Question 2 How effective is domperidone at reducing symptoms of gastro-oesophageal reflux in infants?
- Correspondence to Bryonnie Scott, Department of Paediatrics, Leeds Teaching Hospitals, Leeds YO8 6LL, UK
- Received 15 April 2012
- Accepted 17 April 2012
As a paediatric registrar you are presented with a 4-month-old baby in paediatric outpatients. His mum describes him as having large vomits up to 30 min after most feeds, and recently he has seemed reluctant to feed. He is exclusively breast fed, and is gaining weight appropriately and otherwise thriving. He has previously been investigated for his vomiting and you conclude that he has gastroeosphageal reflux without complications (GOR).
In the first instance, positioning and thickened, small frequent feeds would be advised. The latter suggestion is not without difficulty for mothers exclusively breast feeding. As the mother is describing vomiting at least six times per day, but no signs suggestive of pain at present, could a pro-kinetic alone such as domperidone improve these symptoms?
Structured clinical question
For infants with symptoms of GOR [patient], is domperidone [intervention] as effective as or more effective than standard therapy [comparison] at improving symptoms [outcome]?
Medline (1948–March 2011) and Embase (1947–March 2011) were searched using the keywords (infant or child) and (gastro-oesophageal reflux or gastro-oesophageal reflux) and (domperidone or Motilium), limited to the English language.
The Cochrane Database of Systematic Reviews was also searched.
A total of 137 articles were found of which seven were relevant: six papers were identified after review of abstracts, and a further paper from referenced citations. These are summarised in table 2.
GOR in infancy is a common disorder encountered in paediatric practice. The reported prevalence varies widely from 20% to 67% in the literature,8 the prognosis is very good, and prevalence peaks at 4 months, after which the majority of children are symptom free by 1–2 years of age.9 When treatment is needed, positioning and thickening of feeds is the recommended first line management in most clinical practice, as suggested by ESPGAN.10 Since the hypothesised pathogenesis for GOR in young babies includes inappropriate relaxation of the lower oesophageal sphincter,11 ,12 pro-kinetics to intensify oesophageal peristalsis and accelerate gastric emptying have also been used widely. Prior to its withdrawal, cisapride was the mainstay of this treatment despite equivocal evidence of efficacy,13 with other drugs such as domperidone and erythromycin adopted later.
Investigation in this area is made problematic by the suggestions that objective measurements such as oesophageal pH monitoring and manometry correlate poorly with symptoms14 and that outcomes measured by symptom reporting are liable to subjectivity. In addition, there are variable definitions of what constitutes abnormal vomiting or regurgitation.
The ‘best fit’ answer to this clinical question found was the systematic review undertaken by Pritchard et al in 2005.1 They identified four randomised control trials (three of which were also identified by the above search strategy). The trials are described as having variable methodological quality and heterogeneous populations, interventions and outcome measures recorded. The results presented for each study were also disparate. The authors conclude that there is no robust evidence for the efficacy of domperidone for the treatment of GOR in young children and suggest further study is required. To address our clinical question the individual studies were examined, and information for the last trial5 extracted from abstracts and data presented in the systematic review.
Of the first two studies, Bines et al2 randomly allocated 17 children to domperidone or placebo for a 4-week period. They found significant changes (p<0.01, 25% decrease) in the pH metric recordings for reflux episodes but no other measured, or symptom, changes. Carroccio et al3 randomly allocated 80 children to one of four groups relevant to this question. In the group which received domperidone and a placebo, they report a significant reduction (from a median of median 59 to a median of 48.5; p<0.009) in measured reflux episodes but no decrease for any other pH metric data or symptom reporting.
De Loore et al4 randomly allocated 47 children to receive domperidone, metoclopramide or placebo, and found improvements in symptom score and ‘nausea’ for both drugs investigated (p<0.001) compared to placebo; no details are provided regarding symptom assessment (including that of nausea in infants). Clara et al 5 reported that 93% of 37 infants and children randomly allocated to domperidone had symptomatic improvement compared to placebo over a 2-week period (p<0.05); no details on symptom assessment, methodology or blinding beyond allocation were presented .
Two small and dissimilar trials were undertaken after the review by Pritchard et al. Cresi et al 7 recorded 24 h oesophageal pH monitoring data in inpatient neonates openly allocated to domperidone or no treatment. They found no significant improvement in objective outcome measures. In contrast, Hegar et al 6 found improvement in symptoms and some oesophageal pH metric data in infants with GOR treated for 4 weeks with domperidone openly compared with cisapride.
With the addition of the two more recent studies to those identified by Pritchard et al1 in 2005, the results are still somewhat disparate. Study design, definitions of diagnosis, use of controls, objectivity of outcome measures, use of concomitant treatments and the duration and dose of drug treatments used all differ, and, where applicable, there is little analysis of ‘infants’ as a subgroup. The studies present little convincing evidence for the efficacy of domperidone in infants with GOR. The only consensus between these studies is that no adverse events were reported; however, since the total sample size is small and the accumulated treatment period short at 608 patient weeks (excluding Cresi et al7 from ‘patient weeks’ as their patients were only treated for 24 h), this information cannot be taken as proof of safety.
In answer to this clinical question, given current evidence, it is not possible to support the use of domperidone in the treatment of uncomplicated gastroeosphageal reflux in infancy. The numbers found here are too small to engender confidence regarding the safety of domperidone in this usually benign and self-limiting condition. The advice in our current scenario has to be that the best management options are the use of positioning, small frequent feeds and thickeners, with reassurance and a ‘watch and wait’ policy also reasonable.
Clinical bottom line
There is insufficient evidence for the efficacy of domperidone in uncomplicated infantile gastro-oesophageal reflux without complications.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.