8.8% orthopedic, no neurological). LOS was 5±7d with an overall mortality of 2.5%. The most common incident day of the week for bicyclist trauma was Friday at 24%, compared to 15% of all other traumas (p=0.03). Of these patients, 13% were age 0-6y, 49% were age 7-11y, and 39% were 12-17y (p<0.001). Younger children were significantly more likely to require an abdominal operation (20% in 0-6y vs 6.8% in 7-11y vs 3.2% in 12-17y, p=0.05). Mortality rates were similar at 0%, 2.3%, and 3.2% for the three age groups, respectively (p=NS).

Conclusion: The mortality rate associated with bicyclist injury in children is comparable to that of the  overall pediatric trauma population. Orthopedic injury is the most frequent overall indication for surgery, yet the youngest children more often required an abdominal operation. This description of pediatric bicyclist injury patterns may help with the development of more specific preventive measures and anticipate injuries and outcomes according to age group.

67.11 Emergency Department Visits and Readmissions Among Children After Gastrostomy Tube Placement

Posted on December 22, 2014

A. Goldin2, K. Heiss3, M. Hall4, D. Rothstein5, P. Minneci6, M. Blakely7, S. Shah9, S. Rangel10, L. Berman12, C. Snyder11, C. Vinocur12, M. Browne8, M. Raval3, M. Arca13  2Seattle Children’s Hospital,Seattle, WA, USA 3Emory University School Of Medicine,Atlanta, GA, USA 4Children’s Hospital Association,Overland Park, KANSAS, USA 5Women And Children’s Hospital Of Buffalo,Buffalo, NEW YORK, USA 6Nationwide Children’s Hospital,Columbus, OH, USA 7Vanderbilt University Medical Center,Nashville, TN, USA 8Lurie Children’s Hospital,Chicago, ILLINOIS, USA 9Cincinnati Children’s Hospital Medical Center,Cincinnati, OH, USA 10Children’s Hospital Boston,Boston, MA, USA 11Children’s Mercy Hospital- University Of Missouri Kansas City,Kansas City, MO, USA 12Nemours Alfred DuPont Hospital For Chldren,Wilmington, DE, USA 13Children’s Hospital Of Wisconsin,Milwaukee, WI, USA

Introduction:

Gastrostomy tubes (GT) are devices placed to supplement hydration, nutrition, and medication administration in infants and children. Anecdotally, children with these GT’s utilize health resources frequently.  We wanted to quantify the post-operative needs of this population by identifying the incidence of 30-day post-discharge emergency department (ED) visits and hospital readmissions following GT placement.  We wanted to identify the GT-related reasons for visits and readmissions.

Methods:
This multicenter retrospective cohort study used data from the Pediatric Health Information System. Patients <18 years of age who were discharged between January 1, 2010 and December 31, 2012, with an ICD-9-CM procedure code for GT placement during the index hospitalization were included. Subjects were classified as having the GT placed on the date of admission (scheduled gastrostomy cohort) or later in the hospital course (unscheduled gastrostomy cohort). Factors significantly associated with ED revisits and hospital readmissions within 30-days of hospital discharge were identified using multivariable logistic regression.  IRB-approved validation studies were conducted in five institutions on a randomly generated subset of patients to confirm the accuracy of exposures and outcomes.

Results:
During the study period, 15,642 patients had a GT placed; 67% were ≤ one year old, 25% had the GT placed on the day of admission, and 72% had ≥1 chronic comorbid condition (CCC).  Overall, 8.6% of all patients had an ED visit within 30 days of hospital discharge, and 3.9% of all patients were readmitted through the ED with GT-related issues. GT-related events associated with these visits included infection (27%), mechanical complication (22%), and replacement (19%).  In multivariable analysis, timing of GT placement (scheduled vs. unscheduled) was not associated with either ED revisits or hospital readmission. Hispanic ethnicity, non-Hispanic black race, and the presence of three or more CCCs were independently associated with ED revisits, while gastroesophageal reflux and not having a concomitant fundoplication at time of GT placement were independently associated with hospital readmission.

Conclusion:
GT placement is associated with high rates of ED revisits and hospital readmissions in the first 30 days after hospital discharge.  The association of non-modifiable risk factors such as race/ethnicity and medical complexity is an initial step towards understanding this population so that interventions can be developed to decrease these potentially preventable occurrences. 
 

« Previous Page
Next Page »