BACKGROUND AND AIMS: Narrow band imaging (NBI) endoscopy system enhances visualization of microvasculature and mucosal patterns. This study assessed the utility of NBI in patients with gastroesophageal reflux disease (GERD) symptoms. METHODS: Patients with and without GERD symptoms completed 2 validated GERD questionnaires prior to enrollment. The distal esophagus was examined by standard white light endoscopy followed by NBI. The features seen only by NBI were compared between GERD patients and controls. RESULTS: Overall, 80 patients (50 GERD, 30 controls) were eligible for final analysis (mean age, 58.4 years; males, 93.7%; white, 82.5%). A significantly higher proportion of patients with GERD had increased number (OR, 12.6; 95% CI: 3.7-42; P < .0001), dilatation (OR, 20; 95% CI: 6.1-65.3; P < .0001), tortuosity of intrapapillary capillary loops (IPCLs) (OR, 6.9; 95% CI: 2.5-19; P < .0001), presence of microerosions (P < .0001), and increased vascularity at the squamocolumnar junction (OR, 9.3; 95% CI: 1.9-43.6; P = .001) compared with controls. On multivariate analysis, increased number (OR, 5.5; 95% CI: 1.4-21.6) and dilatation (OR, 11.3; 95% CI: 3.2-39.9) of IPCLs were the best predictors for diagnosing GERD. The maximum, minimum, and average number of IPCLs/field were significantly greater in the GERD group compared with controls (P < .0001). Although the interobserver agreement for the various NBI findings was very good, the intraobserver agreement was modest. CONCLUSIONS: NBI endoscopy may represent a significant improvement over standard endoscopy for the diagnosis of GERD. These preliminary findings including inter- and intraobserver agreement need to be evaluated in future prospective, controlled, and blinded GERD trials.
BACKGROUND AND AIMS: Narrow band imaging (NBI) endoscopy system enhances visualization of microvasculature and mucosal patterns. This study assessed the utility of NBI in patients with gastroesophageal reflux disease (GERD) symptoms. METHODS:Patients with and without GERD symptoms completed 2 validated GERD questionnaires prior to enrollment. The distal esophagus was examined by standard white light endoscopy followed by NBI. The features seen only by NBI were compared between GERD patients and controls. RESULTS: Overall, 80 patients (50 GERD, 30 controls) were eligible for final analysis (mean age, 58.4 years; males, 93.7%; white, 82.5%). A significantly higher proportion of patients with GERD had increased number (OR, 12.6; 95% CI: 3.7-42; P < .0001), dilatation (OR, 20; 95% CI: 6.1-65.3; P < .0001), tortuosity of intrapapillary capillary loops (IPCLs) (OR, 6.9; 95% CI: 2.5-19; P < .0001), presence of microerosions (P < .0001), and increased vascularity at the squamocolumnar junction (OR, 9.3; 95% CI: 1.9-43.6; P = .001) compared with controls. On multivariate analysis, increased number (OR, 5.5; 95% CI: 1.4-21.6) and dilatation (OR, 11.3; 95% CI: 3.2-39.9) of IPCLs were the best predictors for diagnosing GERD. The maximum, minimum, and average number of IPCLs/field were significantly greater in the GERD group compared with controls (P < .0001). Although the interobserver agreement for the various NBI findings was very good, the intraobserver agreement was modest. CONCLUSIONS: NBI endoscopy may represent a significant improvement over standard endoscopy for the diagnosis of GERD. These preliminary findings including inter- and intraobserver agreement need to be evaluated in future prospective, controlled, and blinded GERD trials.