Orogastric or nasogastric (NG) tube feedings are required by as many as 20% of hospitalized children each day, and an increasing number of children are receiving NG tube feedings at home, where an initially inexperienced family member may need to reinsert the tube at each feeding. Numerous studies have shown the risk of misplacing NG tubes, and factors that predispose to tube misplacement are especially prevalent in children. However, the methods that appear in the literature for predicting the insertion-length are based on relatively untested assumptions, and the relative accuracy of the various prediction methods are not known. The objective of the proposed research is to identify external measurements for predicting the insertion lengths of orogastric and nasogastric gavage (NG) tubes in children that give clinically accurate predictions virtually all of the time and that are reliable and simple enough for use both by hospital and home care nurses and by parents and other home care providers. The proposed research will determine the relationship between the external distances recommended in the literature and the actual distance to the stomach measured in endoscopy and esophageal manometry subjects. The research will also investigate the fit between the external distances and the distance to the floor of the stomach at the greater curvature in children undergoing endoscopies. The data will be collected at two major medical centers to ensure that the sample size is large enough to represent the age range in children 1 month to 18 years of age. The difference in nasal and oral insertions and in the internal endoscopy and manometry measurements will be compared, and the reliabilities of the measurements and the tendency for rater differences will be determined. The research will test the assumption that a single method will give close predictions in virtually all children and will identify ways to improve the predictions. The predictive adequacy of the proposed methods will be compared in a separate validation sample to be collected in the fourth year, and the probable position of tubes inserted the predicted distances will be described. Regression modeling and regression model validation methods will be used in the analyses. This research will alow clinicians to determine which prediction method is likely to give the closest predictions for a given child. The clinician will also know the normal range of errors and the chance of making a gross in error in the insertion-distance using each method.
