pediatric gastroesophageal reﬂux clinical practice ......phagus with or without regurgitation and...

Copyright © 2009 by Lippincott Williams & Wilkins.Unauthorized reproduction of this article is prohibited.

Pediatric Gastroesophageal Reflux Clinical Practice Guidelines:Joint Recommendations of the North American Society of

Pediatric Gastroenterology, Hepatology, and Nutrition and theEuropean Society of Pediatric Gastroenterology, Hepatology,

and Nutrition

Co-Chairs: !Yvan Vandenplas and yColin D. RudolphCommittee Members: zCarlo Di Lorenzo, §Eric Hassall, jjGregory Liptak,

!Lynnette Mazur, #Judith Sondheimer, !!Annamaria Staiano, yyMichael Thomson,zzGigi Veereman-Wauters, and §§Tobias G. Wenzl

!UZ Brussel Kinderen, Brussels, Belgium, {Division of Pediatric Gastroenterology, Hepatology, and Nutrition,Children’s Hospital of Wisconsin, Medical College of Wisconsin, Milwaukee, WI, USA, {Division of Pediatric Gastroenterology,

Nationwide Children’s Hospital, The Ohio State University, Columbus, OH, USA, §Division of Gastroenterology,Department of Pediatrics, British Columbia Children’s Hospital/University of British Columbia, Vancouver, BC, Canada,

jjDepartment of Pediatrics, Upstate Medical University, Syracuse, NY, USA, !Department of Pediatrics, University of Texas HealthSciences Center Houston and Shriners Hospital for Children, Houston, TX, USA, #Department of Pediatrics, University of Colorado

Health Sciences Center, Denver Colorado, !!Department of Pediatrics, University of Naples ‘‘Federico II’’, Naples, Italy,{{Centre for Paediatric Gastroenterology, Sheffield Children’s Hospital, Western Bank, Sheffield, United Kingdom,

{{Pediatric Gastroenterology & Nutrition, Queen Paola Children’s Hospital-ZNA, Antwerp, Belgium, and§§Klinik für Kinder- und Jugendmedizin, Universitätsklinikum der RWTH Aachen, Aachen, Germany

ABSTRACT

Objective: To develop a North American Society of PediatricGastroenterology, Hepatology, and Nutrition (NASPGHAN) andEuropean Society of Pediatric Gastroenterology, Hepatology,and Nutrition (ESPGHAN) international consensus on thediagnosis and management of gastroesophageal reflux andgastroesophageal reflux disease in the pediatric population.Methods: An international panel of 9 pediatricgastroenterologists and 2 epidemiologists were selected byboth societies and developed these guidelines based on theDelphi principle. Statements were based on systematicliterature searches using the best available evidence fromPubMed, Cumulative Index to Nursing and Allied HealthLiterature, and bibliographies. The committee convened inface-to-face meetings 3 times. Consensus was achieved for

all recommendations through Nominal Group Technique, astructured, quantitative method. Articles were evaluatedusing the Oxford Centre for Evidence-based Medicine Levelsof Evidence. Using the Oxford Grades of Recommendation, thequality of evidence of each of the recommendations made by thecommittee was determined and is summarized in appendices.Results: More than 600 articles were reviewed for this work.The document provides evidence-based guideline for thediagnosis and management of gastroesophageal reflux andgastroesophageal reflux disease in the pediatric population.Conclusions: This document is intended to be used in dailypractice, for thedevelopmentoffutureclinicalpracticeguidelines,and as a basis for clinical trials. JPGN 49:498–547, 2009. KeyWords: Clinical practice guidelines—Diagnostic tests—Gastroesophageal reflux (GER)—Gastroesophageal refluxdisease (GERD)—Therapeutic modalities. # 2009 byEuropean Society for Pediatric Gastroenterology, Hepatology,and Nutrition and North American Society for PediatricGastroenterology, Hepatology, and Nutrition

SYNOPSIS

This synopsis contains some essentials of the guide-lines, but does not convey the details, nuances, andcomplexity of the issues addressed in the complete

Received May 27, 2009; accepted May 31, 2009.Address correspondence and reprint requests to Colin D. Rudolph,

MD, PhD, Professor of Pediatrics & Vice Chair for Clinical Affairs,Chief Pediatric Gastroenterology, Hepatology and Nutrition, Divisionof Pediatric Gastroenterology, Hepatology and Nutrition, Children’sHospital of Wisconsin, Medical College of Wisconsin, Milwaukee, WI,USA (e-mail: [email protected]).

Carlo Di Lorenzo, Eric Hassall, Gregory Liptak, Lynnette Mazur,Judith Sondheimer, Annamaria Staiano, Michael Thomson, Gigi Veere-man-Wauters, and Tobias G. Wenzl contributed equally to the devel-opment of these guidelines. Abstract adapted by Gregory Liptak.

Authors’ disclosures are listed in Appendix D.

Journal of Pediatric Gastroenterology and Nutrition49:498–547 # 2009 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition andNorth American Society for Pediatric Gastroenterology, Hepatology, and Nutrition

498

mailto:[email protected]


guidelines, and therefore can be interpreted only withreference to the full article.

Rationale

The purpose of these guidelines is to provide pedia-tricians and pediatric subspecialists with a commonresource for the evaluation and management of patientswith gastroesophageal reflux (GER) and gastroesopha-geal reflux disease (GERD). These guidelines are notintended as a substitute for clinical judgment or as aprotocol for the management of all of the pediatricpatients with GER and GERD.

Methods

‘‘Pediatric Gastroesophageal Reflux Clinical PracticeGuidelines: Joint Recommendations of the North Amer-ican Society of Pediatric Gastroenterology, Hepatology,and Nutrition (NASPGHAN) and the European Societyof Pediatric Gastroenterology, Hepatology, and Nutrition(ESPGHAN)’’ is a document developed by a committeeof 9 pediatric gastroenterologists from NASPGHAN andESPGHAN and 2 pediatric epidemiologists from theUnited States. Using the best-available evidence fromthe literature, the committee critically evaluated currentdiagnostic tests and therapeutic modalities for GERand GERD.

Definitions and Mechanisms

GER is the passage of gastric contents into the eso-phagus with or without regurgitation and vomiting. GERis a normal physiologic process occurring several timesper day in healthy infants, children, and adults. Mostepisodes of GER in healthy individuals last


Barium Contrast Radiography

This test is not useful for the diagnosis of GERD but isuseful to confirm or rule out anatomic abnormalities ofthe upper gastrointestinal (GI) tract that may causesymptoms similar to those of GERD.

Nuclear Scintigraphy

The standards for interpretation of this test arepoorly established. According to limited publishedliterature, scintigraphy may have a role in the diagnosisof pulmonary aspiration in patients with chronic andrefractory respiratory symptoms. A negative test doesnot rule out possible pulmonary aspiration of refluxedmaterial. Gastric emptying studies by themselves donot confirm the diagnosis of GERD and are recom-mended only in individuals with symptoms of gastricretention. Nuclear scintigraphy is not recommended forthe routine evaluation of pediatric patients withsuspected GERD.

Esophageal and Gastric Ultrasonography

These tests are not recommended for the routineevaluation of GERD in children.

Tests on Ear, Lung, and Esophageal Fluids

Evaluation of middle ear or pulmonary aspirates forlactose, pepsin, or lipid-laden macrophages have beenproposed as the tests for GERD. No controlled studieshave proven that reflux is the only reason thesecompounds appear in ear or lung fluids, and nocontrolled studies have shown that the presence ofthese substances confirms GER as the cause of ear,sinus, or pulmonary disease. Diagnosis of duodeno-gastroesophageal reflux by detection of bilirubin in theesophagus is not recommended for the routine evalu-ation for possible GERD in children. The role of bilereflux in causing GERD that is resistant to protonpump inhibitors (PPI) therapy has not been estab-lished.

Empiric Trial of Acid Suppression as aDiagnostic Test

Expert opinion suggests that in an older child oradolescent with typical symptoms suggesting GERD,an empiric trial of PPI is justified for up to 4 weeks.However, improvement of heartburn, following treat-ment, does not confirm a diagnosis of GERD becausesymptoms may improve spontaneously or respond by aplacebo effect. There is no evidence to support an empirictrial of acid suppression as a diagnostic test in infants andyoung children where symptoms suggestive of GERD areless specific.

Treatment

Lifestyle Changes in the Infant

Parental education, guidance, and support are alwaysrequired and usually sufficient to manage healthy, thrivinginfants with symptoms likely because of physiologic GER.Milk protein sensitivity is sometimes a cause of unex-plained crying and vomiting in infants. Therefore,formula-fed infants with recurrent vomiting may benefitfrom a 2- to 4-week trial of an extensively hydrolyzedprotein formula that has been evaluated in controlled trials.Use of a thickened formula (or commercial antiregurgita-tion formulae, if available) may decrease visible regurgita-tion but does not result in a measurable decrease in thefrequency of esophageal reflux episodes. Prone position-ing decreases the amount of acid esophageal exposuremeasured by pH probe compared with that measured in thesupine position. However, prone and lateral positions areassociated with an increased incidence of sudden infantdeath syndrome (SIDS). The risk of SIDS outweighs thebenefit of prone or lateral sleep position on GER; there-fore, in most infants from birth to 12 months of age, supinepositioning during sleep is recommended.

Lifestyle Changes in Children and Adolescents

In older children, there is no evidence to support theroutine elimination of any specific food for managementfor GERD. In adults, obesity, large meal volume, and latenight eating are associated with symptoms of GERD.Proneor left side sleeping position and/or elevation of the head ofthe bed may decrease GER, as shown in adult studies.

Pharmacologic Therapies

The major pharmacologic agents currently used fortreatingGERDinchildrenaregastricacid-bufferingagents,mucosal surface barriers, and gastric antisecretory agents.Acidsuppressantagentsarethemainstayoftreatmentforallbut the patient with occasional symptoms. The potentialadverseeffectsofacidsuppression, includingincreasedriskof community-acquired pneumonias and GI infections,need to be balanced against the benefits of therapy.

Histamine-2 Receptor Antagonists (H2RAs)

H2RAs exhibit tachyphylaxis or tolerance but PPIs donot. Tachyphylaxis is a drawback to chronic use. H2RAshave a rapid onset of action and, like buffering agents, areuseful for on-demand treatment.

Proton Pump Inhibitors

For healing of erosive esophagitis and relief of GERDsymptoms, PPIs are superior to H2RAs. Both medicationsare superior to placebo. Administration of long-term acid

500 VANDENPLAS ET AL.

J Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009


suppression without a diagnosis is inadvisable. When acidsuppression is required, the smallest effective dose shouldbe used. Most patients require only once-daily PPI; routineuse of twice-daily dose is not indicated. No PPI has beenapproved for use in infants


indicated (pH monitoring" impedance monitoring,endoscopy); a time-limited (2-week) trial of antisecretorytherapy may be considered but there is a potential risk ofadverse effects. Clinical improvement following empirictherapy may be because of spontaneous symptom resol-ution or a placebo response. The risk/benefit ratio ofthese approaches is not clear.

The Child Older Than 18 Months of Age With ChronicRegurgitation or Vomiting

Physiologic regurgitation, episodic vomiting, or regur-gitation followed by swallowing of refluxate in the mouthare frequent in infants. Whether of new onset or persist-ing from infancy, these symptoms are less common inchildren older than 18 months of age. Although thesesymptoms are not unique to GERD, evaluation to diag-nose possible GERD and to rule out alternative diagnosesis recommended based on expert opinion. Testing mayinclude upper GI endoscopy and/or esophageal pH/MII,and/or barium upper GI series.

Heartburn

Extrapolation from adult data suggests that in olderchildren and adolescents, on-demand therapy with buf-fering agents, sodium alginate, or H2RA, may be usedfor occasional symptoms. Adolescents with typicalsymptoms of chronic heartburn should be treated withlifestyle changes if applicable (diet changes, weight loss,smoking avoidance, sleeping position, no late-nighteating) and a 2- to 4-week trial of PPI. If symptomsresolve, PPI may be continued for up to 3 months. Heart-burn that persists on PPI therapy or recurs after thistherapy is stopped should be further investigated by apediatric gastroenterologist.

Reflux Esophagitis

In pediatric patients with endoscopically-diagnosedreflux esophagitis or established nonerosive reflux dis-ease, PPI for 3 months constitutes initial therapy. Not allreflux esophagitis are chronic or relapsing and therefore,trials of tapering the dose and then withdrawal of PPItherapy should be performed at intervals. Most but not allof the children with chronic-relapsing reflux disease haveone of the GERD-predisposing disorders describedbelow. In most cases of chronic-relapsing esophagitis,symptom relief can be used as a measure of efficacy oftherapy, but in some circumstances, repeat endoscopy ordiagnostic studies may be indicated. Recurrence of symp-toms and/or esophagitis after repeated trials of PPI with-drawal usually indicate that chronic-relapsing GERD ispresent, if other causes of esophagitis have been ruledout. At that point, therapeutic options include long-termPPI therapy or antireflux surgery.

Barrett Esophagus

BE occurs in children with less frequency than it doesin adults. Multiple biopsies documented in relation toendoscopically-identified esophagogastric landmarks areadvised to confirm or rule out the diagnosis of BE anddysplasia. In BE, aggressive acid suppression is advisedby most experts. Symptoms are a poor guide to theseverity of acid reflux and esophagitis in BE, and pHstudies are often indicated to guide treatment. BE per se isnot an indication for surgery. Dysplasia is managedaccording to adult guidelines.

Dysphagia, Odynophagia, and Food Refusal

Dysphagia, or difficulty in swallowing, occurs inassociation with oral and esophageal anatomic abnorm-alities, neurologic and motor disorders, oral and esopha-geal inflammatory diseases, and psychological stressorsor disorders. Of the mucosal disorders, eosinophilicesophagitis is increasingly recognized to be a morecommon cause of dysphagia or odynophagia than GERD,although this finding is not consistently reported in allgeographic regions. Odynophagia, or pain caused byswallowing, must be distinguished from heartburn (sub-sternal pain caused by esophageal acid exposure) anddysphagia. Although odynophagia may be a symptom ofpeptic esophagitis, it is more often associated with otherconditions such as oropharyngeal inflammation, esopha-geal ulcer, eosinophilic esophagitis, infectious esophagi-tis, and esophageal motor disorders. Although GERD isnot a prevalent cause of difficulty in swallowing or painwith swallowing, an evaluation including barium upper-GI series and possibly upper endoscopy should be con-sidered if physical examination and history of disease donot reveal a cause. Therapy with acid suppression withoutearlier evaluation is not recommended. In the infantwith feeding refusal, acid suppression without earlierdiagnostic evaluation is not recommended.

Infants With Apnea or Apparent Life-threateningEvent (ALTE)

In the majority of infants with apnea or ALTEs, GER isnot the cause. In the uncommon circumstance in which arelation between symptoms and GER is suspected orin those with recurrent symptoms, MII/pH esophagealmonitoring in combination with polysomnographicrecording and precise, synchronous symptom recordingmay aid in establishing cause and effect.

Reactive Airways Disease

In patients with asthma who also have heartburn,reflux may be a contributing factor to the asthma. Despitea high frequency of abnormal reflux studies in patients




with asthma who do not have heartburn, there is no strongevidence to support empiric PPI therapy in unselectedpediatric patients with wheezing or asthma. Only 3groups—those with heartburn, those with nocturnalasthma symptoms, and those with steroid-dependentdifficult-to-control asthma—may derive some benefitfrom long-term medical or surgical antireflux therapy.Finding abnormal esophageal pH exposure by esopha-geal pH monitoring, with or without impedance, beforeconsidering a trial of long-term PPI therapy or surgerymay be useful, although the predictive value of thesestudies for this purpose is not established. The relativeefficacy of medical versus surgical therapy for GERD inchildren with asthma is unknown.

Recurrent Pneumonia

Recurrent pneumonia and interstitial lung disease maybe the complications of GER due to aspiration of gastriccontents. No test can determine whether GER is causingrecurrent pneumonia. An abnormal esophageal pH testmay increase the probability that GER is a cause ofrecurrent pneumonia but is not a proof thereof. Nuclearscintigraphy can detect aspirated gastric contents whenimages are obtained for 24 hours after enteral adminis-tration of a labelled meal. Aspiration during swallowingis more common than aspiration of refluxed material. Atrial of nasogastric feeding may be used to excludeaspiration during swallowing as a potential cause ofrecurrent disease. A trial of nasojejunal therapy mayhelp in determining whether surgical antireflux therapyis likely to be beneficial. In patients with severelyimpaired lung function, antireflux surgery may benecessary to prevent further pulmonary damage, despitelack of definitive proof that GER is causative.

Upper Airway Symptoms

The data linking reflux to chronic hoarseness, chroniccough, sinusitis, chronic otitis media, erythema, andcobblestone appearance of the larynx comes mainly fromcase reports and case series. The association of refluxwith these conditions and response to antisecretorytherapy have not been proven by controlled studies.Patients with these symptoms or signs should not beassumed to have GERD without consideration of otherpotential etiologies.

Dental Erosions

An association between GERD and dental erosions isestablished. The severity of dental erosions seems to becorrelated with the presence of GERD symptoms, and, inadults, with the severity of proximal esophageal or oralexposure to an acidic pH. Young children and childrenwith neurologic impairment appear to be at the greatest

risk. Factors other than reflux that cause similar dentalerosions include juice drinking, bulimia, and racial andgenetic factors affecting the characteristics of enameland saliva.

Dystonic Head Posturing (Sandifer Syndrome)

Sandifer syndrome (spasmodic torsional dystonia witharching of the back and opisthotonic posturing, mainlyinvolving the neck and back) is an uncommon butspecific manifestation of GERD. It resolves with antire-flux treatment.

Groups at High Risk for GERD

Certain conditions are predisposed to severe, chronicGERD. These include neurologic impairment, obesity,repaired esophageal atresia or other congenital esopha-geal disease, cystic fibrosis, hiatal hernia, repaired acha-lasia, lung transplantation, and a family history of GERD,BE, or esophageal adenocarcinoma. Although manypremature infants are diagnosed with GERD becauseof nonspecific symptoms of feeding intolerance, apneaspells, feeding refusal, and pain behavior, there are nocontrolled data that confirm reflux as a cause. Althoughreflux may be more common in infants with broncho-pulmonary dysplasia, there is no evidence that antirefluxtherapy impacts upon the clinical course or outcome ofthis condition.

1. RATIONALE

The North American Society for Pediatric Gastroen-terology, Hepatology, and Nutrition (NASPGHAN) pub-lished the first clinical practice guidelines on pediatricgastroesophageal reflux (GER) and gastroesophagealreflux disease (GERD) in 2001 (1). Consensus-basedguidelines on several aspects of GER and GERD weredeveloped in Europe at about the same time but were notofficially endorsed by the European Society of PediatricGastroenterology, Hepatology, and Nutrition (ESP-GHAN) (2,3). In 2007, the Councils of ESPGHANand NASPGHAN established a joint committee toreview, update, and unify these guidelines as a meansof improving uniformity of practice and quality of patientcare (4,5).

The Committee used the 2001 NASPGHAN guidelinesas an outline, adding new sections on certain pediatricpopulations at high risk for GERD. In all deliberations,the Committee attempted to distinguish physiologic GERevents from GERD. Further, in response to evidence thatthe diagnosis of GERD is applied excessively to healthyinfants with bothersome but harmless symptoms ofphysiologic GER (6–9), the Committee reevaluatedthe 2001 diagnostic and therapeutic algorithms to clarifythe distinction between physiologic GER and GERD. In

PEDIATRIC GASTROESOPHAGEAL REFLUX CLINICAL PRACTICE GUIDELINES 503



its recommendations for testing, the Committee con-fronted the ongoing problem that current reflux testsmay identify variations from normal but cannot predictsymptom severity, natural history, or response to therapy.

These guidelines are designed to assist pediatric healthcare providers in the diagnosis and management of GERand GERD. They are intended to serve as general guide-lines and not as a substitute for clinical judgement, or as aprotocol applicable to all of the patients.

2. METHODS

2.1. Selection of Committee Members

The NASPGHAN–ESPGHAN Joint Guideline Committeeincluded 5 European and 4 North American pediatric gastro-enterologists with extensive experience in GER and GERD,selected by their respective societies, and 2 North Americanprimary care pediatricians experienced in clinical epidemiol-ogy. Both pediatric epidemiologists, members of the AmericanAcademy of Pediatrics Section on Epidemiology, were selectedbecause of their contribution to the previous NAPSGHAN-GERD guidelines.

2.2. Guideline Preparation Process

The previous guidelines developed by NASPGHAN (1) andESPGHAN (2,3) were used as the foundation for the currentguidelines. Articles written in English and published betweenMarch 1999 (the date of the previous review) and October 2008were identified using PubMed and Cumulative Index to Nursingand Allied Health Literature. Letters, editorials, case reports,and reviews were eliminated from the initial evaluation.Additional articles were identified by members of the commit-tee, from bibliographies found in other articles, and studyresults in the public domain on the US National Institutes ofHealth Web site. These included review articles as well asarticles that involved the care of adults. A total of 377 articlesrelated to therapy, and 195 articles related to etiology, diagnosis,and prognosis were reviewed for this guideline.

Using the best available evidence from the literature, thecommittee evaluated current diagnostic tests and therapeuticmodalities for GER and GERD. Evidence of a causal relationbetween GER/GERD and several common symptoms or symp-tom complexes was evaluated. Diagnostic tests were evaluated bythe following criteria: ability to confirm a diagnosis of GERD;ability to exclude other diagnoses with similar presentation;ability to detect complications of GERD; and ability to predictdisease severity, natural history of disease and response to treat-ment. Therapy was evaluated considering efficacy, appropriateclinical indications, and potential risks and complications.

The committee convened face-to-face 3 times and hadseveral conference calls. It based its recommendations on itsstudy of the literature review combined with expert opinion andthe evidence available in the adult literature when pediatricevidence was insufficient. Consensus was achieved for all of therecommendations through Nominal Group Technique, a struc-tured quantitative method (10). Articles were evaluated usingthe Oxford Centre for Evidence-based Medicine–Levels ofEvidence (11). Using the Oxford grades of recommendation

(11) the quality of evidence of each of the recommendationsmade by the committee was determined and is summarized inAppendices A to C. Sections of the document were written byindividual committee members then reviewed and edited by aseparate committee member; in most instances both a NASP-GHAN and ESPGHAN member participated in preparing theinitial draft of each section. These sections and other evidencesavailable in previously prepared tables that listed references andgraded the quality of each reference were distributed, thenreviewed and discussed to achieve consensus agreement inconference sessions. The document was then distributed toall of the NASPGHAN members for comment. Further revi-sions were made based on their suggestions following phoneconference and e-mail communications among committeemembers. Complete voting anonymity could not be maintainedthrough the revision process because voting was done by e-mail,but only one of the co-chairs (C.D.R.) was aware of e-mailvotes. Following final committee approval, the document wasendorsed by the Executive Councils of NASPGHAN andESPGHAN.

2.3. Management of Potential Conflict of Interest

Disclosures of potential conflicts of interest of committeemembers or immediate family were documented and sharedwith committee members before the first meeting of the com-mittee and updated before the review of the final document.Disclosures included paid or donated services of any kind,research support, stock ownership or options, and intellectualproperty rights. During the process of preparing the guidelines,the scientific data were reviewed by all of the members of thecommittee, and recommendations were voted on by all of themembers. No section of the document was written solely by any1 member. Chairs or committee members did not require thatany individual be removed from discussions or voting based onpotential conflicts of interest. Potential conflicts of interest arelisted in Appendix D. No industry support was used for theproduction of these guidelines.

3. DEFINITIONS AND MECHANISMS

GER is the passage of gastric contents into the eso-phagus with or without regurgitation and vomiting. GERis a normal physiologic process occurring several timesper day in healthy infants, children, and adults. Mostepisodes of GER in healthy individuals last


infants


heartburn, were the most frequently reported symptomsin children and adolescents with GERD. Cough andanorexia or feeding refusal were more common in chil-dren 1 to 5 years of age than in older children (40).

Symptoms and signs associated with reflux (Table 1) arenonspecific. For example, not all of the children with GERhave heartburn or irritability. Conversely, heartburn andirritability can be caused by conditions other than GER.Regurgitation, irritability, and vomiting are common ininfants with physiologic GER or GERD (14,18,41,42) butare indistinguishable from regurgitation, irritability, andvomiting caused by food allergy (43,44), colic (45,46), andother disorders. The severity of reflux or esophagitis foundon diagnostic testing does not directly correlate with theseverity of symptoms (47–49).

GERD is often diagnosed clinically in adults based ona history of heartburn, defined as substernal, burningchest pain, with or without regurgitation. Recent adultand pediatric consensus guidelines have applied theterms ‘‘typical reflux syndrome’’ or ‘‘reflux chest painsyndrome’’ to this presentation (13,50). Based on expertopinion, the diagnosis of GERD can be made in adoles-cents presenting with typical heartburn symptoms as inadults (49,51–55). However, a clinical diagnosis basedon a history of heartburn cannot be used in infants,children, or nonverbal adolescents (eg, those with NI)because these individuals cannot reliably communicatethe quality and quantity of their symptoms. The verbalchild can communicate pain, but descriptions of quality,intensity, location, and severity generally are unreliableuntil at least 8 and possibly 12 years of age (56–60).

As in adults, individual symptoms in children gener-ally are not highly predictive of findings of GERD byobjective studies. For example, in a study of irritableinfants younger than 9 months of age, regurgitation >5times per day had a sensitivity of 54% and specificity of71% for a reflux index (RI) >10% by esophageal pH

testing, where as feeding difficulties had a sensitivity of75% and specificity of 46% (61). A similar poor corre-lation of symptoms and esophageal acid exposure wasobserved during an omeprazole-treatment study inirritable infants; similar reductions in crying occurredin both treated and untreated infants and the extent ofreduction crying did not correlate with extent ofreduction of the RI in the treated patients (46).

Because individual symptoms do not consistently cor-relate with objective findings or response to medicaltreatment, parent or patient-reported questionnairesbased on clusters of symptoms have been developed.Orenstein et al (51,62) developed a diagnostic question-naire for GERD in infants. A score of>7 (of 25 possible)on the initial instrument demonstrated a sensitivity of0.74 and specificity of 0.94 during primary validation.The questionnaire has undergone several revisions (54).The questionnaire has been shown to be reliable fordocumentation and monitoring of reported symptoms.However, when applied to a population in India, it had asensitivity and specificity of only 43% and 79%, respect-ively compared with pH monitoring results (52). Inanother study of infants referred for symptoms of refluxdisease and controls, the questionnaire had sensitivityand specificity of 47% and 81% for a RI >10% and 65%and 63% for a reflux index>5%. The questionnaire scorefailed to identify 26% of the infants with GERD. Thescore was positive in 17 of 22 infants with normalbiopsies and pH studies and in 14 of 47 infants withnormal pH studies. No single symptom was significantlyassociated with esophagitis (49). In another study, thequestionnaire was unable to identify a group of infantsresponsive to proton pump inhibitor (PPI) therapy (9).Thus, no symptom or cluster of symptoms has beenshown to reliably predict complications of reflux or topredict those infants likely to respond to therapy.

A 5-item questionnaire developed for children 7 to16 years of age had a sensitivity of 75% and specificityof 96% compared with pH monitoring during primaryvalidation (63). No subsequent independent confirmatoryvalidation has been performed. Other diagnostic question-naires, such as the GERD symptom questionnaire (53),have not been compared with objective standards likeendoscopy, pH monitoring, or esophageal multiple intra-luminal impedance (MII) monitoring. Some researchershave used questionnaires to monitor symptoms of childrenduring GERD therapy (64). Whether this method is pre-ferable to monitoring individual symptoms is uncertain.Although daily symptom diaries are frequently used inadults to monitor the effects of therapy, these have not beenvalidated in children.

4.2. Esophageal pH Monitoring

Intraluminal esophageal pH monitoring measuresthe frequency and duration of acid esophageal reflux

TABLE 2. Warning signals requiring investigation in infantswith regurgitation or vomiting

Bilious vomitingGastrointestinal bleeding

HematemesisHematochezia

Consistently forceful vomitingOnset of vomiting after 6 months of lifeFailure to thriveDiarrheaConstipationFeverLethargyHepatosplenomegalyBulging fontanelleMacro/microcephalySeizuresAbdominal tenderness or distensionDocumented or suspected genetic/metabolic syndrome




episodes. Most commercially available systems include acatheter for nasal insertion with 1 or more pH electrodes(antimony, glass or ion-sensitive field effect) arrayedalong its length and a system for data capture, analysis,and reporting. Slow electrode response times (antimonybeing the slowest) do not alter the assessment of totalreflux time substantially but may affect the accuracy ofcorrelation between symptoms and reflux episodes (65).Esophageal pH monitoring is insensitive to weakly-acidand non-acid reflux events. Recently, wireless sensorsthat can be clipped to the esophageal mucosa duringendoscopy have allowed pH monitoring without a nasalcannula for up to 48 hours. Placement of wireless elec-trodes requires sedation or anesthesia, and comfort hasbeen an issue in some studies (66–68). The size ofcurrent wireless electrodes precludes their use in smallinfants. Benefits, risks, and indications for wireless elec-trode monitoring have not been fully defined in children.Data on reproducibility of conventional and wireless pHstudies are contradictory (68–72).

By convention, a drop in intraesophageal pH


diagnosis other than GERD (88,89). Esophageal pHmonitoring is useful for evaluating the efficacy of anti-secretory therapy. It may be useful to correlate symptoms(eg, cough, chest pain) with acid reflux episodes, and toselect those children with wheezing or respiratory symp-toms where acid reflux may be an aggravating factor. Thesensitivity and specificity of pH monitoring are notwell established.

4.3. Combined Multiple Intraluminal Impedance(MII) and pH Monitoring

MII is a procedure for measuring the movement offluids, solids, and air in the esophagus (102). It is arelatively new technology that provides a more detaileddescription of esophageal events with a more rapidresponse time than current pH monitoring technology.MII measures changes in the electrical impedance (ie,resistance) between multiple electrodes located along anesophageal catheter. Esophageal impedance tracings areanalyzed for the typical changes in impedance caused bythe passage of liquid, solid, gas, or mixed boluses. If theimpedance changes of a liquid bolus appear first in thedistal channels and proceed sequentially to the proximalchannels, they indicate retrograde bolus movement, thatis GER. The direction and velocity of a bolus can becalculated using the defined distance between electrodesand the time between alterations in the impedance patternof sequential electrode pairs. The upward extent of thebolus and the physical length of the bolus can also beevaluated (103). MII can detect extremely small bolusvolumes (104).

MII and pH electrodes can and should be combined ona single catheter. The combined measurement of pH andimpedance (pH/MII) provides additional information asto whether refluxed material is acidic, weakly acidic, ornonacidic (105–109). Recent studies have found variablereproducibility (110,111). Evaluation of MII recordingsis aided by automated analysis tools (112). But untilthe currently available automatic analysis software hasbeen validated, a visual reading of the data is required.Normal values for all of the age groups have not yet beenestablished (113).

The risks and side effects of MII are low and the sameas those of isolated pH monitoring. The combination ofpH/MII with simultaneous monitoring of symptomsusing video-polysomnography or manometry has provenuseful for the evaluation of symptom correlationsbetween reflux episodes and apnea, cough, other respir-atory symptoms, and behavioral symptoms (23,24,114–116). The technology is especially useful in the post-prandial period or at other times when gastric contentsare nonacidic. The relation between weakly-acid refluxand symptoms of GERD requires clarification. Measure-ment of other parameters such as SI or SAP may be ofadditional value to prove symptom association with

reflux, especially when combined with MII (117).Whether combined esophageal pH and impedancemonitoring will provide useful measurements that varydirectly with disease severity, prognosis, and response totherapy in pediatric patients has yet to be determined.

4.4. Motility Studies

Esophageal manometry measures esophageal peristal-sis, upper and lower esophageal sphincter pressures, andthe coordinated function of these structures during swal-lowing. Although esophageal manometry has been animportant tool in studying the mechanisms of GERD,GERD cannot be diagnosed by esophageal manometry.Manometric studies were critical in identifying TLESRas a causative mechanism for GERD (21). A variety ofnonspecific esophageal motor-abnormalities have beenfound in children with developmental delay and NI, agroup at high risk for severe GERD (118). Esophagealmotor-abnormalities are also common in patients withesophagitis (119,120). In these 2 situations esophagealmotor dysfunction may be a secondary phenomenonrelated to esophagitis as it has been observed to resolveupon treatment of esophagitis (119). Recent studiesindicate that there is no role for manometry in predictingoutcome of fundoplication (121). Manometric studies arealso important in confirming a diagnosis of achalasia orother motor disorders of the esophagus which may mimicGERD (122).

Esophageal manometry may be abnormal in patientswith GERD but the findings are not sufficiently sensitiveor specific to confirm a diagnosis of GERD, nor to predictresponse to medical or surgical therapy. It may be usefulin patients who have failed acid suppression and whohave negative endoscopy to search for a possible motilitydisorder, or to determine the position of the LES to placea pH probe. Manometric studies are useful to confirm adiagnosis of achalasia or other motor disorders of theesophagus that may mimic GERD.

4.5. Endoscopy and Biopsy

Upper gastrointestinal (GI) endoscopy allows directvisual examination of the esophageal mucosa. Mucosalbiopsies enable evaluation of the microscopic anatomy(123). Macroscopic lesions associated with GERD includeesophagitis, erosions, exudate, ulcers, strictures, HH, areasof possible esophageal metaplasia, and polyps. Althoughendoscopy can detect strictures, subtle degrees of narrow-ing may be better shown on barium contrast study, duringwhich the esophagus can be distended with various tech-niques, such as a radio-opaque pill, and barium-soakedbread or marshmallows. Malrotation and achalasia cannotbe diagnosed by endoscopy. These and other anatomic andmotility disorders of the esophagus are better evaluated bybarium radiology or motility studies.




Recent global consensus guidelines define reflux eso-phagitis as the presence of endoscopically-visible breaksin the esophageal mucosa at or immediately above the GEjunction (13,50,124). Evidence from adult studiesindicates that visible breaks in the esophageal mucosaare the endoscopic signs of greatest interobserverreliability (125–127). Operator experience is an import-ant component of inter-observer reliability (128,129).Mucosal erythema or an irregular Z-line is not a reliablesign of reflux esophagitis (126,127). Grading the severityof esophagitis, using a recognized endoscopic classifi-cation system is useful for evaluation of the severity ofesophagitis and response to treatment. The Hetzel–Dentclassification (125) has been used in several pediatricstudies (29,130,131), where as the Los Angeles classifi-cation (124) is generally used for adults, but is suitablealso for children. The presence of endoscopically-normalesophageal mucosa does not exclude a diagnosis ofNERD or esophagitis of other etiologies (93,132,133).

The diagnostic yield of endoscopy is generally greaterif multiple samples of good size and orientation areobtained from biopsy sites that are identified relativeto major esophageal landmarks (28,123,134). Severalvariables have an impact on the validity of histologyas a diagnostic tool for reflux esophagitis (133,135).These include sampling error because of the patchydistribution of inflammatory changes and a lack instandardization of biopsy location, tissue processing,and interpretation of morphometric parameters.Histology may be normal or abnormal in NERD becauseGERD is an inherently patchy disease (133,136). Histo-logic findings of eosinophilia, elongation of papillae (retepegs), basal hyperplasia, and dilated intercellular spaces(spongiosis) are neither sensitive nor specific for refluxesophagitis. They are nonspecific reactive changes thatmay be found in esophagitis of other causes, or in healthyvolunteers (49,89,132,133,135,137–141). Recent studieshave shown considerable overlap between the histologyof reflux esophagitis, and EoE (93,94,132,142). Manyhistologic parameters are influenced by drugs used totreat esophagitis or other disorders.

GERD is likely the most common cause of esophagitisin children, but other disorders such as EoE, Crohndisease, and infections also cause esophagitis (Table 3)(132). EoE and GERD have similar symptoms and signs,and can be best distinguished by endoscopy with biopsy.A key difference, endoscopically, is that EoE is generallynot an erosive disease, but has its own typical endoscopicfeatures such as speckled exudates, trachealization of theesophagus, or linear furrowing. In up to 30% of cases,however, the esophageal mucosal appearance is normal(93). When EoE is considered as a part of the differentialdiagnosis, it is advisable to take esophageal biopsies fromthe proximal and distal esophagus (93). Mucosal eosi-nophilia may be present in the esophageal mucosa inasymptomatic infants


positive predictive value of the upper GI series rangefrom 29% to 86%, 21% to 83%, and 80% to 82%,respectively, when compared with esophageal pHmonitoring (151–157). The brief duration of the upperGI series produces false-negative results, where as thefrequent occurrence of nonpathological reflux during theexamination produces false-positive results.

Therefore, routine performance of upper GI series todiagnose reflux or GERD is not justified (158). However,the upper GI series is useful to detect anatomic abnorm-alities such as esophageal stricture, HH, achalasia,tracheoesophageal fistula, intestinal malrotation, orpyloric stenosis, which may be considered in the differ-ential diagnosis of infants and children with symptomssuggesting GERD.

4.7. Nuclear Scintigraphy

In gastroesophageal scintigraphy, food or formulalabeled with 99Technetium is introduced into the stomachand areas of interest—stomach, esophagus, and lungs—are scanned for evidence of reflux and aspiration. Thenuclear scan evaluates only postprandial reflux anddemonstrates reflux independent of the gastric pH. Scin-tigraphy can provide information about gastric emptying,which may be delayed in children with GERD (159–161). A lack of standardized techniques and the absenceof age-specific norms limit the value of this test. Sensi-tivity and specificity of a 1-hour scintigraphy for thediagnosis of GERD are 15% to 59% and 83% to 100%,respectively, when compared with 24-hour esophagealpH monitoring (162–165). Late postprandial acidexposure detected by pH monitoring may be missed withscintigraphy (166).

Gastroesophageal scintigraphy scanning can detectreflux episodes and aspiration occurring during or shortlyafter meals, but its reported sensitivity for microaspira-tion is relatively low (167–169). Evidence of pulmonaryaspiration may be detected during a 1-hour scintigraphicstudy or on images obtained up to 24 hours after admin-istration of the radionuclide (170). A negative test doesnot exclude the possibility of infrequently occurringaspiration (168). One study of children with refractoryrespiratory symptoms found that half had scintigraphicevidence of pulmonary aspiration (169). However,aspiration of both gastric contents and saliva also occursin healthy adults during deep sleep (171,172).

Gastric emptying studies have shown prolonged half-emptying times in children with GER. Delayed gastricemptying may predispose to GERD, but tests of gastricemptying are not a part of the routine examination ofpatients with suspected GERD, but may be importantwhen symptoms suggest gastric retention (173–176).

Nuclear scintigraphy is not recommended in the rou-tine diagnosis and management of GERD in infantsand children.

4.8. Esophageal and Gastric Ultrasonography

Ultrasonography is not recommended as a test forGERD but can provide information not available throughother technology. Ultrasonography of the GE junctioncan detect fluid movements over short periods of timeand thereby can detect nonacid reflux events. It can alsodetect HH, length and position of the LES relative to thediaphragm, and magnitude of the gastroesophagealangle of His. Barium upper GI series can provide thesame information. When compared with the results of24-hour esophageal pH testing as a diagnostic test forGERD, the sensitivity of color Doppler ultrasound per-formed for 15 minutes postprandially is about 95% witha specificity of only 11%, and there is no correlationbetween reflux frequency detected by ultrasound andreflux index detected by pH monitoring (177,178).Intraluminal esophageal ultrasound is used in adults toevaluate esophageal wall thickness and muscle short-ening, parameters that vary with inflammation, scarring,and malignancy (179). At present, there is no role forultrasound as a routine diagnostic tool for GERD inchildren.

4.9. Tests on Ear, Lung, and Esophageal Fluids

Recent studies have suggested that finding pepsin, agastric enzyme, in middle ear effusions of children withchronic otitis media indicates that reflux is playing anetiologic role (180–183). One recent study showed norelation between the presence of pepsin in the middle earand symptoms of GERD (184), and this relation has notbeen validated in controlled treatment trials. Similarly,the presence of lactose, glucose, pepsin, or lipid filledmacrophages in bronchoalveolar lavage fluids has beenproposed to implicate aspiration secondary to reflux as acause of some chronic pulmonary conditions (185–187).No controlled studies have proven that reflux is the onlyreason these compounds appear in bronchoalveolarlavage fluids or that reflux is the cause of pulmonarydisease when they are present.

Continuous monitoring of bilirubin in the esophagushas been suggested as a means of detecting esophagealreflux of duodenal juice or duodenogastroesophagealreflux . Duodenal juice components appear to damagethe esophagus in a pH dependent manner (188). Twouncontrolled pediatric case series have suggested thatduodenogastroesophageal reflux produced GERD thatwas refractory to therapy with PPIs (189,190). One studyindicated that therapy with PPI decreased the esophagealdamage caused by duodenogastroesophageal reflux(190). At present, there is insufficient evidence to recom-mend continuous monitoring of the esophagus for bilir-ubin in the routine evaluation of GERD. The role of bilereflux in children resistant to PPI treatment has notbeen established.




4.10. Empiric Trial of Acid Suppression as aDiagnostic Test

In adults, empiric treatment with acid suppression, thatis, without diagnostic testing, has been used for symp-toms of heartburn (191), chronic cough (192,193), non-cardiac chest pain (194), and dyspepsia (195). However,empiric therapy has only modest sensitivity and speci-ficity as a diagnostic test for GERD depending upon thecomparative reference standard used (endoscopy, pHmonitoring, symptom questionnaires) (196), and theappropriate duration of a ‘‘diagnostic trial’’ of acidsuppression has not been determined. A meta-analysisevaluating pooled data from 3 large treatment trialsamong the adults with NERD showed that 85% of thepatients who had symptom resolution after 1 week of PPItreatment remained well for the entire 4 weeks of PPItreatment, thus ‘‘confirming’’ the diagnosis of GERD(197). However, 22% of the patients who had noimprovement after 1 week of treatment did improve bythe fourth week of treatment. An uncontrolled trial ofesomeprazole therapy in adolescents with heartburn,epigastric pain, and acid regurgitation showed completeresolution of symptoms in 30% to 43% by 1 week, but theresponders increased to 65% following 8 weeks of treat-ment (55). Another uncontrolled treatment trial of pan-toprazole in children ages 5 to 11 years of age reportedgreater symptom improvement at 1 week with one 40-mgdose compared with one 10-mg or 20-mg dose (64). After8 weeks all of the treatment groups improved. Similarimprovement in symptoms over time has been observedin adults with erosive esophagitis (198,199). One study ofinfants with symptoms suggestive of GERD who weretreated empirically with a PPI showed no efficacy overplacebo (9).

The treatment period required to achieve uniformtherapeutic responses with PPI therapy probably varieswith disease severity, treatment dose and specific symp-toms or complications (200). The 2-week ‘‘PPI test’’lacks adequate specificity and sensitivity for use inclinical practice. In an older child or adolescent withsymptoms suggesting GERD, an empiric PPI trial isjustified for up to 4 weeks. Improvement following treat-ment does not confirm a diagnosis of GERD becausesymptoms may improve spontaneously or respond by aplacebo effect. There is no evidence to support an empirictrial of pharmacologic treatment in infants and youngchildren as a diagnostic test of GERD.

5. TREATMENT

Management options for physiologic GER and forGERD discussed in this section include lifestyle changes,pharmacologic therapy, and surgery. Lifestyle changes ininfants with physiologic GER include nutrition, feeding,and positional modifications. In older children and ado-

lescents, lifestyle changes include modification of dietand sleeping position, weight reduction, and smokingcessation.

Medications for use in GERD include agents to buffergastric contents or suppress acid secretion. Agents affect-ing GI motility are discussed. Surgical therapy includesfundoplication and other procedures to eliminate reflux.

5.1. Lifestyle Changes

Parental education, guidance, and support are alwaysrequired and usually sufficient to manage healthy, thriv-ing infants with symptoms likely because of physiologicGER.

5.1.1. Feeding Changes in Infants

About 50% of the healthy 3- to 4-month old infantsregurgitate at least once a day (16,18) and up to 20% ofcaregivers in the United States seek medical help for thisnormal behavior (16). Breast-fed and formula-fed infantshave a similar frequency of physiologic GER, althoughthe duration of reflux episodes measured by pH probemay be shorter in breast-fed infants (201–203).

A subset of infants with allergy to cow’s milk proteinexperience regurgitation and vomiting indistinguishablefrom that associated with physiologic GER (9,69,142,204–206). In these infants, vomiting frequency decreasessignificantly (usually within 2 weeks) after the elimin-ation of cow’s milk protein from the diet, and reintroduc-tion causes recurrence of symptoms (206,207). Studiessupport the use of extensively hydrolyzed or amino acidformula in formula-fed infants with bothersome regur-gitation and vomiting for trials lasting up to 4 weeks(206–208). Cow’s milk protein and other proteins, passinto human breast milk in small quantities. Breast-fedinfants with regurgitation and vomiting may thereforebenefit from a trial of withdrawal of cow’s milk and eggsfrom the maternal diet (209,210). The symptoms of infantreflux are almost never so severe that breast-feedingshould be discontinued. There are no studies specificallyevaluating soy protein allergy in infants with regurgita-tion and vomiting, or the role of soy protein–basedformula in the treatment of infants with regurgitation.Moreover, there are no data on allergy to possibleformula thickening agents such as rice cereals.

One study in infants showed that large volume feed-ings promote regurgitation, probably by increasing thefrequency of TLESR, and reduced feeding volume anddecreased reflux frequency (211). Severe reduction infeeding volume during an extended period may deprivethe infant of needed energy and adversely affect weightgain. Infants with inadequate weight gain because oflosses by regurgitation may benefit from increasing theenergy density of formula when volume or frequency offeedings is decreased as a part of therapy.




Adding thickening agents such as rice cereal toformula does not decrease the time with pH


Esophageal pH and combined pH/MII monitoringshow that reflux is quantitatively similar in the left-side-down and prone positions. Measured reflux in these2 positions is less than in the right-side-down and supinepositions (234,245–247). Two impedance studies ofpreterm infants found that postprandial reflux was greaterin the right-side-down than in the left-side-down position(173,235). Based on these findings, one study recom-mended that infants be placed right-side-down for thefirst hour after feeding to promote gastric emptying andthen switched to left-side-down thereafter to decreasereflux (173). These findings notwithstanding, it is import-ant to note that side-lying is an unstable position for aninfant who may slip unobserved into the prone position.Bolstering an infant with pillows to maintain a side-lyingposition is not recommended (248).

5.1.3. Lifestyle Changes in Children and Adolescents

Lifestyle changes often recommended for children andadolescents with GER and GERD include dietary modi-fication, avoidance of alcohol, weight loss, positioningchanges, and cessation of smoking. Most studies inves-tigating these recommendations have been performed inadults, thus their applicability to children of all ages isuncertain. A review of lifestyle changes in adults withGERD concluded that only weight loss improved pHprofiles and symptoms (249). Although alcohol, choco-late, and high-fat meals reduce LES pressure, only a fewstudies have evaluated the impact of these factors onsymptoms. Tobacco smoke exposure is associated withincreased irritability in infants, yet neither tobacco noralcohol cessation has been shown to improve esophagealpH profiles or symptoms. One uncontrolled study (250)found that a low-carbohydrate diet reduced distal eso-phageal acid exposure and improved symptoms in obeseindividuals with GERD. Gastric bypass surgery signifi-cantly improved symptoms of GERD in obese adults(251). Another study (252) detected more overnightreflux in adults eating a late evening meal than in adultseating an earlier evening meal. The difference wasespecially obvious in overweight adults.

Current evidence generally does not support (or refute)the use of specific dietary changes to treat reflux beyondinfancy. Expert opinion suggests that children and ado-lescents with GERD should avoid caffeine, chocolate,alcohol, and spicy foods if they provoke symptoms (253–264). In an overweight individual, weight loss doesdecrease reflux, and is therefore recommended (250–252,265–267). Smoking should be avoided in those withGERD because it has been linked to adenocarcinoma ofthe esophagus in adults (268,269). Three studies haveshown that chewing sugarless gum after a meal decreasesreflux (270–272). It is not known whether any lifestylechanges have an additive benefit in children or adoles-cents receiving pharmacological therapy.

The effectiveness of positioning for treatment of GERand GERD in children older than 1 year of age has notbeen studied. It is unclear whether the benefits of pos-itional therapy identified in adults and infants


histopathology scores occurred only in the cimetidine-treated group. Another randomized study in 24 childrenwith mild-to-moderate esophagitis demonstrated thatnizatidine (10 mg $ kg%1 $ day%1) was more effective thanplacebo for the healing of esophagitis and symptom relief(288). There are case series providing additional supportfor the efficacy of H2RA in infants and children (289–294). Although no randomized controlled studies inchildren demonstrate the efficacy of ranitidine or famo-tidine for the treatment of esophagitis, expert opinion isthat these agents are as effective as cimetidine andnizatidine. Extrapolation of the results of a large numberof adult studies to older children and adolescents suggeststhat H2RA may be used in these patients for the treatmentof GERD symptoms and for healing esophagitis,although H2RA are less effective than PPI for bothsymptom relief and healing of esophagitis (283,295,296).

The fairly rapid tachyphylaxis that develops withH2RA is a drawback to chronic use. In some infants,H2RA therapy causes irritability, head banging, head-ache, somnolence, and other side effects which, if inter-preted as persistent symptoms of GERD, could result inan inappropriate increase in dosage (293). H2RA,particularly cimetidine, are associated with an increasedrisk of liver disease (297,298), and cimetidine withgynecomastia (299). Other adverse effects of suppressionof gastric acid are discussed in the section on PPIs.

5.2.2. Proton Pump Inhibitors

PPIs inhibit acid secretion by blocking Naþ–Kþ

ATPase, the final common pathway of parietal cell acidsecretion, often called the proton pump. Studies in adultshave shown that PPIs produce higher and faster healingrates for erosive esophagitis than H2RAs, which in turnare better than placebo (122). The superior efficacy ofPPIs is largely because of their ability to maintainintragastric pH at or above 4 for longer periods and toinhibit meal-induced acid secretion, a characteristic notshared by H2RAs. In contrast with H2RAs, the effect ofPPIs does not diminish with chronic use. The potentsuppression of acid secretion by PPIs also results indecrease of 24-hour intragastric volumes, thereby facil-itating gastric emptying and decreasing volume reflux(300). Despite their efficacy in the management of acid-related disorders, PPIs have limitations as a consequenceof their pharmacologic characteristics. They must betaken once a day, before breakfast and must be protectedfrom gastric acid by enteric coatings. Bioavailability ofPPIs is decreased if they are not taken before meals.However, taking the medications before meals effectivelydelays absorption and onset of their antisecretory effect.Most available PPIs are therefore regarded as ‘‘delayedrelease’’ preparations. Achievement of maximal acidsuppressant effect can take up to 4 days (301). However,a summary of adult data suggests that PPIs can also be

used for ‘‘on-demand’’ treatment of symptoms (302).One commercially available ‘‘immediate-release’’ PPI isuncoated omeprazole with added bicarbonate (302).There are no data available concerning its use in children.Dexlansoprazole MR is said to be less dependent onbeing taken on an empty stomach. This new medicationhas 2 delayed-release mechanisms, and therefore a longerduration of acid suppression (303). The clinical import-ance of this modification has yet to be determined. Thereare no pediatric clinical trials and the drug is notapproved for use in children.

PPIs currently approved for use in children in NorthAmerica are omeprazole, lansoprazole, and esomepra-zole. At this moment, in Europe, only omeprazole andesomeprazole are approved. No PPI has been approvedfor use in infants


are 4 main categories of adverse effects related to PPIs -idiosyncratic reactions, drug–drug interactions, drug-induced hypergastrinemia, and drug-induced hypochlor-hydria. Idiosyncratic side effects occur in up to 14% ofchildren taking PPIs (28,311,312). The most common areheadache, diarrhea, constipation, and nausea, each occur-ring in 2% to 7%. These may resolve with decreased doseor change to a different PPI. Parietal cell hyperplasia(313,314) and occasional fundic gland polyps (315) arebenign changes resulting from PPI-induced acid suppres-sion and hypergastrinemia. Enterochromaffin-like cellhyperplasia is also a result of acid suppression. A pro-spective study monitoring patients treated for up to2 years (316), and retrospective studies of patients treatedup to 11 years (28), have found only mild grades ofenterochromaffin-like cell hyperplasia. A recent retro-spective study using sensitive staining techniques (317)showed ECL hyperplasia in the gastric body of almosthalf of children receiving long-term PPI continuously fora median of 2.84 years (up to 10.8 years); the hyperplasiawas of the lowest 2 grades (not clinically significant), andno patient developed atrophic gastritis or carcinoidtumors.

Increasing evidence suggests that hypochlorhydria,that is, acid suppression, associated with H2RA or PPImay increase rates of community-acquired pneumonia inadults and children, gastroenteritis in children, and can-didemia and necrotizing enterocolitis in preterm infants(318–322). In 1 study, PPI but not H2RA was associatedwith bacterial enterocolitis in adults. Doubling of the PPIdose increased the risk (323). Infants treated with PPI in astudy (9) had a significantly higher rate of all adverseeffects compared with the placebo group. Lower respir-atory tract infections were the most frequent among theseadverse effects, although the difference in respiratorytract infection rate between treated and placebo groupsdid not achieve statistical significance. PPIs have beenshown to alter the gastric and intestinal bacterial flora inadults (324). The effect of PPI therapy on the flora ofinfants and childen, or the consequences of any altera-tions have not been evaluated.

Other adverse effects have been reported in elderlypatients on chronic PPI therapy, such as deficiency ofvitamin B12 and increased incidence of hip fractures(325,326) but these findings have not been corroboratedby recent studies (327,328). In a retrospective casereview, 18 cases of biopsy-proven PPI-induced acuteinterstitial nephritis causing acute renal failure werereported, and the authors suggest this entity may gounrecognized as ‘‘unclassified acute renal failure’’(329). PPIs are considered to be the commonest causeof acute interstitial nephritis in adults (330). Thisadverse effect is considered to be an idiosyncraticreaction, more frequent in the elderly. No childhoodcases have been described. Animal studies suggest thatacid suppression may predispose to the development of

food allergy (331) but this remains to be confirmed byhuman studies.

5.2.3. Prokinetic Therapy

Cisapride is a mixed serotonergic agent that facilitatesthe release of acetylcholine at synapses in the myentericplexus, thus increasing gastric emptying and improvingesophageal and intestinal peristalsis. Clinical studies ofcisapride in children with GERD showed significantreduction in the RI (332) but with less consistent reductionin symptoms (333,334). After cisapride was found toproduce prolongation of the QTc interval on electrocar-diogram, a finding increasing the risk of sudden death(335), its use was restricted to limited-access programssupervised by a pediatric gastroenterologist and to patientsin clinical trials, safety studies or registries.

Domperidone and metoclopramide are antidopaminer-gic agents that facilitate gastric emptying. Metoclopra-mide has cholinomimetic and mixed serotonergic effects.Metoclopramide and placebo equally reduced symptomscores of infants with reflux. Metoclopramide did reducethe RI on pH probe examination but did not normalize it(336). A meta-analysis of 7 RCTs of metoclopramide indevelopmentally healthy children 1 month to 2 years ofage with symptoms of GER found that metoclopramidereduced daily symptoms and the RI but was associatedwith significant side effects (215). Metoclopramide com-monly produces adverse side effects in infants and chil-dren, particularly lethargy, irritability, gynecomastia,galacctorhea, and extrapyramidal reactions and hascaused permanent tardive dyskinesia (337–340). Arecent systematic review of studies on domperidone(341) identified only 4 RCTs in children, none providing‘‘robust evidence’’ for efficacy of domperidone inpediatric GERD. Domperidone occasionally causesextrapyramidal central nervous system side effects (342).

Bethanechol, a direct cholinergic agonist, studiedin a few controlled trials has uncertain efficacy anda high incidence of side effects in children withGERD (338,343,344). Erythromycin, a dopamine-receptor antagonist, is sometimes used in patients withgastroparesis to hasten gastric emptying. Its role intherapy of GER and GERD has not been investigated.

Baclofen is a g-amino-butyric-acid receptor agonistthat reduces both acid and nonacid reflux in healthyadults and in adults with GERD (345). In children, itwas shown to accelerate gastric empyting for 2 hoursafter dosing, without any deleterious effect on LESresting pressure or esophageal peristalsis (346). In asmall group of children with GERD and NI, it wasreported to decrease the frequency of emesis (347).Although no side effects were noted in one study, baclo-fen is known to cause dyspeptic symptoms, drowsiness,dizziness, fatigue, and to lower the threshold for seizures.Such side effects preclude its routine use (348).




Currently, there is insufficient evidence to justify theroutine use of cisapride, metoclopramide, domperidone,bethnechol, erythromycin, or baclofen for GERD(215,333,341,349,350).

5.2.4. Other Agents

Antacids directly buffer gastric contents, thereby redu-cing heartburn and healing esophagitis. On-demand useof antacids may provide rapid symptom relief in somechildren and adolescents with NERD (351). Althoughthis approach appears to carry little risk, it has not beenformally studied in children. Intensive, high-dose antacidregimens (eg, magnesium hydroxide and aluminumhydroxide; 700 mmol/1.73 m2/day) are as effective ascimetidine for treating peptic esophagitis in children ages2 to 42 months (352,353). No studies of antacids to datehave used combined esophageal pH/MII to assess out-come. Prolonged treatment with aluminum-containingantacids significantly increases plasma aluminum ininfants (354,355), and some studies report plasmaaluminum concentrations close to those that have beenassociated with osteopenia, rickets, microcytic anemia,and neurotoxicity (356–358). Milk-alkali syndrome, atriad of hypercalcemia, alkalosis and renal failure canoccur due to chronic- or high-dose ingestion of calciumcarbonate. Although these side effects are less commonthan they were in the era before acid suppressive drugs(359), all of the antacid buffering agents should be usedwith particular caution in infants and young children.Because safe and convenient alternatives are availablethat are more acceptable to patients, chronic antacidtherapy is generally not recommended for patients withGERD.

Most surface protective agents contain either alginateor sucralfate. Alginates are insoluble salts of alginic acid,a component of algal cell walls. In older studies of alginicacid therapy in pediatric patients with GERD, the liquidpreparations used also contained buffering agents, mak-ing it difficult to isolate the effect of the surface protec-tive agent itself (360–363). Efficacy in these studies hasvaried widely. In 1 clinical study, a commercial liquidpreparation containing only sodium–magnesium algi-nate significantly decreased the mean frequency andseverity of vomiting in infants compared with placebo(364). Another placebo-controlled study of this prep-aration in infants showed that although symptomsimproved with therapy, the only objective change oncombined pH/MII evaluation was a marginal decreasein the height of reflux in the esophagus (365). Alginate isalso available as tablets, and is useful for on-demandtreatment of symptoms.

Sucralfate is a compound of sucrose, sulfate, andaluminum which, in an acid environment, forms a gelthat binds to the exposed mucosa of peptic erosions. Inadults, sucralfate decreased symptoms and promoted

healing of nonerosive esophagitis (366). The onlyrandomized comparison study in children demonstratesthat sucralfate was as effective as cimetidine for treat-ment of esophagitis (367). The available data areinadequate to determine the safety or efficacy of sucral-fate in the treatment of GERD in infants and children,particularly the risk of aluminum toxicity with long-term use.

None of the surface agents is recommended as a soletreatment for severe symptoms or erosive esophagitis.

5.3. Surgical Therapy

Fundoplication decreases reflux by increasing the LESbaseline pressure, decreasing the number of TLESRs andthe nadir pressure during swallow induced relaxation,increasing the length of the esophagus that is intraab-dominal, accentuating the angle of His and reducing aHH if present (24,368). Fundoplication usually elimin-ates reflux, including physiologic reflux (369). Fundo-plication does not correct underlying esophageal clear-ance, gastric emptying, or other GI dysmotility disorders(21,24,370–373).

Most of the literature on surgical therapy in childrenwith GERD consists of retrospective case series in whichdocumentation of the diagnosis of GERD and details ofprevious medical therapy are deficient, making it difficultto assess the indications for and responses to surgery(374–377). Children with underlying conditions predis-posing to the most severe GERD (Section 5.1.1) comprisea large percentage of most of the surgical series, furtherconfounding efforts to determine the benefits versus risksof surgical antireflux procedures in specific patient popu-lations. The absence of systematic postoperative evalu-ation, including objective testing with pH or impedancestudies and endoscopy, further complicates the assess-ment of surgical outcomes in most series (368,372,378).

In general, outcomes of antireflux surgery have beenmore carefully evaluated in adults than in children. In 1study (379), at a mean of 20 ("10) months after surgery,61% of the adults were satisfied with their outcome; 32%were taking medications for heartburn, 11% requiredesophageal dilatation, and 7% had repeat surgery. Thisstudy found that a substantial number of patients under-went fundoplication for questionable reasons. In anotherstudy of patients relieved of typical reflux symptomspostoperatively, up to two thirds developed new symp-toms postoperatively, including excessive gas, abdominalbloating, increased flatus, dysphagia, difficulty witheructation, and vomiting (378–381). In a large multi-center controlled study, 62% of the adults were takingPPIs for reflux symptoms 7 years after antireflux surgery(381). In another study, 37% of adults were takingantireflux medications at a mean of 5.9 years followingantireflux surgery (382). Another study showed a sim-ilarly high surgical failure rate (383).




A large open RCT compared the efficacy and safety oflaparoscopic fundoplication versus esomeprazole (20 mgqd) for treatment of adults with GERD (384). Short-termoutcomes were reported in an interim analysis of data at3 years. More than 90% of both the surgically andmedically treated adults showed good to excellent symp-tom control; 10% of the surgical group had dysphagiawhere as dysphagia was uncommon in the medicallytreated group. Quality of life measures were similar inboth groups (384). Death related to open or laparoscopicsurgery occurs. In adults, the mortality of the first oper-ation is reported to be between 1 in 1000 to 1 in 330(385–387).

In children who were operated on, those with NI havemore than twice the complication rate, 3 times themorbidity and 4 times the reoperation rate of childrenwithout NI (388). Other studies show similar data (389–391). One case series with a follow-up period of 3.5 yearsreported that more than 30% of children with NI hadmajor complications or died within 30 days of antirefluxsurgery (392). Twenty-five percent of those patients hadoperative failure and 71% had a return of 1 or morepreoperative symptoms within 1 year of surgery.

Children with repaired EA also have a high rate ofoperative failure (393,394), although not as high as thosewith NI. Recurrence of pathologic reflux after antirefluxsurgery in children with NI or EA may not be obvious anddetection often requires a high index of suspicion,repeated evaluation over time, and use of more than 1test (391,394).

In a recent retrospective review of 198 children, 74%of whom had underlying disorders, two thirds had GERDsymptoms or required medical treatment for GERDwithin 2 months of antireflux surgery (374). Fundoplica-tion in early infancy has a higher failure rate thanfundoplication performed later in childhood (395,396),and appears to be more frequent in children with associ-ated anomalies (396).

The impact of antireflux surgery on hospitalization forreflux-related events, especially adverse respiratoryevents, was reviewed using a large administrative data-base (397). A significant reduction in the number ofadverse respiratory events was observed in the yearfollowing surgery in those operated at


had recurrent symptoms requiring a repeat procedure 2to 24 months postoperatively. Three years after surgery,9 patients (56%) were taking no antireflux medication.Longer-term studies in adults have shown little orno difference in procedure time or failure rate betweenendoluminal and surgical antireflux procedures(414,415). In some studies, sham-operated patients havedone as well as operated patients (416,417). Otherendoscopic GERD treatments have not been studied inchildren (368).

The annual number of antireflux operations has beenon the increase in the United States, especially in childrenyounger than 2 years of age (375,406). In contrast, inadults, rates of fundoplication are declining in the UnitedStates and have dropped 30% from their peak in 1999(378). The greatest decline is in teaching hospitals and inyoung adult patients.

Antireflux surgery may be of benefit in children withconfirmed GERD who have failed optimal medicaltherapy, or who are dependent on medical therapy overa long period of time, or who are significantly nonad-herent with medical therapy, or who have life-threateningcomplications of GERD. Children with respiratory com-plications including asthma or recurrent aspirationrelated to GERD are generally considered most likelyto benefit from antireflux surgery when medical therapyfails but additional study is required to confirm this.Children with underlying disorders predisposing to themost severe GERD are at the highest risk for operativemorbidity and operative failure. Before surgery it isessential to rule out non-GERD causes of symptoms,and ensure that the diagnosis of chronic-relapsing GERDis firmly established. It is important to provide familieswith appropriate education and a realistic understandingof the potential complications of surgery, includingsymptom recurrence.

6. EVALUATION AND MANAGEMENT OF THEPEDIATRIC PATIENT WITH SUSPECTED GERD

The following sections describe the relation betweenreflux and several common signs, symptoms or symptomcomplexes of infants and children. The evaluationsappropriate to establish a diagnosis of GERD and recom-mendations for management in each case are outlined.Recommendations are based on the available evidenceand the consensus opinion of the members of theGuidelines Committee.

6.1. Recurrent Regurgitation and Vomiting

The practitioner’s challenge is to distinguish regurgi-tation and vomiting caused by reflux or reflux diseasefrom vomiting caused by numerous other disorders(Table 4). This can be confusing because reflux episodessometimes trigger vomiting, a coordinated autonomic

and voluntary motor response causing forceful expulsionof gastric contents. Vomiting associated with reflux isprobably a result of the stimulation of pharyngeal sensoryafferents by refluxed gastric contents (418,419). Labora-tory and radiographic investigation may be necessary toexclude other causes of vomiting.

TABLE 4. Differential diagnosis of vomiting in infants andchildren

Gastrointestinal obstructionPyloric stenosisMalrotation with intermittent volvulusIntestinal duplicationHirschsprung diseaseAntral/duodenal webForeign bodyIncarcerated hernia

Other gastrointestinal disordersAchalasiaGastroparesisGastroenteritisPeptic ulcerEosinophilic esophagitis/gastroenteritisFood allergyInflammatory bowel diseasePancreatitisAppendicitis

NeurologicHydrocephalusSubdural hematomaIntracranial hemorrhageIntracranial massInfant migraineChiari malformation

InfectiousSepsisMeningitisUrinary tract infectionPneumoniaOtitis mediaHepatitis

Metabolic/endocrineGalactosemiaHereditary fructose intoleranceUrea cycle defectsAmino and organic acidemiasCongenital adrenal hyperplasia

RenalObstructive uropathyRenal insufficiency

ToxicLeadIronVitamins A and DMedications—ipecac, digoxin, theophylline, etc

CardiacCongestive heart failureVascular ring

OthersPediatric falsification disorder (Munchausen syndrome by proxy)Child neglect or abuseSelf-induced vomitingCyclic vomiting syndromeAutonomic dysfunction




6.1.1. The Infant With Uncomplicated Recurrent Regurgitation

In the infant with recurrent regurgitation or spitting, athorough history (Table 5) and physical examination withattention to warning signals suggesting other diagnoses(Table 1) is generally sufficient to establish a clinicaldiagnosis of uncomplicated infant GER (Fig. 1). Thetypical presentation of uncomplicated infant GER iseffortless, painless regurgitation in a healthy-appearingchild with normal growth—the so-called happy spitter.Intermittently, an episode of vomiting, even forcefulvomiting may occur. Irritability may accompany regur-gitation and vomiting; however, in the absence of otherwarning symptoms, it is not an indication for extensivediagnostic testing. An upper GI series or other diagnostictests is not required unless other diagnoses such as GIobstruction are suspected. Recurrent regurgitation due toGER generally decreases over the first year, resolving at12 to 18 months of age (17,18). If ‘‘warning signs’’ forGERD or other diagnoses are present, or if regurgitationis not resolving by 12 to 18 months of age, consultationwith a pediatric gastroenterologist is recommended.

Generally, only parental education, anticipatory gui-dance, and modification of feeding composition, fre-quency and volume are necessary for the managementuncomplicated infant GER (208,420). Overfeedingexacerbates recurrent regurgitation and should beavoided (211). In some infants with persistent regurgita-tion, a thickened or commercial antiregurgitation formulamay help control the frequency of regurgitation (Section4.1.1). There is no evidence that antisecretory or promo-tility agents improve physiologic infant regurgitation.Prone positioning is not recommended because of itsassociation with SIDS. Because regurgitation is some-times the sole manifestation of cow’s milk protein allergyin healthy looking infants (420,421), a 2-week trial of

TABLE 5. History in the child with suspectedgastroesophageal reflux disease

Feeding and dietary historyAmount/frequency (overfeeding)Preparation of formulaRecent changes in feeding type or techniquePosition during feedingBurpingBehavior during feeding

Choking, gagging, cough, arching, discomfort, refusalPattern of vomiting

Frequency/amountPainForcefulBlood or bileAssociated fever, lethargy, diarrhea

Medical historyPrematurityGrowth and developmentPast surgery, hospitalizationsNewborn screen resultsRecurrent illnesses, especially croup, pneumonia, asthmaSymptoms of hoarseness, fussiness, hiccupsApneaPrevious weight and height gainOther chronic conditions

MedicationsCurrent, recent, prescription, nonprescription

Family psychosocial historySources of stressMaternal or paternal drug usePostpartum depression

Family medical historySignificant illnessesFamily history of gastrointestinal disordersFamily history of atopy

Growth chart including height, weight, and head circumferenceWarning signals (Table 2)

FIG. 1. Approach to the infant with recurrent regurgitation andvomiting.




protein hydrolysate– or amino acid–based formula, ora trial of milk-free diet for the breast-feeding motheris appropriate in infants not responding to previousmanagement.

6.1.2. The Infant With Recurrent Regurgitationand Poor Weight Gain

The infant with recurrent regurgitation and poorweight gain should not be confused with the ‘‘happyspitter’’ described in Section 6.1.1. Whereas the historyand physical examination may be identical, poor weightgain is not typical of uncomplicated infant GER and is acrucial warning sign that alters clinical management.

Because there are no well-controlled studies evaluat-ing diagnostic or therapeutic strategies for these infants,the following approach is based on expert opinion(Fig. 2). A feeding history should be obtained that

includes an estimate of energy offered and ingestedper day, an estimate of energy loss through regurgitation,a description of formula preparation and feeding sche-dule, an assessment of breast milk sufficiency, and adescription of infant sucking and swallowing behavior.Parents should be advised not to reduce intake to the pointof energy deprivation in the attempt to prevent regurgita-tion. If problems identified by history seem to explainthe symptoms and can be addressed, close outpatientmonitoring of weight gain will determine whether furtherevaluation is indicated.

If chronic regurgitation and inadequate weight gainpersist after observation and despite adequate energyintake, evaluation for causes of failure to thrive compa-tible with the history is mandatory. Among possibleetiologies in infancy are infections (especially urinarytract), food allergy, anatomic abnormalities, neurologicdisorders, metabolic disease, and neglect or abuse(Table 4). A 2- to 4-week trial of extensively hydrolyzedor amino-acid based formula is appropriate. Dependingon the results of investigations and response to dietarymanagement, the infant should be referred to a pediatricspecialist. Hospitalization for observation and testing isappropriate in some infants with persistent failure tothrive. Nasogastric or nasojejunal feeding is occasionallynecessary to achieve weight gain in the infant with noother clear explanation for poor weight gain (231).

6.1.3. The Infant With Unexplained Crying and/orDistressed Behavior

Irritability and regurgitation are nonspecific symptomsthat occur in healthy infants and are associated with awide range of physiologic and pathologic conditions. Forexample, exposure to environmental factors, such astobacco smoke may result in irritability in infants(422,423). Healthy young infants fuss or cry an averageof 2 hours daily. There is substantial individual variationand some healthy infants cry as much as 6 hours per day.Likewise there is variation in parental perceptions regard-ing the severity and duration of crying and its importance.The amount of daily crying typically peaks at 6 weeks ofage (424,425). As with fussing, sleeping patterns ofhealthy infants show great individual and maturationalvariation as do parental expectations for sleep behavior(426).

The concept that infant irritability and sleep disturb-ances are manifestations of GER is largely extrapolatedfrom adult descriptions of heartburn and sleep disturb-ances that improve with antacid therapy (50,427–429).Although 1 study in infants showed a correlation betweeninfant grimacing and episodes of reflux (430) multipleother studies have shown no relation between crying andGERD determined by esophageal pH testing (61,84,140,431) or the presence of esophagitis (47,84). Some smalldescriptive studies have evaluated pH probe studies in

FIG. 2. Approach to the infant with recurrent regurgitation andweight loss.




infants with irritability and sleep disturbance. One com-pared infants with normal and abnormal pH probe studiesand found a slight increase in nighttime waking, delayedonset of sleep and greater daytime sleeping in those withabnormal pH probe studies (432). Another study foundno increase in sleep disturbance in infants with abnormalesophageal pH tests (431). One dual pH probe studyshowed slightly poorer proximal acid clearance incolicky infants, but no abnormality in other parameters(433). Recently, a study of colicky infants found abnor-mal pH test results only in those with excessive regur-gitation or feeding difficulties (61).

There are few studies addressing the appropriate man-agement of infants with irritability and reflux symptoms.One study showed a greater decrease in crying time ininfants treated with a 1-mg/kg dose of famotidine than ininfants given 0.5 mg/kg. Although the authors concludedfrom this study that famotidine was effective in treatinginfant crying, differences in age between treatmentgroups, absence of placebo control and a lack of differ-ence between the treatment group

pediatric gastroesophageal reﬂux clinical practice ......phagus with or without regurgitation and...

Documents