583 THE INTER- AND INTRA-OBSERVER RELIABILITY OF A NOVEL VIDEO-FLUOROSCOPY PROTOCOL FOR GASTRO-ESOPHAGEAL JUNCTION ANATOMY POST-FUNDOPLICATION
Abstract Despite modifications in surgical technique, the occurrence of troublesome dysphagia after fundoplication remains difficult to predict. Objective measurements on video-fluoroscopy swallow study (VFSS) may hold the key to understanding post-fundoplication dysphagia. This study reports the inter- and intra-observer reliability of video-fluoroscopic anatomical measurements of the post-fundoplication gastro-esophageal junction (GEJ). Methods Thirty-one patients underwent structured VFSS 6–12 months after laparoscopic total or partial fundoplication. VFSS protocol included six views: standing AP, 2x standing oblique, 2x prone oblique (POb), and prone oblique with continuous drinking. The primary observer recorded two sets of 11 variables of GEJ anatomy (3 mo interval between set 1- set 2). Further datasets (one each) were obtained from two medical students trained in two 2-hour sessions. Inter-observer reliability was determined from datasets of three observers and intra-observer reliability from primary observer. Intraclass correlation coefficients (ICC) two-way mixed-effects model was used (ICC agreement: 0.40–0.59 “fair”; 0.60–0.74 “good”; 0.75–1.00 “excellent”). Results ICC for inter-observer reliability was good-excellent in 47/66 measurements. All measures of maximal esophageal diameter, maximal esophageal diameter cf. wrap opening diameter and posterior distal esophageal angulation were good-excellent (ICC range 0.74–0.78; 0.84–0.91; 0.68–0.80 respectively). Four parameters recorded 83% good-excellent ICC: wrap closing diameter, minimal wrap diameter, GEJ anterior displacement and degree of axis deviation (GEJ cf. esophagus) (ICC range 0.50–0.85; 0.58–0.93; 0.56–0.79; 0.41–0.77 respectively). Absolute and axial wrap length showed low reproducibility: “good” ICC in 25% of measurements. POb measurements were most reliable: good-excellent ICC 86% of times. Intra-observer reliability was excellent in 98% of measurements (ICC range 0.74–0.99). Conclusion Following a structured protocol, VFSS measurements of GEJ post-fundoplication anatomy are reliable between observers except for measures of absolute and axial wrap length. Measures from the POb view showed the greatest inter-observer reliability and may have higher clinical utility. Further studies using structured VFSS are warranted to determine the correlation between troublesome dysphagia and anatomical changes of the gastro-esophageal junction post-fundoplication.