In reply
We disagree with many of the concerns that Numans raised and would like to discuss the meta-analysis by him and his colleagues.1 Understanding the physiologic features of nonerosive reflux disease would have identified the limitation of the PPI test in patients with GERD that resulted in attenuated conclusions. In these trials, the PPI test is compared with a combination of 24-hour esophageal pH testing and upper endoscopy. As a result, patients excluded by the PPI test are those with functional heartburn (typical heartburn but normal endoscopy and 24-hour pH monitoring). However, as Numans et al1 have found, the sensitivity of the PPI test in GERD was relatively high, but the specificity was low. This is primarily because a subset of patients with functional heartburn responded to PPI therapy. Up to 50% of patients with functional heartburn respond to treatment with PPIs once a day,2 with an even better response to higher doses of PPIs.3 In fact, the meta-analysis by Numans et al1 provides support for the PPI test because of its capability of diagnosing functional heartburn that is responsive to antireflux medications. This was clearly not taken into consideration.