su1055 similar clinical characteristics of familial and sporadic inflammatory bowel disease

2
patients age 50-59 years. Methods: We performed a retrospective registry study of all first time colonoscopies performed in patients between the ages of 40 to 59 years. Exclusion criteria included having a family history of CRC, personal history of inflammatory bowel disease, weight loss or iron deficiency anemia as indication for colonoscopy, incomplete colonoscopy and poor bowel preparation (Aronchik= poor, sub-optimal or Ottawa =/ .10). Information obtained included demographics, BMI, as well as endoscopic and pathologic reports. The variables were analyzed using multivariate regression analysis. Results: A total of 2,607 colonoscopies were included in the analysis (1,911 ages 50-59 and 696 ages 40- 49). Sixty six percent (1252/1911) of patients with age 50-59 years and 61% (425/696) of patients with age 40-49 years had a BMI =/ . 25. The ADR for patients 40 to 49 and 50 to 59 years were 16.3% and 27% respectively. Patients with a BMI =/ . 25 had an increased ADR only after the age of 50 years (20% vs. 29% in 50-59 years & 16.5% vs. 16.3% in 40 to 49 years). On logistic regression analysis, a BMI =/ . 25 increased the risk for adenoma detection in patients age 50 to 59 years (OR 1.54 CI 1.19 to 1.99 / p=0.0009), but no difference was noted in patients age 40 to 49 years (OR 0.92 CI 0.56 to 1.50 / p=0.74). Conclusion: Patients aged 50-59 with a BMI =/ . 25 have an increased risk for detection of adenomas compared with patients with a BMI ,25. This risk was not appreciated in population younger than 50 years. Su1052 Rate of Colonoscopy Has Decreased in Patients With Diverticular Bleeding but the Mortality Has Remained the Same: A Nationwide Analysis of Trends 2004- 2008 Muhammad Ali, Shahryar Ahmad, Nilay Kumar, Gagan Kumar, Kia Saeian Introduction: Diverticular bleeding is a major cause of lower gastrointestinal bleeding (LGIB) and leads to morbidity and mortality among hospitalized patients. Colonoscopy can be performed as a diagnostic as well a therapeutic procedure. If bleeding is uncontrolled, a colectomy may need to be performed. There is conflicting data on whether trends in these procedures have changed over time. With our study, we wanted to 1) identify procedure trends in diverticular bleeding patients including colonoscopy and colectomy, 2) determine if this has had any impact on the mortality. Materials and Methods: Using Nationwide Inpatient Sample (NIS) from year 2004-2008, we identified patients admitted with a diverticu- lar bleed using appropriate ICD-9 codes. Among these patients, the rate of colonoscopy and colectomy was determined using appropriate procedure codes. Mortality was also calculated for these admissions. Results: There was an overall decrease in the number of patients admitted with diverticular bleeding from 2004-2008. The rate of colonoscopy decreased from 71.06% to 65.72% (p ,0.001) among patients admitted with diverticular bleeding. The rate of colectomy stayed about the same from 3.55% to 3.40% (p=0.66) for this group. Overall mortality remained unchanged (1.87%-1.84%) among patients admitted with diverticulosis. On subgroup analysis, mortality remained statistically unchanged when looked at those undergoing colectomy or colonoscopy separately (10.6%-9.8% and 1.18%-1.25%). Other demographical variables such as age groups, gender, insurance distribution and discharge disposition remained the same over 2004-2008. Conclusion: In this nationally representative sample, we found that from 2004-2008, the rate of colonoscopy in patients admitted with diverticular bleeding has decreased but has not impacted the morality. Further studies to characterize the factors involved are warranted. Su1053 Prevalence of Functional Bowel Disorders and Faecal Incontinence in Primary Health-Care Seekers: An Australian Experience Kheng-Seong Ng, Natasha Nassar, Marc A. Gladman Introduction: Functional bowel disorders (FBDs) and faecal incontinence (FI) are chronic and debilitating disorders that represent significant management problems. Whilst information regarding their prevalence in the community is available, there is a relative paucity of clinic- based studies, which may provide useful information in health-care seekers. Additionally, little is known about the prevalence of FBD based on the Rome III criteria or the severity of FI in Australian populations. Therefore, the aim of this study was to measure the prevalence of FBDs and FI (and its severity) among Australian primary health-care seekers using objective, standardised criteria. Methods: A cross-sectional survey was conducted amongst primary health-care seekers in Sydney, Australia. Participants were prospectively recruited on a consecutive basis as they entered clinic waiting rooms and completed a self-administered questionnaire designed to diagnose FBDs (irritable bowel syndrome [IBS], constipation, functional bloating, and functional diarrhoea) using the Rome III criteria. Further, the severity of FI was assessed using the modified Wexner incontinence score. In addition to measuring disease prevalence, data were modelled to identify associations with medical and surgical histories and health-care utilisation. Results: 396 of 596 subjects (66.4%) approached agreed to participate. Demographic and socio-economic characteristics of the participants were reflective of the general population of Sydney. IBS, constipation, functional bloating, and functional diarrhoea were diagnosed in 11.1%, 8.1%, 6.1%, and 1.5% of subjects, respec- tively. FI was present in 12.1% of subjects with the majority (93%) reporting mild to S-387 AGA Abstracts moderate incontinence (modified Wexner score ,17). Subjects with FI were 4 times more likely to have had previous anal surgery (OR3.80, 95%CI 1.55-9.33), 4 times more likely to have IBS (OR3.80, 95%CI 1.82-7.93), and 3 times more likely to have urinary incontinence (OR3.24, 95%CI 1.73-6.08). Previous obstetric trauma was not associated with FI (P=0.79). Subjects with IBS were 9 times more likely to have had a previous colonoscopy (OR9.32, 95%CI 3.10-28.04), 3 times more likely to report anxiety or depression (OR3.42, 95%CI 1.78-6.58), and 3 times more likely to have had a previous cholecystectomy (OR3.11, 95%CI 1.16-8.37). Whilst there was a tendency for diabetic subjects to be constipated, this association failed to reach significance (P=0.07). Conclusions: FI and FBDs appear to be prevalent conditions amongst Australian primary health-care seekers. Furthermore, the needs of affected patients would seem to be complex given their co-existing symptoms/conditions, suggesting that detailed assessment is required to adequately address their needs. These findings have implications for health service planning and provision. Su1054 Prevalent Round Stage Shift in the National Bowel Cancer Screening Programme in Wales; Data From the First 3 Years in At a Single Screening Centre Joanna Hurley, John Green, A. B. Hawthorne, Jared Torkington, Sunil Dolwani Introduction: Colorectal cancer screening is based on early detection of cancers and removal of premalignant polyps though this adenoma to carcinoma sequence is thought to progress over several years. The Bowel Screening Programme in Wales based on guaiac FOBt and colonoscopy for individuals resting positive began roll-out in October 2008, with the aim of reducing mortality through cancer detection at an early stage. The aim of this study was to investigate whether screen detected cancers in Cardiff and the Vale of Glamorgan demonstrated any shift in the stage of cancer during the first three years of screening (initial prevalent round). Methods: Data was collected prospectively to compare the staging of colorectal cancer diagnosed in the BCSP with cancers diagnosed in the non-screening popula- tion in the same geographical region from 1st October 2008 to 31st December 2011. All information was cross checked with Cancer Registry data. Results: Screen detected cancer was found in 69 individuals (44 male, 25 female), with a positive predictive value of colonoscopy (after positive FOB testing) of 8.7%. Complete clinical staging was available for all 69 individuals; two patients did not undergo surgical resection due to the presence of metastases after radiological staging. There were 696 non-screening detected cancers during the same time period. For the purposes of this analysis, polyp cancers (cancer that was removed by endoscopic means at the time of colonoscopy/flexible sigmoidoscopy) were included in Duke's stage A, except for one polyp cancer that required subsequent surgical resection and was staged as Duke's C1. The results are shown in table 1. Three-quarters of cancers diagnosed in the BCSP were Dukes A or B, compared to 44.1% in the non-screening population. Of Duke's D cancers, only 2.8% were diagnosed through screening, with 27% diagnosed in the non screening population (p ,0.0001). Conclusion: This data strongly supports significant stage shift of colorectal cancer even within the initial prevalent round in this single Bowel Cancer Screening centre in Wales that the benefits of screening may be demonstrable in outcomes at a relatively early stage of the program. Table 1. Stage of cancer diagnosed with comparable data from non-screening population Su1055 Similar Clinical Characteristics of Familial and Sporadic Inflammatory Bowel Disease SooK Hee Chung, Soo Jung Park, Lee Hye Sun, Jae Hee Cheon, Sung Pil Hong, Tae Il Kim, Won Ho Kim Background: Inflammatory bowel disease (IBD) is caused by complex interaction between genetic and environmental factors. In spite of several studies of familial IBD, we do not have sufficient information about the clinical characteristics in familial IBD. The aim of this study is to investigate differences of clinical characteristics and disease courses between familial and sporadic IBD patients. Methods: We obtained clinical data on Crohn's disease (CD, 691 cases) and ulcerative colitis (UC, 1113 cases) in Severance Hospital, Yonsei University College of Medicine from Jan. 2005 to Feb. 2012. Seventeen patients (2.5%) with CD and 27 patients (2.4 %) with UC identified to have familial history of IBD. For each control case, 3 times age- and sex-matched CD and UC patients who had no family history of IBD were randomly selected based on the electronic medical data. We compared the clinical characteristics and disease course of familial CD/UC patients with sporadic CD/UC patients, respectively. Results: There were no significant differences of age at diagnosis, age at symptom onset, main symptom at diagnosis (abdominal pain, diarrhea, weight loss, fever, abdominal mass, hematochezia), extraintestinal manifestation, location, behavior of disease disease activity, number of hospi- talization, number of operation, operation type, causes of operation and operation rate, number of relapse, medical treatment with 5-ASA, oral steroid, and azathioprine between familial and sporadic CD and familial and sporadic UC patients, respectively. Median (min- max) follow up peroids after diagnosis of familal CD patients and sporadic CD patients were 84 (24-312) months and 36 (8-240) months, respectively (P=0.008). But there was no significant difference of follow up peroids after diagnosis of familial UC patients and sporadic UC patients. Familial CD patients more frequently used anti TNF agents than sporadic CD patients (3/17, 17.6 % vs 0/21, 0 %, P=0.014). Longer follow up period of disease might result in more usage of anti-TNF antibody in familial CD patients than sporadic CD patients. AGA Abstracts

Upload: won-ho

Post on 31-Dec-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

patients age 50-59 years. Methods: We performed a retrospective registry study of all firsttime colonoscopies performed in patients between the ages of 40 to 59 years. Exclusioncriteria included having a family history of CRC, personal history of inflammatory boweldisease, weight loss or iron deficiency anemia as indication for colonoscopy, incompletecolonoscopy and poor bowel preparation (Aronchik= poor, sub-optimal or Ottawa =/ .10).Information obtained included demographics, BMI, as well as endoscopic and pathologicreports. The variables were analyzed using multivariate regression analysis. Results: A totalof 2,607 colonoscopies were included in the analysis (1,911 ages 50-59 and 696 ages 40-49). Sixty six percent (1252/1911) of patients with age 50-59 years and 61% (425/696) ofpatients with age 40-49 years had a BMI =/ . 25. The ADR for patients 40 to 49 and 50to 59 years were 16.3% and 27% respectively. Patients with a BMI =/ . 25 had an increasedADR only after the age of 50 years (20% vs. 29% in 50-59 years & 16.5% vs. 16.3% in 40to 49 years). On logistic regression analysis, a BMI =/ . 25 increased the risk for adenomadetection in patients age 50 to 59 years (OR 1.54 CI 1.19 to 1.99 / p=0.0009), but nodifference was noted in patients age 40 to 49 years (OR 0.92 CI 0.56 to 1.50 / p=0.74).Conclusion: Patients aged 50-59 with a BMI =/ . 25 have an increased risk for detectionof adenomas compared with patients with a BMI ,25. This risk was not appreciated inpopulation younger than 50 years.

Su1052

Rate of Colonoscopy Has Decreased in Patients With Diverticular Bleeding butthe Mortality Has Remained the Same: A Nationwide Analysis of Trends 2004-2008Muhammad Ali, Shahryar Ahmad, Nilay Kumar, Gagan Kumar, Kia Saeian

Introduction: Diverticular bleeding is a major cause of lower gastrointestinal bleeding (LGIB)and leads to morbidity and mortality among hospitalized patients. Colonoscopy can beperformed as a diagnostic as well a therapeutic procedure. If bleeding is uncontrolled, acolectomy may need to be performed. There is conflicting data on whether trends in theseprocedures have changed over time. With our study, we wanted to 1) identify proceduretrends in diverticular bleeding patients including colonoscopy and colectomy, 2) determineif this has had any impact on the mortality. Materials and Methods: Using NationwideInpatient Sample (NIS) from year 2004-2008, we identified patients admitted with a diverticu-lar bleed using appropriate ICD-9 codes. Among these patients, the rate of colonoscopy andcolectomy was determined using appropriate procedure codes. Mortality was also calculatedfor these admissions. Results: There was an overall decrease in the number of patientsadmitted with diverticular bleeding from 2004-2008. The rate of colonoscopy decreasedfrom 71.06% to 65.72% (p,0.001) among patients admitted with diverticular bleeding.The rate of colectomy stayed about the same from 3.55% to 3.40% (p=0.66) for thisgroup. Overall mortality remained unchanged (1.87%-1.84%) among patients admitted withdiverticulosis. On subgroup analysis, mortality remained statistically unchanged when lookedat those undergoing colectomy or colonoscopy separately (10.6%-9.8% and 1.18%-1.25%).Other demographical variables such as age groups, gender, insurance distribution anddischarge disposition remained the same over 2004-2008. Conclusion: In this nationallyrepresentative sample, we found that from 2004-2008, the rate of colonoscopy in patientsadmitted with diverticular bleeding has decreased but has not impacted the morality. Furtherstudies to characterize the factors involved are warranted.

Su1053

Prevalence of Functional Bowel Disorders and Faecal Incontinence in PrimaryHealth-Care Seekers: An Australian ExperienceKheng-Seong Ng, Natasha Nassar, Marc A. Gladman

Introduction: Functional bowel disorders (FBDs) and faecal incontinence (FI) are chronic anddebilitating disorders that represent significant management problems. Whilst informationregarding their prevalence in the community is available, there is a relative paucity of clinic-based studies, which may provide useful information in health-care seekers. Additionally,little is known about the prevalence of FBD based on the Rome III criteria or the severityof FI in Australian populations. Therefore, the aim of this study was to measure the prevalenceof FBDs and FI (and its severity) among Australian primary health-care seekers usingobjective, standardised criteria. Methods: A cross-sectional survey was conducted amongstprimary health-care seekers in Sydney, Australia. Participants were prospectively recruitedon a consecutive basis as they entered clinic waiting rooms and completed a self-administeredquestionnaire designed to diagnose FBDs (irritable bowel syndrome [IBS], constipation,functional bloating, and functional diarrhoea) using the Rome III criteria. Further, the severityof FI was assessed using the modified Wexner incontinence score. In addition to measuringdisease prevalence, data were modelled to identify associations with medical and surgicalhistories and health-care utilisation. Results: 396 of 596 subjects (66.4%) approached agreedto participate. Demographic and socio-economic characteristics of the participants werereflective of the general population of Sydney. IBS, constipation, functional bloating, andfunctional diarrhoea were diagnosed in 11.1%, 8.1%, 6.1%, and 1.5% of subjects, respec-tively. FI was present in 12.1% of subjects with the majority (93%) reporting mild to

S-387 AGA Abstracts

moderate incontinence (modified Wexner score ,17). Subjects with FI were 4 times morelikely to have had previous anal surgery (OR3.80, 95%CI 1.55-9.33), 4 times more likelyto have IBS (OR3.80, 95%CI 1.82-7.93), and 3 times more likely to have urinary incontinence(OR3.24, 95%CI 1.73-6.08). Previous obstetric trauma was not associated with FI (P=0.79).Subjects with IBS were 9 times more likely to have had a previous colonoscopy (OR9.32,95%CI 3.10-28.04), 3 times more likely to report anxiety or depression (OR3.42, 95%CI1.78-6.58), and 3 times more likely to have had a previous cholecystectomy (OR3.11,95%CI 1.16-8.37). Whilst there was a tendency for diabetic subjects to be constipated, thisassociation failed to reach significance (P=0.07). Conclusions: FI and FBDs appear to beprevalent conditions amongst Australian primary health-care seekers. Furthermore, the needsof affected patients would seem to be complex given their co-existing symptoms/conditions,suggesting that detailed assessment is required to adequately address their needs. Thesefindings have implications for health service planning and provision.

Su1054

Prevalent Round Stage Shift in the National Bowel Cancer ScreeningProgramme in Wales; Data From the First 3 Years in At a Single ScreeningCentreJoanna Hurley, John Green, A. B. Hawthorne, Jared Torkington, Sunil Dolwani

Introduction: Colorectal cancer screening is based on early detection of cancers and removalof premalignant polyps though this adenoma to carcinoma sequence is thought to progressover several years. The Bowel Screening Programme in Wales based on guaiac FOBt andcolonoscopy for individuals resting positive began roll-out in October 2008, with the aimof reducing mortality through cancer detection at an early stage. The aim of this studywas to investigate whether screen detected cancers in Cardiff and the Vale of Glamorgandemonstrated any shift in the stage of cancer during the first three years of screening (initialprevalent round). Methods: Data was collected prospectively to compare the staging ofcolorectal cancer diagnosed in the BCSP with cancers diagnosed in the non-screening popula-tion in the same geographical region from 1st October 2008 to 31st December 2011. Allinformation was cross checked with Cancer Registry data. Results: Screen detected cancerwas found in 69 individuals (44 male, 25 female), with a positive predictive value ofcolonoscopy (after positive FOB testing) of 8.7%. Complete clinical staging was availablefor all 69 individuals; two patients did not undergo surgical resection due to the presenceof metastases after radiological staging. There were 696 non-screening detected cancersduring the same time period. For the purposes of this analysis, polyp cancers (cancer thatwas removed by endoscopic means at the time of colonoscopy/flexible sigmoidoscopy) wereincluded in Duke's stage A, except for one polyp cancer that required subsequent surgicalresection and was staged as Duke's C1. The results are shown in table 1. Three-quarters ofcancers diagnosed in the BCSP were Dukes A or B, compared to 44.1% in the non-screeningpopulation. Of Duke's D cancers, only 2.8% were diagnosed through screening, with 27%diagnosed in the non screening population (p ,0.0001). Conclusion: This data stronglysupports significant stage shift of colorectal cancer even within the initial prevalent roundin this single Bowel Cancer Screening centre in Wales that the benefits of screening maybe demonstrable in outcomes at a relatively early stage of the program.Table 1. Stage of cancer diagnosed with comparable data from non-screening population

Su1055

Similar Clinical Characteristics of Familial and Sporadic Inflammatory BowelDiseaseSooK Hee Chung, Soo Jung Park, Lee Hye Sun, Jae Hee Cheon, Sung Pil Hong, Tae IlKim, Won Ho Kim

Background: Inflammatory bowel disease (IBD) is caused by complex interaction betweengenetic and environmental factors. In spite of several studies of familial IBD, we do not havesufficient information about the clinical characteristics in familial IBD. The aim of this studyis to investigate differences of clinical characteristics and disease courses between familialand sporadic IBD patients. Methods: We obtained clinical data on Crohn's disease (CD, 691cases) and ulcerative colitis (UC, 1113 cases) in Severance Hospital, Yonsei University Collegeof Medicine from Jan. 2005 to Feb. 2012. Seventeen patients (2.5%) with CD and 27 patients(2.4 %) with UC identified to have familial history of IBD. For each control case, 3 timesage- and sex-matched CD and UC patients who had no family history of IBD were randomlyselected based on the electronic medical data. We compared the clinical characteristics anddisease course of familial CD/UC patients with sporadic CD/UC patients, respectively. Results:There were no significant differences of age at diagnosis, age at symptom onset, main symptomat diagnosis (abdominal pain, diarrhea, weight loss, fever, abdominal mass, hematochezia),extraintestinal manifestation, location, behavior of disease disease activity, number of hospi-talization, number of operation, operation type, causes of operation and operation rate,number of relapse, medical treatment with 5-ASA, oral steroid, and azathioprine betweenfamilial and sporadic CD and familial and sporadic UC patients, respectively. Median (min-max) follow up peroids after diagnosis of familal CD patients and sporadic CD patients were84 (24-312) months and 36 (8-240) months, respectively (P=0.008). But there was nosignificant difference of follow up peroids after diagnosis of familial UC patients and sporadicUC patients. Familial CD patients more frequently used anti TNF agents than sporadic CDpatients (3/17, 17.6 % vs 0/21, 0 %, P=0.014). Longer follow up period of disease mightresult in more usage of anti-TNF antibody in familial CD patients than sporadic CD patients.

AG

AA

bst

ract

s

AG

AA

bst

ract

sConclusion: Clinical characteristics between familial IBD and sporadic IBD patients do notseem to be significantly different. A family history of IBD does not seem to be an importantpredictive factor affecting clinical characteristics or disease course. The duration of diseasemight be more important factor predictive of disease course rather than the familial historyof IBD. Further large scaled prospective studies are warranted to compare clinical characteris-tics and disease course of familial IBD with sporadic IBD patients.

Su1056

Low Utilization of Surveillance Colonoscopy Among Medicare Patients WithInflammatory Bowel Disease Prior to Colorectal Cancer DiagnosisYize R. Wang, John R. Cangemi, Edward V. Loftus, Michael F. Picco

Background: Patients with inflammatory bowel disease (IBD) have an increased risk forcolorectal cancer (CRC). One prior study showed low utilization of surveillance colonoscopyin patients with 8 or more years of ulcerative colitis (UC) in an integrated health care deliverysystem (Velayos FS et al, Gastroenterology 2010;139:1511-8). Aims: To study the use ofsurveillance colonoscopy among Medicare beneficiaries with IBD in the 2-year period priorto their CRC diagnosis. Data and Methods: Our study sample included all Medicare beneficia-ries in the SEER-Medicare linked database who were diagnosed with CRC during 2001-2005 and had 3 or more physician visits with an ICD-9 diagnosis code for IBD (Crohn'sdisease 555.x; UC 556.x) prior to their CRC diagnosis. Medicare beneficiaries over 85 yearsold, without Part B coverage or enrolled in HMOs were excluded. Colonoscopy performedwithin 6 months of CRC diagnosis was defined as diagnostic colonoscopy and excluded.Colonoscopy performed in the 2-year period of 6 to 30 months prior to CRC diagnosis wasdefined as surveillance colonoscopy. The chi-square test and multivariate logistic regressionwere used in statistical analysis. Results: Of the 241 Medicare beneficiaries with IBD anddiagnosed with CRC during 2001-2005, 92 (38.2%) patients underwent at least 1 surveillancecolonoscopy within the 2 years prior to their CRC diagnosis. The use of surveillancecolonoscopy was similar between Crohn's disease (28/86, 32.6%) and UC (64/155, 41.3%)(p=0.18). In multivariate logistic regression, factors positively associated with utilization ofsurveillance colonoscopy included younger age and personal history of colon polyps (bothp,0.05). Conclusions: Utilization of surveillance colonoscopy was low among Medicarebeneficiaries with IBD in the 2-year period prior to their CRC diagnosis.

Su1057

The Burden of Comedication Among Inflammatory Bowel Disease PatientsJessie P. Buckley, Michael Kappelman, Jeffery K. Allen, Susan A. VanMeter, Suzanne Cook

Background: Polypharmacy increases the risk of drug interactions and is of growing concernin the chronically ill, including patients with inflammatory bowel disease (IBD). Furthermore,the use of non-IBD medications is an indirect measure of comorbidity and disease burden.We aimed to characterize the prevalence and predictors of non-IBD medication use, and tocompare use of these medications among IBD patients to the general population. Methods:Using health insurance claims data contained within the Thomson Reuters MarketScandatabase, we conducted a cross-sectional study of commercially-insured individuals withcontinuous enrollment in Medicare or an employer-sponsored health plan between January2009 and December 2010. IBD patients were identified through diagnosis codes and IBDmedication dispensings, and matched to 5 individuals without a diagnosis of IBD on thebasis of age, sex, and region. We estimated the prevalence of dispensed prescriptions foranalgesics (narcotics; non-narcotics), psychiatric drugs (anxiolytics, hypnotics, and sedatives;antidepressants), and other commonly used drug classes as defined by the Anatomic Thera-peutic Classification system. Predictors of non-IBD medication dispensings and comparisonsof drug use between IBD patients and the general population were evaluated by logisticregression. Results: The 10 most common drug classes dispensed were similar amongIBD patients and the general population (Table 1). However, drug class dispensings wereproportionally higher among IBD patients compared to the general population [narcoticanalgesics (48.1% versus 34.1%); non-narcotic analgesics (12.8% versus 8.1%); anxiolytics,hypnotics, and sedatives (25.8% versus 16.7%); and antidepressants (28.3% versus 19.4%)].Crohn's disease patients were more likely to be prescribed analgesic and psychiatric drugsthan ulcerative colitis patients. Use of IBD maintenance medications was associated withreduced prescriptions for analgesic and psychiatric medications. The prevalence of non-narcotic analgesic and psychiatric drug use was higher among women whereas narcotic usewas more common among men. In addition, gastrointestinal surgery and hospitalizationwere strong predictors of narcotic use. Conclusions: Patients with IBD, particularly Crohn'sdisease, have increased use of analgesic and psychiatric medications. Use of IBD maintenancemedications is associated with reduced use of these agents. IBD providers should be awareof polypharmacy and its potential for drug interactions. Furthermore, use of these medicationsmay be a useful surrogate for disease activity and/or co-morbidity, and should be adjustedfor in pharmacoepidemiological studies.

S-388AGA Abstracts

Su1058

Clinical Epidemiological Analysis of Co-Infection At Acute Diarrheal DiseaseAccelerates the Controversy of Bacteria Bearing Histo-Blood Group Antigen(HBGA) Leading to Viral Infection SusceptibilityHitoshi Nakajima, Sho Kijima, Hidenori Kido, Tadashi Maeda, Toshiyasu Watanabe, TaitoMiyazaki, Yoshihisa Urita, Motonobu Sugimoto

[Background] There have been many reports that some people have genetically determinedreceptor, they presents some susceptibility to Norovirus (NV). Histo-blood group antigens(HBGAs) is supposed to play a role of NV receptor and leads to susceptibility of NV. Westart this study formulating a hypothesis that co-infection in the gastrointestinal tract couldbe related to the mechanism except genetics. [Subjects & Methods] We recruited total 261cases of acute gastroenteritis who all presented diarrhea symptom since Jan 2011, aged from14 to 83 years. A male/ female ratio 148:113. Patients' stool sample were collected andexamined for antigens of Norovirus (NV), Rotavirus (RV), and Adenovirus (AV) by thecommercialized kits. At the same time, stool culture for pathogenic bacteria got started usingsame stool sample. Single infection (each bacteria, each virus), co-infection between virusand virus, and co-infection between virus and bacteria were statistically analyzed. [Results]Fifty six per 261 subjects (21.5 %) were positive for only virus, 75/261(28.7 %) were positivefor only bacteria. Both of bacteria and virus were found in 33/261(12.6 %). Detected bacteriawere pathogenic Escherichia coli (Ec) species, Campylobacter species (Cp), Salmonella specie(Sa) etc, and Ec species were most frequently detected 61/261 (15.8%). Mean ages of infectedpatients were 50±24.6, 36.3±15.8, 38.6±18.3, respectively. RV infection 80/261 (30.7%,co-infection 28.8%), AV infection 78/261 (29.1%, co-infection 29.4%), and NV infection51/261(19.5%, co-infection 25.4%) were observed. NV co-infection with bacteria were 10/51 (19.6%) with Ec, 5/51(9.8%) with Ca. Statistical analysis revealed co-infection of bacteriaand NV is not significant. Referring to ABO histo-blood group type and NV infection,frequency was not significantly increased although the cumulated number of O group patientswas largest among ABO blood groups. . [Conclusion] The co-infection of bacteria and virusoccurred frequently at gastrointestinal tract, and this result suggests that there could beanother pathway leading to an infectious diarrheal disease except a correlation between aperson's genetically determined HBGAs expression and their susceptibility to enteric viruses,especially NV.

Su1059

Findings on Surveillance Colonoscopy in Patients With Serrated Lesions ofthe ColonYuk-Fai Lam, Wai-Kay Seto, Teresa Tong, Milky Tang, Wai K Leung

BACKGROUND: Serrated lesions of the colon are now believed to be the precursors ofabout one-third of colorectal cancer. However, the optimal surveillance interval for patientswith different types of serrated lesions of the colon is not well defined. AIM: To determinethe prevalence of and time to recurrent colonic polyps/adenoma in patients with differentserrated lesions of the colon. METHODS: We identified patients who were diagnosed tohave serrated adenoma and hyperplastic polyp ≥5mm in diameter between January 2008and June 2011 in our colonoscopy database. All patients were invited to have surveillancecolonoscopy. Patients were divided into four groups according to the baseline serratedlesions: serrated adenoma alone (SA), large ( ≥10mm) hyperplastic polyps alone (LHP),medium-sized (5-9mm) hyperplastic polyps alone (MHP) and patients with concurrentserrated lesions and adenoma in the colon (CSA). The proportion of patients and the timeto recurrent colonic polyps/adenoma on surveillance colonoscopy were determined accordingto their baseline serrated lesions. RESULTS: 186 patients with baseline serrated lesions wereidentified. Surveillance colonoscopy was completed in 105 (56.5%) patients with baselineserrated lesions (15 SA, 9 LHP, 43 MHP and 38 CSA). The median time to surveillancecolonoscopy was 33.7 months. The median age of the four groups were: 59 (range 34-83)in the SA group, 74 (42-76) in the LHP group, 55 (26-81) in the MHP group and 68 (42-82) years in the CSA group, respectively (P ,0.001). The proportions of patients withrecurrent colonic polyps (serrated lesions and adenoma) on surveillance colonoscopy were46.7% (SA), 55.6% (LHP), 30.2% (MHP) and 60.5% (CSA), respectively (P = 0.049). Therewas a significant difference on the time to polyps recurrence among the four groups asshown in the Figure, with the lowest rate in the MHP group (Log Rank; P = 0.006). However,the respective proportion of patients found to have colonic adenoma on surveillance was20.0% (SA), 33.3% (LHP), 25.6% (MHP) and 52.6% (CSA) (P =0.04). A significant differenceon time to adenoma detection was also demonstrated (Log rank, P = 0.009), with the highestproportion in the CSA group. CONCLUSION: Patients with baseline SA and LHP alonehave a high polyp recurrence rate on surveillance colonoscopy, but the adenoma detectionrates of these patients on surveillance colonoscopy are lower than those with concurrentadenoma at baseline. These findings may have implications on the optimal screening intervalfor patients with different baseline serrated lesions.