decrease as much as 50%. However, no reduction in procedure time was seen with EUSs requiring FNA. Indications Mean time attending only mins (95% CI) Mean time including fellow mins (95% CI) Percentage increase in time p value abdominal pain, CBD dilation, chronic/recurrent pancreatitis 8.6 (8.3-8.9) 19.0 (17.9-20.1) 120% 0.0001 gastric or esophageal submucosal lesions 10.9 (8.7-9.7) 18.8 (10.9-17.0) 73% 0.0001 esophageal and rectal cancer staging 6.56 (6.1-7.0) 12.1 (11.1-13.2) 84% 0.0001 celiac plexus block/ neurolysis 11.6 (9.5-13.8) 19.1 (13.6-17.5) 64% 0.0001 pancreatic cyst 11.3 (10.5-12.1) 21.4 (19.7-23.1) 90% 0.0001 pancreatic mass 27.5 (26.1-29.0) 38.6 (29.7-32.8) 40% 0.0001 mediastinal or intra- abdominallymphadenopathy 25.5 (23.5-27.4) 35.3 (28.1-31.5) 38% 0.0001 Su1306 Safety, Quality and Comfort of Upper Gastrointestinal Endoscopy With Conscious Sedation After Fasting for Two Hours Angelica T. Koeppe 2,3 , Marcio Lubini 3 , Karine T. Piazzetta 1 , Leticia Pereira 1 , Nilton M. Bonadeo 3,1 , Iran Moraes, Jr 3 , Julio C. Stobbe 3,1 , Fernando Fornari* 1,2 1 Sistema Digestório, School of Medicine, Universidade de Passo Fundo- RS, Brazil, Passo Fundo, Brazil; 2 Programa de Pós-Graduação: Ciências em Gastroenterologia e Hepatologia, UFRGS, Porto Alegre, Brazil; 3 Gastrointestinal Endoscopy, Hospital São Vicente de Paulo, Passo Fundo, Brazil Background: Upper gastrointestinal (GI) endoscopy is widely used in the clinical practice. An empty stomach is required to ensure quality and safety. For this reason, endoscopy is conventionally performed after 8 or more hours of fasting, which is potentially related to discomfort to the patients. In recent studies, surgical cholecystectomy has been described to be safe after two hours of fasting on liquids. Aims: To assess safety, quality and comfort of upper GI endoscopy after two hours of fasting. Methods: In this prospective study, patients who were referred for elective endoscopy and accepted to participate were randomly assigned to perform the procedure after fasting for 8 (F8) or 2 hours (F2). F2 patients were instructed to avoid food for 6 hours and drink 200 ml of a liquid protein supplement (300 Kcal) two hours before the endoscopy. Four skilled endoscopists carried out the endoscopies, blinded to patient fasting status. Safety was defined by the endoscopists as presence/absence of regurgitation of gastric contents into the esophagus following intubation, as well as occurrence of aspiration. Quality was graded by the endoscopists using a Likert scale (1 = poor10 = excellent), whereas comfort was assessed by questioning the patients about presence/absence of anxiety, discomfort, hunger, thirst, nausea, and weakness. Results: A total of 115 patients were interviewed. Of these, 17 (14.8%) were excluded: 8 denied to participate, 6 had unstable clinical conditions, 2 had morbid obesity, and one patient had anti-reflux surgery. Ninety-eight patients were studied (aging 48.5  16.5 years, 60.2% women): 50 (51%) were F2 and 48 F8. Regurgitation of gastric contents into the esophagus did not differ significantly between F2 and F8 (26% vs. 19%; P = 0.471). There was no case of aspiration in both groups. Quality was slightly but significantly higher in F8 compared to F2 [median (IQR25-75%): 9 (9-10) vs. 9 (7.75-10); P = 0.010]. In comparison to F8, a higher proportion of F2 patients classified the procedure as more comfortable in terms of anxiety (F2: 8% vs. 25%; P = 0.029), discomfort (18% vs. 42%; P = 0.010), hunger (44% vs. 67%; P = 0.024), and weakness (22% vs. 42%; P = 0.034). Conclusions: Our data suggest that upper GI endoscopy after two hours of fasting on liquids is as safe as the endoscopy with the conventional 8-hour fasting. It was classified by patients as more comfortable. However, endoscopists found it with a slightly lower quality. Further data and discussion are needed to decide whether this shorter fasting period can be assumed in the clinical practice of upper GI endoscopy. Su1307 Percutaneous Gastrostomy Tube Insertion by Gastroenterologists, Surgical Endoscopists, and Interventional Radiologists: Which Is Safer? Nyla Hazratjee* 2 , Markus Agito 3 , Raghu C. Vanama 1 , Rocio Lopez 1 , Maged K. Rizk 1 , John J. Vargo 1 , Prabhleen Chahal 1 1 Gastroenterology, Cleveland Clinic, Cleveland, OH; 2 Internal Medicine, Cleveland Clinic, Cleveland, OH; 3 Internal Medicine, Akron General Medical Center, Akron, OH Background: Percutaneous endoscopic gastrostomy tube insertion is a preferred ,less invasive method of gastrostomy tube placement. However,surgeons and interventional radiologists also routinely perform percutaneous gastrostomy tube(PGT)insertion.With escalating emphasis on quality control and reimbursement,a comparison of complications and outcomes in these three PGT placement groups is imperative. Aims: Our objective was to compare complications and outcomes of PGT insertion by gastroenterologists, surgical endoscopist, and interventional radiologist. Methods: All adult patients who underwent an initial PGT inserted at our institution within the period of 1/1/2009 to 1/1/2011 were identified, and categorized into 3 groups based on specialty of the performing physician: gastroenterology (GI), surgery (SG) or Interventional Radiology (IR). Early ( 14 days after PGT) and late (15 days to 6 months after PGT) complications of PGT including minor (dislodged tube, tube leaks, infection requiring wound care) and major (aspiration, bleeding, infection requiring antibiotics, peritonitis, peritonitis requiring laparotomy, perforation, and perforation requiring laparotomy and death) were identified. Descriptive statistics were computed for all variables. A univariable analysis and multivariable logistic regression were performed to assess differences in occurrence of complications between the 3 groups. All patients were followed up to 6 months after procedure. Results: A total of 790 patients (523 PGT by SG, 197 PGT by GI, 70 PGT by IR) were identified. There was no significant difference in age, gender, and BMI. Furthermore, coagulopathy, prophylactic use of antibiotics, and presence of anti-platelet and anticoagulants within 7 days of procedure were not different between the 3 groups. After adjusting for confounding variables, compared to those who had PGT performed by GI or SG, patients who had PGT by IR were more likely to have minor complications including early tube leaks(17.1% IR vs 1.3% SG vs 1.5% GI p.001) , late tube leaks( 8.6% IR vs 2.0% GI p=.027), and late local infections requiring wound care only(8.6% IR vs 2.1% GI p=.008) (Figure 1) and major complications including early medically managed peritonitis(1.4% IR vs 0% SG p=0.006) and early perforation requiring laparotomy(2.9% IR vs 0.19% SG p=.002) (Figure 2). Mortality within 6 months after procedure was significantly higher in patients who underwent PGT by IR (Figure 3). Conclusions: In this first and large US quaternary center study, PGT placement by interventional radiologists is associated with higher morbidity and mortality compared to PGT placement by gastroenterologists and surgical endoscopist. Su1308 Unexpected High Rate of Malignant Seeding by Percutaneous Endoscopic Gastrostomy (PEG) Implantation: Final Results of a Prospective Study? Mark Ellrichmann* 1 , Petr Sergueev 1 , Johannes Bethge 1 , Victor Kataev 1 , Alexander Arlt 1 , Perdita Wietzke-Braun 1 , Theodoros Topalidis 2 , Annette Fritscher-Ravens 1 1 Medical Department I, University Hospital Schleswig-Holstein, Campus Kiel, Kiel, Germany; 2 Institute of Cytology, Hannover, Germany Introduction: Insertion of a percutaneous endoscopic gastrostomy (PEG) is a standard procedure in many patients with malignancies of the head and neck and upper gastrointestinal (GI) tract to ensure enteral feeding. Due to the insertion technique, the gastric secure plate gets in contact with superficial tumor cells. Case reports and retrospective analyses report of metastases at the PEG insertion site of the abdominal wall, but no prospective studies are available. Since modern treatments with intention to cure are available survival rates are improving. Thus, late PEG complications gain significance. Aim: To prospectively and systematically quantify the rate of early and late malignant cell seeding at the abdominal wall pull-through site at the time of the PEG placement and to evaluate possible risk factors that predispose to malignant cell seeding. Methods: 42 consecutive patients with head and neck cancer (ENT) and esophageal malignancy were included in this prospective study. Standard pull-technique was used for PEG placement using a norm gastroscope. Immediately after the PEG procedure, brush cytology was taken from the PEG tubing as well as from the transcutaneous incision site. 3 months after PEG insertion a brush cytology was taken from the PEG outlet site at the abdominal wall. Cytology samples were rated according to the modified PAP (Papanicolaou)-classification. Results: 13 patients with esophageal-,, one with esophageal infiltration of lung cancer and 26 with ENT cancer received a PEG without any immediate complication. Subsequent cytology of the brushing of tube and incision sites proved malignant cells present on cytology in 9 of the 40 cases (22.5%). All of these cases were squamous cell cancer. There was no malignant seeding in adeno-carcinomas. Abstracts AB287 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 4S : 2012 www.giejournal.org