Testing Placement of Gastric Feeding Tubes in Infants

Abstract

Background Inadvertent positioning of a nasogastric tube in the lung can cause serious complications, so identifying methods to detect improperly inserted tubes is imperative. Objectives To compare the sensitivity, specificity, and negative and positive predictive values of 4 pH cut points (< 4.0, < 4.5, < 5.0, and < 5.5) in differentiating gastric and tracheal aspirates under various treatment conditions and to explore the utility of a pepsin assay for distinguishing between gastric and tracheal aspirates. Methods Gastric and tracheal aspirates were collected from critically ill infants undergoing mechanical ventilation who had nasogastric or orogastric feeding tubes. Aspirates were tested with colorimetric pH indicators and a rapid pepsin assay. Information about treatment conditions was obtained from medical records. Results Two hundred twelve gastric aspirates and 60 tracheal aspirates were collected from 212 patients. Sensitivity was highest and specificity was lowest at the gastric aspirate pH cut point of less than 5.5. Positive predictive values were 100% at all pH cut points less than 5.0. Negative predictive values were higher at the pH cut point of less than 5.0 than at cut points less than 4.5. A higher percentage of pepsin-positive readings was found in gastric aspirates (88.3%) than in tracheal aspirates (5.4%). Conclusion For a desired positive predictive value of 100%, a pH cut point of less than 5.0 provides the best negative predictive values, regardless of gastric acid inhibitor administration and feeding status. The pepsin assay is promising as an additional marker to distinguish gastric from tracheal aspirates.