Steve Serrao

Endocuff assisted colonoscopy significantly increases sessile serrated adenoma detection in veterans

BACKGROUND Colorectal cancer (CRC) is the second most common cause of cancer related deaths in the United States. Colonoscopy is the gold standard for the detection of CRC. There are many colonoscopy quality measures and among these the adenoma detection rate (ADR) has demonstrated a significant impact in reducing mortality from CRC. The primary aim of our study was to compare ADR and distribution of polyp type in patients undergoing Endocuff-assisted colonoscopy (EAC) versus standard colonoscopy (SC) in a VA system. METHODS Retrospective data was collected from 496 patients who underwent routine screening, surveillance and diagnostic colonoscopies either via SC from January 6, 2014 through March 12, 2014 or EAC from September 24, 2014 through February 19, 2015. A total of 54 patients were excluded based on a personal history of CRC and prior resection, incomplete colonoscopy due to poor bowel preparation, and removal or loss of Endocuff (EC). Primary outcomes measured and compared were ADR and types of polyps found. RESULTS The overall ADR in the EAC group was higher at 59.91% versus 50.66% for SC, accounting for a 9% increase (P=0.0508). EAC was able to detect a total of 59 sessile serrated adenoma/polyps (SSA/Ps) compared to SC only detecting 8 (P≤0.0001). There was a significant increase in the SSA/P detection rate with EAC at 15% versus 3% in the SC group (P≤0.0001). CONCLUSIONS EAC significantly increases the detection of SSA/P and has shown a trend in improving ADR in our veteran population.

Sa1882 Exploring Predictors of in-Hospital Mortality in Dieulafoy's Lesions of the Stomach and Intestine

Introduction: Dieulafoys lesions (DL) are a rare and often unrecognized cause of obscure, and sometimes significant upper gastrointestinal (GI) hemorrhage. There are currently no population-based studies evaluating mortality risk associated with the locality of DL. We used a national database to assess the association between in-hospital mortality and demographic, co-morbidities and intervention variables, among patients with DL of the stomach and intestine. Methods: Using National Inpatient Sample (NIS) data between January 2004 and December 2009, simple and multiple logistic regression analyses were conducted to assess the effect of the various covariates onmortality. Demographic covariates explored in this analysis included gender, age, race/ethnicity, income, and type of insurance. Co-morbidities of interest include atrial fibrillation (AFIB), coronary heart disease (CAD), congestive heart failure (CHF), acute renal failure (ARF), chronic kidney disease (CKD), end stage renal disease (ESRD) and coagulopathies (COAG). Additionally, the Charlson-Deyo Index (CDI) was used to measure the burden of co-morbidities. Interventions include packed red cell transfusion (PCT), endoscopic control of gastric hemorrhage (EGD), small bowel endoscopy (SBE) and intensive level of care (ICU). Results: We identified 4,652 patients with a primary diagnosis of DL, out of which 88% were located in the stomach and 12% were found in the small and large intestine. The overall percentage mortality of DL is around 3%. There was no significant difference in mortality between DL of the stomach and intestine. When compared to 2004, there is a significant decrease in mortality from year to year. Multiple logistic regression, modeling mortality as the outcome variable, was statistically significant for patients with age greater than 85, odds ratio (OR) =5.30 and patients with ARF OR= 4.21. About 1,335 patients or 29% of the total population had small bowel endoscopy and was found to be associated with an OR=1.78. About 70% or 3290 patients had EGD and 60% or 2786 patients had PCT. About 6% of the patients had ICU level of care, as defined by the use of mechanical ventilation (6%), central venous catheterization (13%) and hemodialysis (3%). If patient had ICU level of care, it was associated with OR=16.8 of death. Coagulopathies were not found to be significantly associated with death. Conclusion: This is the first population-based study that explored factors associated with in-hospital mortality for DL. Among demographic variables, only age greater than 85 was associated with significantly higher mortality. Disparity in care was not a factor since gender, race/ethnicity, income and type of insurance had no bearing on mortality. Mortality however, was significantly associated with the development of ARF and if the patient was admitted to ICU level of care.

Sa1279 Exploring the Risk of Septicemia Among Patients Diagnosed With Clostridium difficile

Introduction: Clostridium difficile infection (CDI) represents an increasing public health problem as it is a primary cause of antibiotic-induced diarrhea and colitis. In the United States, CDI affects millions of patients each year and represents an annual cost of over

1030 Socio Economic Status and Type of Insurance Affects In-hospital Mortality From Peptic Ulcer Bleeding

1 billion. A more in-depth understanding of C. difficile colonization is necessary to improve therapeutics. In C. difficile pathogenesis, antibiotic disruption of the gut microbiota provides an open niche and invading C. difficile enters a colonization phase, which includes bacterialhost interaction, mucus adhesion and toxin production. Mucus oligosaccharides serve as both a fuel source and as binding sites for a number of bacteria. Previous studies have demonstrated that C. difficile is unable to cleave terminal mucus oligosaccharides. We hypothesize that an altered gut microbiota in CDI patients cleaves terminal mucus oligosaccharides exposing/producing bacterium and toxin binding sites. Furthermore we hypothesize that cleaved oligosaccharides can be foraged by C. difficile and used for proliferation. Methods& results: CDI patients presented with increased Bacteroidetes and decreased Firmicutes stool microbiota. In addition, CDI biopsies exhibited decreased N-acetylgalactosamine and increased terminal galactose mucus oligosaccharides. Terminal galactose residues have been shown to be the toxin A binding site in animal models, but lack of terminal galactose residues in the human colon have led to the hypothesis that the toxin A receptor must be different in humans. Our data indicates that terminal galactose residues may represent the human toxin A receptor which is upregulated in CDI. No changes were observed in mucus fucose or mannose levels. CDI patients did present with decreased mucus MUC2, with no changes in MUC1. These data demonstrate a unique mucus oligosaccharide composition in CDI patients. In vitro C. difficile BAA-1870 had enhanced binding to mucus extracted from CDI patients compared to healthy patient mucus, indicating the presence of an alternative binding epitope in the mucus of CDI patients. Furthermore, C. difficile BAA-1870 grown in TYGwas able to use fucose,mannose, galactose, N-acetylgalactosamine, N-acetylglucosamine, and sialic acid oligosaccharides for growth, although growth varied depending on pH and Na.Conclusions: These data demonstrate that CDI patients exhibit an altered gut microbiota with corresponding altered mucus oligosaccharide and MUC2 composition. In vitro C. difficile is capable of using multiple oligsoaccharides for growth which may represent a factor in the colonization phase.

In-hospital weekend outcomes in patients diagnosed with bleeding gastroduodenal angiodysplasia: a population-based study, 2000 to 2011

Background: The Canadian North Helicobacter pylori (CANHelp) Working Group conducts community Helicobacter pylori projects to address public concerns about health risks from this infection in Arctic Canada, where H.pylori prevalence and stomach cancer rates are elevated. At the request of community leaders of Fort McPherson, Northwest Territories (population~800, ~95% Aboriginal), the ongoing Fort McPherson H. pylori Project launched in 2012 to investigate the disease burden related to H.pylori infection and identify strategies for reducing related health risks. Methods: A local planning committee guided the design and implementation of the project, which includes six components: surveys of risk factors, urea breath test (UBT) screening for H. pylori infection, upper gastrointestinal endoscopy with biopsies collected and histopathology, treatment, knowledge exchange, and policy development. During 2012-13, all residents of Fort McPherson were invited to participate in UBT screening and questionnaire-based risk factor interviews; residents ≥15 years of age were invited to undergo endoscopy in temporary endoscopy units in the local health centre. Participants could also enrol in a randomized trial comparing two 10-day H.pylori therapies: sequential therapy (ST) consisted of a proton pump inhibitor (PPI) and amoxicillin for days 1-5, followed by a PPI, clarithromycin, and metronidazole for days 6-10; quadruple therapy (QT) consisted of a PPI, bismuth, metronidazole, and tetracycline for days 1-10. Treatment outcomes were classified by UBT at >=10 weeks after treatment. Results: To date, 226 residents, aged 4-98 years, have consented to participate. Parental consent and childs assent were obtained for residents <17 years of age. Of the 226 project participants, 180 have completed risk factor interviews, 217 had a UBT (positivity=59%), 53 had endoscopy with biopsies collected, 71 consented to treatment and 60 enrolled in the treatment trial, with 24 to date completing a post-treatment UBT. Of the 53 participants who had endoscopy, gastroenterologists identified gastritis in 15%, gastric ulcer in 4%, gastric erosion in 11%, duodenitis in 4%, duodenal erosion in 2%, and esophagitis in 11%. Histopathology (Sydney classification) of gastric biopsies from 53 Fort McPherson H. pylori Project participants is shown in table 1. Preliminary findings from the treatment trial to date are: 86% treatment success for participants randomized to ST (12/14; 95% CI 57%-98%) and 100% treatment success for participants randomized to QT (10/10; 95% CI 74%-100%). Discussion: These results add to evidence that shows Arctic Aboriginal communities to be disproportionately affected by H. pylori infection. The high prevalence of moderate-severe gastritis shows that public concern over risks from H. pylori infection is warranted. Histology Results