10.05 Key findings from a prospective trauma registry at a Regional Hospital in Southwest Cameroon

Posted on January 25, 2017

O. C. Nwanna-Nzewunwa1, A. C. Mefire2,3, V. V. Siysi2,3, I. Feldhaus1, R. A. Dicker1, C. Juillard1  1University Of California – San Francisco,Center For Global Surgical Studies,San Francisco, CA, USA 2University Of Buea,Faculty Of Health Sciences,Buea, , Cameroon 3Regional Hospital Limbe,Limbe, , Cameroon

Introduction:
Trauma is a leading cause of morbidity and mortality worldwide. Data characterizing the burden of trauma in low- and middle-income countries (LMIC), such as Cameroon, are limited. Regular, prospective injury surveillance can address the shortcomings of existing hospital administrative logs. This study aims to characterize trauma as seen at the emergency department of a Regional Referral Hospital and assess the completeness of trauma registry data relative to administrative records. 

Methods:

From January 2008 to October 2013, we prospectively captured data on injured patients using a strategically designed, context-relevant trauma registry instrument filled out by clinicians during their normal course of work. Indicators around patient demographics, injury characteristics, delay in accessing care, and treatment outcomes were recorded. Descriptive, bivariate, and multivariate statistical analyses were conducted. A two-sample test of proportions was used to compare this trauma registry data with historical data from each of the following sources: 1) administrative records at the same facility, 2) a Central Hospital trauma registry, and 3) the central hospital administrative records.

Results:

A total of 5,617 patients with mean age of 26.04 years (95% C.I.) presented to the hospital with an injury; 67% were male. Students (27%) were the most affected occupation category. Road traffic injuries (RTIs) comprised the leading cause of injury (56%), followed by assault (22%), and domestic accidents (13%). Overall, proportions of missing data typically ranged from 0.5% to 8.2% with the exception of RTS and type of collision variables. Relative to the administrative records, the trauma registry at the Regional Referral Hospital showed a significantly higher level of documentation of patient occupation, vital signs, injury severity score (ISS), Glasgow Coma Score (GCS) and anatomical location, transport time, treatment outcome (p<0.001). Improvement in documentation was highest for transport time (88%), ISS (84%), GCS (59%) and blood pressure (58%) but respiratory rate  had the lowest improvement rate (24%) . The Regional Hospital registry had lower documentation rates for vital signs, ISS, and GCS than the Central Hospital registry (p<0.001).

Conclusion:

Implementation of a context-appropriate trauma registry by providers is feasible in developing settings. It can provide valuable, high-quality data that can inform trauma care quality improvement efforts and policy development. Study findings point towards the need for injury prevention interventions and policies that will prioritize high-risks groups, such as those aged 20-29 years, men, and those working in occupations requiring frequent road travel. 

10.04 Police Transport of Bluntly Injured Trauma Patients in Philadelphia, 2006-2015

Posted on January 25, 2017

C. E. Sharoky2, E. J. Kaufman2, S. F. Jacoby2, P. M. Reilly2, D. N. Holena2  2Hospital Of The University Of Pennsylvania,Traumatology, Surgical Critical Care And Emergency Surgery,Philadelphia, PA, USA

Introduction:  Police transport (PT) of penetrating trauma patients has been promoted as a method of decreasing pre-hospital times for patients with life-threatening hemorrhage, and is part of police policy in Philadelphia. We hypothesized that bluntly injured patients would also undergo PT, particularly if injured by MVC or assault, and that outcomes would be worse for PT vs. ambulance transport.

Methods:  We used Pennsylvania Trauma Outcomes Study registry data from 2006-2015 to identify bluntly injured adult patients transported to all 8 trauma centers in Philadelphia. PT was compared to ambulance transport, excluding transfers, burn patients, and private transport. We compared demographics, mechanism, and injury outcomes between PT and ambulance transport patients. We used multivariable logistic regression to identify independent predictors of PT. 

Results: Of 30,501 bluntly injured patients, 327 (1.1%) were transported by police and 29,903 (98.9%) by ambulance. PT patients were younger, more often male (83.5 vs. 62.3%, p <0.001), and black (56.7 vs. 37.1%, p < 0.001). More PT patients were injured by assault, and fewer were injured at home. PT patients had lower mortality (2.1 vs. 6.1%). Of PT patients, 9.8% had spine injuries, and 40.1% had head injuries (AIS ≥ 2). PT patients were more often discharged to jail (32.8 vs. 1.4%). In multivariable logistic regression, black race was associated with increased odds of PT. Age ≥ 50 was protective, as was female sex and Injury Severity Score ≥ 9. Compared to fall, all other mechanisms were associated with increased odds of PT, particularly assault (Table 1).

Conclusion: PT affects a small minority of blunt trauma patients, and did not appear associated with worse outcomes. However, PT patients included many who could have benefited from proven, pre-hospital interventions such as supplemental oxygen for brain injury or spine stabilization. Further research is needed to understand the indications for PT within the trauma care system and the impact of PT on patients.
 

10.03 Impact of Triage Guidelines on Pre-hospital Triage Errors

Posted on January 25, 2017

P. P. Parikh1, P. J. Parikh4, B. Guthrie4, L. Mamer4, M. Whitmill1, T. Erskine2, R. Woods1, J. Saxe3  1Wright State University,Department Of Surgery,Dayton, OH, USA 2Ohio Department Of Public Safety,Emergency Medical Services,Columbus, OH, USA 3Marian University,Surgery,Indianapolis, IN, USA 4Wright State University,Department Of Biomedical Industrial, And Human Factors Engineering,Dayton, OH, USA

Introduction:
The American College of Surgeons (ACS) developed the National Field Triage Decision Scheme (NFTDS) that has been adapted by many trauma centers in the nation. Some states have modified the NFTDS in an effort to make it more relevant to their population. However, quantitative evidence of the efficacy of these guidelines for a regional trauma is unclear. The objective of this study is to evaluate how the NFTDS and state guidelines compared against observed rates, and against a statistically derived-model, on triage errors.  

Methods:
This study included the state trauma data (EMS data merged with Trauma Registry) from Ohio Department of Public Safety (ODPS).  The EMS Incident Reporting System (EMSIRS) 2 data were used for the study that ranged from 2008-2012, which accounted for a total of 4914 patient records.  Out of these, for 817 patient records “triage protocol” was used for trauma triage decisions. Both the Ohio state and NFTDS were simulated using these data. A predictive model based on multivariate logistic regression was developed and validated using a train-test approach (AUC=0.76). We used the ISS method to identify triage errors for these three approaches.  

Results:
As shown in Table 1, the model performed better than NFTDS and state level guidelines by significantly reducing over-triage (OT) rates for the same under-triage (UT) rates, and vice versa.  Some factors that were significant and thus added in the statistical model, but not present in the NFTDS were; blunt injury, severe pain, complaint in head, chest, abdomen, and whole body, and change in responsiveness. The observed OT and UT rates were 38.31% and 4.04%, respectively for the state of Ohio.  The statistical model performed similar to observed rates.  However, since every region has their own version of “protocol,” many such versions existed in the state of Ohio.  The model, therefore, may help standardize the triage decision scheme in the state.

Conclusion:
Use of national and state guidelines would have resulted in significantly higher OT rates.  The statistical model has many other factors that such guidelines do not have, which might have resulted in better performance. It might be helpful revising these guidelines to include those factors.  Further, standardizing the “triage protocol” in the state would certainly help identify factors that affect OT and UT rates, and help improve the performance statewide. 
 

10.02 Post-Anticipated Discharge Length of Stay: An Actionable Target for Quality Improvement in Trauma

Posted on January 25, 2017

Z. G. Hashmi1, K. Florecki1, B. Covey1, D. Codling1, K. Sweeney1, L. Smith1, T. McNay1, H. Park1  1Sinai Hospital Of Baltimore,Department Of Surgery,Baltimore, MD, USA

Introduction:

Prior research cites non-clinical reasons, such as delays in rehabilitation facility placements, as the predominant cause for prolonged length of stay (LOS) in trauma. However, the currently used total LOS metric is non-informative to specifically measure delays due to these non-clinical reasons. The objective of this study is to evaluate post-anticipated discharge length of stay (PAD-LOS) as a more actionable QI and benchmarking metric in trauma. We hypothesize that the PAD-LOS will provide a better target to improve efficiency in transitions to post-acute care.

Methods:

All adult (≥16 years) trauma patients with blunt/penetrating injuries discharged by the Trauma Service at a Level II Trauma Center between October 2015 and May 2016 were included. Post-anticipated discharge length of stay (PAD-LOS) was defined as the LOS after the attending surgeon’s decision to discharge the patient, marking the completion of active clinical care. Patient demographic and injury characteristics were identified using the institutional trauma registry. Inpatient cost estimates were calculated using the Healthcare Cost and Utilization Project (HCUP) Nationwide Inpatient Sample (NIS) 2013.

Results:

A total of 185 trauma patients were identified. Most of the patients were young (median age 45, interquartile range 28-65 years), males (62%) with blunt injuries (77%). More than one third of the patients (69/185) had non-home discharge dispositions, while 22% were uninsured. Complete PAD-LOS data was recorded for 84 patients. Overall, PAD-LOS accounted for 28.5% (104/364 patient days) of in-patient time with an estimated attributable cost of $213,164. Non-home discharges had higher PAD-LOS, with discharges to inpatient rehabilitation requiring an additional 2 days of non-clinical hospital stay (Figure 1).

Conclusion:

Post-anticipated discharge length of stay readily identifies groups at high risk for prolonged LOS and provides an actionable target to directly address delays in discharge due to non-clinical reasons. At individual trauma centers, PAD-LOS can help drive quality improvement (QI) initiatives to improve discharge efficiency. System-wide, PAD-LOS benchmarking may enable separate, more informed quality of care comparisons between trauma centers by differentiating prolonged LOS attributable to clinical and non-clinical reasons and help implement targeted QI initiatives.  

 

10.01 ~~Simulation Based Testing of Operative Distractions during Laparoscopic Cholecystectomies

Posted on January 25, 2017

J. Sujka1, W. Havron1, K. Safcsak1, I. Bhullar1  1Orlando Regional Medical Center,Department Of Surgical Education,Orlando, FL, USA

Introduction:
~~Advances in technology have improved access to physicians and increased demand for immediate response regarding patient care. These interruptions have become accepted even in the operating room. The purpose of this study was to use simulation based operative distractions during laparoscopic cholecystectomies (LC) to determine if these interruptions affected operative time, operative safety, and appropriate management of pager issue.

Methods:
~~Utilizing a Simbionix Lap Mentor II simulator we tested 10 surgical resident volunteers (PGY 2-5) with previous experience in LC. Each resident performed 2 training modules to familiarize themselves with the simulator followed by 6 simulated LC; 3 with interruptions (INT) and 3 with no interruptions (NO-INT). The order of the 6 LC was randomized for each resident. During the INT cholecystectomy a page interruption was sent to the resident pager which was answered by the examiner. Effective management of the pager-issue was determined on a pass/fail basis based on appropriate questions asked and effective resolution of issue. These pages were timed to coincide with the dissection of critical view of safety and clipping of the cystic duct. Data was also collected by the simulator for the following operative endpoints: time to completion of LC, safety and efficiency of cautery, instrument movement, total cautery time, and complications of common bile duct injury and gallbladder perforation.

Results:
~~10 surgical residents were tested: 2 PGY-II, 3 PGY-III, 3 PGY-IV, and 2 PGY-V. There was no significant difference in the total time (minutes) required to complete the LC between the two groups, (NO-INT vs. INT, 6.3 ± 2.7 vs. 6.2 ± 2.2, p=0.9). There were no major complications of common bile duct injury in either group. Although a higher number of gallbladder perforation complications occurred in the NO-INT group this did not reach significance. There was no significant difference between the two groups for the operative end points: safety and efficiency of cautery, average left or right instrument movement (cm/sec), and total cautery time (secs). However, the residents failed to adequately address and resolve the patient issues of the pager interruption. The patient on the floor suffered and did not receive adequate care.  Nine out of the 10 residents failed to ask the appropriate questions and 9 of the 10 residents failed to effectively manage the patient problem. The mean failure rate for questioning was 45% and for resolving the issue 28%.

Conclusion:
~~Although pager interruptions did not affect the operative patient (time to completion or complications), there were significant failures in the appropriate evaluation and management of the pager issue and floor patient. Consideration for diversion of floor patient care issues to fellow residents not in the operating room to improve quality and safety of patient care requires further study.
 

09.20 Misdiagnosing Pediatric Appendicitis: Clinical, Economic, and Socioeconomic Implications

Posted on January 25, 2017

G. Dubrovsky2, J. Rouch2, N. Huynh2, S. Friedlander1, Y. Lu1, S. L. Lee1,2  1Harbor-UCLA,Surgery And Pediatrics,Torrance, CA, USA 2UCLA,Surgery And Pediatrics,Los Angeles, CA, USA

Introduction: ~~Misdiagnosing appendicitis may lead to unnecessary surgery. The study evaluates the risk factors for negative appendectomies, as well as the clinical, financial and socioeconomic consequences of negative appendectomy (NA) across 3 states.

Methods: ~~Data were obtained from the California, New York and Florida State Inpatient Databases 2005-2011. Patients (<18 years) who underwent non-incidental appendectomies (n=156,660) were evaluated with hierarchical and multivariate negative binomial regression analyses on outcomes including hospital cost, length of stay (LOS) and associated morbidity.

Results:~~Overall rates of negative appendicitis (NA), where a normal appendix was found at operation, have decreased over time (3.2% in 2005 to 1.8% in 2011, p <0.01) (Fig 1). Perforated appendicitis (PA) has also decreased (25.6% in 2005 to 24.1% in 2011, p <0.01). This reflects an increase in true acute non-perforated appendicitis (ANPA). However, certain subpopulations are at higher risk for undergoing surgery for NA. Age (<5 years), whites, females, use of laparoscopy, having private insurance and care at a low volume hospital were all significant risk factors for NA. In contrast, significant risk factors for PA include age (<5 years), African Americans, males, having open surgery and having public or no insurance. Compared to ANPA, NA patients are associated with increasing hospital stay, greater cost and higher morbidity. LOS for patients with NA showed a 5.6% increase over time (3.0 to 3.3 days), relative to a 13.7% decrease (2.0 to 1.7 days) for patients with ANPA, p <0.01 (Fig 1). Hospital costs averaged over time for NA are greater than ANPA ($6,926 vs $6,492 per patient, p <0.01) and morbidity is significantly higher (2.5% for NA vs 1.3% for ANPA, p <0.01). An estimated average of $4.6 million in total hospital costs/year resulted from admissions related to NA in California, New York and Florida. Of the three states, California had higher rates of NA and PA, as well as a significantly higher complication rate and median cost.

Conclusion:~~Despite a low incidence, NA is associated with longer LOS, higher cost and greater morbidity than ANPA. Whites, females and privately insured patients were associated with higher NA rates while those at an increased risk for PA were African Americans, males and those with public or no insurance. The highest rate of NA and PA was in California, which also had the highest median cost and the highest rate of complications. Further research is needed to understand what drives such disparities and to inform efforts to improve quality of hospital care across all groups of patients.

 

09.19 Safety of Peri-operative Ketorolac Administration in Pediatric Appendectomy

Posted on January 25, 2017

R. M. Dorman1,2, H. Naseem2, G. Ventro1,2, D. H. Rothstein1,2, K. Vali1,2  1State University Of New York At Buffalo,Department Of Surgery,Buffalo, NY, USA 2Women And Children’s Hospital Of Buffalo,Department Of Pediatric Surgery,Buffalo, NY, USA

Introduction: We sought to evaluate the impact of peri-operative ketorolac administration on outcomes in pediatric appendectomy. 

Methods:  The Pediatric Health Information System database was queried for patients aged 5-17 with a primary diagnosis of appendicitis and a primary procedure of appendectomy during the period 2010-2014. Patients with procedures suggesting incidental appendectomy, those records with data quality issues, deaths, and extra-corporeal membrane oxygenation (ECMO) were excluded. Variables recorded included age, sex, race, ethnicity, discharge year, complex chronic conditions (CCC), geographic region, intensive care unit admission, mechanical ventilation, and whether appendicitis was coded as complicated. The exposure variable was ketorolac administration on the day of or day after operation. The primary outcomes of interest were any surgical complications during the initial encounter, postoperative length of stay (LOS), total cost for the initial visit, any readmission to ambulatory, observation, or inpatient status within 30 days, and readmission with a diagnosis of peritoneal abscess or other postoperative infection or with transabdominal drainage performed.

Results: 78,926 were included in the analysis cohort.  Mean age was 11.4 years (SD 3.3 years), the majority were male (61%), white (70%), and non-Hispanic (65%). Few had a CCC (3%) or required mechanical ventilation (2%) or an ICU admission (1%).  Patients with complicated appendicitis comprised 28% of the cohort. Most (73%) received ketorolac on postoperative day 0-1; those with complicated appendicitis were less likely to receive ketorolac. In all, 2.6% of the cohort had a surgical complication during the index visit, 4.3% were readmitted within 30 days, and 2% had a post-operative infection or transabdominal drainage (1% in the uncomplicated group, 5% in the complicated group). Median post-operative LOS was 1 day and mean cost was $9,811 ±  $9,509. On bivariate analysis, ketorolac administration was associated with a decrease in same-visit surgical complications (p=0.004) and cost ($459 decrease, p<0.001) but was not associated with readmission, post-operative LOS, or post-operative infection. On multivariate analysis, ketorolac administration was associated with a significant decrease in any complication (adjusted odd ratio 0.89, 95% C.I. 0.80-0.99) and cost (ANOVA p<0.001) but was not associated with readmission, post-operative LOS, or post-operative infection.  

Conclusion: Based on a large, contemporary data set of children’s hospitals, ketorolac administration in the immediate post-operative period after appendectomy for appendicitis is common and was not associated with an increase in post-operative LOS, post-operative infection, or any-cause 30-day readmission. Ketorolac was, however, independently associated with a lower overall rate of post-operative complications and cost in this population. 

09.18 Impact of Insurance and Economic status on care for Pyloric Stenosis

Posted on January 25, 2017

K. M. Herremans1, A. Yohann1, J. A. Taylor1, S. D. Larson1, D. Solomon1, S. Islam1  1University Of Florida,Pediatric Surgery,Gainesville, FL, USA

Purpose: Hypertrophic pyloric stenosis (HPS) is a common condition in children with a 1:500 incidence. Patients typically present with forceful, repetitive emesis which leads to dehydration and failure to thrive. Delay in presentation is associated with a longer pre operative resuscitation and potentially longer hospital length of stay. The purpose of this study was to investigate the relationship between insurance and economic status with a delay in presentation for HPS.

 

Methods: A retrospective review of patients presenting with HPS was performed over a 5 year period. Patients who did not have the main procedure performed at our institution were excluded. Data regarding demographics, clinical history, time to presentation, and hospital course were collected. Insurance status was assessed, and patients without that information were excluded from final analysis. Zip code related census data was used as a proxy for economic status. The cohort was then divided by time to presentation and severity of hypochloremia. Students t test, Fisher’s exact test, and the Mann Whitney U test were used for comparative analysis.

 

Results: 167 patients were found, of which 155 met inclusion criteria. Overall, 84% patients had Medicaid. The cohort was divided into symptom duration less than or greater than 10 days, as well as by insurance status, and initial chloride level greater than or less than 95. There was no difference in racial or gender distribution among the groups, and the income data was not significantly different either. Commercial insurance patients had a shorter duration of symptoms than Medicaid cases (5.4 vs. 11.2 days), had higher chloride levels (99.4 vs. 91.5), were significantly younger (32.3 vs. 40.9 days), and spent half a day less in the hospital 3.27 vs. 3.85 days).

 

Conclusions: Patients with commercial insurance were significantly more likely to present earlier in the course, have a younger age, as well as a higher chloride level. In addition, patients with Medicaid tended to have more weight loss at presentation and stayed in the hospital longer. Despite the inability of census based income data to show any significance, these data would suggest that access to care may play an important role in timeliness of care for HPS. 

 

09.17 Factors affecting Complications in Tunnelled line Placement in Children

Posted on January 25, 2017

N. Laconi1, S. Islam1  1University Of Florida,Gainesville, FL, USA

 

Introduction:

A large number of children require long term central venous access for a variety of reasons. Tunneled lines are the most common kind of access used in these patients as they can last for longer duration. These catheters can be associated with complications and result in substantial morbidity. The purpose of this study was to better understand the factors associated with complications with tunneled central lines in children

Methods:

After obtaining IRB approval, the hospital database was searched for all central lines placed in patients aged 0-18 years of age over a 5 year period. Patients with PICC and non tunneled central lines were excluded. Data regarding demographics, indications, line type, complications and outcomes were collected and compiled. The cohort was divided into those who had any complication vs. those who did not. Students t test and fischers exact test were used where appropriate and a p value of less than 0.05 was considered significant.

Results:

594 children had tunneled lines placed. Overall mean age was 7.7 years, 52% were male and 43% were placed for cancer. A majority were placed in the subclavian vein and were single lumen. Completely subcutaneous ports were also the most common. 164 cases had a complication (27%), with 82 occlusions and 58 infections. Patients with complications were younger (p=0.002), and had a lower absolute neutrophil count at the time of line placement. Patients with complications were also more likely to have a non neoplastic diagnosis, more frequent access, and have Medicaid for their insurance. There was no difference in complications whether the subclavian or internal jugular were used, nor which service placed it

Conclusions:

In a large cohort of children with long term central access, there was a substantial overall complication rate. Younger patients, with a lower neutrophil count, and with a non neoplastic diagnosis were more likely to have a problem. Significantly more patients with complications had Medicaid for insurance, which is a proxy for poverty. This analysis will help to improve outcomes for children with long term central lines. 

09.16 Postoperative Complications in Children with Congenital Malformations – A NSQIP-Pediatric Analysis

Posted on January 25, 2017

C. L. Kvasnovsky1, J. Salazar1, J. Y. Chun1  1University Of Maryland Medical Center,Department Of Surgery,Baltimore, MD, USA

Introduction:  

Limited data exist to assess the increased risk conferred to children with congenital malformations (CM) undergoing abdominal surgery. Children with CM are presumed to have worse outcomes postoperatively. We sought to quantify the risk of postoperative complications in children with CM, after controlling for co-morbidities. 

Methods:  

The 2012-2014 National Surgical Quality Improvement Program-Pediatric (NSQIP-P) databases were queried to identify patients with and without CM, as defined by the American College of Surgeons, undergoing the 20 most common abdominal procedures in General Surgery. We assessed univariate associations between co-morbidities in patients with CM. Length of stay was compared using the Wilcoxon rank sum test.

 

Multivariate logistic regression to assess the odds of complication, controlling for co-morbid and operative conditions. We compared different wound classifications, using clean wounds as a reference. We assessed for collinearity between comorbidities by testing for variance inflation factors on linear regression. 

Results:

Over the study period, 46,368 children underwent abdominal surgery, including 7752 (16.7%) with a congenital malformation. Children with CM were more likely to have other pre-existing comorbidities, including cardiac risk factors (32.5% vs 6.6% of patients without CM, P<0.0001) and structural pulmonary disease (15.6% vs 2.3%, P<0.0001).

 

Patients with CM had a longer median hospital length of stay (median 3 days, interquartile range [IQR] 1-14, as compared with median 2 days, IQR 1-4, P<0.0001). The majority of patients, (59.0%) underwent urgent or emergent procedures, while patients with CM were more likely to undergo elective procedures (70.2%, P<0.0001).

 

On univariate analysis, the presence of a CM was associated with all complications. For instance, patients with CM had 1.9 the odds of readmission (CI 1.7-2.1, P<0.0001) and 2.7 the odds of wound dehiscence (CI 2.0-3.6, P<0.0001).

 

On multivariate logistic regression, the presence of a congenital malformation was a strong predictor of several complications, from urinary tract infection (P=0.01), post-operative sepsis (P=0.0002), need for reoperation during index stay (P<0.0001), and hospital readmission (P=0.0003, Table). There was no collinearity, allowing for complete analysis of complications. 

However, even after controlling for surgical contamination and other risk factors, there was no correlation between the presence of a congenital malformation and superficial wound infection (P=0.16), deep space infection (P=0.76), or post-operative bleeding episode (P=0.10). 

 

Conclusion:There was no increased postoperative risk conferred by CM for many important outcomes, such as bleeding and wound infection. Future work will focus on the risks associated with individual malformations, to further aid in preoperative risk assessment and family discussions. 

09.15 Non-Accidental Burns in Kids: What Are the Risk Factors?

Posted on January 25, 2017

R. P. Barker1, K. B. Savoie2, R. C. Passaro1, J. W. Eubanks2, R. F. Williams2  1Univeristy Of Tennessee Health Science Center,College Of Medicine,Memphis, TN, USA 2Univeristy Of Tennessee Health Science Center,Department Of General Surgery And Pediatrics,Memphis, TN, USA

Introduction:
Non-accidental burns account for up to 20% of all non-accidental trauma and have been associated with increased septic complications, longer lengths of stay, more operations, and higher morbidity. Identification of these burns can be difficult; therefore, we sought to identify risk factors associated with non-accidental burns.

Methods:
After institutional review board approval, a retrospective chart study of all patients from 2011-2013 with confirmed or suspected diagnoses of non-accidental burns were identified through the institutional trauma data bank. These patients were then matched 2:1 with burn patients who had no suspicion for abuse based on gender, race, TBSA, and date of burn. Individual charts were reviewed and data was abstracted for basic demographics, injury characteristics, and parameters related to non-accidental trauma. Standard statistical analysis was performed.

Results:
A total of 94 patients were identified; 33 of these were either suspected or confirmed non-accidental cases. The remaining 61 were burn cases with no suspicion for abuse. Non-accidental cases presented at younger ages than accidental cases (median age 1.89 vs 8.42, p <0.0001) and were more likely to present with clinical signs of shock (median SBP 101 vs 124, p = 0.0004, median HR 133 vs 103, p = 0.004, median RR 28 vs 22, p = 0.001). Non-accidental cases were more likely to require ICU admissions (32% vs 7%, p = 0.05). Hospital length of stay was longer for non-accidental cases compared to accidental cases (median days 3 vs. 0, p <0.0001).  A surgery consultation was required for 87.9% of non-accidental cases vs. 50.8% of accidental cases (p <0.0004). Non-accidental cases were more likely to involve the perineum (36.4% vs. 8.2%, p =0.0007) and the feet (42.4% vs. 14.8%, p =0.0033) when compared to accidental burns.  Accidental burns were more likely to involve the upper extremities (37.7% vs. 18.2%, p =0.05) compared to non-accidental burns.  While only 47.5% of accidental cases required admission, 93.9% of non-accidental cases were admitted (p <0.0001). Non-accidental cases were more likely to have underlying social concerns when compared to accidental cases. These included a single parent (45.5% vs. 21.7%, p =0.02) and a previous Department of Children Services (DCS) encounter (27.3% vs. 1.7%, p <0.0001). Trauma activation was initiated in 24% of non-accidental cases and none of the accidental matched cases (p <0.0001).

Conclusion:
Suspected and confirmed non-accidental burn cases are overall more severe than accidental burn cases and require more hospital resources. Younger patients with burns to the perineum or feet who live in a single parent home or have a previous DCS encounter are at increased risk for non-accidental burns and may benefit from early social work consultation or transfer to a burn center for further evaluation. 
 

09.14 Bariatric Surgery in Adolescents: Factors Contributing to Type of Surgery and Treatment Cost

Posted on January 25, 2017

O. Nunez Lopez1, D. Jupiter2, D. Adhikari2, R. S. Radhakrishnan1,3, K. A. Bowen-Jallow1  1University Of Texas Medical Branch,Surgery,Galveston, TX, USA 2University Of Texas Medical Branch,Preventive Medicine And Community Health,Galveston, TX, USA 3University Of Texas Medical Branch,Pediatrics,Galveston, TX, USA

Introduction:

Despite the increasing epidemic rates of adolescent obesity, the use of bariatric surgery in adolescents has plateaued since 2003. Sex and race disparities contribute to the underutilization of bariatric surgery. The use of different types of bariatric surgery has changed over time. In order to better understand the underutilization of adolescent bariatric surgery, we set out to identify potential factors that can impact treatment cost and influence the type of bariatric procedure used in adolescents.

Methods:

We used the Kid’s Inpatient Database, a nationwide population-based survey from 2006, 2009, and 2012. Adolescents (age 10-19 years) with a primary diagnosis of obesity who underwent bariatric surgery were identified. Univariate and bivariate analysis were computed. Multinomial logistic and linear regression were used to determine the association of the predictor variables with type of bariatric procedure, treatment cost and length of hospital stay (LOS), respectively. Income was represented by quartiles (Q1-lowest, Q4-highest), self-pay status included self-pay/uninsured patients, other payer status included federal and non-federal programs.

Results:

1,799 adolescents underwent bariatric surgery. The majority of the subjects were female 77% (n=1,379). Mean age was 18 ± 1 years. Whites represented 60% (n=1,076), Blacks 13% (n=234), Hispanics 20% (n=359) and other races 7% (n=130). The most commonly performed procedure was gastric bypass (GB) (56%, n=993), followed by sleeve gastrectomy (SG) (23%, n=429) and adjustable gastric banding (AGB) (21%, n=377). Several sociodemographic characteristics are associated with specific type of bariatric procedure (Table 1). Hispanics were less likely to undergo AGB (OR 0.5; 95% CI 0.3-0.8); self-pay patients were less likely to undergo GB (OR 0.5; 95% CI 0.3-0.7), and patients with other payer type were more likely to undergo GB (OR 6; 95% CI 2.4-14.9) and AGB (OR 10.1; 95% CI 3.6-28.7), all as compared to SG. LOS was not affected by the variables analyzed. Overall, treatment cost was decreased by low income (Q1, Q2), teaching status and large hospital size. Stratification by type of surgery showed that GB cost was decreased by low income (Q1, Q2); AGB cost was reduced by female sex and large hospital size; and SG cost was reduced by large hospital size and teaching status.

Conclusion:

Primary payer, hospital region and teaching status play a role in the type of procedure performed. Income, teaching status, and hospital size are determinants of treatment cost. Understanding factors associated with the use of suboptimal procedures can identify opportunities for change of practice. Identifying factors that decrease treatment cost can improve access to surgical care.

09.13 Cardiovascular Risk Factors In Long-term Pediatric Burn Survivors

Posted on January 25, 2017

G. Hundeshagen1,4, R. P. Clayton1,4, V. N. Collins3, D. N. Herndon1,4, L. K. Branski1,4, M. P. Kinsky2  1University Of Texas Medical Branch,Surgery,Galveston, TX, USA 2University Of Texas Medical Branch,Anesthesiology,Galveston, TX, USA 3University Of Texas Medical Branch,School Of Medicine,Galveston, TX, USA 4Shriners Hospitals For Children,Galveston,Galveston, TX, USA

Introduction: There is little data describing long-term sequelae of pediatric burn injury on cardiovascular disease (CVD). A recent retrospective cohort analysis of more than 10.000 pediatric burn patients over 30 years showed increased admission rates and prolonged hospitalization times for cardiovascular and circulatory disease in this patient population1. In an ongoing prospective trial we are assessing cardiovascular risk factors of long-term pediatric burn survivors. The following data collection focuses on specific and potentially modifiable risk factors of CVD.

Methods:  Former pediatric burn patients greater than 3 years post injury were prospectively enrolled and screened for established CVD risk factors: elevated blood pressure is defined as >140mmg systolic or >90mmg diastolic, smoking is defined as smoking status at the time of survey, overweight is defined as BMI>25 and obesity is defined as BMI>30, total body fat percentage is measured using whole body bone densitometry (DEXA). All results are presented as mean ± SD or count and percentage.

Results: We included 64 patients (39male, 29 female) in this study. Age was 21±5 years, age at burn was 9±6 years, elapsed time between burn injury and assessment was 12±5 years, burn size was 62±19% total body surface area (TBSA). Four patients (6.25%) had evidence of hypertension (average prevalence in this age group, 5.4%), mean systolic blood pressure was 118±5mmg, mean diastolic blood pressure was 72±10mmg. Twenty-six patients (41%) were overweight (average prevalence in this age group 35%), 7 patients (11%) were obese (average prevalence in this age group 20%). While mean BMI was 24±5, mean total body fat percentage was 31±8% which is above the threshold for obesity in adolescents and adults. Six patients (9%) were actively smoking at the time of survey (average prevalence in this age group 27%). 

Conclusion: Our results to date suggest a similar incidence of hypertension. Interestingly, smoking risk was lower in this cohort. On the other-hand, despite only a moderate increase in BMI, total body fat percentage was higher than age match controls. The considerable prevalence of overweight and obesity in this patient collective raises questions about the long-term persistence of the burn-induced hyperdynamic state. Our results have limitations, other CVD risk factors e.g., hyperlipidemia was not measured and indices of systolic and diastolic dysfunction, which could impact hospitalization1, are not reported to date.

Reference: 1 Duke, Janine M., et al. "Long-term Effects of Pediatric Burns on the Circulatory System." Pediatrics 136.5 (2015): e1323-e1330.

 

09.12 Exploring Regional Variability in Utilization of Antireflux Surgery in Children

Posted on January 25, 2017

H. L. Short1, W. Zhu1, C. McCracken2, C. Travers2, L. Waller1, M. V. Raval1  1Emory University School Of Medicine,Pediatric Surgery,Atlanta, GA, USA 2Emory University School Of Medicine,Pediatrics,Atlanta, GA, USA

Introduction:  There is significant variation surrounding the indications, surgical approaches, and outcomes for children undergoing antireflux procedures (ARPs) resulting in differences in geographic utilization in the United States. Our purpose was to examine regional and state-level trends in utilization of ARPs and to determine if overall surgical utilization was correlated with use of laparoscopy. 

Methods:  A cross-sectional analysis of the 2009 Kid’s Inpatient Database was performed to identify patients <21 years old with one of the following diagnoses of interest: esophagitis, esophageal stricture, dysphagia, aspiration, apnea, failure to thrive, Barrett’s esophagus, gastroesophageal reflux (GERD), esophageal ulcer, or hiatal hernia. We then determined which of these patients underwent an ARP and the surgical approach utilized (laparoscopic versus open). A mixed effects model was used to determine which regions and individual states were high utilizers of surgery and to identify patient and hospital factors associated with open versus laparoscopic procedures.

Results: Of the 148,959 patients with one or more of the diagnoses of interest, 4,848 (3.3%) patients underwent an ARP with 2376 (49%) undergoing a laparoscopic procedure.  GERD was the most common indication (79%) for ARPs.  Older children (ages 11-20 years) had lower ARP utilization compared to children <1 year old (Odds Ratio (OR) 0.37; 95% Confidence Interval (CI) 0.33-0.40). The Northeast and Midwest had the lowest overall utilization of surgery (2.5%), compared to the West (3.8%) and South (3.8%).  After adjustment for age, case-mix, and surgical approach and allowing for state specific utilization rates, regional variation persisted with the West and the South demonstrating close to 2 times the odds of undergoing an ARP compared to the Northeast (Table).  Surgical utilization rates appeared to be independent of state-level case volume with some of the highest case volume states (Florida, New York and Ohio) having surgical utilization rates below the national rate.  In the West, the use of laparoscopy appeared to correlate with overall utilization of surgery, while surgical approach was not correlated with ARP use in the South. 

Conclusion: Significant regional variation in ARP utilization exists that cannot be explained entirely by differences in patient age, patient race/ethnicity, case-mix, and surgical approach.  In order to decrease variation in care and potentially improve care, further research is warranted to delineate local factors driving surgical utilization of ARP in children.  Consensus guidelines regarding indications and appropriateness for use are needed.     

 

09.11 Evaluation of Postoperative Fever in Children

Posted on January 25, 2017

K. S. Corkum1, J. E. Grabowski1, C. J. Hunter1, T. B. Lautz1  1Ann & Robert H. Lurie Children’s Hospital Of Chicago,Division Of Pediatric Surgery/Department Of Surgery,Chicago, IL, USA

Introduction: Early postoperative fever is an extremely common occurrence in both adult and pediatric patients. Adult data suggest that fever workup is unnecessary for most patients in the first 24-48 hours after a major operation. Data in the pediatric population is extremely limited, and as a result, many children with early postoperative fever may undergo unnecessary “pan-culturing”. The aims of this study were to describe the incidence of early postoperative fever following elective inpatient operations in children, determine current utilization of laboratory and radiologic tests in these children with fever, and assess the frequency with which early postoperative fever workup yields a positive result.

Methods: A retrospective analysis of all surgical subspecialty patients at our institution undergoing an inpatient or observational stay surgery between 2011 and 2015 was performed using our electronic medical record (EMR). The EMR query identified 18,612 distinct patient operative encounters of which 6,943 met our inclusion criteria of elective admission type, surgery performed on hospital day zero or one, and an identified wound class of I-IV. Of those encounters, 2,128 had a documented postoperative fever (>100.5 F within 0-2 days post procedure). The EMR was also queried for all blood cultures, urinalysis, urine cultures, respiratory viral panels (RVP) and chest radiographs during that time period and the generated data was then cross-referenced against our cohort of patient operative encounters based on medical record number and date of service using IBM SPSS.

Results: 2,128 patients (30.6%) developed an early postoperative fever, including 761 on POD0, 1422 on POD1, and 1157 on POD2. Urinalysis was tested in 450 (21.1%) and was positive in 89 (19.8% of patients tested and 4.2% of all with fever). Urine culture was tested in 479 (22.5%) and was positive in 72 (15.0% of patients tested and 3.4% of all with fever). Of patients with a positive urine culture, 90.2% (65/72) had an indwelling urinary catheter at the time of surgery. Blood culture was performed in 453 (21.2%), but only 3 patients, all with a central venous catheter, had clinically significant positive cultures. Overall, chest radiographs were performed in 853 (40.1%), and 26 (3.0%) were read as concerning for pneumonia, 326 (15.3%) non-infectious, 420 (19.7%) normal, and 81 (3.8%) were obtained to confirm line, tube or hardware placement. RVP was performed in 52 (3.1%) and was positive in 20 (38.5% of patients tested and 1.2% of all with fever).

Conclusion: Similar to adult patients, early postoperative fever is extremely common in the pediatric surgical population, and rarely associated with a positive blood, urine, respiratory culture and/or chest radiograph suggestive of an infectious source. Workup for early postoperative fever in the pediatric surgical population should be applied selectively, based on patient history, severity of illness, and clinical assessment.

 

09.10 Factors Influencing Replacing an Infected Central Line Catheter in Children

Posted on January 25, 2017

M. M. Nourian1, A. L. Schwartz1, A. Stevens1, E. R. Scaife1, B. T. Bucher1  1University Of Utah,Division Of Pediatric Surgery,Salt Lake City, UT, USA

Introduction:

The optimal time to reinsert Tunneled central venous catheters (tCVC) after a documented catheter related blood stream infection (CLABSI) is not well defined. In infants and children in particular, this leads to additional hospital stay and increased health care utilization. The goal of this study is to identify risk factors for children who develop persistent bacteremia after tCVC removal and therefore would not be a candidate for immediate tCVC replacement.

Methods:

We performed a retrospective cohort analysis of children with a tCVC associated CLABSI from the electronic medical records at a tertiary care children’s hospital from 2000-2016. All children with a tCVC (Broviac or Port-a-Cath) and documented CLABSI were included in our analysis. Our primary outcome was persistent bacteremia after removal of the tCVC as defined by positive blood cultures after tCVC removal. Salient patient demographic and clinical factors were extracted from the medical record. Statistical significance was defined as p<0.05.

Results:
From 2000-2016 there were 4,735 patients who had a tCVC placed. Of those patients 78 (1.6%) had a documented CLABSI and tCVC removed. The majority of patients were white (68%) and male (53%) with an average age of 6.5 years. Most of tCVC placed were Broviac catheters (82%) compared to Port-a-Cath, and the median (IQR) lifespan of the lines placed was 70 (30-167) days. The majority of patients had a history of malignancy (53%) and approximately 36% were treated with chemotherapy 30 days prior to the documented CLABSI. In addition, 42% of patients were placed on home TPN with a history of GI failure rate of 30%. The most common causative organism for a CLABSI was S. epidermidis (28%) followed by Pseudomonas species (15%). Sixteen patients (20%) had persistent bacteremia after line removal. Compared to patients who cleared the bacteremia, those with persistent bacteremia were similar in age, race, and history of malignancy. Patients with persistent bacteremia were more likely to have a history of GI Failure (57% vs 23%, p=0.01) and MRSA bacteremia (13% vs 0%, p=0.04). There was no significant difference between the cleared and persistent bacteremia in gram positive or negative, polymicrobial, or fungal CLABSIs. The majority (78%) of children required temporary CVC placement on average 3.9 days after removal of infected tCVC. 

Conclusion:

We have identified several risk factors for persistent bacteremia after tCVC removal including history of GI failure and MRSA bacteremia. This study supports the practice of early replacement of tCVC after removal for bacteremia in low risk patients. Early replacement could save on average a minimum of 2-3 hospital days in infants and children with tCVC associated CLABSI.

09.09 Trends in Neonatal Surgical Outcomes in Children’s Versus Non-Children’s Hospitals

Posted on January 25, 2017

H. L. Short1, A. Savinkina1, R. M. Patel2, M. V. Raval1  1Emory University School Of Medicine,Pediatric Surgery,Atlanta, GA, USA 2Emory University School Of Medicine,Neonatology,Atlanta, GA, USA

Introduction:  Newborns undergoing surgery represent one of the most fragile patient populations and require specialized care.  Our purpose was to examine trends in neonatal surgical outcomes between children’s and non-children’s hospitals (CH and NCH).

Methods:  A cross-sectional, retrospective review of the 2000, 2003, 2006, 2009 and 2012 Kid’s Inpatient Database (KID) was performed to identify all neonatal surgical cases of necrotizing enterocolitis (NEC), patent ductus arteriosis (PDA), esophageal atresia/tracheoesophageal fistula (EA/TEF), congenital diaphragmatic hernia (CDH), and gastroschisis/omphalocele (GAS/OMP).  Mortality rates, length of stay (LOS) and hospital costs at CH and NCH were compared.

Results: We identified 48,149 patients who underwent a surgical procedure to correct one of the diagnoses of interest during the neonatal period.  During the 12-year study period the incidence of all diagnoses increased.  The majority of patients (73%) were treated at NCH, however the proportion of children treated at CH increased 12%, from 18.4% to 30.3%, during the study period.  Overall mortality decreased from 14.9% in 2000 to 12.6% in 2012. This improvement is largely due to an improvement in mortality at CH from 17.2% in 2000 to 10.9% in 2012, while mortality in NCH remained stable at about 14% (Figure). Mortality was consistently lower at CH than NCH for 4 of 5 diagnoses, excluding NEC for all study years and CDH in 2006. From 2000 to 2012, overall mean LOS increased from 44 to 57 days and this trend was similar in CH and NCH.  However, when individual diagnoses were examined LOS was longer in CH than NCH every year for all diagnoses except PDA. After adjustment for inflation, there was a two-fold increase in cost per day for all diagnoses from $5,015/day in 2000 to $10,508 in 2012.  Cost per day was higher for each diagnosis at CH compared to NCH in each year. In 2000 these neonatal conditions cost $1.2 billion (22% at CH) and in 2012 cost increased to $7.6 billion (35% at CH).

Conclusion: Mortality among neonates undergoing surgery is improving at CH and is stable at NCH.  However, LOS and costs are consistently higher at CH than NCH.  In order to optimize outcomes and contain costs for these fragile patients the observed trends warrant further investigation.

 

09.08 Geographic Variation in Prompt Access to Care for Children Involved in Motor Vehicle Crashes

Posted on January 25, 2017

L. L. Wolf1,2, R. Chowdhury1, J. Tweed3, L. Vinson3, E. Losina4, A. H. Haider1,2, F. G. Qureshi3,5  1Center For Surgery And Public Health, Brigham And Women’s Hospital,Department Of Surgery,Boston, MA, USA 2Brigham And Women’s Hospital,Division Of Trauma, Burns, And Surgical Critical Care,Boston, MA, USA 3Children’s Medical Center Of Dallas, Part Of Children’s Health,Dallas, TX, USA 4Orthopaedic And Arthritis Center For Outcomes Research, Brigham And Women’s Hospital,Department of Orthopedic Surgery,Boston, MA, USA 5University Of Texas Southwestern Medical Center,Department Of Surgery,Dallas, TX, USA

Introduction: Unintentional injury is the leading cause of death in children 1-19 years and motor vehicle crashes (MVCs) are the most common cause of unintentional injury in this group. Previous research has demonstrated substantial variability in pediatric trauma resources at the state level. However, it is unclear how these differences in resources may affect actual access to care for pediatric trauma patients. We sought to examine the impact of state in which the crash occurred on prompt access to care for children involved in MVCs.

Methods: Using the 2010-2014 Fatality Analysis Reporting System, we assembled a cohort of child passengers (<15y) involved in a fatal MVC, defined as a crash occurring on a U.S. public road and resulting in ≥1 death (adult or pediatric) within 30 days. We included children requiring transport from the crash scene to a hospital for medical care for whom data on time of hospital arrival were available. Our primary outcome was time to first hospital, defined as a binary variable (>1h or ≤1h). We used multivariable logistic regression to establish the state-level variability of the percentage of children whose time to hospital was >1h, after adjusting for injury severity (no injury, possible injury, suspected minor injury, suspected severe injury, fatal injury, injury of unknown severity), mode of transport (Emergency Medical Services (EMS) air, EMS ground, other), and rural roads. We described variability by state, dividing states into quartiles according to the percentage of children delivered to the first hospital >1h after the fatal MVC.

Results: We identified 12,152 child passengers involved in a fatal MVC from 2010-2014; 4,672 (38.4%) required transport for medical care from the scene of the MVC. Of those not transported, 1,424 (19.0%) died at the scene. Of those transported, median time to first hospital was 1h (IQR: [1,1]; range: [0,87]). The percent of children that experienced transport times >1h varied significantly by state, from 0.0 in Rhode Island to 100.0 in Florida, Idaho, Indiana, and Michigan (p<0.001). While we observed striking state-level variation in transport times, there were no clear regional patterns (Figure).

Conclusions: Time to hospital varied greatly by U.S. state for children requiring transport for medical treatment from the scene of a fatal MVC. State-level resources for EMS services and the availability of pediatric trauma care may contribute to the availability of prompt trauma care for children involved in fatal MVCs. These results provide critical data to inform state-level trauma care planning.

09.07 Pediatric Falls in the Greater Los Angeles Area: Targeting Populations for Injury Prevention

Posted on January 25, 2017

J. A. Zagory1, M. Mallicote1, H. Arbogast1, J. Upperman1, M. Fenlon1, A. Jensen1  1Children’s Hospital Los Angeles,Surgery,Los Angeles, CA, USA

Introduction:  Children suffering fall related injury sustain significant morbidity and mortality and require significant resource utilization. Currently, there are no injury prevention programs in the greater Los Angeles area targeted to prevent pediatric falls. We aim to identify injury patterns, resource utilization, and complication rates in children sustaining falls. We hypothesize that falls from significant height occur in children as they become mobile, and an at-risk age can be identified for injury prevention targeting.

Methods:  We conducted a 10-year retrospective review (2004-2014) of a prospectively collected countywide trauma database for children (age <18y) who sustained injury related to a fall mechanism.  Mechanism codes were utilized to identify <15 ft/low-risk (LR) and >15ft/high-risk (HR) falls for comparison.  Statistical analysis was conducted with independent t-test or χ2 as appropriate.

Results: 4451 children sustained LR and 229 HR falls. HR falls were more likely to be younger (4.3±3.7 v 5.6±4.4) , non-white (11.3% v 12.2%), have greater Injury Severity Score (ISS)(7±7.5 v 6±4.2), and sustain injury to their head (skull/facial fracture, intracranial hemorrhage, closed head injury)(70% v 43%)(p<0.001). Of all falls in children under 1y, less than 1.5% were HR, while in children 1-4y 8% of falls were HR. Children 1-4y represent 62% of all HR falls. Resource utilization (computed tomography, length of stay, Intensive Care Unit admission, intubation) and complications (decubitus ulcer, need for respiratory support, pneumonia) were higher in HR (p<0.0001) (Table). Two deaths were in LR and due to non-accidental trauma.

Conclusion: Non-white children 1-4y are at especially high risk for falls from a significant height, with higher ISS, resource consumption, and complications. Injury prevention education should be part of well-child visits as children begin to ambulate, and structured targeted community-based programs should be developed and paired with health policy efforts.

 

09.05 Socioeconomic Status Affects Time to Treatment of Pediatric Well-Differentiated Thyroid Cancer

Posted on January 25, 2017

E. F. Garner1, I. I. Maizlin1, K. W. Gow2, M. Goldfarb3, M. Langer4, M. V. Raval5, J. G. Nuchtern6, S. A. Vasudevan6, J. J. Doski7, A. B. Goldin2, E. A. Beierle1  3John Wayne Cancer Institute/Providence St. John’s Medical Center,Department Of Sugery,Santa Monica, CA, USA 4Maine Medical Center,Division Of Pediatric Surgery,Portland, ME, USA 5Emory University School Of Medicine,Division Of Pediatric Surgery/Department Of Surgery,Atlanta, GA, USA 6Baylor College Of Medicine,Division Of Pediatric Surgery/Department Of Surgery,Houston, TX, USA 7University Of Texas Health Science Center At San Antonio, San Rosa Children’s Hospital,Division Of Pediatric Surgery/Department Of Surgery,San Antonio, TX, USA 1University Of Alabama,Division Of Pediatric Surgery/Department Of Surgery,Birmingham, Alabama, USA 2University Of Washington,Division Of Pediatric Surgery/Department Of Surgery,Seattle, WA, USA

Introduction:  Well-differentiated thyroid cancer (WDTC) is the most common endocrine malignancy in children. Adult literature has demonstrated socioeconomic disparities in adults undergoing thyroidectomy. However, few studies have looked at the effects of socioeconomic status on the management of pediatric thyroid cancer. We sought to determine if children with lower socioeconomic status experience delays in diagnosis and management of their thyroid cancer. 

Methods:  Patients <21 years of age with well-differentiated thyroid cancer (WDTD) were reviewed from the National Cancer Data Base (NCDB) from 1998-2013. Three socioeconomic surrogate variables were identified: insurance type, median income in the patient’s ZIP code, and percent of people with no high school degree in the patient’s ZIP code. Chi-square and pool-variance t-tests were then used to compare tumor characteristics, intervals from diagnosis to staging and diagnosis to treatment, as well as clinical outcome variables within each of the socioeconomic surrogate variables, while controlling for the effect of age, race and gender.

Results: A total of 9585 children with WDTC were reviewed; 8696 (90.72%) with papillary thyroid cancer and 889 (9.28%) with follicular thyroid cancer. 7914 (82.6%) were female. In multivariate analysis, lower income (p = 0.041, Hazard Ratio [HR] = 1.98, 1.88 and 1.68 in each successive quartile compared to highest one), lower educational quartile (p<0.001, HR=1.86, 1.50 and 1.12, compared to highest quartile) and insurance status (p < 0.001, HR = 2.26 for uninsured and HR = 1.46 for government insurance, as compared to private insurance) were associated with higher stage at diagnosis. Furthermore, lower income quartile was associated with a longer time from diagnosis to treatment (p <0.002). Similarly, uninsured children had a longer time from diagnosis to treatment (28 days) compared to those with government (19 days) or private (18 days) insurance (p < 0.001). However, despite diagnosis at a higher stage and requiring longer time between diagnosis and treatment, there was no significant difference in either overall survival or rate of complications (demonstrated by unplanned readmission rates) based on any of the socioeconomic surrogate variables.  

Conclusion: Children from lower income families and those lacking insurance experienced a longer period from diagnosis to treatment of their WDTC. These patients also presented with higher stage disease. These data suggest that a disparity exists in access to care for children from low-income families.  Although these findings did not translate into worse outcomes for WDTC, future efforts should focus on reducing these differences. 

 

« Previous Page
Next Page »