Reply

Abstract

of the Journal. The article was difficult to follow and confusing to me in the documentation of gastroesophageal reflux disease (GERD) and in the selection and follow-up of patients. The article begs the question of whether the patients were studied consecutively and whether there were any patients with preoperative esophageal pH studies who died before their follow-up esophageal pH study could be performed after percutaneous endoscopic gastrostomy (PEG). The reader also needs to know how many of the authors' patients studied died in the first year after PEG.The study performed has serious problems in terms of methodology and use of esophageal pH monitoring. The placement of the esophageal pH probe 5 cm proximal to the lower esophageal sphincter (LES) and also being at the T7 to T8 level is impossible, unless all of their patients were of the same size and were the size of normal adults If the pH probe was truly positioned 5 cm proximal to the LES, then the probe was in different segments of the esophagus for small and large children. The reflux index and the DeMeester score for analyzing the esophageal pH recording are inaccurate in children for evaluating the presence of GERD because they both have a high false-negative rate (40% to 50%). 2 This means that the authors missed one half of the patients with actual GERD pre-and post-PEG. Furthermore, I have never seen, in my 25 years of doing esophageal pH monitoring, a patient whose GERD improved after PEG or Stamm gastrostomy, unless a lesser-curve gastrostomy was performed to simulate the effect of a Boerema gastropexy.The authors use the upper limit of normal of 5% for the reflux index. This upper limit of normal for the reflux index is for the entire 24-hour esophageal pH study. However, the authors seem to apply inappropriately this upper limit of normal of 5% to individual segments of the esophageal pH recording (ie, postprandial, daytime, nighttime). The authors should state the published studies that establish the normal range in children for the other parameters used to define GERD, such as "(1) greater than 50 episodes in which the pH is less than 4.0, (2) greater than 3 episodes in which pH was less than 4.0 for greater than 5 minutes during the 24 hour study, and (3) symptom index of 50% or more." I feel that this article's conclusions are misleading in the setting of an unclear definition and documentation of GERD provided by the authors. The study by Grunow et al 3 is a more reliable study of the risk for developing GERD in children post-PEG. The method of esophageal pH monitoring and analysis used by Grunow et al does not have the pitfalls associated with the analysis of GERD used by Samuel and Holmes.