We read with great interest the study by Cheng et al,1 who investigated patients with gastroesophageal reflux disease (GERD) refractory to antisecretory therapy by means of endoscopy and impedance-pH monitoring offtherapy and concluded that roughly half of these patients referred for testing actually underwent investigations and received medications with no evidence of GERD; they were affected by functional heartburn, functional disorders other than heartburn, or by undetermined disorders. These data confirm previous studies2,3 on the importance of investigating refractory patients and also emphasize the need of stopping antisecretory therapy to reduce overuse of proton pump inhibitors.4 However, we believe that the decision taken by the authors of using the impedance-pH parameters (ie, number of reflux episodes) rather than the symptom association analysis for the categorization of nonerosive reflux disease patients requires further comments. Although the number of reflux episodes has been shown as a reproducible parameter in patients investigated twice with impedance-pH testing, its role for the diagnosis and management of patients with GERD is still unknown.\ud
To the best of our knowledge, to date, only 1 retrospective study focused its attention on this issue. Frazzoni et al5 reported that one-third of their refractory patients with remission of persistent symptoms at 3-year follow-up after\ud
surgical fundoplication had an abnormal distal number of total refluxes as the only preoperative impedance-pH finding. This suggests that this parameter is promising for GERD management, but further outcome studies are\ud
mandatory. On the other hand, a recent study by Patel et al6 emphasized the importance of the impedancedetected reflux-symptom association to predict a positive outcome after medical or surgical therapy. Thus, unless further data will be available on the importance of reflux numbers and because of the emerging advantages provided by impedance to symptoms analysis, we believe such approach should be preferred when investigating patients with suspected refractory GERD, particularly when the pretest probability of GERD diagnosis is low.\ud
Moreover, we think that the 95th percentile value of 73, set as the upper limit of normality in a previous article on healthy volunteers, should be regarded with caution. Recently, the French Group of Neuro-Gastroenterology\ud
modified their original cutoff from 75 to 53.7 The explanation provided by these authors was that after a decade of impedance-pH experience, a more careful manual evaluation of impedance-pH tracings permitted to reduce significantly the number of false-positive episodes detected by current available software. Interestingly, in 2006 we published our set of normal values, and we found a 95th percentile value of 54, which is in line with that observed by the French Group of Neuro-Gastroenterology.8 As a consequence, a considerable proportion of patients classified by the authors as “hypersensitive” or “functional” could instead be subjects with a no...