Su1854 Inter-Rater Agreement of Functional Luminal Imaging Probe (FLIP) Topographic Measurements for Eosinophilic Esophagitis

Abstract

Purpose: Abnormal esophageal distensibility has been demonstrated in patients with eosinophilic esophagitis (EoE) when measured with the functional luminal imaging probe (FLIP). Although FLIP offers an objective measurement of esophageal cross-sectional area (CSA) with relation to pressure response, interpretation requires tedious manual analysis that may be altered by vascular, respiratory and esophageal contractile artifact. FLIP topography provides a methodology to improve the representation of the qualitative and anatomical variability of esophageal distensibility in EoE patients by utilizing filtering techniques so that the entire data set can be presented in a single image (Figure). The aim of this study was to assess the inter-rater agreement and accuracy of FLIP topography in trainees and experienced gastroenterologists. Methods: Esophageal body FLIP topography and pressure plots of 6 normal controls and 24 patients with histologically confirmed EoE were included in the study. Along with the 30 study FLIP topographic and pressure plots, a short teaching file was sent to 2 medical residents with experience with esophageal function testing, 2 recent fellow graduates, and 2 senior gastroenterologists. The teaching file contained 3 example FLIP topography plots with descriptions of the assessment criteria. The studies were scored by each rater for the following criteria: 1.) Normal Distensibility, defined as an index value (minimal CSA/ pressure) at maximal distention . 10 (Yes/No), 2.) Maximal Constricting CSA .250 mm2, 150-249 mm2, 75-149 mm2, or ,75 mm2. The Kappa statistic was utilized to assess for inter-rater agreement within each training group. The variability between all 6 raters was also computed by intraclass correlation coefficient. Accuracy was also determined by comparison with results obtained from an experienced esophagologist who was able to utilize both real-time imaging and the FLIP topography. Results: The inter-rater agreement using FLIP topography to assess esophageal distensibility and maximal constricting diameter was good to excellent within each training group and across all raters (Table). Accuracy was .90% among all raters for distensibility, but lower (60-83%) for maximal constricting CSA (Table). Conclusions: FLIP topography offers a method with high inter-rater agreement across various levels of clinical experience in assessment of mechanical esophageal properties of EoE patients. High accuracy of esophageal distensibility assessment appears feasible with only minimal orientation to FLIP analysis. Remaining inaccuracy likely stems from varying interpretation of the color-pressure representation that could be remedied with development of analytic software tools to increase objectivity of the topographic measurements. Table. Results of FLIP inter-rater agreement and accuracy