Search

Primary biliary cholangitis, formerly known as primary biliary cirrhosis, is a chronic, autoimmune, and cholestatic disease ameliorating the biliary epithelial system causing fibrosis and end-stage liver disease, over time. Patients range from an asymptomatic phase early in the disease course, to symptoms of decompensated cirrhosis later in its course. This review focuses on the current consensus on the epidemiology, diagnosis, and management of patients with primary biliary cholangitis. We also discuss established medical management as well as novel and investigational therapeutics in the pipeline for management of PBC.

BACKGROUND & AIMS The rising prevalence of obesity and diabetes mellitus (DM) among hepatitis C virus (HCV) patients contributes to concurrent nonalcoholic fatty liver disease (NAFLD). We aim to evaluate the impact of concurrent obesity or DM on waitlist survival and probability of liver transplantation (LT) among adults with chronic HCV awaiting LT. METHODS Using 2003-2013 United Network for Organ Sharing data, we evaluated the impact of obesity and DM among adults with chronic HCV awaiting LT: non-obese, non-DM vs. obese, non-DM (obese) vs. non-obese, DM (DM) vs. obese and DM. Overall, LT waitlist survival and probability of receiving LT were evaluated using Kaplan-Meier and multivariate logistic regression models. RESULTS From 2003-2013, there were 43 478 new LT waitlist registrants with chronic HCV (21.0% with HCC, 79% without HCC). Obesity was associated with lower probability of receiving LT (OR, 0.91; 95% CI, 0.85-0.97; P < 0.01), and lower probability of waitlist mortality (OR, 0.80; 95% CI, 0.72-0.89; P < 0.001) when compared to non-obese patients. DM among HCV patients did not impact probability of waitlist survival or receiving LT. When evaluating post-LT survival, compared to non-obese, non-DM patients, obese HCV patients had significantly lower post-LT mortality (HR 0.86; 95%CI, 0.81-0.92; P < 0.001); whereas, HCV patients with DM had significantly higher post-LT mortality (HR, 1.22; 95% CI, 1.12-1.33; P < 0.001). CONCLUSION Among adults with chronic HCV awaiting LT in the US, obesity is associated with lower probability of receiving LT, but did not impact waitlist survival. DM among chronic HCV patients did not impact waitlist survival or probability of LT.

Rising rates of obesity, diabetes mellitus (DM), and nonalcoholic fatty liver disease (NAFLD) among chronic hepatitis C (HCV) patients may contribute to higher hepatocellular carcinoma (HCC) risk. To perform a systematic review evaluating the impact of DM, body mass index (BMI), or steatosis on HCC risk among chronic HCV patients. A structured keyword search of PubMed from January 1, 2001, to July 1, 2014, was performed to identify original articles evaluating the association of DM, BMI, or steatosis with HCC among adults with chronic HCV. Studies involving HCV patients co-infected with human immunodeficiency virus, hepatitis B virus, or other chronic liver diseases with the exception of NAFLD were excluded. Quality assessment utilized the Newcastle–Ottawa scale. Nine studies (seven cohorts, two case–controls) met inclusion criteria for the final analysis. Five of seven studies analyzing DM demonstrated significantly increased HCC risk associated with concurrent DM with effect sizes ranging from HR 1.73 (95 % CI 1.30–2.30) to RR 3.52 (95 % CI 1.29–9.24). One of three studies analyzing BMI demonstrated a significant association with HCC risk (BMI ≥ 30.0 vs. BMI

GOALS:: To identify an association between prior weight loss surgery (WLS) and acetaminophen-induced acute liver failure (ALF). BACKGROUND:: WLS, which has increased in proportion to the global rise of obesity, alters the absorption and metabolism of many drugs including acetaminophen (APAP) and may predispose to toxicity. No study has identified an association between prior WLS and APAP-ALF. STUDY:: We retrospectively reviewed a cohort of patients who presented to our center with ALF. We identified 101 patients who presented to our center with ALF between January 2009 and December 2011. All patients were prospectively enrolled into a database using consensus criteria. A history of WLS was obtained through a retrospective chart review. RESULTS:: Fifty-four patients (53.5%) had APAP-ALF and 47 (46.5%) had ALF caused by other etiologies. A prior history of WLS was present in 9 of the 54 patients with APAP-ALF versus 0 of the 47 with non-APAP-ALF (P=0.003). Patients with APAP-ALF and prior WLS did not have higher rates of factors commonly associated with APAP overdose, including depression, alcohol abuse, intent to cause self-harm, or use of APAP-narcotic combination drugs. CONCLUSIONS:: A history of WLS may predispose to hepatotoxicity and ALF caused by acetaminophen. Language: en

The impact of obesity on survival following liver transplantation is unclear, and existing studies report conflicting results. Our current study aims to further delineate the impact of obesity using population-based registry data from the USA.All US adult liver transplant recipients from 2003 to 2012 were evaluated using the United Network for Organ Sharing registry. The impact of obesity on survival following liver transplantation was further stratified into class I obesity [body mass index (BMI) 30.0-34.9 kg/m(2)], class II obesity (BMI 35.0-39.9 kg/m(2)), and class III obesity (BMI ≥ 40 kg/m(2)) and evaluated using Kaplan-Meier methods and multivariate Cox proportional hazards models.Overall, 57,255 patients with chronic liver disease underwent liver transplantation, among which 32.9 % had BMI ≥ 30 kg/m(2). While patients in all obesity classes had similar survival to patients with BMI 18.0-24.9 kg/m(2), the presence of concurrent diabetes mellitus resulted in significantly lower post-transplant survival. After multivariate regression, post-transplant survival in patients with class II obesity (HR 0.97; 95 % CI 0.89-1.05) or class III obesity (HR 0.99; 95 % CI 0.90-1.09) was not significantly lower than patients with BMI 18.0-24.9 kg/m(2), but diabetes mellitus was independently associated with lower post-transplant survival (HR 1.29; 95 % CI 1.21-1.36).In conclusion, obesity alone was not associated with lower post-transplant survival. However, DM, either alone or comorbid with obesity, is associated with significantly greater post-transplant mortality.

Background & Aims: A number of factors have been identified that influence the yield of screeningcolonoscopy. The perceived tolerability of bowel preparation has not been studied as a predictor of quality outcomes in colonoscopy. We aimed to characterize the association between patient-perceived tolerability of bowel preparation and polyp detection during colonoscopy.Methods: We performed a cross-sectional cohort study of 413 consecutive adult patients presenting foroutpatient colonoscopy at two outpatient endoscopy centers at our institution. We developed a standardized questionnaire to assess the patient's experience with bowel preparation. Bowel preparation quality was measured using the validated Ottawa scale and colonoscopic findings were recorded for each patient. The primary outcome was polyp detection and the secondary outcome was the quality of bowel preparation.Results: Patient-reported clarity of effluent during bowel preparation correlated poorly with Ottawa score during colonoscopy, k=0.15. Female gender was an independent risk factor for a poorly tolerated bowel prep (OR 3.93, 95% CI 2.30 - 6.72, p

A young woman presented with a febrile illness in the third trimester of pregnancy. Laboratory investigation revealed severe acute hepatitis with thrombocytopenia and coagulopathy. Liver injury progressed despite emergent caesarian section and delivery of a healthy infant. Therefore, therapeutic plasma exchange (TPE) was performed on three consecutive days post-partum for a presumed diagnosis of acute liver failure (ALF) associated with pregnancy due to hemolysis, elevated liver enzymes, and low platelets (HELLP) or acute fatty liver of pregnancy (AFLP). Treatment with TPE was followed by biochemical and clinical improvement but during her recovery herpes simplex virus type 2 (HSV-2) infection was diagnosed serologically and confirmed histologically. Changes in the immune system during pregnancy make pregnant patients more susceptible to acute HSV hepatitis, HSV-related ALF, and death. The disease is characterized by massive hepatic inflammation with hepatocyte necrosis, mediated by both direct viral cytotoxicity and the innate humoral immune response. TPE may have a therapeutic role in acute inflammatory disorders such as HSV hepatitis by reducing viral load and attenuating systemic inflammation and liver cell injury. Further investigation is needed to clarify this potential effect. The roles of vigilance, clinical suspicion, and currently accepted therapies are emphasized.

Rising rates of obesity, diabetes mellitus (DM), and nonalcoholic fatty liver disease among patients with chronic hepatitis C virus infection (HCV) may contribute to more rapid disease progression. To evaluate the impact of concurrent obesity, DM, and steatosis on disease progression among HCV patients. A systematic review using structured keyword search of MEDLINE and EMBASE from January 1, 2001, to July 1, 2014, was performed to identify original articles evaluating the association of obesity, DM, and steatosis with advanced fibrosis (AF) among adults with chronic HCV. Studies involving HCV patients coinfected with human immunodeficiency virus, hepatitis B virus, hepatocellular carcinoma, or other chronic liver diseases were excluded. Quality assessment utilized Newcastle–Ottawa Scale. Twenty cohort studies met inclusion criteria for analyses. Obesity was associated with increased risk of AF in seven studies with effect size ranging from OR 1.08 to 7.69. However, four studies did not demonstrate a significant association between obesity and AF. The presence of advanced steatosis among HCV patients was associated with increased risk of AF in 12 studies (OR 1.80–14.3). Concurrent DM was associated with increased risk of AF in six studies (OR 2.25–9.24). Thirteen studies were good quality, and seven studies were fair quality. Concurrent DM and steatosis are associated with increased risk of AF among chronic HCV patients. The majority of studies demonstrated significant associations of obesity with AF. Targeted interventions to optimize management of obesity-related diseases among HCV patients may help mitigate HCV disease progression.

Oxidative stress is increased in patients with metabolic syndrome (MS). Antioxidants, including carotenoids, are decreased in MS. We hypothesized that a low skin carotenoid score (SCS), calculated using resonance Raman spectroscopy, would correlate with the presence of MS. We retrospectively reviewed consecutive patients referred for dietary assessment between 2010 and 2012. For each patient, a nutrition history, medical history, and SCS were recorded. χ 2 and Student t test were used to determine factors associated with MS. Multivariate logistic regression was used to identify factors associated with MS. One hundred fifty-five patients were included. The mean age was 54.1 ± 13.1 years, and the mean body mass index was 28.3 ± 6.1 kg/m 2 . Metabolic syndrome was present in 43.9% of patients. The mean SCS was 28 084 ± 14 006 Raman counts (RC), including 23 058 ± 9812 RC for patients with MS and 32 011 ± 15 514 RC for patients without MS ( P = .0001). In a multivariate analysis, SCS less than 25 000 RC (odds ratio, 3.71; 95% confidence interval, 1.36-10.7; P = .01) was independently associated with MS. A higher number of MS components was associated with a progressively lower SCS ( P = .004). In a consecutive sample of patients referred for dietary assessment, a noninvasively measured SCS was lower among patients with MS.

The ICRP has recently recommended that the occupational exposure limit for the lens of the eye be reduced to 20 mSv in a year, averaged over defined periods of 5 years, with no single year exceeding 50 mSv. There has been concern amongst some groups of individuals, particularly interventional cardiologists and radiologists as well as relevant professional bodies, that implementation of these recommendations into UK law will adversely affect working patterns. However, despite a number of informative European studies, there is currently little UK dosimetry data available upon which judgements can effectively be based. In order to address this knowledge gap, Public Health England has carried out a small, targeted survey of UK lens doses to medical staff undertaking procedures likely to involve the highest levels of radiation exposure. Two out of a total of 61 individuals surveyed had projected annual doses which could be close to 20 mSv, measured outside lead glasses. Use of protective equipment was generally good; however, lead glasses were only used by 9 participants. The results of this survey suggest that compliance with the ICRP recommendations is likely to be possible for most individuals in the UK medical sector.

Purpose Same session endosonography (EUS) immediately prior to scheduled endoscopic retrograde cholangiopancreatography (ERCP) may eliminate the need for ERCP and its associated risks in pregnant patients with no evidence of choledocholithiasis on EUS. In patients with choledocholithiasis, EUS provides information regarding the location, size and number of stones present, which helps guide biliary interventions and confirm stone clearance without the use of fluoroscopy. Methods We retrospectively identified 10 pregnant patients referred to our tertiary endoscopy center for suspected choledocholithiasis between June 2008 and January 2012. All patients underwent same-session EUS-based ERCP. Results Of 10 pregnant patients managed with EUS-guided ERCP, six were found to have common bile duct stones and went on to ERCP. Four patients with no evidence of choledocholithiasis on EUS did not undergo ERCP. Patients with confirmed choledocholithiasis underwent ERCP without the use of fluoroscopy using the additional information provided by EUS. Conclusions Same-session EUS immediately prior to scheduled ERCP may eliminate the need for ERCP and its risks in pregnant patients with no evidence of choledocholithiasis on EUS. In patients with confirmed choledocholithiasis, EUS provided additional information regarding the location, number and size of bile duct stones, which enabled the successful clearance of the bile duct without the use of fluoroscopy.

A young woman presented with a febrile illness in the third trimester of pregnancy. Laboratory investigation revealed severe acute hepatitis with thrombocytopenia and coagulopathy. Liver injury progressed despite emergent caesarian section and delivery of a healthy infant. Therefore, therapeutic plasma exchange (TPE) was performed on three consecutive days post-partum for a presumed diagnosis of acute liver failure (ALF) associated with pregnancy due to hemolysis, elevated liver enzymes, and low platelets (HELLP) or acute fatty liver of pregnancy (AFLP). Treatment with TPE was followed by biochemical and clinical improvement but during her recovery herpes simplex virus type 2 (HSV-2) infection was diagnosed serologically and confirmed histologically. Changes in the immune system during pregnancy make pregnant patients more susceptible to acute HSV hepatitis, HSV-related ALF, and death. The disease is characterized by massive hepatic inflammation with hepatocyte necrosis, mediated by both direct viral cytotoxicity and the innate humoral immune response. TPE may have a therapeutic role in acute inflammatory disorders such as HSV hepatitis by reducing viral load and attenuating systemic inflammation and liver cell injury. Further investigation is needed to clarify this potential effect. The roles of vigilance, clinical suspicion, and currently accepted therapies are emphasized.

Colonoscopy is associated with a reduction in colorectal cancer (CRC) incidence and mortality [1], although this association may be limited to cancers arising in the distal colon [2]. Recently, a prospective randomized controlled trial showed that flexible sigmoidoscopy reduces CRC mortality [3], but a comparable prospective randomized trial for colonoscopy is currently in phase III and will not be completed until 2036 [4]. Nevertheless, the American Cancer Society and the US Preventive Services Task Force recommend screening colonoscopy every 10 years starting at age 50 for patients with average risk of CRC. Likewise, Medicare has covered a majority of the costs of this screening procedure since 2001 for patients at average risk. The molecular basis for the adenoma–carcinoma sequence in CRC is well described [5]. More recently it was proposed that a genetic pathway characterized by microsatellite instability (MSI) accounts for a number of sporadic CRC [6]. Many of these lesions arise from nonpolypoid lesions in the proximal colon [7]. Although proximal cancers occur less frequently than distal ones, they may more frequently result from a missed precursor lesion at colonoscopy, as an association between interval cancers and both the presence of MSI and proximal location has been reported [8]. This evidence represents a significant challenge to colonoscopists. Distal CRC represents the low-hanging fruit, so to speak, in that its precursor lesions are more numerous, more polypoid and reachable with a shorter, cheaper and safer procedure. Removal of a sufficient number of precursor lesions to significantly reduce mortality from proximal CRC may require a more consistent level of training and expertise than is currently reported in the literature. In some of the retrospective trials that failed to show a reduction in CRC mortality from proximal lesions removed at colonoscopy, gastroenterologists comprised a minority of the colonoscopists [2]. In an effort to further lower CRC mortality, many approaches to quality improvement in colonoscopy have been proposed including standards for withdrawal time, withdrawal technique, operator experience, optical enhancement and measurement of adenoma detection rate. However, none of these quality standards will succeed if the colonoscopist cannot consistently visualize an adequate amount of the colonic mucosa. In the end, the quality of the bowel preparation may determine whether screening colonoscopy, rather than flexible sigmoidoscopy, continues to be so universally recommended and reimbursed. It remains to be seen how to best administer bowel preparation before colonoscopy so that the colonoscopist has greatest opportunity to reduce CRC mortality. We have known for decades that both polyethylene glycol (PEG) and bile take approximately 1 h to reach the cecum [9, 10]. It is the goal of a bowel preparation, then, to deliver a sufficient quantity of fluid to the cecum to cleanse the entire colon, allow enough time for the cleansing agent to pass completely through the colon, but not wait so long that the proximal colon is again coated in a film of bilious secretions. With respect to the latter two goals, timing is everything. There have been a number of recent studies, including the present one, which investigated the relationship between timing and quality of bowel preparation [11–18]. In this issue of Digestive Diseases and Sciences, Eun et al. [11] draw an association between the quality of bowel preparation and the interval between the time of last PEG intake and the start of colonoscopy. In this single-center E. W. Holt (&) M. S. Verhille California Pacific Medical Center, San Francisco, CA, USA e-mail: edholt@gmail.com

BACKGROUND: Colorectal cancer (CRC) is the third most common cancer, as well as the third most common cause of cancer deaths in the United States. Bisphosphonates are antiresorptive agents widely used in treating and preventing osteoporosis. Several large studies have recently reported a reduced risk of CRC with bisphosphonate use, however no meta-analysis on the subject exists. AIM: We aimed to conduct a meta-analysis of existing observational studies and thus provide a quantitative estimate of the association between bisphosphonate use and the risk of CRC. METHODS: We followed the Meta-Analysis of Observational Studies in Epidemiology (MOOSE) guidelines in performing our systematic review. A search was conducted through Medline, PubMed, Embase, and Current

EUS-Based ERCP Reduces Risk Exposure in Pregnant Patients With Suspected Choledocholithiasis Sheba Vohra*, Edward W. Holt, Yasser M. Bhat, Janak N. Shah, Steve Kane, Kenneth F. Binmoeller Internal Medicine, California Pacific Medical Center, San Francisco, CA; Gastroenterology, California Pacific Medical Center, San Francisco, CA; Interventional Endoscopy Service, California Pacific Medical Center, San Francisco, CA Background: ERCP is routinely used to evaluate and treat suspected choledocholithiasis (CDL). However, in the pregnant patient, radiation exposure during ERCP poses a risk to the fetus. Endoscopic ultrasound (EUS) is highly sensitive in the detection of CDL, and is able to determine the size and number of stones. We evaluated the utility of EUS prior to ERCP in pregnant patients with suspected CDL. Aim: To determine if same-session EUS-based ERCP in pregnant patients with suspected CDL is associated with a reduction in ERCPrelated complications. Methods: We retrospectively identified pregnant patients referred to our center for suspected CDL between 2008 and 2011. EUS, using a curved linear array echoendoscope, was performed in all patients to confirm CDL and to determine the size and number of stones. Patients with confirmed CDL underwent ERCP, without fluoroscopy when possible. Deep selective bile duct cannulation was confirmed by aspiration of yellow bile and stone extraction was accomplished after sphincterotomy was performed. The number of extracted stones was counted and compared to the number of stones identified by EUS. Cholangioscopy was performed at the discretion of the endoscopist to confirm CBD clearance. Patient charts were reviewed to determine indications, procedure times, radiation use, clinical course and complications. Results: Ten pregnant women with suspected CDL were included in the study. The mean age was 30.7 years (range 20-40) and the mean gestation was 24 weeks (range 11-31). All 10 patients were symptomatic (acute abdominal pain 10, elevated transaminases 9, gallstone pancreatitis 3). Four patients had no evidence of stones at EUS and were managed conservatively with uneventful outcomes. Five patients with CDL and one patient with a pancreatic duct stone at EUS went on to same-session ERCP. In 4 of 5 patients with CDL, the number of stones extracted matched the number of stones seen at EUS. Cholangioscopy was used to clear the common bile duct in the patient with mismatched stone counts. The sixth patient with a pancreatic duct stone underwent stent placement with a 5 French stent. No fluoroscopy was used in 5 patients and a 1-second fluoroscopic image without radiograph was required in the patient with a pancreatic duct stone to confirm wire position prior to stent insertion. The average endoscopy time was 43 minutes for the ERCP group vs. 11 minutes for the EUS group (p 0.01). There were no complications in any of the cases. Conclusion: EUS prior to ERCP eliminated the need for ERCP and its attendant risks in 4 of 10 pregnant patients with suspected CDL. ERCP with stone extraction was successful without the use of fluoroscopy in all patients with bile duct stones. When available, same-session EUS should precede ERCP in the treatment of pregnant patients with suspected CDL. Sa1504 Endoscopic Ultrasound-Guided Fine Needle Aspiration and Biopsy (EUS-FNAB) Using a Novel 25-Gauge Core Biopsy Needle: Optimizing the Yield of Both Cytology and Histology Takuji Iwashita*, Yousuke Nakai, Jason B. Samarasena, Do Hyun Park, John G. Lee, Kenneth J. Chang H.H. Chao Comprehensive Digestive Disease Center, University of California, Irvine, Orange, CA Background: Novel 19 and 22-gauge core biopsy needles which feature a hollowed-out reverse bevel to trap core are now available. Based on initial publications on the 19-gauge core biopsy needle, the current “best practice” for high core acquisition is to utilize only 3-5 to-and-fro needle movements along with high suction (10cc). A new 25-gauge core biopsy needle (CBN25) has been developed. Whether this same FNAB technique is most effective for CBN25 is unknown. Aim: To determine the optimal FNAB technique to obtain core specimen and increase histological and cytological yields using CBN25. Patients and Methods: CBN25 was used for consecutive patients presenting with a solid lesion for EUS-FNA between May and October 2011. After EUS evaluation, the lesion was punctured using CBN25 with either one of two FNA techniques. In the first 3 months, 3-5 to-and-fro movements with 10cc-syringe suction (ST, standard technique) was used and in the latter 5 months 10-20 to-and-fro movements with minimal negative pressure by pulling the stylet slowly (MT, modified technique) was used. The specimen was expelled onto a glass slide and all visible core were placed in formalin after measurement of core specimen lengths. The remainder of the non-core specimen was submitted for cytology. Per protocol there was a 4-pass stopping rule which was salvaged using a standard 25 gauge FNA needle. Data was analyzed retrospectively. Results: 60 patients (21 female; median age 67.5 (range 35-90)) underwent EUS-FNAB using CBN25 with ST in 20 and MT in 40. Punctured lesions and final diagnoses are summarized in table 1 and 2. FNAB with CBN25 was successful in all cases. Median number of FNAB passes was 4 (range 3-4). A total of 236 passes using CBN25 was performed with ST in 80 passes and MT in 156 passes. The yield of visible core was 24% (19/80) in ST compared to 79% (123/156) in MT (P 0.001, t-test). The mean core length was 1.6 mm (SD, 1.7) in ST and 8.0 mm (SD, 9.0) in MT. (P 0.001). Among patients with a final diagnosis of a neoplastic lesion (59 patients), the yield of histological diagnosis was 55% (11/20) in ST as compared to 90% (35/39) in MT (P 0.01, Chi-squared test). The cytological yield showed a positive diagnosis in 70% (14/20) in ST as compared to 92% (36/39) in MT (P 0.05). Overall diagnostic yields were 75% (15/20) and 95% (37/39) in ST and MT, respectively (P 0.05). 7 patients had salvage FNA (4 positive, 3 false negative). No complications were seen in any of the patients. Conclusion: After technique modification, EUS-FNAB using CBN25 had a histologic yield of 90% and an overall diagnostic yield of 95%. This needle and technique appear optimum for EUS tissue acquisition.

that could only have been obtained from ER. Overall, 65 % of significant findings were missed on the RR when the endoscopy report was not available in the EMR at the time of the RR, as compared to 25% when the ER was accessible at the time of the RR (OR 0.179, p 0.01). Conclusions: 1. Radiology interpretation of post ERCP spot films does not guide decision-making during this dynamic test and adds little to the endoscopists’ read of the study. 2. The significant error rate in RR is likely due to an interpretation based solely on spot films, whereas endoscopists have access to dynamic intraprocedural fluoroscopic and endoscopic images. 3. Radiologists have fewer discrepancies in their reads when the endoscopy report is accessible in the EMR at the time of their reads. Sensitiviy of Radiology Read Findings

Complications of Gastrointestinal Endoscopy in 85,391 Procedures Alberto Espino*, Ximena Garcia, Macarena Mac-Namara, Hugo Richter, Fernando Pimentel, Francisco Biel, Camila Robles, Matias Callejas, Allan C. Sharp, Andres Donoso, Roberto Candia, Robinson G. Gonzalez, Nicolas Jarufe, Marco Arrese, Manuel Alvarez-Lobos, Oslando Padilla Gastroenterology, Pontifical Catholic University of Chile, Santiago, Chile; Digestive Surgery, Pontifical Catholic University of Chile, Santiago, Chile; The Digestive Endoscopy Center of University Clinical Hospital, Pontifical Catholic University of Chile, Santiago, Chile; Public Health, Pontifical Catholic University of Chile, Santiago, Chile Background: Complications are inherent to GI endoscopy (GIE) and do not necessarily imply endoscopist’s negligence. They may occur even using highest standards of practice. Objectives: To analyze the frequency and severity of complications occurring within 30 days after of the GIE at a single university hospital in Chile. Methods: We reviewed the records about patients who underwent GIE from January 2001 through May 2011. Results: A total of 85,391 GIE were evaluated. Procedures: 46,928 (55%) esophagogastroduodenoscopy (EGD); 27,993 (32.8%) diagnostic colonoscopies; 1427 (1.7%) polypectomies; and other procedures (hemostasis, variceal band ligation (VBL), foreign-body removal, dilation, stents, PEG, ERCP, EUS and double balloon endoscopy) 9043 (10.5%). A total of 299 complications were associated with GIE (59 % female, mean age 63 years, range 5 99). The overall complications rate was 0.35% (cardiopulmonary (CP) 0.1%, bleeding 0.07%, perforation 0.06%, infection 0.04%, pancreatitis 0.03% and other). The overall complication rate was higher in therapeutic procedures (TP) vs diagnostic procedures (DP) (2.7% v/s 0.16%, p 0.0001). The percentage of severe complications was higher in TP vs DP (52.3% vs 28.4%, p 0.0001). The overall complication rate for EGD was 0.14% (CP 0.07%, perforation 0.017%, bleeding 0.019%); diagnostic colonoscopy, 0.27% (CP 0.1%, perforation 0.06%, bleeding 0.02%); and polypectomy, 1.8% (CP 0.14%, perforation 0.28%, bleeding 0.98%). A total of 15 deaths occurred (overall rate 0.018%, 83% in TP). The overall mortality rate was higher in TP vs DP (0.2% v/s 0.003%, p 0.0001). The mortality rate for PEG was 0.7%; VBL 0.4%; ERCP 0.2%; diagnostic colonoscopy 0.004%; EGD 0.004%; and polypectomy 0%. Conclusions: GIE is associated with complications and mortality. The severity and risk of complications are higher in therapeutic procedures. These risks should be clearly explained to patients and their family before the procedure.