Statistics from Altmetric.com
We read with interest the recent article by Del Buono et al evaluating the effect of sodium and magnesium alginate (Gaviscon) on gastro-oesophageal reflux (GOR) in infants.1 It provides an objective assessment of the effects of a drug widely used in the treatment of paediatric GOR by means of a double blind drug versus placebo trial, in which the effects of each treatment were evaluated by means of the simultaneous application of multichannel intraluminal impedance and pH-metry (MII/pH). The authors show that Gaviscon reduces this height, probably because it increases the viscosity of the gastric content and hence acts in the same way as thickened feeding.2 They also found that fewer acid reflux episodes occurred after Gaviscon, though the difference was not significant. By contrast with the evidence produced in other studies,3 therefore, these results seem to suggest that Gaviscon Infant has little effect on GOR when assessed in objective terms. It is, however, possible that the significance of some of the differences they observed has been weakened by the influence of sleep and wakefulness on GOR episodes.4
During 53 MII/pH 24 h monitorings in infants with GOR symptoms at our centre, we noted a significant difference between the number of episodes during wakefulness (535 hours) and sleep (450 hours): 3.2±4.1 episodes per hour versus 1.8±3.3 (p < 0.001; CI 0.93±1.87). Del Buono et al gave six milk meals plus drug or placebo according to a 3+3 schedule. If we suppose that a infant sleeps 12 hours a day, then the probability that “sleep” and “wakefulness” periods were equally distributed between Gaviscon and placebo in each of their 20 patients can be no more than 50%. This probability drops even further if account is taken of the fact that infants with GOR sleep less than normal. The difference in GOR frequency between sleep and wakefulness, coupled with the asymmetrical distribution of these phases, constitutes a “confounding factor” responsible for great variability of all the frequency data (number of episodes per hour, number of acid episodes, number of postprandial episodes per hour, etc), whereas it may have little influence on GOR “quality” (duration, pH, and height).
We thus believe that assessment of efficacy of the treatment of GOR by means of MII/pH requires longer observation periods (for example, 24 h placebo versus 24 h drug), or at all events consideration of the influence of sleep and wakefulness on GOR episodes.
Pre-published book reviews
Book reviews that have been accepted for publication but have not yet been published in the print journal can be viewed online at http://adc.bmjjournals.com/misc/bookreviews.shtml
Competing interests: none declared