EVIDENCE-BASED ANSWER
The literature on pediatric reflux can be divided into studies addressing clinically apparent reflux (vomiting or regurgitation) and reflux as measured by pH probe or other methods (TABLES 1 AND 2). Sodium alginate reduces vomiting and improves parents’ assessment of symptoms (strength of recommendation [SOR]: B, small randomized controlled trial [RCT]). Formula thickened with rice cereal decreases the number of postprandial emesis episodes in infants with gastroesophageal reflux disease (GERD) (SOR: B, small RCT).
There are conflicting data on the effect of carob bean gum as a formula thickener and its effect on regurgitation frequency (SOR: B, small RCTs). Metoclopramide does not affect vomiting or regurgitation, but is associated with greater weight gain in infants over 3 months with reflux (SOR: B, low-quality RCTs).
Carob bean gum used as a formula thickener decreases reflux as measured by intraluminal impedance but not as measured by pH probe (SOR: B, RCT). Omeprazole and metoclopramide each improve the reflux index as measured by esophageal pH probe (SOR: B, RCT).
Evidence is conflicting for other commonly used conservative measures (such as positional changes) or other medications for symptomatic relief of infant GERD. There is very limited evidence or expert opinion regarding breastfed infants, particularly with regard to preservation of breastfeeding during therapy.
TABLE 1
Interventions that affect vomiting or regurgitation
| INTERVENTION | TRIAL DESCRIPTION | EFFECT |
|---|---|---|
| Carob bean gum* | Unblinded crossover RCT (n=14 infants w/regurgitation). Reflux episodes measured by intraluminal impedance and visual regurgitation score.5 | Improved. |
| 0.4 g/100 cc | Carob bean gum: 15 regurgitations/342 hrs. | |
| Standard formula: 68 | ||
| P<.0003 | ||
| RCT, thickened vs. standard formula (n=20). | No improvement. | |
| Outcome: regurgitation score, parental diary.6 | Thickened formula: 2.2≠ 1.92 regurgitation score. Control formula: 3.3≠ 1.16. | |
| P=.14 | ||
| Crossover RCT (n=24). Formula thickened with carob bean gum vs rice cereal. | Improved. | |
| Outcomes: symptom scores and emesis episodes.7 | Both groups showed improved symptom scores and decreased emesis, but carob bean gum was superior to rice cereal-thickened formula. | |
| Sodium alginate† | Double-blind multicenter RCT of alginate vs placebo added to formula or breast milk (n=88). Intention-to-treat analysis.9 | Improved. |
| 225 mg/115 cc | Funded by manufacturer. 25% dropout rate. Breastfed infants included, but results not reported separately. | Alginate: from 8.5 vomiting/regurgitation episodes to 3 per 24 h. |
| or | Placebo: from 7 episodes to 5 per 24 h. | |
| 450 mg/225 cc | P=.009 | |
| Rice cereal | RCT of thickened vs unthickened formula (n=20). Emesis episodes per 90-min postprandial period.4 | Improved. |
| (see also Carob emesis bean gum, above) | Thickened formula: 1.2 +/- 0.7 episodes per 90 minutes postprandial | |
| Placebo: 3.9 +/- 0.9 emesis episodes | ||
| P=0.015 | ||
| Metoclopramide | Crossover RCT (n=30). Metoclopramide vs placebo for 7 days. Mean daily symptom count (included vomiting and regurgitation).10 | No improvement. |
| 0.1 mg/kg 4 times daily | Placebo: Symptom count for | |
| Placebo 6.5 1.3 per day | ||
| Metoclopramide 5.6 1.2 | ||
| P=.19 | ||
| Subgroup analysis infants >3 mo showed greater weight gain for treated infants. | ||
| *Used in the UK (Instant Carobel); not widely available in US | ||
| †Available in UK as Gaviscon Infant. | ||
TABLE 2
Interventions that affect pH probe/measured reflux
| INTERVENTION | DESCRIPTION | EFFECT |
|---|---|---|
| Carob bean gum* | Unblinded crossover RCT (n=14 infants w/regurgitation). Reflux episodes measured by intraluminal impedance and visual regurgitation score. Limitations: unblinded; small sample size; no breastfed infants included.5 | Improved. |
| 0.4 g/100 cc | Carob bean gum: 536 episodes in 342 hours. Placebo: 647 episodes. P<.02 | |
| RCT, thickened vs standard formula. | No improvement. | |
| Reflux meas. by 24-h pH probe.6 | Reflux index for thickened formula, 11.1 ± 6.1. Standard formula, 13.2 ± 4.7. P=.41 | |
| Rice cereal | RCT of thickened vs unthickened formula | No improvement. Thickened formula group: |
| (n=20). Reflux measured by scintigraphy.4 | 26.8 ± 5.8 episodes per 90 min postprandial period. Unthickened formula group: 27.9 ± 4.0. P=NS. | |
| Infant seat at 60° | RCT, positioning in infant seat vs prone. | Worsened. Infant seat: 16 ± 2.4 episodes |
| Episodes of reflux measured by pH probe.3 | in 2 h. Prone position: 10 ± 2.3 episodes. | |
| P=.002 | ||
| Head of bed at 30° | Crossover RCT (n=90). Prone position vs prone/head of bed elevated to 30°. Number and length of reflux episodes, measured by pH probe.8 | No improvement. Head-elevated 6.2 ± 0.6 episodes per 2 h. Flat prone 7.8 ± 0.8 episodes per 2 h. P=NS. |
| Head-elevated 17.1 ± 2.4 minutes longest episode. Flat prone 17.9 ± 2.2 minutes. P=NS. | ||
| Pacifier use | RCT (n=48). Seated vs prone position, with or without pacifier; reflux episodes meas. by pH probe.3 | Prone: Worsened from 7.2 ± 1.1 episodes in 2 h without pacifier to 12.8 ± 2.3 w/pacifier. P=.04. |
| Omeprazole | RCT (n=30 irritable infants with reflux or esophagitis). Reflux index (% of time pH <4) meas. by pH probe and “cry/fuss time.”11 | Irritability unchanged. Improved pH: |
| (Infants 5–10 kg: 10 mg/d; infants >10 kg: 10 mg bid) | Omeprazole: Reflux index –8.9% ± 5.6. | |
| Placebo: Reflux Index –1.9% ± 2. P<.001. | ||
| Metoclopramide | Crossover RCT (n=30). Metoclopramide vs placebo for 7 days. Reflux index measured by pH probe. Wide confidence intervals.10 | Improved reflux index. Metoclopramide: |
| (0.1 mg/kg 4 times daily) | 10.3% (95% CI, 2.4–22.8). Placebo: 13.4% (95% CI, 2.8–30.5). P<.001 |
Evidence summary
Regurgitation (“spitting up”) and gastroesophageal reflux are common in infants. In a cross-sectional survey of 948 parents of healthy infants aged 0 to 13 months, regurgitation occurred daily in half of infants from birth to 3 months old, peaked to 67% at age 4 months, and was absent in 95% by age 12 months.1 Gastroesophageal disease (GERD) is characterized by refractory symptoms or complications (pain, irritability, vomiting, failure to thrive, dysphagia, respiratory symptoms, or esophagitis) and occurs in the minority of infants with reflux.2 This distinguishes the “happy spitter,” whose parents may simply require reassurance, from infants who require treatment.