1178 Accuracy of Glucose Breath Testing for Small Intestine Bacterial Overgrowth (SIBO) Using Endoscopy Aspirate Cultures as a Reference Standard: A Meta-Analysis

Abstract

INTRODUCTION: Small Intestinal Bacterial Overgrowth (SIBO) is associated with abnormally high bacterial counts in the small intestine. SIBO is under-diagnosed and there are significant limitations with currently available testing methods. METHODS: Two research databases (Google Scholar and Web of Science) were used to find relevant published literature using keywords including glucose breath test accuracy, endoscopy aspirate culture and small intestine bacterial overgrowth. Eleven studies with data on patients tested with both a glucose breath test and endoscopy aspirate culture were found. Fixed-effects and random-effects models were fit to the eleven studies’ data using the Metafor (Meta-Analysis Package for R). The mean positive and negative percent agreement across studies is estimated as the weighted average of each study’s percent agreement, where the weights are the inverse of the standard errors, squared, plus the variance of the random study effects. RESULTS: The homogeneity of effects across studies was tested with a Chi-square test and homogeneity of effects was rejected for both positive and negative percent agreement (c2 = 91.5 and 107.5, respectively, df = 10, P-value < 0.0001) (Table 1). The mean percent agreement across studies was estimated with both the fixed-effects and the random-effects models (Table 2). Figure 1 presents each study’s estimate of the positive and negative percent agreement, respectively, and their associated 95% confidence intervals, as well as the overall estimate and its 95% confidence interval. CONCLUSION: Heterogeneity was found in the study designs and the implementation of breath tests such as the differences in dose of the substrate (50g or 75-100g) and the duration of the breath test (120∼240min). Meanwhile, the endoscopy aspirate culture is also not completely standardized, as the amount of liquid collected, the site of collection and the technical details of the microbiological tests may differ. This heterogeneity in practice of the two tests certainly contributes to the heterogeneity in positive and negative percentage agreement observed between the two tests. The random effects models’ estimates of 54% positive agreement and 76% negative agreement indicate moderate to poor agreement between the breath test and endoscopy aspirate culture. Given these limitations, there is an unmet need for novel tools to evaluate patients suspected of SIBO.