Review question
What is the best and safest treatment for babies and children with gastro-oesophageal reflux?
Key messages:
- the evidence for medications for babies with gastro-oesophageal reflux/reflux disease is very uncertain;
- for children with gastro-oesophageal reflux disease, the evidence is very uncertain regarding the effects of proton pump inhibitors. There was no adequate evidence to draw conclusions regarding other medications.
What is gastro-oesophageal reflux?
Gastro-oesophageal reflux happens when stomach contents come back up into the oesophagus (food pipe). Most babies (under 1 year) grow out of reflux symptoms, but does medicine help? Children (older than 1 year) can have reflux just like adults. Reflux can be normal ('physiological reflux'), but in babies and children, it can cause symptoms, including pain or weight loss, as the oesophagus becomes inflamed (oesophagitis). Bothersome symptoms of reflux are called 'gastro-oesophageal reflux disease' (GORD).
How is gastro-oesophageal reflux treated?
Medicines can thicken the stomach contents (alginates), neutralise stomach acid (ranitidine, omeprazole, lansoprazole), or help the stomach to empty faster (domperidone, erythromycin).
What did we want to find out?
We wanted to learn the best way to reduce reflux in babies and children. We wanted to see if medicines help infants and children to feel better (symptom scores), heal the oesophagus (which is checked by using endoscopy, where a tiny camera is put down the oesophagus), or lower the time the oesophagus is exposed to stomach acid. We also investigated whether the medicines were safe by considering the harmful or unwanted effects reported in the studies.
What did we do?
We searched for studies testing gastro-oesophageal reflux medicines in babies and children. We included all studies comparing these medicines, or comparing them to an inactive medicine (placebo). We assessed results which are important to doctors, nurses, and parents, and performed our own analysis of the results. We rated our confidence in the evidence, based on factors such as study methods and sizes.
What did we find?
We found 36 suitable studies (involving 2251 babies and children), conducted worldwide, with most in the USA. The largest study recruited 268 babies, the smallest, 16 children. Fifteen studies compared an active medicine to placebo; 8 compared one active medication to another; and 11 studies gave the same medication at different doses. We found useable outcome information in 14 of the 36 studies. The remaining studies either did not report outcomes we were interested in or did not report them in a way we could analyse. We could not combine the results of any studies because they were too different (in terms of how long they followed participants up and the outcomes they investigated) to use in a meaningful way.
Key results
Babies. There is no clear effect on symptoms or measured acidity (one measure is reflux index, which is the percentage of time in 24 hours the oesophagus is exposed to stomach acid) between babies given omeprazole or placebo. One study (30 babies) showed cry/fuss time went down from 287 to 201 minutes/day in the placebo group and 246 to 191 minutes/day in the omeprazole group. Reflux index changed in the omeprazole group from 9.9% to 1.0% in 24 hours, and in the placebo group from 7.2% to 5.3%. One study (76 babies) showed that omeprazole and ranitidine may have a similar benefit for symptoms after 2 weeks: symptom scores (higher scores mean worse symptoms) in the omeprazole group dropped from 51.9 to 2.4, and in the ranitidine group, from 47 to 2.5. In one study of 52 newborn babies, esomeprazole appeared to show no reduction in the number of symptoms (184.7 to 156.7) compared to placebo (183.1 to 158.3). None of the studies reported harmful events or results about changes to babies' oesophaguses.
Children. In children older than 1 year of age, no studies assessed medical treatment versus placebo. Proton pump inhibitors (PPIs), which block stomach acid production, at different doses may provide little to no improvements in symptoms or oesophagus healing. In one study (127 children), both lower-weight and higher-weight children given rabeprazole at lower and higher doses had both minimal – probably unimportant – changes in symptom scores and endoscopic scores (which indicate whether healing of the oesophagus has occurred). Pantoprazole may or may not improve symptom scores in children aged 1 to 5 years by week 8: there was no difference between lower and higher dosing in one study (60 children). Studies investigating other medications did not report enough information for us to assess their results properly.
Quality of the evidence
We are not confident in the evidence, which was mainly based on single studies with few babies and children. Several studies had pharmaceutical company help with manuscript writing. The question of how best to treat children with disabilities, and whether any PPIs are better than other medicines remain. The evidence is current to 17 September 2022.
There is very low-certainty evidence about symptom improvements and changes in pH indices for infants. There are no summary data for endoscopic changes. Medications may or may not provide a benefit (based on very low-certainty evidence) for infants whose symptoms remain bothersome, despite nonmedical interventions or parental reassurance. If a medication is required, there is no clear evidence based on summary data for omeprazole, esomeprazole (in neonates), H₂antagonists, and alginates for symptom improvements (very low-certainty evidence). Further studies with longer follow-up are needed.
In older children with GORD, in studies with summary data extracted, there is very low-certainty evidence that PPIs (rabeprazole and pantoprazole) may or may not improve GORD outcomes. No robust data exist for other medications.
Further RCT evidence is required in all areas, including subgroups (preterm babies and children with neurodisabilities).
Gastro-oesophageal reflux (GOR) is characterised by the regurgitation of gastric contents into the oesophagus. GOR is a common presentation in infancy, both in primary and secondary care, affecting approximately 50% of infants under three months old. The natural history of GOR in infancy is generally of a self-limiting condition that improves with age, but older children and children with co-existing medical conditions can have more protracted symptoms. The distinction between gastro-oesophageal reflux disease (GORD) and GOR is debated. Current National Institute of Health and Care Excellence (NICE) guidelines define GORD as GOR causing symptoms severe enough to merit treatment. This is an update of a review first published in 2014.
To assess the effects of pharmacological treatments for GOR in infants and children.
For this update, we searched CENTRAL, MEDLINE, Embase, and Web of Science up to 17 September 2022. We also searched for ongoing trials in clinical trials registries, contacted experts in the field, and searched the reference lists of trials and reviews for any additional trials.
We included randomised controlled trials (RCTs) that compared any currently-available pharmacological treatment for GOR in children with placebo or another medication. We excluded studies assessing dietary management of GORD and studies of thickened feeds. We included studies in infants and children up to 16 years old.
We used standard methodology expected by Cochrane.
We included 36 RCTs involving 2251 children and infants. We were able to extract summary data from 14 RCTs; the remaining trials had insufficient data for extraction. We were unable to pool results in a meta-analysis due to methodological differences in the included studies (including heterogeneous outcomes, study populations, and study design).
We present the results in two groups by age: infants up to 12 months old, and children aged 12 months to 16 years old.
Infants
Omeprazole versus placebo: there is no clear effect on symptoms from omeprazole. One study (30 infants; very low-certainty evidence) showed cry/fuss time in infants aged three to 12 months had altered from 246 ± 105 minutes/day at baseline (mean +/- standard deviation (SD)) to 191 ± 120 minutes/day in the omeprazole group and from 287 ± 132 minutes/day to 201 ± 100 minutes/day in the placebo group (mean difference (MD) 10 minutes/day lower (95% confidence interval (CI) -89.1 to 69.1)). The reflux index changed in the omeprazole group from 9.9 ± 5.8% in 24 hours to 1.0 ± 1.3% and in the placebo group from 7.2 ± 6.0% to 5.3 ± 4.9% in 24 hours (MD 7% lower, 95% CI -4.7 to -9.3).
Omeprazole versus ranitidine: one study (76 infants; very low-certainty evidence) showed omeprazole may or may not provide symptomatic benefit equivalent to ranitidine. Symptom scores in the omeprazole group changed from 51.9 ± 5.4 to 2.4 ± 1.2, and in the ranitidine group from 47 ± 5.6 to 2.5 ± 0.6 after two weeks: MD -4.97 (95% CI -7.33 to -2.61).
Esomeprazole versus placebo: esomeprazole appeared to show no additional reduction in the number of GORD symptoms compared to placebo (1 study, 52 neonates; very low-certainty evidence): both the esomeprazole group (184.7 ± 78.5 to 156.7 ± 75.1) and placebo group (183.1 ± 77.5 to 158.3 ± 75.9) improved: MD -3.2 (95% CI -4.6 to -1.8).
Children
Proton pump inhibitors (PPIs) at different doses may provide little to no symptomatic and endoscopic benefit.
Rabeprazole given at different doses (0.5 mg/kg and 1 mg/kg) may provide similar symptom improvement (127 children in total; very low-certainty evidence). In the lower-dose group (0.5 mg/kg), symptom scores improved in both a low-weight group of children (< 15 kg) (mean -10.6 ± SD 11.13) and a high-weight group of children (> 15 kg) (mean -13.6 ± 13.1). In the higher-dose groups (1 mg/kg), scores improved in the low-weight (-9 ± 11.2) and higher-weight groups (-8.3 ± 9.2). For the higher-weight group, symptom score mean difference between the two different dosing regimens was 2.3 (95% CI -2 to 6.6), and for the lower-weight group, symptom score MD was 4.6 (95% CI -2.9 to 12).
Pantoprazole: pantoprazole may or may not improve symptom scores at 0.3 mg/kg, 0.6 mg/kg, and 1.2 mg/kg pantoprazole in children aged one to five years by week eight, with no difference between 0.3 mg/kg and 1.2 mg/kg dosing (0.3 mg/kg mean −2.4 ± 1.7; 1.2 mg/kg −1.7 ± 1.2: MD 0.7 (95% CI -0.4 to 1.8)) (one study, 60 children; very low-certainty evidence).
There were insufficient summary data to assess other medications.