Cochrane Corner: Pharmacological Treatment of Gastro-Oesophageal Reflux in Children

Abstract

Reflux is commonly cited as the cause of common issues in babies, such as vomiting and crying. Though it is often self-limiting, in many cases, anti-reflux medications and feed thickeners are prescribed to infants and young children. Their use is controversial, and it is not clear whether they are effective. Data shown are trends in community prescribing for total quantity of thickener (Gaviscon infant and Instant Carobel) sachets/live births in England in the previous year (Figure 1); total items of infant anti-reflux medications prescribed/live birth in England in the previous year (Figure 2) [1].

This Cochrane Corner aims to assess the evidence for pharmacological agents in treating GORD in infants and children.

GOR is common [2], and GORD incidence in children is estimated at 0.84 per 1000 person years [3]. Several pharmacological treatments are available to treat reflux, including PPIs and H2 receptor antagonists, and they are being prescribed more frequently in recent years [4]. This Cochrane review aimed to establish which were beneficial in children split into subgroups of infants and older children. In addition to the cost of prescribing these medications, it is important to consider the potential side effects, which include infection risk, electrolyte abnormalities and fractures [5]. Similarly, whilst prokinetics such as domperidone, cisapride and erythromycin are effective in managing reflux, they have been associated with the very rare but serious risks of drug-induced long-QT syndrome, Torsades de Pointes and sudden cardiac death, due to a combination of non-modifiable (age, family history of arrhythmias, genetic conditions and co-morbidities) and modifiable risk factors (electrolyte imbalances, concurrent use of other medications). Thus, use of these prokinetics is cautioned against, though cardiac complications of domperidone have not been noted in children [6].

This Cochrane Review analysed 36 randomised control trials involving 2251 infants and children, out of which 22 trials were excluded due to insufficient data for extraction. The sample sizes of the studies included were small (range = 10–268), and due to methodological differences between the studies, a meta-analysis could not be performed. The trials often used subjective questionnaires, and blinding the trials was a challenge. Although blinding is less of an issue for trials using objective data such as 24-h pH impedance studies and endoscopic findings, the subjective outcome assessments used in these trials are vulnerable to detection bias in the absence of blinding. Such questionnaires do have the benefit of aiding clinicians with decision-making as they are patient-centred.

Though one trial [7] performed endoscopic studies at baseline, they did not repeat them after the intervention; therefore, no comparisons could be made. In another case, pH indices were used [8]; however, the outcome was not reported in sufficient detail to allow for the extraction of summary statistics to assess efficacy and implications for practice. Suggestions for future research include improved trial design to ensure that there is sufficient detail to allow for extraction of summary statistics and use of measurable outcomes such as improvement in endoscopic findings and pH indices, which could be assessed both before and after the intervention to allow for direct comparison. It would also be beneficial to evaluate the studies according to the GRADE criteria.

Despite this, the rationale for using acid blockers would remain unchanged as the pH of the gastric contents in infants is not that acidic due to the low acidity found in milk, and therefore, the rationale for using blockers is limited. However, as most studies are now impedance studies, they can also report on non-acidic reflux.

There was also a conflict of interest as many of the studies had pharmaceutical involvement with the manuscript writing process. In the future, studies should be written without pharmaceutical influence.

A number of the studies also assessed the efficacy of ranitidine, which is no longer available in the European Union (EU) after it was suspended in 2019 as it contained levels of an impurity called NMDA [9]. It is therefore recommended that future research focus on pharmacological interventions that are available in the EU, such as PPIs.

A.D. and N.B. contributed equally to the writing of this manuscript.

The authors declare no conflicts of interest.