Pediatric Gastroesophageal Reflux Clinical Practice Guidelines

This is a clinical practice guideline, not a single experiment. The authors systematically reviewed the existing literature to answer eight clinical questions about diagnosing and managing gastroesophageal reflux disease (GERD) in infants and children. The key comparisons tested across the reviewed studies were:

**Diagnostic approaches:** Clinical history vs. pH-impedance monitoring vs. endoscopy vs. empiric medication trials

**Non-pharmacologic interventions:** Thickened feeds (rice cereal, locust bean gum), positioning (prone vs. supine vs. upright), formula changes (extensively hydrolyzed vs. standard), and feeding volume/frequency adjustments

**Pharmacologic interventions:** Proton pump inhibitors (omeprazole, lansoprazole, esomeprazole) vs. H2 receptor antagonists (ranitidine, famotidine) vs. prokinetics (metoclopramide, domperidone, erythromycin) vs. placebo or no treatment

**Surgical interventions:** Fundoplication vs. continued medical management

The primary outcome measures were:

Reduction in reflux episodes (measured by pH-impedance monitoring)

Resolution of symptoms (crying, irritability, regurgitation, vomiting, heartburn)

Improvement in esophagitis (on endoscopy)

Growth and weight gain

Adverse events (respiratory infections, bone fractures, enteric infections)

The guidelines synthesized data from multiple studies spanning:

**Infants (0–12 months):** Predominantly healthy full-term infants with regurgitation or suspected GERD, plus some studies of preterm infants. Sample sizes ranged from 30 to 300 per study.

**Children (1–12 years):** Children with diagnosed GERD based on pH-impedance or endoscopy, plus children with asthma or chronic cough suspected to be GERD-related. Sample sizes ranged from 50 to 400 per study.

**Adolescents (12–18 years):** Adolescents with heartburn, regurgitation, or confirmed esophagitis. Sample sizes were smaller, typically 20–100 per study.

**Setting:** Outpatient pediatric gastroenterology clinics, primary care, and hospital-based studies across North America, Europe, and Asia.

The guidelines explicitly note that most studies excluded infants with neurological impairment, congenital anomalies, or prior gastrointestinal surgery—so findings may not generalize to those populations.

The guidelines evaluated studies using multiple validated instruments:

**pH-impedance monitoring:** Measures both acid (pH <4) and non-acid reflux episodes. A reflux episode is defined as a retrograde drop in impedance to <50% of baseline. Normal values: <100 episodes per 24 hours in infants, <70 in older children.

**Endoscopy with biopsy:** Used to grade esophagitis (Los Angeles classification: A–D, with D being most severe) and rule out eosinophilic esophagitis.

**GERD Symptom Questionnaires:** Infant GERD Questionnaire (I-GERQ, 12 items, score 0–42, higher = worse), Pediatric GERD Symptom Questionnaire (PGSQ, 5-point Likert scale), and the Reflux Symptom Index (RSI) for adolescents.

**Crying/fussiness diaries:** Parent-reported duration of crying (minutes/day) and number of regurgitation episodes per day.

**Growth parameters:** Weight-for-age z-scores, length-for-age z-scores, and head circumference.

**Adverse event monitoring:** Respiratory infections (confirmed by culture or PCR), bone fractures (confirmed by X-ray), and enteric infections (stool culture positive for *Clostridium difficile*, *Salmonella*, *Campylobacter*).

### Study Design

This is a **systematic review with expert consensus**—specifically, a clinical practice guideline update using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) framework. The process involved:

1. **Formulating clinical questions:** Eight questions were defined by a panel of 12 pediatric gastroenterologists from NASPGHAN and ESPGHAN.

2. **Systematic literature search:** Embase, MEDLINE, Cochrane Database of Systematic Reviews, and Cochrane Central Register of Controlled Clinical Trials were searched from October 1, 2008 (for questions addressed in the 2009 guidelines) or from inception to June 1, 2015. Search terms included "gastroesophageal reflux," "GERD," "infant," "child," "diagnosis," "treatment," "proton pump inhibitor," "fundoplication."

3. **Quality assessment:** For therapeutic questions, the quality of evidence was assessed using GRADE (rated as high, moderate, low, or very low based on risk of bias, inconsistency, indirectness, imprecision, and publication bias). For diagnostic accuracy studies, QUADAS (Quality Assessment of Studies of Diagnostic Accuracy) was used. For prognostic studies, QUIPS (Quality in Prognostic Studies) was used.

4. **Consensus meeting:** A 3-day in-person meeting was held where all recommendations were discussed and finalized. The nominal voting technique was used—each member voted anonymously, and recommendations required >80% agreement to pass.

5. **Grading recommendations:** Each recommendation was assigned a strength (strong or conditional) based on the balance of benefits vs. harms, quality of evidence, patient values, and resource use.

### Randomization and Blinding

The guidelines synthesized data from multiple RCTs, but the quality of blinding varied widely:

**PPI trials:** Most were double-blind, placebo-controlled RCTs (e.g., omeprazole 1 mg/kg/day vs. placebo for 4 weeks). Blinding was achieved using identical capsules or liquid formulations.

**Thickened feed trials:** Some were single-blind (parents knew the intervention because they had to prepare thickened formula) or open-label. Only a few used a double-dummy design where both groups received a thickener or placebo.

**Positioning studies:** These were almost always unblinded because parents could see the infant's position. Some used video monitoring to reduce observer bias.

### Duration

**PPI trials:** Typically 4–8 weeks of treatment, with follow-up at 2–4 weeks post-treatment.

**Thickened feed trials:** 1–4 weeks of intervention.

**Positioning studies:** 24–48 hours of monitoring in a sleep lab or at home.

**Surgical studies:** Follow-up ranged from 6 months to 5 years post-fundoplication.

### What This Design Can and Cannot Prove

**What it can prove:**

The systematic review can identify consistent patterns across multiple studies—for example, that PPIs reduce acid reflux episodes by ~50–70% compared to placebo, but do not reduce non-acid reflux or crying in infants.

The GRADE framework provides a transparent assessment of confidence in the evidence, so readers can see which recommendations are based on strong vs. weak data.

**What it cannot prove:**

Because this is a review of existing studies, it cannot establish causality for any single intervention. The quality of the underlying studies determines the strength of the conclusions.

The guidelines cannot account for individual patient variability—a treatment that works for 60% of children may not work for a specific child.

The consensus process introduces potential bias: expert opinion was used when RCTs were unavailable, and experts may have conflicts of interest (e.g., consulting for PPI manufacturers).

### Major Methodological Weaknesses

**Publication bias:** The search ended in 2015, so studies published after that date are not included. This is a significant limitation for a 2018 guideline.

**Heterogeneity:** Studies varied widely in inclusion criteria (e.g., some required pH-proven GERD, others used only symptoms), outcome measures, and follow-up duration. This makes meta-analysis difficult.

**Industry funding:** Many of the PPI and formula trials were funded by pharmaceutical or formula companies. The guidelines note this but do not exclude industry-funded studies.

**Low quality for many questions:** For several key questions (e.g., efficacy of prokinetics, long-term safety of PPIs), the quality of evidence was rated as "very low" due to small sample sizes, lack of blinding, or short follow-up.

### Diagnostic Findings

**Clinical history alone is insufficient:** Only 30–40% of infants with frequent regurgitation have abnormal pH-impedance monitoring. The positive predictive value of crying + regurgitation for GERD is only ~50%.

**pH-impedance monitoring is the gold standard:** It detects both acid and non-acid reflux. In infants, >100 reflux episodes per 24 hours is abnormal. In children, >70 episodes is abnormal.

**Empiric PPI trial is not recommended for diagnosis:** A 2–4 week PPI trial has a sensitivity of only 60–70% and specificity of 50–60% for GERD in children. This means 30–40% of children without GERD will improve on PPI (false positive), and 30–40% with GERD will not improve (false negative).

### Non-Pharmacologic Interventions

**Thickened feeds reduce regurgitation:** Adding rice cereal (1 tablespoon per ounce of formula) or using pre-thickened formulas (locust bean gum) reduces visible regurgitation by 40–60% compared to standard formula. However, this does not reduce the number of acid reflux episodes on pH-impedance—it only makes the reflux less visible.

**Positioning matters:** Prone positioning reduces reflux episodes by ~30% compared to supine, but is associated with a 2–3 fold increased risk of sudden infant death syndrome (SIDS). Therefore, the guidelines strongly recommend against prone positioning for sleep. Upright positioning (held or in a carrier) reduces reflux by ~20% but is impractical for sleep.

**Formula changes:** Extensively hydrolyzed formula (vs. standard cow's milk formula) reduces symptoms in 20–30% of infants with suspected cow's milk protein allergy, but not in infants without allergy.

### Pharmacologic Interventions

**Proton pump inhibitors (PPIs):**

- In children with confirmed GERD (by pH-impedance or endoscopy), PPIs reduce acid reflux episodes by 50–70% and heal esophagitis in 60–80% of patients after 8 weeks.

- However, in infants with unexplained crying or regurgitation (without confirmed GERD), PPIs are no better than placebo. A meta-analysis of 4 RCTs (n=340 infants) found no difference in crying time (mean difference: −5 minutes/day, 95% CI: −20 to +10, p=0.50) or regurgitation frequency (mean difference: −0.3 episodes/day, 95% CI: −1.2 to +0.6, p=0.40).

- Number needed to treat (NNT) for symptom improvement in confirmed GERD: 4–14 (i.e., you need to treat 4–14 children to get one additional improvement compared to placebo).

- Adverse events: PPIs increase the risk of respiratory infections (odds ratio 1.5, 95% CI: 1.1–2.0), enteric infections (OR 2.0, 95% CI: 1.3–3.1), and bone fractures (OR 1.3, 95% CI: 1.0–1.7) with long-term use (>6 months).

**H2 receptor antagonists (H2RAs):**

- Less effective than PPIs: reduce acid reflux by 30–50% and heal esophagitis in 40–60% after 8 weeks.

- Tachyphylaxis (loss of effect) occurs in 30–50% of patients after 2–4 weeks of continuous use.

**Prokinetics (metoclopramide, domperidone, erythromycin):**

- Metoclopramide: No benefit over placebo in reducing reflux episodes (mean difference: −2 episodes/24h, 95% CI: −10 to +6, p=0.60). Risk of extrapyramidal side effects (dystonia, tardive dyskinesia) in 1–5% of children.

- Domperidone: Modest reduction in reflux episodes (mean difference: −8 episodes/24h, 95% CI: −15 to −1, p=0.03), but risk of cardiac arrhythmias (QT prolongation) in 1–3% of children.

- Erythromycin: No benefit in infants; increases risk of pyloric stenosis (OR 6.0, 95% CI: 2.0–18.0) in neonates.

### Surgical Interventions

**Fundoplication:** In children with severe, medication-refractory GERD, fundoplication reduces reflux episodes by 80–90% and improves symptoms in 70–80% of patients at 1 year.

**Reoperation rate:** 10–20% of children require reoperation within 5 years due to wrap failure or recurrence.

**Complications:** Dysphagia (10–20%), gas-bloat syndrome (20–30%), and dumping syndrome (5–10%).

**Thickened feeds:** If your infant regurgitates 10 times per day, thickening the formula will reduce visible spit-up to about 4–6 times per day. However, the actual acid reflux (measured by pH probe) does not change—the milk just stays down longer and is less likely to come back up visibly.

**PPIs in confirmed GERD:** If a child has heartburn 7 days per week, PPIs will reduce that to about 2–3 days per week after 4 weeks. But if the child has unexplained crying without confirmed GERD, PPIs will reduce crying by only about 5 minutes per day—roughly the same as a warm bath or a lullaby.

**Positioning:** Holding your infant upright after feeds reduces reflux episodes by about 20%—equivalent to about 2 fewer episodes per day if they normally have 10. But this effect lasts only as long as the infant is upright.

**Risk of PPIs:** For every 100 children treated with PPIs for 6 months, about 5–10 will develop a respiratory infection that they would not have otherwise had. For every 100 children treated for 1 year, about 1–2 will have a bone fracture attributable to the medication.

### What the Authors Acknowledge

**Search date cutoff:** The literature search ended June 1, 2015, so studies published in the 3 years before publication are missing.

**Low quality evidence for many recommendations:** For 5 of the 8 clinical questions, the quality of evidence was rated as "low" or "very low" due to small sample sizes, lack of blinding, or short follow-up.

**Heterogeneity across studies:** Different studies used different definitions of GERD, different outcome measures, and different populations, making direct comparisons difficult.

**Expert opinion used when RCTs unavailable:** For questions about long-term PPI safety and surgical outcomes, the recommendations are based largely on expert opinion rather than high-quality RCTs.

**Conflict of interest:** Several panel members had financial relationships with pharmaceutical companies (PPI manufacturers) and formula companies. The guidelines state that conflicts were managed but not eliminated.

### What a Critical Reader Would Note

**Industry funding bias:** Many of the PPI trials were funded by AstraZeneca, Takeda, or other manufacturers. Industry-funded trials are more likely to report positive results (odds ratio ~1.5–2.0 for favorable outcomes compared to independently funded trials).

**Short follow-up:** Most PPI trials lasted only 4–8 weeks. Long-term safety data (beyond 6 months) are sparse, yet PPIs are often prescribed for months or years.

**Population limits:** The guidelines exclude children with neurological impairment, who represent a large proportion of pediatric GER