LAUSR

Abstract Background: Capsule endoscopy (CE) is a useful tool for patients with obscure gastrointestinal bleeding (OGIB), but positive finding rate differs among trials, which may be attributable to the difference in patients’ background. Objectives: To evaluate the predictive factors associated with positive findings on CE. Methods: Consecutive patients with OGIB who underwent CE between March 2004 and May 2015 at a single university hospital were enrolled. Patients’ clinical factors and CE data were reviewed retrospectively, and we evaluated the relationship between clinical factors and positive findings by univariate and multivariate logistic regression analyses. Results: Five hundred and seventy-eight patients were included in the analysis. Positive CE findings were obtained in 284 patients (49.1%). In multivariate analysis, low hemoglobin level (odds ratio (OR), 1.142 per 1 g/dL decrease; p < .001), Charlson comorbidity index (CCI) score (OR, 1.170 per 1 point increase; p = .002), and non-steroidal anti-inflammatory drug (NSAID) use (OR, 1.640; p = .044) were associated with an increased prevalence of positive findings. As for components of CCI, malignant tumor (OR, 1.839; p = .017) was associated with the positive findings. Conclusions: OGIB patient with a low-hemoglobin level, complex and severe comorbidities, and NSAID use should receive CE.