To the Editor: We thank Dr Indrio and colleagues for their letter, which makes a number of salient points in reference to our manuscript on the effects of omeprazole in premature neonates with pathological esophageal acid exposure (1). We begin by reiterating that the principal aim of our study was to assess the safety and efficacy of omeprazole for the reduction of acid gastroesophageal reflux in premature infants with gastroesophageal reflux disease (GERD). Indrio et al correctly point out that the appropriateness or otherwise of the cutoff criteria of reflux index >5% to indicate pathological esophageal acid exposure is a matter of conjecture. Most pediatric centres would routinely use 5%; however, a fair reading of the limited literature that reports 24-hour pH studies in healthy infants indicates that higher cutoff criteria of 10% may be more appropriate. It is, however, not true that 8 of the infants enrolled in our study had a reflux index of between 5% and 7%. As our Figure 1 clearly shows, 7 of the 10 infants had a reflux index >10% on placebo, with 3 >20% (average 19% overall). Hence, measured acid exposure off-therapy was highly significant and not equivocal. When on proton pump inhibitor (PPI) therapy, the reflux index was reduced to ≤5% in 9 of the 10 infants (average, 4.9%). We, therefore, demonstrated that a once-daily dose of 0.7 mg/kg omperazole “normalised” the reflux index, the precise criteria applied had no impact on this major finding or the conclusions drawn from this study. The lack of symptomatic improvement, despite normalisation of esophageal acid exposure levels, was both disappointing and counterintuitive. We reiterate that this negative finding needs to be treated with caution due to limitations in the study design related to the accuracy of recording of symptomatic episodes and the 1-week treatment period which may have been too short. We do not believe the lack of symptomatic change relates to our specific diagnostic criteria because stratification of our patients, based on severity of esophageal acid exposure on placebo and/or magnitude of change in esophageal acid exposure with therapy, still failed to show significant symptomatic changes. We do not agree with the suggestion that the lack of a clinical response to omeprazole may have related to age at the time of study and whether our patients were technically preterm or term. The postmenstrual age of our patients ranged from 34 to 40 weeks, and the suggestion that specific reflux index criteria be developed across this age range seems futile given the high level of variability in this parameter. This is due to a number of factors that may influence the reflux index including feed type, feed frequency and volume, nasogastric tube placement, and xanthine therapy. Indeed, this highlights yet another limitation of reflux index as a diagnostic parameter. The most obvious explanation for the lack of symptomatic response relates to the fact that a “proven” association of symptoms with gastroesophageal reflux episodes was not a selection criterion. GERD in infants is a complex disorder due to the range of clinical presentations, and the causal relationship between significant acid exposure and clinical symptoms is often difficult to prove. Rather than relying on measurement of acid exposure based on arbitrary reflux index criteria, the ability to demonstrate a temporal association between individual reflux events and the onset of symptomatic episodes would, in our view, be more diagnostically useful. Intraluminal pH-impedance monitoring allows detection of all reflux independently of pH and, as already demonstrated in adults on and off PPI therapy, improves the accuracy of reflux-symptom associations (2,3). Similar studies in infants have yet to be published; however, given the increased prevalence of nonacid bolus reflux in infants due to frequent milk/formula feeding, it is our expectation that pH-impedance monitoring will improve the diagnostic yield over and above pH monitoring alone. The present lack of appropriate diagnostic tests for infants presenting with putative gastroesophageal reflux–related symptoms is contributing to the global expansion of empirical use of PPI in infants 0 to 6 months of age. This is despite the prevalence of so-called physiological reflux in this age group and the paucity of clear evidence-based guidance on appropriate dosing levels to achieve normalisation of acid exposure and clinical efficacy for reflux-related symptoms. Our study addresses the former with respect to the drug omeprazole; however, other PPIs need to be similarly tested. There is an urgent need to address the latter issue of clinical efficacy by way of randomised controlled trials. These trials must link PPI therapy with symptomatic improvement in patients specifically selected based on symptomatic, rather than pH-based, criteria. We would encourage all of those working in this field to take up this important challenge.