3-minute Withdrawal Group (n[99) 6-minute Withdrawal Group (n[101) p- value Demographics and Patient Characteristics Age, years- Mean (std dev) 61.0 (8.97) 61.5 (8.4) 0.94 Gender- % male 65.7% 68.3% 0.69 Indication Screening, n(%) 61 (61.6) 68 (67.3) 0.40 Surveillance- history of polyps or cancer, n(%) 33 (33.3) 30 (29.7) 0.58 Family history of colorectal cancer, n(%) 4 (4.0) 5 (5.0) 0.73 Iron deficiency anemia, n(%) 3 (3.0) 4 (4.0) 0.70 Bright red blood per rectum, n(%) 2 (2.0) 1 (1.0) 0.56 Overall Adenoma Miss Rate 48.5% 21.8% 0.00 Adenoma Miss Rate by Location Right 51.4% 25.0% 0.01 Transverse 43.2% 17.1% 0.01 Left 54.6% 21.7% 0.02 Adenoma Miss Rate by Size !Z5mm 49.4% 24.7% 0.001 6-9mm 50.0% 20.0% 0.09 OZ10mm 25.0% 0.0% 0.10 230 Incidence of Advanced Neoplasia in Individuals With Untreated Diminutive Adenomas: a Longitudinal Study Yosuke Otake* 1 , Yasuo Kakugawa 1 , Minori Matsumoto 1 , Chihiro Tsunoda 1 , Taku Sakamoto 1 , Takeshi Nakajima 1 , Takahisa Matsuda 1 , Yutaka Saito 1 , Yukio Muramatsu 2 1 Endoscopy division, National Cancer Center Hospital, Tokyo, Japan; 2 Screening Technology and Development Division, National Cancer Center, Research Center for Cancer Prevention and Screening, Tokyo, Japan Background and Aim: The recommendations for surveillance colonoscopy intervals depend not only on the number and size but also on the histopathological findings of adenomas detected and treated. On the other hand, to reduce post-test referral and improve efficiency, computed tomographic colonography (CTC) or colon capsule endoscopy (CCE) for screening is recommended every 5 years, irrespective of whether these tests are negative or only diminutive polyps have been detected. However, the long-term outcome of patients with untreated diminutive adenomas is still unknown. This study aimed to determine the cumulative incidence of advanced neoplasia in asymptomatic patients who were found to have diminutive adenomas at initial screening colonoscopies but were not referred for removal. We also aimed to evaluate the suitability of this non-referral strategy by comparing these results to the incidence in patients not found to have adenomas at initial screening colonoscopies. Method: Asymptomatic patients who underwent initial screening colonoscopy be- tween February 2004 and March 2008 were included in this study. Follow-up colo- noscopy within 5 years was recommended for asymptomatic patients in whom diminutive adenomas ! 5 mm were detected at the initial screening colonoscopy but who were not referred for removal and for asymptomatic patients without ad- enomas. Appropriate therapy was recommended for patients in whom adenomas R 5 mm or cancer was detected. Advanced neoplasia was defined as an invasive cancer, a high-grade dysplasia, or a large adenoma R 10 mm. We retrospectively investi- gated the cumulative incidence of advanced neoplasia in patients without adenomas (Group A) and in those with adenomas ! 5mm (Group B) at initial screening co- lonoscopy. To precisely differentiate non-neoplastic lesions from neoplastic one without biopsy, we routinely used chromoendoscopy with magnification. Results: A total of 2,775 patients (2,070 in Group A and 705 in Group B) were included. The median colonoscopy interval was 60 months in both groups, and the cumulative incidence of advanced neoplasia was 1.7% (35/2070) in Group A and 2.8% (20/705) in Group B (p Z 0.059). Among the Group B patients, the cumulative incidence of advanced neoplasia was significantly higher in patients with R3 diminutive ade- nomas at the initial screening (10.7%; 7/65) than in patients with 1 or 2 diminutive adenomas at the initial screening (2.0%; 13/640). Conclusion: The 5-year risk of advanced neoplasia was not increased even when diminutive adenomas detected at screening colonoscopy had been untreated in patients with 1 or 2 diminutive ade- nomas. However, patients with R3 diminutive adenomas detected at screening colonoscopy that were left untreated had a significantly higher risk of neoplasia development. This approach can be applied to colorectal cancer screening using CTC or CCE. 295 Impact on Maximum Insertion Pain During Diagnostic and Screening Colonoscopy With on Demand Sedation: a Two Center Randomized Controlled Trial of Air Insufflation, Water Immersion and Water Exchange Sergio Cadoni* 1 , Stefano Sanna 2 , Paolo Gallittu 1 , Mariangela Argiolas 2 , Viviana Fanari 2 , Laura Porcedda 2 , Matteo Erriu 3 , Felix W. Leung 4,5 1 Digestive Endoscopy Unit, S. Barbara Hospital, Iglesias, Italy; 2 Digestive Endoscopy Unit, N. S. di Bonaria Hospital, San Gavino Monreale, Italy; 3 Department of Surgical Sciences, University of Cagliari, Cagliari, Italy; 4 Gastroenterology, Sepulveda Ambulatory Care Center, Los Angeles, CA; 5 David Geffen School of Medicine, University of California, Los angeles, CA Background: Support for unsedated colonoscopy best exemplifies investment in pa- tient-centered care. Maximum insertion pain limits success. When separately compared to air insufflation (AI), water-aided colonoscopy (WAC: WI, water im- mersion; WE, water exchange) significantly reduces pain. WE (water removal during insertion) and WI (water removal during withdrawal) are different and comparative effectiveness in lowering pain is unknown. Objectives: In 2 registered two-center RCTs (NCT01781650, 01780818) in diagnostic and screening patients we performed a head-to-head comparison of AI, WI and WE. We test the hypothesis that WE produces significantly lower pain scores than AI and WI. Patients: Consecutive diagnostic and screening patients, on-demand sedation based on patient request due to pain. Methods: AI, WI and WE were implemented as previously described. Maximum pain score [visual analog scale (0Znone, 10Zmax)] reported to an un- blinded observer was used to determine if minimal sedation (midazolam 2-5 mg) was needed on patient request. Recollected maximum pain was reported similarly at discharge to a blinded observer. Main Outcome Measurements: Maximum pain score during insertion. Results: See Tables. Demographic and baseline data (not shown) indicated even randomization. Based on similar protocol, the two data sets were combined. 576 patients were randomly allocated to AI (nZ193), WI (nZ197), or WE (nZ186). Mean maximum pain scores (95% CI) of insertion reported to unblinded observer during colonoscopy, and to blinded observer before discharge were: AI 4.1 (4.5-4.8), 1.9 (1.5-2.2); WI 3.3 (2.9-3.7), 1.8 (1.4-2.2); and WE 2.5 (2.2- 2.8), 1.3 (1.0-1.5) (P !0.0005s and 0.04s), respectively. Pearson correlation coef- ficient for these parameters was 0.6 (p !0.0005s). Correct patient guesses about method was 33%. Logistic regression shows method (WE), male gender and BMI as significant predictors of maximum insertion pain score during procedure. Secondary outcomes: proportion completing without sedation, less need for abdominal compression or loop reduction, and willingness to repeat were in favor of WAC methods. WE shows better colon preparation. Although intention-to-treat (ITT) cecal intubation rate favors AI (pZ0.038); final rates were comparable. WI had the shortest ITT cecal intubation time. Limitation: Unblinded colonoscopists. Conclusions: Patient blinding was adequate (only 33% correctly guessed method). The significant pain score correlation validated use of unblinded maximum inser- tion pain scores as primary outcome. WAC methods significantly reduce pain and sedation requirement without compromising procedural outcomes. WAC enhances acceptance of unsedated colonoscopy by patients (higher willingness to repeat without sedation). Uniquely, the unblinded and blinded maximum insertion pain scores of WE were significantly lower than those of AI and WI. AB126 GASTROINTESTINAL ENDOSCOPY Volume 79, No. 5S : 2014 www.giejournal.org Abstracts