G A A b st ra ct s registry includes demographic data, baseline histology, endoscopic findings, date and number of treatment sessions, ablation outcomes, and complications. Eradication was considered complete (CEIM) if all esophageal biopsies obtained at least 12 months after enrollment demonstrated no intestinal metaplasia. Characteristics associated with CEIM on bivariate analysis (p<0.20) were included in a logistic regression model to identify independent predictors of CEIM. Parametric tests were used for statistical comparisons. RESULTS: Among 5,539 patients who received RFA for BE, 2166 (39%) currently have biopsies obtained 12 months or longer after enrollment. In this population, CEIM was achieved in 72%. Compared to patients who achieved CEIM, patients with residual BE were more likely in bivariate analysis to have: 1) dysplastic BE at baseline (70% vs 51%); 2) longer segment length (mean 5.9 vs 3.7 cm); and 3) required more RFA treatment sessions (3.3 vs 2.6). In multivariate analysis, dysplastic BE at baseline (OR 1.72, 1.40 2.11), greater segment length (OR 1.15 per additional cm, 1.12 1.19), and increased RFA session count (OR 1.14 per additional session, 1.07 1.22) were independently associated with incomplete eradication. The odds of failure to achieve CEIM was higher in advanced neoplasia (high-grade dysplasia, intramucosal carcinoma, invasive cancer) [OR 2.07, 1.62 2.64] compared to early neoplasia (indefinite dysplasia, low-grade dysplasia) [OR 1.61, 1.26 2.06]. Notably, likelihood for achieving CEIM was not influenced by race, sex, age, or treatment at a community-based, as opposed to academic practice. CONCLUSIONS: In the largest reported cohort of patients treated with RFA for BE, likelihood for complete eradication of IM was negatively associated with dysplastic BE at baseline, segment length, and number of RFA sessions. These findings should be considered in generating treatment protocols and counseling patients before treatment.