91.14 Pancreatic Adenocarcinoma in Southwest Native Americans: Disparities in Treatment and Survival

A. Greenbaum1, E. Alkhalili1, R. Rodriguez1, J. O’Neill1, O. Estrada Munoz1, F. Qeadan2, O. Myers3, I. Nir1, K. Morris1  1University Of New Mexico HSC,Surgery,Albuquerque, NM, USA 3University Of New Mexico HSC,Internal Medicine,Albuquerque, NM, USA 2University Of New Mexico HSC,Pathology,Albuquerque, NM, USA

Introduction:   Native Americans (NA) have a higher incidence of and mortality from biliary tract cancers, though demonstrate lower pancreatic cancer incidence than non-Hispanic Whites (NHW).  In this study, we examined the treatment and outcomes of pancreatic adenocarcinoma in Southwest NA.  We hypothesized there would be no differences when comparing NA to NHW and Hispanics (H), which compromise the three main ethnic groups in our state.

Methods:   A retrospective chart review was performed of all patients diagnosed with pancreatic adenocarcinoma and treated at a university National Cancer Institute (NCI) Comprehensive Cancer Center between January 2002 and July 2016. Data extracted included patient demographics, AJCC 7th edition staging at presentation, tumor resectability, treatment modalities offered and received, clinical outcomes and survival data.   We employed multivariable logistic regression to determine the odds ratios (OR) and 95% confidence intervals (CI).  Student’s t-Test and ANOVA tests were used to compare means of continuous values.  Chi-square tests were used to assess associations among nominal variables. Overall survival (OS) was examined using Kaplan-Meier analyses. P-values less 0.05 were considered significant.

Results: A total of 457 patients met inclusion criteria.  Our final cohort included 240 (52.5%) NHW, 186 (40.7%) H and 31 (6.8%) NA patients.  After adjusting for age and sex there were no significant differences between ethnic groups in overall stage at presentation, presence of unresectable disease or distant metastases.  All groups were offered surgery and received radiation therapy at similar rates.  NHW (OR 2.41, 1.11 – 5.25. p 0.026) and H (OR 2.37, 1.08-5.24, p=0.032) were more likely to receive chemotherapy than NA at any stage of their treatment and for unresectable disease (OR 2.80, 1.13-6.88; p=0.025 and OR 2.48, 1.00-6.18; p=0.05).  Kaplan-Meier models revealed no significant difference in OS between the three ethnic groups (median OS 12, 13 and 6 months in NHW, H and NA respectively; p=0.224).  However, a significantly larger percentage of NA died within 1 month of diagnosis (25%) compared to 7.5% NHW (OR of being alive after 1 month 4.1, CI 1.56-10.90; p=0.004) and 9.1% H (OR 3.3, 1.24-8.88; p=0.017).     There were no major differences in the number of comorbid conditions or Charlson Comorbidity Index scores between ethnic groups (mean scores 4.46 NHW, 4.58 H and 4.84 NA; p=0.63).

Conclusion:

Southwest NA diagnosed with pancreatic adenocarcinoma are less likely to receive chemotherapy and are significantly more likely to die within 30 days of diagnosis than NHW and H.  The latter may be due to more biologically aggressive disease, though no differences in medical comorbidities, stage at presentation or overall survival were noted.  Larger studies are needed to examine whether cultural factors and access to care due to financial or geographic constraints contribute to these findings.

 

91.13 Undiagnosed Malignancy in Patients Receiving a Surgical Evaluation at an Urban Tertiary Care Center

M. R. Egyud1, M. Plocienniczak2, C. James2, T. Sachs1,2, T. Dechert1,2  1Boston Medical Center,Department Of Surgery,Boston, MA, USA 2Boston University School Of Medicine,Boston, MA, USA

Introduction: The nature of Trauma and Acute Care Surgery (TACS) demands surgeons be able to diagnose and treat a spectrum of disease in complex patients with multi-organ involvement. The proportion of these patients with undiagnosed malignancy is poorly understood. We sought to evaluate the role TACS plays in the early surgical management of patients presenting with undiagnosed malignancy.

Methods: We reviewed records of all patients at an urban tertiary care center, evaluated by TACS for potential operative intervention between 01/2005 and 09/2015. Patients were selected if a malignancy was diagnosed during admission (Cohort A) or within a year of discharge based on findings from the index admission (Cohort B). Cohorts were compared by demographics, type of insurance, comorbidities, operation(s) performed, and hospital course.

Results: We identified 247 patients, with 54% in Cohort A (n=134) and 46% in Cohort B (n=113).  The majority of patients (> 80%) in both cohorts used Medicare, Medicaid, or lacked insurance. There were 21 distinct malignancies identified in Cohort A, with the majority (n=97, 72%) being of gastrointestinal origin. 26 distinct malignancies were identified in Cohort B. Cohort A patients tended to present with more advanced cancers, while Cohort B were earlier stage (Figure 1). Of the Cohort A patients, 111 (83%) required an operation related to their malignancy, 61 (55%) of whom needed an urgent operation.

Conclusion: Patients presenting to TACS may have an undiagnosed malignancy, many of whom are indigent or poor.  A portion of these patients will require an urgent operation. In addition, TACS surgeons often initiate workup for suspicious findings and coordinate care to ensure these patients receive appropriate evaluation in a timely fashion. The underserved are known to present at later stage and have less access to screening and prevention, and further study is needed to improve outcomes in this patient population.

 

91.12 Trends and Disparities in Thyroid Cancer within a Health System

O. Moaven1, R. Xie1, J. Richman1, D. Naftel1, J. K. Kirklin1, H. Chen1  1University Of Alabama at Birmingham,Department Of Surgery,Birmingham, Alabama, USA

Introduction: To improve the quality of care provided to the patients with thyroid cancer, we sought to examine and identify the trends and potential disparities in thyroid cancer diagnosis, various subtypes and patient demographics.

Methods: We assessed electronic medical records of the patients with the diagnostic codes for thyroid cancer in 2011-2016 at our institution. Patient demographics, specific type of cancer, insurance status, time to documented specific diagnosis and follow up visits were studied.  Racial and gender disparities in the documented specific cancer diagnosis, time to specific diagnosis, and time to the first follow-up visit after the initial unspecific diagnosis were examined using univariate and multivariate cox regression modeling. 

Results: There were 1640 patients with a documented diagnosis of thyroid cancer, and the cumulative incidence of thyroid cancer per 100,000 patients per year was highest in white females (60) followed by white males (42), black females (30), and black males (6, p<0.0001). After controlling for age, insurance type, and initial unspecific diagnosis year, white males had a higher likelihood of having a follow up clinic visit within 1 month after initial presentation (62% vs. 56%; HR: 1.19, p=0.01), as well as a documented designation with a specific diagnosis within 6 months (19% vs. 14%; HR: 1.45, p=0.01), when compared to those of white females. These differences were not significant when comparing whites vs. blacks.

Conclusion: There is a gender disparity, but no racial disparity, in a comprehensive workup and diagnosis of thyroid cancer within our health system. Identifying these inequalities is the first essential step in developing a health system that provides equal access to care.

91.11 Surgical Deliveries in Sub-Saharan Africa: Two Methodologies to Estimate Need

K. Garber1,2, R. Groen4,5, O. Ogedengbe3, D. DeUgarte1, A. Kushner2,5  5Surgeons OverSeas (SOS),NEW YORK, NY, USA 1University Of California, Los Angeles,Department Of Surgery,Los Angeles, CA, USA 2Johns Hopkins Bloomberg School Of Public Health,Baltimore, MD, USA 3University Of Lagos,Department Of Obstetrics And Gynecology,Lagos, , Nigeria 4Johns Hopkins Hospital,Department Of Obstetrics And Gynecology,Baltimore, MD, USA

Introduction:  Access to surgical delivery (c-section) is critical for reducing maternal and neonatal mortality rates, which are key aims of the UN Sustainable Development Goals 2030. The Lancet Commission on Global Surgery also recognized c-sections as an important benchmark procedure. Despite the importance, data on actual numbers of procedures performed in sub-Saharan Africa (SSA) is limited. The goal of this study was to estimate the met and unmet need for c-sections in SSA to help plan interventions and inform future policy decisions.

Methods:  Two approaches were used to estimate c-section need. First, annual births for each SSA country in 2015 were calculated using population data from the UN Population Database and crude birth rates from the World Bank Development Indicators. Then, using the World Health Organization optimal c-section rate of 10% and a recently published (Molina et al 2015) higher optimal rate of 19%, the range of c-section need for each country and region was estimated. To determine unmet need, the most recently published actual c-section rates for these countries were obtained (Bertran et al 2016) and raw totals for performed cesarean sections were calculated. Then these numbers were subtracted from the overall need. To support the findings, a second methodology to recalculate SSA c-section needs was used. Data on the prevalence of major maternal conditions (hemorrhage, obstructed labor, sepsis, hypertension, abortion) were obtained from the Institute for Health Metrics and Evaluation 2013 Global Burden of Disease study. Using previously calculated surgical incidence rates for various diseases (Hider et al 2015), the number of needed operations for each maternal condition was estimated, excluding abortions, on the assumption that the overwhelming majority of surgeries for these conditions would be c-sections. Results were summed for each country to provide overall estimates of need.

Results: Using optimal c-section rates of 10-19%, we estimated that 3.6-6.8 million c-sections are needed annually in SSA, with at least 1.8 million, and as many as 4.8 million, of those operations currently unmet. Using 2013 GBD data, a similarly estimated 5.2 million c-sections are needed, well within the range of the first estimate. Regionally, the largest met and unmet needs were seen in Eastern and Western Africa.

Conclusion: A large need and correspondingly large unmet need exist for c-sections in SSA. These data can help inform policymakers of the magnitude of obstetric surgical need in the region as they seek policies to improve maternal and neonatal health. As efforts to improve surgical care in low and middle-income countries increase, planning for and devoting resources to undertake c-sections must be a priority.

91.10 Disparities in Incidence of CRC between Hispanics and Whites: a 10-year SEER database study

J. E. Koblinski1, J. Jandova1, V. Nfonsam1  1University Of Arizona,Department Of Surgery,Tucson, AZ, USA

Introduction:
Overall incidence of colorectal cancer (CRC) has shown a decreasing trend in the last three decades. However, There has been an increasing incidence of CRC in the younger population (<50). It has previously been shown that racial disparities exist in the incidence of CRC. In addition, CRC was the second most commonly diagnosed cancer in Hispanics in 2012. The aim of our study was to assess the trends in incidence of early- (EO) and late-onset (LO) CRC in Hispanics and compare them to White patients.

Methods:
Between 2000 and 2010, we abstracted the national estimates for Hispanic and White patients diagnosed with colon and rectal cancer using the Surveillance, Epidemiology, and End Result (SEER) database. We distinguished between EO and LO cases and then analyzed incidence trends, mortality, gender and stage of disease. Linear regression was performed to compare the trends.

Results:
The overall incidence of CRC increased by 48% in Hispanics while the overall incidence decreased by 12% in Whites (P<0.0001). There was an alarming 80% increase in incidence of EO CRC in Hispanics and a 22% increase in Whites. As expected, there was an observed 19% decrease in incidence of LO CRC in Whites. Surprisingly, 38% increase in incidence of LO CRC was found in Hispanics (P<0.0001). Both Hispanics and Whites showed a higher percentage of distant CRC tumors for both age groups. Neither in Hispanic nor White patients was there any deviation in overall trend between males and females. 

Conclusion:
Although there is an overall decrease in incidence of CRC in Whites, there is an alarming increase in overall incidence of CRC in Hispanics. While incidence of EO CRC is increasing in both races, incidence of LO CRC is increasing in Hispanics but not in Whites. These data suggest that particular policies should be implemented to address these disparities.  

91.09 Hospital Assessment: Examining the Surgical System in Amazonas, Brazil.

J. E. Dos Santos Souza1, S. Saluja2,4, J. Amundson2,3, R. V. Ferreira1, I. Citron2, P. H. Gomes1, J. Correia1, C. Costa1, N. Alonso5,6, M. Shrime2,7  1Universidade Estadual Do Amazonas,Faculdade De Medicina,Manaus, AMAZONAS, Brazil 2Harvard School Of Medicine,Program In Global Surgery And Social Change,Boston, MA, USA 3University Of Miami,Miller School Of Medicine,Miami, FL, USA 4Weill Cornell Medical College,Department Of Surgery,New York, NY, USA 5Universidade De Sáo Paulo,Craniofacial Surgery Unit, Division Of Plastic Surgery, Department Of Surgery,Sáo Paulo, SÁO PAULO, Brazil 7Massachusetts Eye And Ear Infirmary,Department Of Otology And Laryngology And Office Of Global Surgery,Boston, MA, USA

Introduction:  Five billion people lack access to safe and affordable surgical, anesthetic and obstetric care when needed. In 2015, the Lancet Commission on Global Surgery – an academic global consortium – summarized the state of surgical care internationally. The Commission proposed six indicators for evaluating surgical systems. To assess the health of a national surgical system, a mixed-methods qualitative and quantitative Hospital Assessment Tool (HAT) has been developed. The tool will be used in Brazil’s largest and most rural state, Amazonas, to identify priority areas for system improvement and health policy changes, as perceived by local patients and providers. The deployment of the tool involves a partnership between Harvard Medical School and local collaborators at Universidade do Estado de Amazonas (UEA). The aim is to apply this validated tool to a broad range of settings worldwide.

Methods:  An initial pilot of the HAT was undertaken in Cabo Verde, Ethiopia, and India. The tool was then adjusted and validated by 18 experts (Delphi consensus). Over six months, the HAT will be deployed by researchers from UEA at hospitals in 20 municipalities across the state. To select which municipalities to assess, municipalities performing surgery were stratified by population quartile and selected at random within each stratum. At each site, the UEA team will gather quantitative survey data and qualitative interviews. Interview transcriptions will subsequently be evaluated using framework analysis. A selection of sites will undergo repeat data collection at 6-week intervals by a separate team to assess inter-rater and inter-temporal validity.

Results: To date the investigators have visited 6 of 20 target hospitals, with data collection projected to finish by late 2016. Inhalation general anesthesia is available at 1/6 hospitals; IV sedation, spinal and regional anesthesia is available at 3/6 hospitals. Blood bank services are available at 5/6 hospitals, with average time to access less than 30 minutes at 4/5 hospitals. No hospital reported use of the WHO safe surgery checklist. 2/6 hospitals performed procedures other than cesarean section in past 6 months, and 1/6 in the past 30 days. Only 1/6 hospitals reported continuous vital sign monitoring in the PACU. 2/6 hospitals have internet.

Conclusion: This project provides the framework for a successful partnership engaging local stakeholders in meaningful research to influence their own regional surgical agenda. Preliminary quantitative results show a significant lack of basic tools to perform safe surgery across the municipalities of Amazonas.

 

91.08 Characteristics of Patients Presenting to the Emergency Department for Diagnosis of Colon Cancer

D. Weithorn1, G. Umadat1, P. Friedmann1, R. Narang1, R. Huang1, R. Levine1, H. In1  1Albert Einstein College Of Medicine,Surgery,Bronx, NY, USA

Introduction: Patients with colorectal cancer who initially present through the Emergency Department (ED) for colorectal cancer diagnosis have worse outcomes, including poorer stage-adjusted prognosis.  Colonoscopy has been associated with improved survival but has not been studied in the context of persons presenting through the ED. We aimed to examine the characteristics of patients who get diagnosed with colorectal cancer through a visit to the ED, including prior colonoscopy and symptoms.

 

Methods: Patients diagnosed with colorectal cancer in one year (2013) in a single urban academic institution were analyzed.  A detailed retrospective chart review was conducted to identify if the first presentation that lead to the cancer diagnosis was through the ED (ED-Dx), and for colonoscopy history prior to cancer diagnosis. Kaplan-Meier Analysis was used to examine survival. Differences between persons presenting to the ED and not were compared using univariate and multivariate analyses.

 

Results: We identified 226 patients with newly diagnosed colorectal cancer eligible for analysis. 40% of patients had cancer diagnosed through a visit to the ED. Colonoscopy information was available for 72% of patients. About half of these patients had history of colonoscopy prior to their cancer diagnosis. ED-Dx patients were more likely to be either younger than 50 (13% vs 9%) or older than 80 (34% vs 19%) and less likely to be 50-65 years old (16% vs 36%, p=0.005). They presented at an advanced stage (40% vs 15%, P<0.001) and were less likely to have had a prior colonoscopy (20% vs 48%, p<0.001). ED-Dx more commonly presented with symptoms (89% vs. 56%, P<0.001), and “pain” was the most common symptom (47% vs 19%, P<0.001).  ED-Dx had significantly poorer 18-month survival (94% vs 81%, p=0.006). On multivariate analysis adjusting for all variables, we found ED-Dx to be less likely to have had a prior colonoscopy (OR 0.24, CI 0.1 to 0.6), more likely to have had symptoms (OR 4.33, CI 1.69 to 11.1) and have stage IV cancer (OR 8.13, CI 2.44 to 27.1). Patients with Medicare were more likely to be ED-Dx compared to those with private insurance (OR 4.68, CI 1.01 to 21.7).

 

Conclusion: Outcomes of patients with ED-Dx are poor. Identifying the health patterns and clinical attributes of these persons represents an opportunity to develop programs to improve the outcomes of cancer patients. Decreased utilization of colonoscopy during routine health care for patients diagnosed through the ED suggest that these patients may be underutilizing health care services, including cancer screening. Based on our observations, a key modifiable factor may be the increased utilization of colorectal screening.

91.07 Insurance status and choice of surgical therapy in newly diagnosed breast cancer patients

E. C. Feliberti1, R. R. Perry1, R. C. Britt1, J. C. Collins1, E. Feliberti1  1Eastern Virginia Medical School,Surgery,Norfolk, VA, USA

Introduction: Safety net programs aim to minimize disparities in the treatment of breast cancer patients. We hypothesize that differences in the use of breast conservation therapy (BCT) persist in uninsured women despite access to a multidisciplinary clinic.

Methods: A retrospective review of a prospective database was performed on consecutive newly diagnosed female breast cancer patients treated at an academic surgical department form 2001 to 2015. Patients were stratified by insurance status at time of breast cancer diagnosis.

Results: A total of 523 patients were identified meeting the inclusion criteria, 85 without and 438 with medical insurance. The uninsured cohort were younger (mean age: 46.8 vs 58.1, p<0.01) and had a higher proportion of African-American women (68.7% vs 40.4%, p<0.01). Tumor size was similar between the 2 groups (Mean size 2.2 cm vs. 2.0, p=0.8). BCT was selected less often in the uninsured cohort (50.6% vs 64.6%, p=0.02). Differences in the use of BCT in the uninsured were significant in women aged 50 and older (42.*5 vs 68.4%, p<0.01) and for tumors larger than 2 cm (30% vs 54.4%, p<0.01).

Conclusion: Insurance status affects choice of surgical therapy in newly diagnosed breast cancer patients despite access to a safety net program. Increasing tumor size and age play a significant role in the decreased use of BCT.

 

91.06 Anatomic Location of High-Grade Dysplasia from Adenomatous Polyp of the Colon among Black Patients

P. H. Lam1, I. D. Nwokeabia2, A. C. Obirieze3, S. C. Onyewu3, B. S. Li3, N. Enwerem3, G. Ortega3, T. M. Fullum3, W. A. Frederick3, L. L. Wilson3  1Cedars-Sinai Medical Center,Los Angeles, CA, USA 2Washington University,St. Louis, MO, USA 3Howard University College Of Medicine,Washington, DC, USA

Introduction:  Black patients have the highest incidence and mortality rates of colon cancer when compared to other racial/ethnic groups. Screening rates for colon cancer are lower in black patients, and studies have shown varying anatomic locations of adenomatous polyps and colon cancers in these patients. Studying the location of these cancers within the colon could help tailor where to screen and which screening test to use. We aim to investigate the anatomic location of high-grade dysplasia from adenomatous polyp among black patients, using a national tumor registry.

Methods:  The Surveillance Epidemiology and End Results database from 1973 to 2008 was utilized. We identified patients with a single primary diagnosis of a high-grade dysplasia arising from adenomatous polyp of the colon using appropriate ICD-O-3 codes. Age and gender-adjusted proportions of proximal vs. distal lesion location were derived for all patients using multivariable regression analysis.

Results: A total of 18,762 patient records, comprising 16,276 (86.8%) white and 2,486 (13.2%) black patients, met the study criteria. The incidence of high-grade dysplasia of the colon has been increasing in black patients over the last three decades (9.9%, 13.0%, and 15.1% for 1973-1989, 1990-1999, and 2000-2008, respectively).

The most common location in the proximal and distal colon was the cecum (16.6%) and sigmoid colon (48.5%). Most patients had distal lesions (60.0%), most of which occurred at the sigmoid colon (48.5%). On multivariate analysis, black patients were 33% less likely to have a distal lesion compared to white patients (OR: 0.67, p<0.001, 95% CI: 0.61-0.73). 

Black patients were more likely not to undergo surgery (4.9% vs. 3.3%), more likely to undergo partial or hemicolectomy (38.3% vs. 33.5%), or total colectomy (1.9% vs.1.3%), but less likely to undergo local excision with pathology (47.1% vs. 51.9%) (all p<0.001). 

Conclusion: Black patients have increasing incidence of high-grade dysplasia arising from adenomatous polyp, lower likelihood of distal lesions of the colon, and more likely not to undergo surgery. This suggests their importance for screening with colonoscopy. Nonetheless, more research on the use of different screening tests, location of cancer found, and surgical preferences among different race/ethnic groups are needed.

91.05 Burden of Sports Injuries among African Adolescents: A Modeling Study

D. G. LeBrun1,2, J. D. Kelly10, S. Wren8,9, D. A. Spiegel3, N. Mkandawire7, R. A. Gosselin11, A. L. Kushner4,5,6  10University Of Pennsylvania,Sports Medicine, Orthopaedic Surgery,Philadelphia, PA, USA 11University Of California – San Francisco,Orthopaedic Surgery,San Francisco, CA, USA 1Harvard School Of Public Health,Epidemiology,Boston, MA, USA 2University Of Pennsylvania,Perelman School Of Medicine,Philadelphia, PA, USA 3Children’s Hospital Of Philadelphia,Orthopaedic Surgery,Philadelphia, PA, USA 4Columbia University College Of Physicians And Surgeons,Surgery,New York, NY, USA 5Johns Hopkins Bloomberg School Of Public Health,International Health,Baltimore, MD, USA 6Surgeons OverSeas,New York, NY, USA 7University Of Malawi,Orthopaedic Surgery,Blantyre, , Malawi 8VA Palo Alto Healthcare Systems,General Surgery,Palo Alto, CA, USA 9Stanford University,Surgery,Palo Alto, CA, USA

Introduction:
Injuries comprise a major portion of the global burden of disease among adolescents. In particular, injuries sustained while playing sports are extremely common with many requiring surgical management. However, the extent to which sports injuries contribute to the burden of injury in low- and middle-income countries (LMICs) is unknown. The goal of this study was to determine the burden of sports injuries among adolescents in Africa as part of a larger effort to estimate the global burden of sports injuries among adolescents.

Methods:
Data from the World Health Organization (WHO) Global School-Based Student Health (GSHS) cross-national surveys was used to estimate the number of adolescents sustaining sports injuries in Africa. WHO-GSHS surveys have been conducted in 16 African countries since 2003. These surveys measured the number of adolescents sustaining serious injuries (defined as injuries that necessitated treatment or caused the child to miss a day of normal activities) within the past year. Nine surveys contained supplemental data on serious injuries attributable to sports. Gender-stratified sports injury rates were calculated based on weighted averages reflecting the relative sample size of each national survey. These rates were subsequently applied to every African country’s adolescent population to estimate country-specific injury rates and continent-wide totals.

Results:
The 9 countries with supplemental data on serious injuries attributable to sports injuries included: Botswana, Ghana, Kenya, Mauritius, Swaziland, Tanzania, Uganda, Zambia, and Zimbabwe. These surveys represented a total of 15,822 males and 18,495 females. In these countries, sports injury rates ranged from 19.7% to 30.4% for males and 6.9% to 20.2% for females. Weighted average sports injury rates for males and females were 23.8% and 13.5%, respectively. When these rates were applied to the adolescent populations of every African country, an estimated 31,303,816 males and 17,295,093 females sustained sports injuries. In total, an estimated 48,598,909 adolescents sustained sports injuries in Africa over a one-year period. 

Conclusion:
By calculating gender-stratified sports injury rates, we estimated that nearly 50 million African adolescents sustained sports injuries over one year. Injuries represent a major burden of disease worldwide and disproportionately affect LMICs, yet there is a paucity of data characterizing injury rates among adolescents in LMICs. The fraction of injuries attributable to sports and the proportion of these injuries necessitating surgical care are poorly understood. Further work will help to more precisely define the burden of sports injuries in LMICs and the role that surgery can play in mitigating this burden.

91.04 Defining Congenital Anomalies in Mongolia

L. F. Goodman1,4, G. Jensen1, T. Nomindelger3, R. Nurjanar2, T. Gantuya3, D. Farmer1  1University Of California – Davis,Surgery,Sacramento, CA, USA 2National Center For Maternal And Child Health,Pediatric General Surgery,Ulaanbaatar, ., Mongolia 3National Center For Maternal And Child Health,Surveillance Department,Ulaanbaatar, ., Mongolia 4Harvard School Of Public Health,Epidemiology,Boston, MA, USA

Introduction:  Neonatal mortality in Mongolia declined from 32 per 1000 live births in 1990 to 7.8 per 1000 in 2015. As  deaths from infectious disease and birth trauma have been reduced, congenital anomalies have become a relatively more important cause of neonatal mortality. This study sought to determine the prevalence at birth of major congenital anomalies, risk factors associated with anomalies and anomaly-associated neonatal mortality, and the proportion of anomalies that are surgically treatable. 

Methods:  The National Center for Maternal and Child Health (NCMCH) has maintained an electronic national database of congenital anomalies since 2014, including ICD10 codes, clinical characteristics, risk factors, and reporting physician response to “Treatable with surgery?” The Center for Health Development (CHD) maintains a nation-wide vital registry. Combining the data, we determined the prevalence at birth of major congenital anomalies, with a particular focus on those that are surgically treatable in Mongolia. We also examined risk factors for anomalies and for neonatal death, including infant characteristics, maternal home, and season of birth, among others. 

Results: Preliminary analyses of the case series of 1,364 infants in the NCMCH registry suggest prevalence at birth of 8.92 major anomalies per 1000 live births in 2014 (95% CI 7.33-8.55) and 8.87 in 2015 (95% CI 8.24-9.54, birth denominator from CHD). Only 265 (19.4 percent) of infants had anomalies diagnosed in the prenatal period. Of 1,364 infants with major anomalies, 234 or 17.2 percent died within the first 28 days of life. Comparing the 234 neonatal mortality cases to the 1,130 alive at 28 days, there was no significant difference in the gender distribution. The group that died had significantly lower mean birth weight, higher proportion of low and very low birth weight, and lower mean gestational age. A larger proportion of those who died were from rural areas, and a smaller proportion of those who died were considered to be treatable with surgery. A larger proportion of those who died had anomalies diagnosed in the prenatal period.

Conclusion: As expected, the neonates who died were smaller, earlier, and more from rural areas. More of those who died were diagnosed in the prenatal period, though this may reflect more severe anomalies, more easily diagnosed with ultrasound. There are many potentially confounding and effect modifying factors that differ among those who died before 28 days and those who did not. Next steps include determining the prevalence at birth of each ICD10 grouping of anomalies, and carefully controlled regression analyses to determine risk factors for anomalies and anomaly-associated neonatal mortality.

 

91.03 Comparative Analsys of Open vs. Laparoscopic Cholecystectomies in El Peten, Guatemala

J. Imran1, A. Ochoa-Hernandez3, J. Herrejon1, S. Huerta1,2  1University Of Texas Southwestern Medical Center,Surgery,Dallas, TX, USA 2VA North Texas Health Care System,Surgery,Dallas, TX, USA 3Hospital Nacional De San Benito,Surgery,San Benito, EL PETEN, Guatemala

Introduction:
While laparoscopic cholecystectomy (LC) is the standard of care in the Western world, many third world countries still perform a number of open cholecystectomies (OC).  We analyzed the outcomes of all patients undergoing cholecystectomy at a referral hospital in El Peten, Guatemala (Hospital Nacional de San Benito: HNSB). Our null hypothesis was that we would find no difference in outcomes between LC and OC.

Methods:  

This a retrospective, single-institution study at HNSB between January 2014 to April 2016 in all consecutive patients who underwent a cholecystectomy during this time period. Differences between LC and OC were analyzed by univariate analysis [(UA): Fisher’s Exact Test for categorical variables and Student’s T-Test for continuous variables]. Clinically relevant factors and those with a p≤0.2 were entered in a logistic regression model with complications and operative time as the dependent variables. The data is expressed as a means±SD. Significance was established at a p≤0.05 (two-sided).

Results

One hundred consecutive charts were reviewed and used in our analysis.  58% of the cholecystectomies were performed via the open technique and 42% using the laparoscopic approach.  There were 42% emergent and 58% elective cholecystectomies. Of the cholecystectomies performed in the elective setting, 47 % were done open. Conversion rate, hospital length of stay (LOS) and re-admission rate was 4%, 4.8 days and 5% respectively.  There were no SSIs, UTIs or pneumonia in this cohort; 30d and 90d mortality was 0%.  Patients who underwent OC vs. LC were of similar age (36.2±16.3 vs. 37.4 ± 16.1 yo; p=0.7), female gender (79% vs. 88% p=0.27), and ASA class (1.2 ± 0.69 vs. 1.36±0.81; p=0.3). Patients undergoing OC had higher average weight (164.5 ± 27.2 vs. 145 ±42.9 lbs; p=0.03).  Patients with biliary colic were more likely to undergo OC (79% vs. 51%; p=0.001) in comparison to patients with acute cholecystitis who were more likely to undergo LC (36% vs. 14%; p=0.02).  At presentation, patients undergoing LC had a higher mean temperature in comparison to OC (37.1 ± 0.24 vs. 36.9 ± 0.15; p=0.02), but had similar WBC count (10.3±5.0 vs. 9.1±3.1; p=0.2).   There was no difference in operative time between patients undergoing OC and LC (65.3±20.6 vs. 61.6±31.0 min; p=0.5). LOS was similar (4.9±5.4 vs. 4.8 ± 3.9 d; p=0.8), as was readmission rate (7.5% vs. 3.7%; p=0.6). Logistic regression analysis did not identify any independent predictors of outcomes.  

Conclusion:
For this study, we accepted the null hypothesis.  However, we cannot exclude a type II error.  Nearly half of the open cholecystectomies performed during the study period were done in the elective setting.  This finding could be further explored as a potential route to train our surgical residents in open cholecystectomy though the creation of a residency exchange program.  

91.02 Scaling-up Surgical Care in Rural Haiti

L. Ward1,4, D. L. Eisenson2, A. Bowder1,3,4, M. Jean Louis1, M. Raymonville1, T. Pauyo4, M. L. Steer4,9, P. E. Farmer4,8, J. G. Meara4,7, S. R. Sullivan2,4,5,6  1Hopital Universitaire De Mirebalais,Surgery,Mirebalais, CENTRAL PLATEAU, Haiti 2Brown University School Of Medicine,Providence, RI, USA 3Medical College Of Wisconsin,Milwaukee, WI, USA 4Harvard School Of Medicine,Program In Global Surgery And Social Change,Brookline, MA, USA 5Rhode Island Hospital,Plastic Surgery,Providence, RI, USA 6Mount Auburn Hospital,Plastic Surgery,Cambridge, MA, USA 7Children’s Hospital Boston,Boston, MA, USA 8Brigham And Women’s Hospital,Boston, MA, USA 9Tufts Medical Center,Boston, MA, USA

Introduction:

Partners In Health and Zanmi Lasante (PIH/ZL) have provided surgical care in the Central Plateau of Haiti since 1996. These efforts slowly grew to include operating rooms at three hospitals within this catchment area and network of clinics. In 2008, there was an effort to increase surgical capacity with visiting surgical specialists. After the 2010 Earthquake, PIH/ZL partnered with the Haitian Ministry of Health (MSPP) to build University Hospital in Mirebalais (UHM), which opened in 2013. Our purpose is to evaluate the impact of scaling-up surgical care over time.

Methods:

We performed an interrupted time series analysis to compare surgical volume over three time periods: (1) 2007-08, a baseline time-period for surgical care, (2) 2008-09, after the scale-up of visiting surgeons within the existing infrastructure of three hospitals, (3) 2014-2015, after opening UHM with scale-up of surgeons and surgical infrastructure. The primary outcome was total number of operations, measured at monthly intervals from October to March in each time period.

Results:

There was a statistically significant increase in the number of operations performed each month since opening UHM: the average number of operations increased by 121.8 from the baseline trend in time period 1 (95% Confidence Interval 66.2 – 177.4, P = 0.001). The most significant increases were seen in procedures relating to maternal health, orthopedic trauma, and endoscopy.

Conclusion:

Increasing surgical capacity within a health care system in rural Haiti requires more than additional visiting surgical specialists. Scaling-up surgical care requires investments in infrastructure, procurement/supply chains, and of course, skilled surgical specialists – UHM invested in all three areas. UHM provides an impressive model of scaling-up surgical care in a resource poor setting such Haiti.

91.01 The Accessibility, Readability, and Quality of Online Resources for Gender Affirming Surgery

C. R. Vargas1, J. A. Ricci3, M. Lee3, A. M. Tobias3, D. A. Medalie2, B. T. Lee3  1Case Western Reserve University School Of Medicine,Plastic Surgery,Cleveland, OH, USA 2MetroHealth Medical Center,Plastic And Reconstructive Surgery,Cleveland, OH, USA 3Beth Israel Deaconess Medical Center,Plastic And Reconstructive Surgery,Boston, MA, USA

Introduction:  The transgender population is disproportionally affected by health disparities related to access to care.  In many communities, transgender specialists are geographically distant and locally available medical professionals may be unfamiliar with unique needs of transgender patients. As a result, use of internet resources for information about gender affirming surgery is particularly important. Inadequate functional health literacy has been found to contribute to poorer health status, increased mortality, less awareness of preventative health measures, decreased understanding of personal medical conditions, greater likelihood of hospitalization, higher health care costs, lack of self-empowerment, less participation in decision-making in the course of care, and overall worse health outcomes. Minority populations are known to be particularly at risk for these health disparities. This study aims to simulate a patient search for online educational material about gender affirming surgery and to evaluate the accessibility, readability, and quality of the resulting information.

Methods:  An Internet search for the term, “transgender surgery” was performed, and the first ten relevant hits were identified.  Readability was assessed using ten established tests: Coleman-Liau, Flesch-Kincaid, FORCAST, Fry, Gunning Fog, New Dale-Chall, New Fog Count, Raygor Estimate, SMOG, and Flesch Reading Ease. Quality was assessed by two independent raters using JAMA criteria and the DISCERN instrument; these indices were plotted graphically for comparison.

Results: Review of 69 search results was required to identify 10 sites with relevant patient information. 97 articles were subsequently collected; overall mean reading level was 14.7. Individual website reading levels ranged from 12.0 to 17.5.  All articles and websites exceeded the recommended 6th grade level. Quality ranged from 0-4 (JAMA) and 35-79 (DISCERN) across websites. When DISCERN quality was plotted against FRE readability, a nonlinear relationship was observed (Figure).  Peak readability correlated with the middle of the quality index and declined at both extremes. Notably, readability of the highest quality resources was low, suggesting limited utility for average readers.

Conclusion: Websites with relevant patient information about gender affirming surgery were difficult to identify from search results. The content of these sites universally exceeded the recommended reading level. A wide range of website quality was noted and may further complicate successful resource navigation for patients  Barriers in access to appropriately written patient information on the internet may contribute to disparities in referral, involvement, satisfaction, and outcomes for transgender patients.

90.20 Unaccounted Readmissions Following Bariatric Surgery

J. K. Canner1, S. Pourzal1, H. AlSulaim1, K. E. Steele1  1Johns Hopkins University School Of Medicine,Department Of Surgery,Baltimore, MD, USA

Introduction:  Hospitals accredited by the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) are required to report readmission rates.  However, such rates may not account for readmissions to other hospitals if follow-up data are incomplete.  We calculated a more accurate readmission rate using a nationally representative database that tracks readmissions at all hospitals within the same state. We then investigated factors associated with readmission to a different hospital.

Methods:  The new Nationwide Readmissions Database (NRD) from the Healthcare Cost and Utilization Project (HCUP) contains data on inpatient hospital stays from 21 states that collect data linkage information sufficient for identifying readmissions. We identified patients in the NRD admitted for elective bariatric surgery in 2013 and collected patient demographics, including age, gender, insurance status, and residence, as well as clinical information such as length of stay, Charlson comorbidity index, and APR-DRG severity score. We calculated the proportion of patients readmitted within 30 days after initial discharge and also recorded the APR-DRG severity and elective status for the readmission and whether the readmission was to the same hospital.

Results: A total of 61,220 NRD patients underwent elective bariatric surgery in 2013.  Of these, 3,860 (6.3%) were readmitted within 30 days.  Of those readmitted, 693 (18.0%) were to a different hospital. Patients readmitted to a different hospital were more likely to be covered by Medicare (OR=1.46; p=0.036) or Medicaid (OR=1.62; p=0.011) than be privately insured, less likely to live in a medium-sized metro area than in a large metro area (OR=0.61; p=0.013), and less likely to have their surgery at a teaching hospital than at a non-teaching hospital (OR=0.71; p=0.021).  Readmission to a different hospital was strongly associated with higher APR-DRG severity (OR=1.69; p=0.001) and non-elective status at readmission (OR=2.28; p=0.002). Patient age, sex, income level, out-of-state residence, comorbidities, type of surgery and length of stay were not associated with location of readmission. These relationships persisted with multivariable analysis, with the exception of Medicare coverage.

Conclusion: Failure to account for readmissions to different hospitals may underestimate readmission rates by at least 18%.  Patients with more severe complications are the most likely to be readmitted to different hospitals.  A better understanding and accounting for all readmissions may improve the care and safety of the bariatric surgical patient. Further research using data sets with more detailed geographic information may reveal the role of distance as a factor in readmission location.

 

90.19 Physiologic Drivers Of Intraoperative Transfusion During Major Gastrointestinal Surgery?

M. Cerullo2, F. Gani2, S. Y. Chen2, J. K. Canner2, W. W. Yang3, S. M. Frank3, T. M. Pawlik1  1Ohio State University,Wexner Medical Center,Columbus, OH, USA 2Johns Hopkins University School Of Medicine,Department Of Surgery,Baltimore, MD, USA 3Johns Hopkins University School Of Medicine,Department Of Anesthesiology And Critical Care Medicine,Baltimore, MD, USA

Introduction:  Current guidelines for transfusion largely focus on nadir hemoglobin (Hb) levels. Hb triggers may not be helpful, however, in defining appropriate intra-operative use of packed red blood cells (PRBCs).  We sought to define the use intra-operative PRBC relative to quantitative physiologic factors at the time of surgery.

Methods:  Prospective perioperative data on patients undergoing major gastrointestinal surgery between 2010 and 2014 were analyzed. Risk of intraoperative transfusion was assessed with multivariable extended Cox models using clinical covariates (e.g. type of surgery, perioperative Hb, coagulation parameters, American Society of Anesthesiologists (ASA) classification, and Charlson co-morbidity), as well as time-varying intraoperative covariates (e.g. continuously-monitored mean arterial pressure [MAP], heart rate, and estimated blood loss [EBL]).

Results: 2,428 patients were identified; 384 (15.8%) patients received an intraoperative transfusion. Higher risk of intraoperative transfusion was associated with preoperative factors including lower Hb (hazard ratio [HR]=1.22, 95% confidence interval [CI]: 1.14-1.30, p<0.001) and higher ASA class (HR=1.55, 95%CI:1.24-1.93, p<0.001). Intraoperative risk factors for transfusion included higher EBL (HR=1.43, 95%CI:1.27-1.62, p<0.001, per 1000mL), as well as lower instantaneous MAP (HR=1.15, 95%CI:1.08-1.22, p<0.001) and higher heart rate (HR=1.30, 95%CI:1.21-1.39, p<0.001). While the majority of patients had a transfusion for a physiologic indication, among the 384 patients transfused, 27.1% of intra-operative transfusions were delivered to patients who never had a physiologic indication (heart rate>100, MAP<65, or a nadir Hb<8) (Figure). 

Conclusion: Physiologic indicators account for considerable variability in intraoperative transfusion practices among patients undergoing major surgery. Up to 27% of patients who received an intraoperative transfusion had no identifiable physiological reason for a transfusion, thereby suggesting possible overutilization of PRBC in a subset of patients. 

 

90.18 Geographic Proximity of High and Low Quality Bariatric Centers; An Opportunity for Regionalization?

A. M. Ibrahim1, A. A. Ghaferi1, J. Thumma1, J. Dimick1  1University Of Michigan,Ann Arbor, MI, USA

Introduction: Responding to reports about the safety of bariatric surgery, leading surgical societies established criteria to create centers of excellence. As a result, many providers underwent changes to obtain accreditation and now nearly all bariatric procedures occur at these centers. Because clinical outcomes are not part of the accreditation process, it unclear if these centers provide high quality care uniformly. Moreover, the geographic availability of centers with high quality outcomes is unknown.  

Methods: A retrospective review of 137,016 patients undergoing bariatric surgery at centers of excellence between 2009-2011. Data was obtained from the Healthcare Cost and Utilization Project – State Inpatient Database which included unique hospital identification numbers in 12 states allowing comparisons across 188 centers of excellence. For each hospital, we evaluated quality by calculating the risk and reliability adjusted serious complications rates within 30 days of the index operation. Variations across centers of excellence nationally as well as within each individual state and hospital service area was assessed.  

Results:  Wide variation in quality exists across bariatric surgery centers of excellence . At the national level, the risk and reliability adjusted serious complication rates at each individual center varied 42 fold ranging from 0.34% to 14.6%. The top and bottom deciles varied 12 fold (top decile 0.5%; bottom decile 6.1%; p<0.005.) Similar variation was seen at the state level as well; California (N=46) ranged from 0.3% to 5.6% and New York (N=35) ranged from 0.5% to 8.6%. For the 47 hospital service areas with a low quality provider (highest quartile of complications), 34 (72%) of them also had an average or high quality provider in the same or adjacent hospital service area.

Conclusions:  Even among centers of excellence for bariatric surgery, wide variation exists in rates of post-operative serious complications. Given that most low quality providers are geographically located near higher performing providers, opportunities for quality improvement through local regionalization should be considered. 

90.17 Routine Type and Screen is Unnecessary for Patients undergoing Thyroid and Parathyroid procedures

L. Anewenah1, M. Asif1, M. Rasouli1, O. Domingo1  1Mercy Catholic Medical Center,Department Of General Surgery,Darby, PENNSYLVANIA, USA

Introduction:

The thyroid and parathyroid glands are among the most highly vascularized tissue beds. As a result, bleeding from these glands can be life threatening. While approximately only 1% of postoperative bleeding is observed in these procedures, it is common practice to perform type and screening of patients undergoing these procedures.

The purpose of our study is to review records of thyroid and parathyroid procedures performed at our Institution from the beginning of 2012 to the end of 2014 and of that number how many went on to need blood transfusion. We also hope to be able to identify the characteristics of the patients who had postoperative bleeding to inform policy regarding type and screening test for patients undergoing these procedures.

Methods:
This is a retrospective study of patients who underwent thyroid and parathyroid procedures. Demographic, laboratory results and surgical related data were obtained from querying our institutional database. Descriptive analysis was performed. 

Results:

A total 62 patients including 46 females (74%) were included in the study. Mean age of patients at the time of surgery was 55 +/- 13 years.

46 thyroid surgeries (34 total thyroidectomy, 10 thyroid lobectomy, one thyroglossal cyst removal and one completion thyroidectomy) and 16 parathyroid surgeries (3 total parathyroidectomies, 3 left and 3 right parathyroidectomies, 5 single parathyroid gland removal, 1 subtotal thyroidectomy, and 1 exploration of parathyroid) were performed. All surgeries were performed under general anesthesia. Estimated blood loss ranged from 5 to 200 milliliters. Median weight of the removed thyroid gland was 26.3 grams ranged from 2.7 to 409 grams. Median weight of the removed parathyroid gland was 1.5 grams ranged from 0.07 to 10.1 grams. Median length of hospital stay was 1 day ranged from 0 to 28 days.

Preoperative International Normalized Ratio (INR) was high in only three patients, which ranged from 1.4 to 1.5. Only 5 patients (8%) had preoperative platelet counts of less than 150,000 platelet per microliter (ranged from 73 to 139 platelet per microliter). Mean preoperative and postoperative hemoglobin were 12.7 +/- 2.2 g/dL and 11.2 +/- 1.7 g/dL, which was statistically significant (p<0.001).

Type and screen was requested in all cases. Only two patients (3.2%) required postoperative transfusion and 2 units of packed cell transfused in each case (table 1)

Conclusion:

Large volume of blood loss requiring intraoperative or postoperative blood transfusion is extremely rare. For the patients that required a transfusion, the type and screen can be done rapidly and the patients safely transfused. It seems, therefore, that routine type and screen is unnecessary in patients undergoing thyroid and parathyroid.

90.16 Attitudes and Practice Patterns in Management of Adhesive Small Bowel Obstruction Among Surgeons

L. W. Thornblade1, A. R. Truitt1, D. R. Flum1, D. C. Lavallee1  1University Of Washington,Department Of Surgery,Seattle, WA, USA

Introduction:  Classic training instructs surgeons to, “never let the sun set on a small bowel obstruction (SBO)” for concern of bowel ischemia. However the routine use of CT scans for ruling out compromised bowel provides the opportunity for trial of non-operative management, allowing time for spontaneous resolution of adhesive SBO. In light of such advances in practice, little is known about how surgeons choose to manage these patients, in particular whether there is an agreed-upon time window for safe non-operative management.

Methods:  Using a case scenario of a patient with CT-scan confirmed adhesive SBO without bowel ischemia, we interviewed a purposive sample of general surgeons practicing in Washington State to understand approaches to clinical management. Interview questions addressed typical practice, use of an oral contrast study, timing of surgery, and use of laparoscopy. We conducted a qualitative analysis to identify themes in practice and attitudes. 

Results: Surgical practice patterns for patients with SBO vary widely. The importance of timely surgeon involvement and serial abdominal exams emerged as themes among most participants. Many participants identified themes of uncertainty about the diagnosis of a complete obstruction. The period of time that surgeons were willing to manage patients non-operatively ranged from 1-10 days. Most surgeons favored open surgery. All surgeons acknowledged a lack of clinical evidence to support appropriate management of patients with SBO.

Conclusion: Interviews with practicing surgeons across a range of practice sites illuminate a changing paradigm away from routine early operative management of patients with adhesive SBO. However, there is no established length for a trial of non-operative management. The surgeon attitudes and practice patterns identified will inform feasibility and design of future prospective randomized studies of patients with non-ischemic adhesive SBO.

 

90.15 Evolution of Laparoscopic Appendectomy: A 12 Year Experience of a Tertiary Care Hospital

M. H. Siddiqui1, R. Sultan1, F. Shaukat2, H. Zafar1  1Aga Khan University Hospital,Surgery/General Surgery,Karachi, Sindh, Pakistan 2Aga Khan University Hospital,Oncology/Radiation Oncology,Karachi, Sindh, Pakistan

Introduction:  Laparoscopic appendectomy has gained tremendous popularity and acceptance in many countries but has not become the standard of care so far. The aim of this study is to assess the outcomes and trends of this procedure in Aga Khan University Hospital, Karachi, Pakistan, over a decade.

Methods: All adult patients who underwent laparoscopic appendectomy from Jan 2004-Dec 2015 were included in the study. Patients’ demographics, operative details like duration of surgery and conversion rate, histopathology and complications were recorded in proforma.

Results: 831 patients were included in the study and trend showing a significant increase in number of laparoscopic appendectomies. 64% of patients were male, median age of 28 years and median hospital stay was 2 days. Mean duration of surgery is 67.47 (SD 25.86) minutes which has significantly improved over time. 7% of the cases were converted to open with decreased conversion rate in recent years.  Negative appendectomy rate was 7.86% in our study which has decreased since advent of FACT. Total complication rate was 6.37%, there is rising trend in overall complications of laparoscopic appendectomy over time.

 

Conclusion: Laparoscopic appendectomy has become the preferred method of choice in our institution. Results showing improvement in terms of duration of surgery, conversion rate and negative appendectomy rates in addition to all previous known advantages of cosmetics, early recovery, decrease in hospital stay. Our results showing increase in number of complications, but it is reflection of the fact that more complex cases are now being attempted and completed laparoscopically.