Pediatric Gastroesophageal Reflux Clinical Practice Guidelines ...


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Pediatric Gastroesophageal Reflux Clinical Practice Guidelines ...

  1. 1. 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 Pediatric Gastroesophageal Reflux Clinical Practice Guidelines: Joint Recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN) Co-Chairs: ÃYvan Vandenplas and yColin D. Rudolph Committee Members: zCarlo Di Lorenzo, §Eric Hassall, jjGregory Liptak, ôLynnette Mazur, #Judith Sondheimer, ÃÃAnnamaria Staiano, yyMichael Thomson, zzGigi Veereman-Wauters, and §§Tobias G. Wenzl ÃUZ 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, ôDepartment 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, CO, USA, ÃÃDepartment of Pediatrics, University of Naples ‘‘Federico II,’’ Naples, Italy, {{Centre for Paediatric Gastroenterology, Sheffield Children’s Hospital, Western Bank, Sheffield, UK, {{Pediatric Gastroenterology & Nutrition, Queen Paola Children’s Hospital-ZNA, Antwerp, Belgium, and §§Klinik fur Kinder- und Jugendmedizin, Universitatsklinikum der RWTH Aachen, Aachen, Germany ¨ ¨ABSTRACTObjective: To develop a North American Society for Pediatric PubMed, Cumulative Index to Nursing and Allied HealthGastroenterology, Hepatology, and Nutrition (NASPGHAN) and Literature, and bibliographies. The committee convened inEuropean Society for Pediatric Gastroenterology, Hepatology, face-to-face meetings 3 times. Consensus was achieved forand Nutrition (ESPGHAN) international consensus on the all recommendations through nominal group technique, adiagnosis and management of gastroesophageal reflux and structured, quantitative method. Articles were evaluatedgastroesophageal reflux disease in the pediatric population. using the Oxford Centre for Evidence-based Medicine LevelsMethods: An international panel of 9 pediatric of Evidence. Using the Oxford Grades of Recommendation, thegastroenterologists and 2 epidemiologists were selected by quality of evidence of each of the recommendations made by theboth societies, which developed these guidelines based on committee was determined and is summarized in appendices.the Delphi principle. Statements were based on systematic Results: More than 600 articles were reviewed for this work.literature searches using the best-available evidence from The document provides evidence-based guidelines for the diagnosis and management of gastroesophageal reflux and Received May 27, 2009; accepted May 31, 2009. gastroesophageal reflux disease in the pediatric population. Address correspondence and reprint requests to Colin D. Rudolph, Conclusions: This document is intended to be used in dailyMD, PhD, Professor of Pediatrics & Vice Chair for Clinical Affairs, practice for the development of future clinical practice guidelinesChief, Pediatric Gastroenterology, Hepatology, and Nutrition, Division and as a basis for clinical trials. JPGN 49:498–547, 2009. Keyof Pediatric Gastroenterology, Hepatology, and Nutrition, Children’s Words: Clinical practice guidelines—Diagnostic tests—Hospital of Wisconsin, Medical College of Wisconsin, 9000 W Gastroesophageal reflux (GER)—Gastroesophageal refluxWisconsin Ave, Milwaukee, WI 53226 (e-mail: disease (GERD)—Therapeutic modalities. # 2009 by Carlo Di Lorenzo, Eric Hassall, Gregory Liptak, Lynnette Mazur, European Society for Pediatric Gastroenterology, Hepatology,Judith Sondheimer, Annamaria Staiano, Michael Thomson, Gigi Veere- and Nutrition and North American Society for Pediatricman-Wauters, and Tobias G. Wenzl contributed equally to the devel-opment of these guidelines. Abstract adapted by Gregory Liptak. Gastroenterology, Hepatology, and Nutrition Authors’ disclosures are listed in Appendix D. 498
  2. 2. PEDIATRIC GASTROESOPHAGEAL REFLUX CLINICAL PRACTICE GUIDELINES 499 SYNOPSIS pathologic acid reflux does not correlate consistently with symptom severity or demonstrable complications. This synopsis contains some essentials of the guide- In children with documented esophagitis, normal eso-lines, but does not convey the details, nuances, and phageal pH monitoring suggests a diagnosis other thancomplexity of the issues addressed in the complete GERD. Esophageal pH monitoring is useful for evaluat-guidelines, and therefore can be interpreted only with ing the efficacy of antisecretory therapy. It may be usefulreference to the full article. to correlate symptoms (eg, cough, chest pain) with acid reflux episodes and to select those infants and children1. RATIONALE The purpose of these guidelines with wheezing or respiratory symptoms in whom GER isis to provide pediatricians and pediatric subspecialists an aggravating factor. The sensitivity, specificity, andwith a common resource for the evaluation and manage- clinical utility of pH monitoring for diagnosis and man-ment of patients with gastroesophageal reflux (GER) and agement of possible extraesophageal complications ofgastroesophageal reflux disease (GERD). These guide- GER are not well established.lines are not intended as a substitute for clinical judgmentor as a protocol for the management of all pediatric 4.3. Combined Multiple Intraluminal Impedancepatients with GER and GERD. (MII) and pH Monitoring This test detects acid, weakly acid, and nonacid reflux episodes. It is superior to2. METHODS ‘‘Pediatric Gastroesophageal Re- pH monitoring alone for evaluation of the temporalflux Clinical Practice Guidelines: Joint Recommen- relation between symptoms and GER. Whether com-dations of the North American Society for Pediatric bined esophageal pH and impedance monitoring willGastroenterology, Hepatology, and Nutrition (NASP- provide useful measurements that vary directly withGHAN) and the European Society for Pediatric Gas- disease severity, prognosis, and response to therapy introenterology, Hepatology, and Nutrition (ESPGHAN)’’ pediatric patients has yet to be a document developed by a committee of 9 pediatricgastroenterologists from NASPGHAN and ESPGHAN 4.4. Motility Studies Esophageal manometry mayand 2 pediatric epidemiologists from the United be abnormal in patients with GERD but the findings areStates. Using the best-available evidence from the not sufficiently sensitive or specific to confirm a diag-literature, the committee critically evaluated current nosis of GERD, nor to predict response to medical ordiagnostic tests and therapeutic modalities for GER surgical therapy. It may be useful to diagnose a motilityand GERD. disorder in patients who have failed acid suppression and who have a normal endoscopy, or to determine the3. DEFINITIONS AND MECHANISMS GER is position of the lower esophageal sphincter to place athe passage of gastric contents into the esophagus with or pH probe. Manometric studies are useful to confirm awithout regurgitation and vomiting. GER is a normal diagnosis of achalasia or other motor disorders of thephysiologic process occurring several times per day in esophagus that may mimic GERD.healthy infants, children, and adults. Most episodes ofGER in healthy individuals last <3 minutes, occur in the 4.5. Endoscopy and Biopsy Endoscopically visi-postprandial period, and cause few or no symptoms. In ble breaks in the distal esophageal mucosa are the mostcontrast, GERD is present when the reflux of gastric reliable evidence of reflux esophagitis. Mucosalcontents causes troublesome symptoms and/or compli- erythema, pallor, and increased or decreased vascularcations. Every effort was made to use these 2 terms pattern are highly subjective and nonspecific findingsstrictly as defined. that are variations of normal. Histologic findings of eosinophilia, elongated rete pegs, basilar hyperplasia,4. DIAGNOSIS and dilated intercellular spaces, alone or in combination, are insufficiently sensitive or specific to diagnose reflux4.1. History and Physical Examination In infants esophagitis. Conversely, absence of these histologicand toddlers, there is no symptom or symptom com- changes does not rule out GERD. Endoscopic biopsyplex that is diagnostic of GERD or predicts response is important to identify or rule out other causes ofto therapy. In older children and adolescents, as in esophagitis, and to diagnose and monitor Barrett esopha-adult patients, history and physical examination may gus (BE) and its sufficient to diagnose GERD if the symptoms aretypical. 4.6. Barium Contrast Radiography This test is not useful for the diagnosis of GERD but is useful to4.2. Esophageal pH Monitoring This test is a confirm or rule out anatomic abnormalities of the uppervalid quantitative measure of esophageal acid exposure, gastrointestinal (GI) tract that may cause symptomswith established normal ranges. However, the severity of similar to those of GERD. J Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  3. 3. 500 NASPGHAN/ESPGHAN PEDIATRIC GER GUIDELINE COMMITTEE4.7. Nuclear Scintigraphy The standards for protein formula that has been evaluated in controlledinterpretation of this test are poorly established. Accord- trials. Use of a thickened formula (or commercial anti-ing to limited published literature, scintigraphy may have regurgitation formulae, if available) may decrease visiblea role in the diagnosis of pulmonary aspiration in patients regurgitation but does not result in a measurable decreasewith chronic and refractory respiratory symptoms. A in the frequency of esophageal reflux episodes. Pronenegative test does not rule out possible pulmonary aspira- positioning decreases the amount of acid esophagealtion of refluxed material. Gastric emptying studies by exposure measured by pH probe compared with thatthemselves do not confirm the diagnosis of GERD and measured in the supine position. However, prone andare recommended only in individuals with symptoms of lateral positions are associated with an increased inci-gastric retention. Nuclear scintigraphy is not recom- dence of sudden infant death syndrome (SIDS). The riskmended for the routine evaluation of pediatric patients of SIDS outweighs the benefit of prone or lateral sleepwith suspected GERD. position on GER; therefore, in most infants from birth to 12 months of age, supine positioning during sleep4.8. Esophageal and Gastric Ultrasonography is recommended.These tests are not recommended for the routine evaluationof GERD in children. 5.1.3. Lifestyle Changes in Children and Adoles- cents In older children, there is no evidence to support4.9. Tests on Ear, Lung, and Esophageal Fluids the routine elimination of any specific food for manage-Evaluation of middle ear or pulmonary aspirates for ment of GERD. In adults, obesity, large meal volume, andlactose, pepsin, or lipid-laden macrophages have been late night eating are associated with symptoms of GERD.proposed as the tests for GERD. No controlled studies Prone or left-side sleeping position and/or elevation ofhave proven that reflux is the only reason these com- the head of the bed may decrease GER, as shown inpounds appear in ear or lung fluids, and no controlled adult studies.studies have shown that the presence of these substancesconfirms GER as the cause of ear, sinus, or pulmonary 5.2. Pharmacologic Therapies The major phar-disease. Diagnosis of duodeno-gastroesophageal reflux macologic agents currently used for treating GERD inby detection of bilirubin in the esophagus is not recom- children are gastric acid–buffering agents, mucosal sur-mended for the routine evaluation for possible GERD in face barriers, and gastric antisecretory agents. Acid-children. The role of bile reflux in causing GERD that is suppressant agents are the mainstay of treatment forresistant to proton pump inhibitors (PPIs) therapy has not all but the patient with occasional symptoms. The poten-been established. tial adverse effects of acid suppression, including increased risk of community-acquired pneumonias and4.10. Empiric Trial of Acid Suppression as a Diag- GI infections, need to be balanced against the benefitsnostic Test Expert opinion suggests that in an older of therapy.child or adolescent with typical symptoms suggestingGERD, an empiric trial of PPIs is justified for up to 5.2.1. Histamine-2 Receptor Antagonists (H2RAs)4 weeks. However, improvement of heartburn, following H2RAs exhibit tachyphylaxis or tolerance but PPIs dotreatment, does not confirm a diagnosis of GERD not. Tachyphylaxis is a drawback to chronic use. H2RAsbecause symptoms may improve spontaneously or have a rapid onset of action and, like buffering agents, arerespond by a placebo effect. There is no evidence to useful for on-demand an empiric trial of acid suppression as a diag-nostic test in infants and young children where symptoms 5.2.2. Proton Pump Inhibitors For healing of ero-suggestive of GERD are less specific. sive esophagitis and relief of GERD symptoms, PPIs are superior to H2RAs. Both medications are superior to5. TREATMENT placebo. Administration of long-term acid suppression without a diagnosis is inadvisable. When acid suppres-5.1. Lifestyle Changes sion is required, the smallest effective dose should be used. Most patients require only once-daily PPI; routine5.1.1. & 5.1.2. Lifestyle Changes in the Infant use of twice-daily doses is not indicated. No PPI has beenParental education, guidance, and support are always approved for use in infants younger than 1 year of age,required and usually sufficient to manage healthy, thriv- and there are special concerns pertaining to prescriptioning infants with symptoms likely because of physiologic of PPIs in infants, as described in the Guideline.GER. Milk protein sensitivity is sometimes a cause ofunexplained crying and vomiting in infants. Therefore, 5.2.3. Prokinetic Therapy Potential adverse effectsformula-fed infants with recurrent vomiting may benefit of currently available prokinetic agents outweigh thefrom a 2- to 4-week trial of an extensively hydrolyzed potential benefits of these medications for treatment ofJ Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  4. 4. PEDIATRIC GASTROESOPHAGEAL REFLUX CLINICAL PRACTICE GUIDELINES 501GERD. There is insufficient evidence of clinical efficacy formula if available) reduces the frequency of overtto justify the routine use of metoclopramide, erythromy- regurgitation and vomiting.cin, bethanechol, cisapride, or domperidone for GERD.Baclofen reduces the frequency of transient relaxations of 6.1.2. Infants With Recurrent Vomiting and Poorthe lower esophageal sphincter (TLESR), but it has not Weight Gain A diagnosis of physiologic GERbeen evaluated in controlled trials for treatment of GERD should not be made in an infant with vomiting and poorin children. weight gain. Expert opinion suggests that initial evalu- ation in an infant with normal physical examination but5.2.4. Other Agents Buffering agents, alginate, poor weight gain should include diet history, urinalysis,and sucralfate are useful on demand for occasional heart- complete blood count, serum electrolytes, blood ureaburn. Chronic use of buffering agents or sodium alginate nitrogen, and serum creatinine. Additional testing shouldis not recommended for GERD because some have be based on suggestive historical details or results ofabsorbable components that may have adverse effects screening tests. Management may include a 2-week trialwith long-term use. Special caution is required in infants. of extensively hydrolyzed formula or amino acid–basedIf long-term use is required, more effective therapy formula to exclude cow’s milk allergy, increased caloricis available. density of formula and/or thickened formula, and edu- cation as to appropriate daily formula volume required5.3. Surgical Therapy Antireflux surgery may be for normal growth. Careful follow-up of interval weightof benefit in selected children with chronic-relapsing change and caloric intake is essential. If managementGERD. Indications include failure of optimized fails to improve symptoms and weight gain, referral to amedical therapy, dependence on long-term medical pediatric gastroenterologist is recommended.therapy, significant nonadherence to medical therapy,or pulmonary aspiration of refluxate. Children with 6.1.3. Infants With Unexplained Crying and/or Dis-respiratory complications, including asthma or recur- tressed Behavior Reflux is not a common cause ofrent aspiration related to GERD, are generally con- unexplained crying, irritability, or distressed behavior insidered most likely to benefit from antireflux surgery otherwise healthy infants. Other causes include cow’swhen medical therapy fails but additional study is milk protein allergy, neurologic disorders, constipation,required to confirm this assumption. Children with and infection (especially of the urinary tract). Followingunderlying disorders predisposing to the most severe exclusion of other causes, an empiric trial of extensivelyGERD are at the highest risk for operative morbidity hydrolyzed protein formula or amino acid–basedand postoperative failure. Before surgery it is essential formula is reasonable in selected cases, although evi-to rule out non-GERD causes of symptoms and ensure dence from the literature in support of such a trial isthat the diagnosis of chronic-relapsing GERD is firmly limited. There is no evidence to support the empiric useestablished. It is important to provide families with of acid suppression for the treatment of irritable infants.appropriate education and a realistic understanding of If irritability persists with no explanation other thanthe potential complications of surgery, including symp- suspected GERD, expert opinion suggests the followingtom recurrence. options: the practitioner may continue anticipatory gui- dance and training of parents in the management of such6. EVALUATION AND MANAGEMENT OF infants with the anticipation of improvement with time;PEDIATRIC PATIENTS WITH SUSPECTED additional investigations to ascertain the relation betweenGERD The following sections describe the relation reflux episodes and symptoms or to diagnose esophagitisbetween reflux and several common signs, symptoms, or may be indicated (pH monitoring Æ impedance monitor-symptom complexes of infants and children. ing, endoscopy); a time-limited (2-week) trial of anti- secretory therapy may be considered, but there is a6.1. Recurrent Regurgitation and Vomiting The potential risk of adverse effects. Clinical improvementpractitioner’s challenge is to distinguish regurgitation and following empiric therapy may be due to spontaneousvomiting caused by GER from vomiting caused by symptom resolution or a placebo response. The risk/numerous other disorders. benefit ratio of these approaches is not clear.6.1.1. Infants With Uncomplicated Recurrent Regur- 6.1.4. The Child Older Than 18 Months of Age Withgitation A history of disease and physical exami- Chronic Regurgitation or Vomiting Physiologicnation, with attention to warning signs, are generally regurgitation, episodic vomiting, or regurgitation fol-sufficient to allow the clinician to establish the diagnosis lowed by swallowing of refluxate in the mouth areof uncomplicated GER. Parental education, reassurance, frequent in infants. Whether of new onset or persist-and anticipatory guidance are recommended. In formula- ing from infancy, these symptoms are less common infed infants, thickened formula (or antiregurgitation children older than 18 months of age. Although these J Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  5. 5. 502 NASPGHAN/ESPGHAN PEDIATRIC GER GUIDELINE COMMITTEEsymptoms are not unique to GERD, evaluation to diag- phagitis is increasingly recognized to be a more commonnose possible GERD and to rule out alternative diagnoses cause of dysphagia or odynophagia than GERD, althoughis recommended based on expert opinion. Testing may this finding is not consistently reported in all geographicinclude upper GI endoscopy and/or esophageal pH/MII, regions. Odynophagia, or pain caused by swallowing,and/or barium upper GI series. must be distinguished from heartburn (substernal pain caused by esophageal acid exposure) and dysphagia.6.2. Heartburn Extrapolation from adult data Although odynophagia may be a symptom of pepticsuggests that in older children and adolescents, on- esophagitis, it is more often associated with other con-demand therapy with buffering agents, sodium alginate, ditions such as oropharyngeal inflammation, esophagealor H2RA may be used for occasional symptoms. Ado- ulcer, eosinophilic esophagitis, infectious esophagitis,lescents with typical symptoms of chronic heartburn and esophageal motor disorders. Although GERD isshould be treated with lifestyle changes if applicable not a prevalent cause of difficulty in swallowing or pain(diet changes, weight loss, smoking avoidance, sleeping with swallowing, an evaluation including barium upperposition, no late night eating) and a 2- to 4-week trial of GI series and possibly upper endoscopy should be con-PPI. If symptoms resolve, PPIs may be continued for up sidered if physical examination and history of disease doto 3 months. Heartburn that persists on PPI therapy or not reveal a cause. Therapy with acid suppression withoutrecurs after this therapy is stopped should be investigated earlier evaluation is not recommended. In the infant withfurther by a pediatric gastroenterologist. feeding refusal, acid suppression without earlier diag- nostic evaluation is not recommended.6.3. Reflux Esophagitis In pediatric patients withendoscopically diagnosed reflux esophagitis or estab- 6.6. Infants With Apnea or Apparent Life-lished nonerosive reflux disease, PPIs for 3 months con- threatening Event In the majority of infants withstitute initial therapy. Not all reflux esophagitis are apnea or apparent life-threatening events (ALTEs), GERchronic or relapsing, and therefore trials of tapering is not the cause. In the uncommon circumstance in whichthe dose and then withdrawal of PPI therapy should be a relation between symptoms and GER is suspected or inperformed at intervals. Most but not all of the children those with recurrent symptoms, MII/pH esophagealwith chronic-relapsing reflux disease have one of the monitoring in combination with polysomnographicGERD-predisposing disorders described below. In most recording and precise, synchronous symptom recordingcases of chronic-relapsing esophagitis, symptom relief may aid in establishing cause and effect.can be used as a measure of efficacy of therapy, but insome circumstances repeat endoscopy or diagnostic stu- 6.7. Reactive Airways Disease In patients withdies may be indicated. Recurrence of symptoms and/or asthma who also have heartburn, reflux may be a con-esophagitis after repeated trials of PPI withdrawal tributing factor to the asthma. Despite a high frequency ofusually indicate that chronic-relapsing GERD is present, abnormal reflux studies in patients with asthma who doif other causes of esophagitis have been ruled out. At that not have heartburn, there is no strong evidence to supportpoint, therapeutic options include long-term PPI therapy empiric PPI therapy in unselected pediatric patients withor antireflux surgery. wheezing or asthma. Only 3 groups—those with heart- burn, those with nocturnal asthma symptoms, and those6.4. Barrett Esophagus BE occurs in children with steroid-dependent difficult-to-control asthma—maywith less frequency than it does in adults. Multiple derive some benefit from long-term medical or surgicalbiopsies documented in relation to endoscopically ident- antireflux therapy. Finding abnormal esophageal pHified esophagogastric landmarks are advised to confirm exposure by esophageal pH monitoring, with or withoutor rule out the diagnosis of BE and dysplasia. In BE, impedance, before considering a trial of long-term PPIaggressive acid suppression is advised by most experts. therapy or surgery may be useful, although the predictiveSymptoms are a poor guide to the severity of acid reflux value of these studies for this purpose has not beenand esophagitis in BE, and pH studies are often indicated established. The relative efficacy of medical versus sur-to guide treatment. BE per se is not an indication for gical therapy for GERD in children with asthmasurgery. Dysplasia is managed according to adult guide- is unknown.lines. 6.8. Recurrent Pneumonia Recurrent pneumonia6.5. Dysphagia, Odynophagia, and Food Refusal and interstitial lung disease may be the complications ofDysphagia, or difficulty in swallowing, occurs in asso- GER due to aspiration of gastric contents. No test canciation with oral and esophageal anatomic abnormalities, determine whether GER is causing recurrent pneumonia.neurologic and motor disorders, oral and esophageal An abnormal esophageal pH test may increase the prob-inflammatory diseases, and psychological stressors or ability that GER is a cause of recurrent pneumonia butdisorders. Of the mucosal disorders, eosinophilic eso- is not proof thereof. Nuclear scintigraphy can detectJ Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  6. 6. PEDIATRIC GASTROESOPHAGEAL REFLUX CLINICAL PRACTICE GUIDELINES 503aspirated gastric contents when images are obtained for acidic pH. Young children and children with neurologic24 hours after enteral administration of a labeled meal. impairment appear to be at the greatest risk. Factors otherAspiration during swallowing is more common than than reflux that cause similar dental erosions includeaspiration of refluxed material. A trial of nasogastric juice drinking, bulimia, and racial and genetic factorsfeeding may be used to exclude aspiration during swal- affecting the characteristics of enamel and saliva.lowing as a potential cause of recurrent disease. A trial ofnasojejunal therapy may help in determining whether 6.11. Dystonic Head Posturing (Sandifer Syndrome)surgical antireflux therapy is likely to be beneficial. In Sandifer syndrome (spasmodic torsional dystonia withpatients with severely impaired lung function, antireflux arching of the back and opisthotonic posturing, mainlysurgery may be necessary to prevent further pulmonary involving the neck and back) is an uncommon butdamage, despite lack of definitive proof that GER specific manifestation of GERD. It resolves withis causative. antireflux treatment.6.9. Upper Airway Symptoms The data linking 7. GROUPS AT HIGH RISK FOR GERDreflux to chronic hoarseness, chronic cough, sinusitis, Certain conditions are predisposed to severe, chronicchronic otitis media, erythema, and cobblestone appear- GERD. These include neurologic impairment, obesity,ance of the larynx come mainly from case reports and repaired esophageal atresia or other congenital esopha-case series. The association of reflux with these con- geal disease, cystic fibrosis, hiatal hernia, repaired acha-ditions and response to antisecretory therapy have not lasia, lung transplantation, and a family history of GERD,been proven by controlled studies. Patients with these BE, or esophageal adenocarcinoma. Although manysymptoms or signs should not be assumed to have GERD premature infants are diagnosed with GERD becausewithout consideration of other potential etiologies. of nonspecific symptoms of feeding intolerance, apnea spells, feeding refusal, and pain behavior, there are no6.10. Dental Erosions An association between controlled data that confirm reflux as a cause. AlthoughGERD and dental erosions has been established. The reflux may be more common in infants with broncho-severity of dental erosions seems to be correlated with pulmonary dysplasia, there is no evidence that antirefluxthe presence of GERD symptoms and, in adults, with the therapy affects the clinical course or outcome of thisseverity of proximal esophageal or oral exposure to an condition. PEDIATRIC GER GUIDELINE 1. RATIONALE that the diagnosis of GERD is applied excessively to healthy infants with bothersome but harmless symptoms The North American Society for Pediatric Gastroen- of physiologic GER (6–9), the committee reevaluated theterology, Hepatology, and Nutrition (NASPGHAN) pub- 2001 diagnostic and therapeutic algorithms to clarify thelished the first clinical practice guidelines on pediatric distinction between physiologic GER and GERD. In itsgastroesophageal reflux (GER) and gastroesophageal recommendations for testing, the committee confrontedreflux disease (GERD) in 2001 (1). Consensus-based the ongoing problem that current reflux tests may identifyguidelines on several aspects of GER and GERD were variations from normal but cannot predict symptomdeveloped in Europe at about the same time but were not severity, natural history, or response to therapy.officially endorsed by the European Society for Pediatric These guidelines are designed to assist pediatric healthGastroenterology, Hepatology, and Nutrition (ESP- care providers in the diagnosis and management of GERGHAN) (2,3). In 2007, the Councils of ESPGHAN and GERD. They are intended to serve as general guide-and NASPGHAN established a joint committee to lines and not as a substitute for clinical judgment, or as areview, update, and unify these guidelines as a means protocol applicable to all patients.of improving uniformity of practice and quality of patientcare (4,5). 2. METHODS The committee used the 2001 NASPGHAN guidelinesas an outline, adding new sections on certain pediatric 2.1. Selection of Committee Memberspopulations at high risk for GERD. In all deliberations,the committee attempted to distinguish physiologic GER The NASPGHAN–ESPGHAN Joint Guideline Com-events from GERD. Furthermore, in response to evidence mittee included 5 European and 4 North American J Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  7. 7. 504 NASPGHAN/ESPGHAN PEDIATRIC GER GUIDELINE COMMITTEEpediatric gastroenterologists with extensive experience in section. These sections and other evidence available inGER and GERD, selected by their respective societies, previously prepared tables that listed references andand 2 North American primary care pediatricians experi- graded the quality of each reference were distributed,enced in clinical epidemiology. Both pediatric epidemio- then reviewed and discussed to achieve consensus agree-logists, members of the American Academy of Pediatrics ment in conference sessions. The document was thenSection on Epidemiology, were selected because of distributed to the entire NASPGHAN membership fortheir contribution to the previous NASPGHAN GERD comment. Further revisions were made based on theirguidelines. suggestions following telephone conference and e-mail communications among committee members. Complete 2.2. Guideline Preparation Process voting anonymity could not be maintained through the revision process because voting was done by e-mail, but The previous guidelines developed by NASPGHAN (1) only 1 of the co-chairs (C.D.R.) was aware of e-mailand ESPGHAN (2,3) were used as the foundation for the votes. Following final committee approval, the documentcurrent guidelines. Articles written in English and pub- was endorsed by the Executive Councils of NASPGHANlished between March 1999 (the date of the previous and and October 2008 were identified using PubMedand Cumulative Index to Nursing and Allied Health 2.3. Management of Potential Conflict of InterestLiterature. Letters, editorials, case reports, and reviewswere eliminated from the initial evaluation. Additional Disclosures of potential conflicts of interest of com-articles were identified by members of the committee from mittee members or immediate family were documentedbibliographies found in other articles and study results in and shared with committee members before the firstthe public domain on the US National Institutes of Health meeting of the committee and updated before the reviewWeb site. These included review articles as well as articles of the final document. Disclosures included paid orthat involved the care of adults. A total of 377 articles donated services of any kind, research support, stockrelated to therapy and 195 articles related to etiology, ownership or options, and intellectual property rights.diagnosis, and prognosis were reviewed for this guideline. During the process of preparing the guidelines, the Using the best-available evidence from the literature, scientific data were reviewed by all of the members ofthe committee evaluated current diagnostic tests and the committee, and recommendations were voted on bytherapeutic modalities for GER and GERD. Evidence all of the members. No section of the document wasof a causal relation between GER/GERD and several written solely by any 1 member. Chairs or committeecommon symptoms or symptom complexes were eval- members did not require that any individual be removeduated. Diagnostic tests were evaluated by the following from discussions or voting based on potential conflicts ofcriteria: ability to confirm a diagnosis of GERD; ability interest. Potential conflicts of interest are listed into exclude other diagnoses with similar presentation; Appendix D. No industry support was used for theability to detect complications of GERD; and ability to production of these guidelines.predict disease severity, natural history of disease, andresponse to treatment. Therapy was evaluated consider- 3. DEFINITIONS AND MECHANISMSing efficacy, appropriate clinical indications, and poten-tial risks and complications. GER is the passage of gastric contents into the eso- The committee convened face to face 3 times and had phagus with or without regurgitation and vomiting. GERseveral conference calls. It based its recommendations on is a normal physiologic process occurring several timesits study of the literature review combined with expert per day in healthy infants, children, and adults. Mostopinion and the evidence available in the adult literature episodes of GER in healthy individuals last <3 minutes,when pediatric evidence was insufficient. Consensus was occur in the postprandial period, and cause few or noachieved for all of the recommendations through nominal symptoms (12). In contrast, GERD is present when thegroup technique, a structured quantitative method (10). reflux of gastric contents causes troublesome symptomsArticles were evaluated using the Oxford Centre for and/or complications (13). Every effort was made to useEvidence-based Medicine Levels of Evidence (11). these 2 terms strictly as defined.Using the Oxford Grades of Recommendation (11), Regurgitation in pediatrics is defined as the passage ofthe quality of evidence of each of the recommendations refluxed gastric contents into the pharynx or mouth andmade by the committee was determined and is summa- sometimes expelled out of the mouth. Regurgitationrized in Appendices A to C. Sections of the document is generally assigned as effortless and nonprojectile,were written by individual committee members, then although it may sometimes be forceful in infants (13).reviewed and edited by a separate committee member; Other terms such as ‘‘spitting-up,’’ ‘‘posseting,’’ andin most instances both a NASPGHAN and an ESPGHAN ‘‘spilling,’’ are considered equivalent to regurgitation.member participated in preparing the initial draft of each Spitting up, which occurs daily in about 50% of the infantsJ Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  8. 8. PEDIATRIC GASTROESOPHAGEAL REFLUX CLINICAL PRACTICE GUIDELINES 505younger than 3 months of age, is the most visible symptom TABLE 1. Symptoms and signs that may be associated withof regurgitation. Regurgitation resolves spontaneously in gastroesophageal refluxmost healthy infants by 12 to 14 months of age (14–18). Symptoms Reflux episodes sometimes trigger vomiting, a coor- Recurrent regurgitation with/without vomitingdinated autonomic and voluntary motor response, caus- Weight loss or poor weight gaining forceful expulsion of gastric contents through the Irritability in infantsmouth. Vomiting associated with reflux is probably a Ruminative behavior Heartburn or chest painresult of the stimulation of pharyngeal sensory afferents Hematemesisby refluxed gastric contents. Rumination refers to the Dysphagia, odynophagiaeffortless regurgitation of recently ingested food into the Wheezingmouth with subsequent mastication and reswallowing. StridorRumination syndrome is a distinct clinical entity with Cough Hoarsenessregurgitation of ingested food within minutes following Signsmeals because of the voluntary contraction of the Esophagitisabdominal muscles (19,20). Esophageal stricture Reflux episodes occur most often during transient Barrett esophagus Laryngeal/pharyngeal inflammationrelaxations of the lower esophageal sphincter (LES) Recurrent pneumoniaunaccompanied by swallowing, which permit gastric Anemiacontents to flow into the esophagus (21–23). A minor Dental erosionproportion of reflux episodes occur when the LES pres- Feeding refusalsure fails to increase during a sudden increase in intra- Dystonic neck posturing (Sandifer syndrome) Apnea spellsabdominal pressure or when LES resting pressure is Apparent life-threatening eventschronically reduced. Alterations in several protectivemechanisms allow physiologic reflux to become GERD:insufficient clearance and buffering of refluxate, delayed symptom descriptions are unreliable in infants and chil-gastric emptying, abnormalities in epithelial restitution dren younger than 8 to 12 years of age, and many of theand repair, and decreased neural protective reflexes of the purported symptoms of GERD in infants and children areaerodigestive tract. In hiatal hernia (HH), all of the nonspecific. The diagnosis of GERD is inferred whenantireflux barriers at the LES (including the crural sup- tests show excessive frequency or duration of refluxport, intraabdominal segment, and angle of His) are events, esophagitis, or a clear association of symptomscompromised (24–27) and transient LES relaxations and signs with reflux events in the absence of(TLESR) also occur with greater frequency (25). Erosive alternative diagnoses.esophagitis by itself may promote esophageal shortening Although many tests have been used to diagnoseand consequent hiatal herniation (25). HH is prevalent in GERD, few studies compare their utility. Importantly,adults and children with severe reflux complications it is not known whether tests can predict an individual(28–31), and hernia size is a major determinant of GERD patient’s response to therapy. Tests are useful to docu-severity (30,32). ment the presence of pathologic reflux or its compli- Significant clusterings of reflux symptoms, HH, ero- cations to establish a causal relation between reflux andsive esophagitis, Barrett esophagus (BE), and esophageal symptoms, to evaluate therapy, and to exclude otheradenocarcinoma occur in families, suggesting some her- conditions. Because no test can address all of theseitability of GERD and its complications (33–37). A large questions, tests must be carefully selected according toSwedish Twin Registry study found an increased con- the information sought, and the limitations of each testcordance for reflux in monozygotic compared with dizy- must be recognized.gotic twins (33). Several other pediatric patient popu-lations appear to be at higher risk for GERD than healthy 4.1. History and Physical Examinationinfants, children, or adolescents. These include individ-uals with neurologic impairment (NI), obesity, certain The major role of the history of disease and physicalgenetic syndromes, esophageal atresia (EA), chronic lung examination in the evaluation of GERD is to excludediseases, and those with a history of premature birth. other more worrisome disorders that present with vomit-These are discussed in Section 7. ing and to identify complications of GERD (Table 2). Typical presenting symptoms of reflux disease in child- 4. DIAGNOSIS hood vary with age and underlying medical condition (13,38); however, the underlying pathophysiology of The diagnosis of GERD is often made clinically based GERD is thought to be similar at all ages includingon the bothersome symptoms or signs that may be the premature infant (23,39). In 1 study, regurgitationassociated with GER (Table 1). However, subjective or vomiting, abdominal pain, and cough but not heartburn J Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  9. 9. 506 NASPGHAN/ESPGHAN PEDIATRIC GER GUIDELINE COMMITTEETABLE 2. Warning signals requiring investigation in infants testing, whereas feeding difficulties had a sensitivity of with regurgitation or vomiting 75% and specificity of 46% (61). A similar poor corre- lation of symptoms and esophageal acid exposure wasBilious vomitingGastrointestinal bleeding observed during an omeprazole treatment study in irri- Hematemesis table infants; similar reductions in crying occurred in Hematochezia both treated and untreated infants, and the extent ofConsistently forceful vomiting reduction in crying did not correlate with extent ofOnset of vomiting after 6 months of lifeFailure to thrive reduction of the RI in the treated patients (46).Diarrhea Because individual symptoms do not consistently cor-Constipation relate with objective findings or response to medicalFever treatment, parent- or patient-reported questionnairesLethargy based on clusters of symptoms have been developed.HepatosplenomegalyBulging fontanelle Orenstein et al (51,62) developed a diagnostic question-Macro/microcephaly naire for GERD in infants. A score of >7 (of 25 possible)Seizures on the initial instrument demonstrated a sensitivity ofAbdominal tenderness or distension 0.74 and specificity of 0.94 during primary validation.Documented or suspected genetic/metabolic syndrome The questionnaire has undergone several revisions (54). The questionnaire has been shown to be reliable for documentation and monitoring of reported symptoms.were the most frequently reported symptoms in children However, when applied to a population in India, it had aand adolescents with GERD. Cough and anorexia or sensitivity and specificity of only 43% and 79%, respect-feeding refusal were more common in children 1 to ively, compared with pH-monitoring results (52). In5 years of age than in older children (40). another study of infants referred for symptoms of reflux Symptoms and signs associated with reflux (Table 1) are disease and controls, the questionnaire had a sensitivitynonspecific. For example, not all of the children with GER and specificity of 47% and 81% for a RI >10% and 65%have heartburn or irritability. Conversely, heartburn and and 63% for a reflux index >5%. The questionnaire scoreirritability can be caused by conditions other than GER. failed to identify 26% of the infants with GERD. TheRegurgitation, irritability, and vomiting are common in score was positive in 17 of 22 infants with normalinfants with physiologic GER or GERD (14,18,41,42) but biopsies and pH studies and in 14 of 47 infants withare indistinguishable from regurgitation, irritability, and normal pH studies. No single symptom was significantlyvomiting caused by food allergy (43,44), colic (45,46), and associated with esophagitis (49). In another study, theother disorders. The severity of reflux or esophagitis found questionnaire was unable to identify a group of infantson diagnostic testing does not directly correlate with the responsive to proton pump inhibitor (PPI) therapy (9).severity of symptoms (47–49). Thus, no symptom or cluster of symptoms has been GERD is often diagnosed clinically in adults based on shown to reliably predict complications of reflux or toa history of heartburn, defined as substernal, burning predict those infants likely to respond to therapy.chest pain, with or without regurgitation. Recent adult A 5-item questionnaire developed for children 7 toand pediatric consensus guidelines have applied the 16 years of age had a sensitivity of 75% and a specificityterms ‘‘typical reflux syndrome’’ or ‘‘reflux chest pain of 96% compared with pH monitoring during primarysyndrome’’ to this presentation (13,50). Based on expert validation (63). No subsequent independent confirma-opinion, the diagnosis of GERD can be made in adoles- tory validation has been performed. Other diagnosticcents presenting with typical heartburn symptoms as in questionnaires, such as the GERD symptom question-adults (49,51–55). However, a clinical diagnosis based naire (53), have not been compared with objective stan-on a history of heartburn cannot be used in infants, dards like endoscopy, pH monitoring, or esophagealchildren, or nonverbal adolescents (eg, those with NI) multiple intraluminal impedance (MII) monitoring.because these individuals cannot reliably communicate Some researchers have used questionnaires to monitorthe quality and quantity of their symptoms. The verbal symptoms of children during GERD therapy (64).child can communicate pain, but descriptions of quality, Whether this method is preferable to monitoring indi-intensity, location, and severity generally are unreliable vidual symptoms is uncertain. Although daily symptomuntil at least 8 and possibly 12 years of age (56–60). diaries are frequently used in adults to monitor the effects As in adults, individual symptoms in children gener- of therapy, these have not been validated in are not highly predictive of findings of GERD byobjective studies. For example, in a study of irritable 4.2. Esophageal pH Monitoringinfants younger than 9 months of age, regurgitation >5times per day had a sensitivity of 54% and specificity of Intraluminal esophageal pH monitoring measures71% for a reflux index (RI) >10% by esophageal pH the frequency and duration of acid esophageal refluxJ Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  10. 10. PEDIATRIC GASTROESOPHAGEAL REFLUX CLINICAL PRACTICE GUIDELINES 507episodes. Most commercially available systems include a A study by Sondheimer (85) showed a different range ofcatheter for nasal insertion with 1 or more pH electrodes normal values for infants. Most of the data, provided in(antimony, glass, or ion-sensitive field effect) arrayed previous sections, pertain to infants, in whom frequencyalong its length and a system for data capture, analysis, of feeding and buffering of refluxate can confound find-and reporting. Slow electrode response times (antimony ings between studies (76). For these reasons, specificbeing the slowest) do not alter the assessment of total ‘‘cutoff’’ values that discriminate between physiologicreflux time substantially but may affect the accuracy of GER and pathologic GERD are suspect; rather, it is likelycorrelation between symptoms and reflux episodes (65). that a continuum exists such that normal ranges should beEsophageal pH monitoring is insensitive to weakly acid regarded as guidelines for interpretation rather thanand nonacid reflux events. Recently, wireless sensors that absolutes. In pH studies performed with antimony elec-can be clipped to the esophageal mucosa during endo- trodes, an RI >7% is considered abnormal, an RI <3% isscopy have allowed pH monitoring without a nasal considered normal, and an RI between 3% and 7% iscannula for up to 48 hours. Placement of wireless elec- indeterminate.trodes requires sedation or anesthesia, and comfort has Abnormal esophageal pH monitoring has not beenbeen an issue in some studies (66–68). The size of shown to correlate with symptom severity in infants.current wireless electrodes precludes their use in small In a study of infants with suspected GERD, an abnormalinfants. Benefits, risks, and indications for wireless elec- pH study (RI >10%) was associated only with pneumo-trode monitoring have not been fully defined in children. nia, apnea with fussing, defecation less than once per day,Data on reproducibility of conventional and wireless pH and constipation (49). An abnormal RI is more frequentlystudies are contradictory (68–72). observed in adults and children with erosive esophagitis By convention, a drop in intraesophageal pH <4.0 is than in normal adults and children or those with none-considered an acid reflux episode. This cutoff was rosive reflux disease (NERD), but there is substantialinitially chosen because heartburn induced by acid per- overlap among groups (79,86,87). In childen with docu-fusion of the esophagus in adults generally occurs at pH mented esophagitis, normal esophageal pH monitoring<4.0 (73). Although interpretation of pH monitoring data suggests a diagnosis other than GERD (88,89). The RI isis simplified by computerized analysis, visual inspection often abnormal in children with difficult-to-controlof the tracing is required to detect artifacts and evaluate asthma and in otherwise healthy infants with dailypossible clinical correlations. Common parameters wheezing (90). Esophageal pH monitoring may be abnor-obtained from pH monitoring include the total number mal in patients with conditions other than GERD, such asof reflux episodes, the number of reflux episodes lasting gastric outlet obstruction, motility disorders, and esopha->5 minutes, the duration of the longest reflux episode, gitis due to other disorders, including eosinophilic eso-and the RI (percentage of the entire record that esopha- phagitis (EoE) (91–94). Although multiple case seriesgeal pH is <4.0). GER events that occur while supine or report the use of esophageal pH monitoring to selectupright or while awake or asleep are often discriminated the children reported to benefit from antireflux surgeryby the automated software used in both adults and (95–99), the reliability of such data to predict improve-children, but the clinical value of such differentiation ment following either medical or surgical antirefluxhas not been established (74–80). therapy has not been established. The RI is the most commonly used summary score. The application of various methods of analysis ofSeveral scoring systems for pH-monitoring studies have esophageal pH-monitoring results, including the symp-been developed (74,75,81), but no system is clearly tom index (SI), symptom sensitivity index (SSI), andsuperior to measuring the RI (82). Normal pediatric symptom association probability (SAP), may help inranges are established for glass and antimony electrodes correlating symptoms with acid reflux. A prospectivebut not for ion-sensitive field effect or wireless technol- study in adults found that when compared with symptomogies. The normal pediatric ranges previously in general improvement following high-dose PPI therapy, the sen-use were obtained using glass electrodes (65,83), but such sitivities of the SI, SSI, and SAP were 35%, 74%, anddata poorly correlate with that obtained by the antimony 65% and specificities were 80%, 73%, and 73%, respect-electrodes now in common use (84). Moreover, normal ively (100). The clinical utility of pH studies and theirdata depend on the definition of a ‘‘normal population.’’ ability to determine a causal relation between specificIn the first study by Vandenplas et al (83), showing a low symptoms (eg, pain, cough) and reflux remain contro-RI in young infants, the definition of ‘‘normal infant’’ versial in adults (101), and are not validated inwas an infant who did not regurgitate or vomit. In the pediatric patients.second study, a ‘‘normal population’’ was defined as an Esophageal pH monitoring provides a quantitativeinfant who had not been treated for reflux (65). Although measure of esophageal acid exposure with establishedthe definition of the first study was biased toward a ‘‘too normal ranges, but the severity of pathologic acid refluxnormal’’ population, the second study included all of the does not correlate consistently with symptom severity oruntreated infants, thus possibly some infants with GERD. demonstrable complications. In childen with documented J Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  11. 11. 508 NASPGHAN/ESPGHAN PEDIATRIC GER GUIDELINE COMMITTEEesophagitis, normal esophageal pH monitoring suggests a SAP may be of additional value to prove symptomdiagnosis other than GERD (88,89). Esophageal pH association with reflux, especially when combined withmonitoring is useful for evaluating the efficacy of anti- MII (117). Whether combined esophageal pH andsecretory therapy. It may be useful to correlate symptoms impedance monitoring will provide useful measure-(eg, cough, chest pain) with acid reflux episodes, and to ments that vary directly with disease severity, prog-select those children with wheezing or respiratory symp- nosis, and response to therapy in pediatric patients hastoms in which acid reflux may be an aggravating factor. yet to be determined.The sensitivity and specificity of pH monitoring are notwell established. 4.4. Motility Studies 4.3. Combined Multiple Intraluminal Impedance Esophageal manometry measures esophageal peristal- and pH Monitoring sis, upper and lower esophageal sphincter pressures, and the coordinated function of these structures during swal- MII is a procedure for measuring the movement of lowing. Although esophageal manometry has been anfluids, solids, and air in the esophagus (102). It is a relati- important tool in studying the mechanisms of GERD,vely new technology that provides a more detailed descrip- GERD cannot be diagnosed by esophageal manometry.tion of esophageal events with a more rapid response time Manometric studies were critical in identifying TLESRthan current pH-monitoring technology. MII measures as a causative mechanism for GERD (21). A variety ofchanges in the electrical impedance (ie, resistance) nonspecific esophageal motor abnormalities have beenbetween multiple electrodes located along an esophageal found in children with developmental delay and NI, acatheter. Esophageal impedance tracings are analyzed for group at high risk for severe GERD (118). Esophagealthe typical changes in impedance caused by the passage of motor abnormalities are also common in patients withliquid, solid, gas, or mixed boluses. If the impedance esophagitis (119,120). In these 2 situations esophagealchanges of a liquid bolus appear first in the distal motor dysfunction may be a secondary phenomenonchannels and proceed sequentially to the proximal chan- related to esophagitis because it has been observed tonels, they indicate retrograde bolus movement, which is resolve upon treatment of esophagitis (119). RecentGER. The direction and velocity of a bolus can be studies indicate that there is no role for manometry incalculated using the defined distance between electrodes predicting outcome of fundoplication (121). Manometricand the time between alterations in the impedance studies are also important in confirming a diagnosis ofpattern of sequential electrode pairs. The upward extent achalasia or other motor disorders of the esophagus thatof the bolus and the physical length of the bolus can also may mimic GERD (122).be evaluated (103). MII can detect extremely small bolus Esophageal manometry may be abnormal in patientsvolumes (104). with GERD, but the findings are not sufficiently sensitive MII and pH electrodes can and should be combined on a or specific to confirm a diagnosis of GERD, nor to predictsingle catheter. The combined measurement of pH and response to medical or surgical therapy. It may be usefulimpedance (pH/MII) provides additional information as to in patients who have failed acid suppression and whowhether refluxed material is acidic, weakly acidic, or have negative endoscopy to search for a possible motilitynonacidic (105–109). Recent studies have found variable disorder, or to determine the position of the LES to placereproducibility (110,111). Evaluation of MII recordings is a pH probe. Manometric studies are useful to confirm aaided by automated analysis tools (112), but until the diagnosis of achalasia or other motor disorders of thecurrently available automatic analysis software has been esophagus that may mimic GERD.validated, a visual reading of the data is required. Normalvalues for all of the age groups have not yet been estab- 4.5. Endoscopy and Biopsylished (113). The risks and side effects of MII are low and the Upper gastrointestinal (GI) endoscopy allows directsame as those of isolated pH monitoring. The combi- visual examination of the esophageal mucosa. Mucosalnation of pH/MII with simultaneous monitoring of biopsies enable evaluation of the microscopic anatomysymptoms using video-polysomnography or manome- (123). Macroscopic lesions associated with GERDtry has proven useful for the evaluation of symptom include esophagitis, erosions, exudate, ulcers, strictures,correlations between reflux episodes and apnea, cough, HH, areas of possible esophageal metaplasia, and polyps.other respiratory symptoms, and behavioral symptoms Although endoscopy can detect strictures, subtle degrees(23,24,114–116). The technology is especially useful in of narrowing may be better shown on barium contrastthe postprandial period or at other times when gastric study, during which the esophagus can be distendedcontents are nonacidic. The relation between weakly with various techniques, such as a radioopaque pillacid reflux and symptoms of GERD requires clarifica- and barium-soaked bread or marshmallows. Malrotationtion. Measurement of other parameters such as SI or and achalasia cannot be diagnosed by endoscopy. TheseJ Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  12. 12. PEDIATRIC GASTROESOPHAGEAL REFLUX CLINICAL PRACTICE GUIDELINES 509and other anatomic and motility disorders of the esopha- TABLE 3. Causes of esophagitisgus are better evaluated by barium radiology or motility Gastroesophageal reflux Graft-versus-host diseasestudies. Eosinophilic esophagitis Caustic ingestion Recent global consensus guidelines define reflux eso- Infections Postsclerotherapy/bandingphagitis as the presence of endoscopically visible breaks Candida albicans Radiation/chemotherapyin the esophageal mucosa at or immediately above the Herpes simplex Connective tissue diseasegastroesophageal junction (13,50,124). Evidence from Cytomegalovirus Bullous skin diseases Crohn disease Lymphomaadult studies indicates that visible breaks in the esopha- Vomiting, bulimiageal mucosa are the endoscopic signs of greatest inter- Pill inducedobserver reliability (125–127). Operator experience isan important component of interobserver reliability(128,129). Mucosal erythema or an irregular Z-line isnot a reliable sign of reflux esophagitis (126,127). the differential diagnosis, it is advisable to take eso-Grading the severity of esophagitis, using a recognized phageal biopsies from the proximal and distal esophagusendoscopic classification system, is useful for evaluation (93). Mucosal eosinophilia may be present in theof the severity of esophagitis and response to treatment. esophageal mucosa in asymptomatic infants youngerThe Hetzel-Dent classification (125) has been used in than 1 year of age (143), and in symptomatic infantsseveral pediatric studies (29,130,131), whereas the Los eosinophilic infiltrate may be because of milk-proteinAngeles classification (124) is generally used for adults, allergy (142).but it is suitable also for children. The presence of There is insufficient evidence to support the use ofendoscopically normal esophageal mucosa does not histology to diagnose or exclude GERD. The primaryexclude a diagnosis of NERD or esophagitis of other role for esophageal histology is to rule out other con-etiologies (93,132,133). ditions in the differential diagnosis, such as EoE, Crohn The diagnostic yield of endoscopy is generally greater disease, BE, and infection. This conclusion concurs withif multiple samples of good size and orientation are that of a global pediatric consensus group that includedobtained from biopsy sites that are identified relative some members of the present committee (E.H., Y.V.,to major esophageal landmarks (28,123,134). Several C.D.R.) (13). When symptoms suggestive of GERD arevariables have an impact on the validity of histology present in adolescents or adults in the absence of erosiveas a diagnostic tool for reflux esophagitis (133,135). esophagitis, the clinical entity is known as NERD. InThese include sampling error because of the patchy NERD, there is no evidence that esophageal histologydistribution of inflammatory changes and a lack in makes a difference to clinical care decisions; that is,standardization of biopsy location, tissue processing, patient treatment is guided by symptoms, whether or notand interpretation of morphometric parameters. Histo- reactive histologic changes are present on biopsy.logy may be normal or abnormal in NERD because At endoscopy, accurate documentation of esophago-GERD is an inherently patchy disease (133,136). Histo- gastric landmarks is necessary for the diagnosis of HHlogic findings of eosinophilia, elongation of papillae (rete and endoscopically suspected esophageal metaplasiapegs), basal hyperplasia, and dilated intercellular spaces (ESEM) (123,134,144–147). This is of particular import-(spongiosis) are neither sensitive nor specific for reflux ance in children with severe esophagitis, in whom land-esophagitis. They are nonspecific reactive changes that marks may be obscured by bleeding or exudate, or whenmay be found in esophagitis of other causes or in healthy landmarks are displaced by anatomic abnormalities orvolunteers (49,89,132,133,135,137–141). Recent studies HH (28,123,134). In these circumstances, a course ofhave shown considerable overlap between the histology high-dose PPIs for at least 12 weeks is advised, followedof reflux esophagitis and EoE (93,94,132,142). Many by a repeat endoscopy, to remove the exudative camou-histologic parameters are influenced by drugs used to flage and better visualize the landmarks (134,148).treat esophagitis or other disorders. When biopsies from ESEM show columnar epi- GERD is likely the most common cause of esophagitis thelium, the term BE should be applied and the presencein children, but other disorders such as EoE, Crohn or absence of intestinal metaplasia specified (13,50).disease, and infections also cause esophagitis (Table 3) Thus, BE may be diagnosed in the presence of only(132). EoE and GERD have similar symptoms and cardia-type mucosa (149,150). BE occurs with greatestsigns and can be best distinguished by endoscopy with frequency in children with underlying conditions puttingbiopsy. A key difference, endoscopically, is that EoE is them at high risk for GERD (see Section 7) (28,31).generally not an erosive disease but has its own typicalendoscopic features such as speckled exudates, trachea- 4.6. Barium Contrast Radiographylization of the esophagus, or linear furrowing. In up to30% of cases, however, the esophageal mucosal appear- The upper GI series is neither sensitive nor specificance is normal (93). When EoE is considered as a part of for diagnosing GERD. The sensitivity, specificity, and J Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  13. 13. 510 NASPGHAN/ESPGHAN PEDIATRIC GER GUIDELINE COMMITTEEpositive predictive value of the upper GI series range 4.8. Esophageal and Gastric Ultrasonographyfrom 29% to 86%, 21% to 83%, and 80% to 82%,respectively, when compared with esophageal pH moni- Ultrasonography is not recommended as a test fortoring (151–157). The brief duration of the upper GI GERD but can provide information not available throughseries produces false-negative results, whereas the fre- other technology. Ultrasonography of the gastroesopha-quent occurrence of nonpathological reflux during the geal junction can detect fluid movements over shortexamination produces false-positive results. periods of time and thereby can detect nonacid reflux Therefore, routine performance of upper GI series to events. It can also detect HH, length and position of thediagnose reflux or GERD is not justified (158). However, LES relative to the diaphragm, and magnitude of thethe upper GI series is useful to detect anatomic abnor- gastroesophageal angle of His. Barium upper GI seriesmalities such as esophageal stricture, HH, achalasia, can provide the same information. When compared withtracheoesophageal fistula, intestinal malrotation, or pylo- the results of 24-hour esophageal pH testing as a diag-ric stenosis, which may be considered in the differential nostic test for GERD, the sensitivity of color Dopplerdiagnosis of infants and children with symptoms suggest- ultrasound performed for 15 minutes postprandially ising GERD. about 95% with a specificity of only 11%, and there is no correlation between reflux frequency detected by ultra- 4.7. Nuclear Scintigraphy sound and reflux index detected by pH monitoring (177,178). Intraluminal esophageal ultrasound is used In gastroesophageal scintigraphy, food or formula in adults to evaluate esophageal wall thickness andlabeled with 99technetium is introduced into the stomach muscle shortening, parameters that vary with inflam-and areas of interest—stomach, esophagus, and lungs— mation, scarring, and malignancy (179). At present, thereare scanned for evidence of reflux and aspiration. The is no role for ultrasound as a routine diagnostic tool fornuclear scan evaluates only postprandial reflux and GERD in children.demonstrates reflux independent of the gastric pH. Scin-tigraphy can provide information about gastric emptying, 4.9. Tests on Ear, Lung, and Esophageal Fluidswhich may be delayed in children with GERD (159–161). A lack of standardized techniques and the absence Recent studies have suggested that finding pepsin, aof age-specific norms limit the value of this test. Sensi- gastric enzyme, in middle ear effusions of children withtivity and specificity of a 1-hour scintigraphy for the chronic otitis media, indicates that reflux is playing andiagnosis of GERD are 15% to 59% and 83% to 100%, etiologic role (180–183). One recent study showed norespectively, when compared with 24-hour esophageal relation between the presence of pepsin in the middle earpH monitoring (162–165). Late postprandial acid expo- and symptoms of GERD (184), and this relation has notsure detected by pH monitoring may be missed with been validated in controlled treatment trials. Similarly,scintigraphy (166). the presence of lactose, glucose, pepsin, or lipid-filled Gastroesophageal scintigraphy scanning can detect macrophages in bronchoalveolar lavage fluids has beenreflux episodes and aspiration occurring during or shortly proposed to implicate aspiration secondary to reflux as aafter meals, but its reported sensitivity for microaspira- cause of some chronic pulmonary conditions (185–187).tion is relatively low (167–169). Evidence of pulmonary No controlled studies have proven that reflux is the onlyaspiration may be detected during a 1-hour scintigraphic reason these compounds appear in bronchoalveolarstudy or on images obtained up to 24 hours after admin- lavage fluids or that reflux is the cause of pulmonaryistration of the radionuclide (170). A negative test does disease when they are present.not exclude the possibility of infrequently occurring Continuous monitoring of bilirubin in the esophagusaspiration (168). One study of children with refractory has been suggested as a means of detecting esophagealrespiratory symptoms found that half had scintigraphic reflux of duodenal juice or duodenogastroesophagealevidence of pulmonary aspiration (169). However, reflux. Duodenal juice components appear to damageaspiration of both gastric contents and saliva also occurs the esophagus in a pH-dependent manner (188). Twoin healthy adults during deep sleep (171,172). uncontrolled pediatric case series have suggested that Gastric emptying studies have shown prolonged half- duodenogastroesophageal reflux produced GERD thatemptying times in children with GER. Delayed gastric was refractory to therapy with PPIs (189,190). One studyemptying may predispose to GERD. Tests of gastric indicated that therapy with PPIs decreased the esopha-emptying are not a part of the routine examination of geal damage caused by duodenogastroesophageal refluxpatients with suspected GERD, but may be important (190). At present, there is insufficient evidence to recom-when symptoms suggest gastric retention (173–176). mend continuous monitoring of the esophagus for bili- Nuclear scintigraphy is not recommended in the rou- rubin in the routine evaluation of GERD. The role of biletine diagnosis and management of GERD in infants reflux in children resistant to PPI treatment has notand children. been established.J Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  14. 14. PEDIATRIC GASTROESOPHAGEAL REFLUX CLINICAL PRACTICE GUIDELINES 511 4.10. Empiric Trial of Acid Suppression as a lescents, lifestyle changes include modification of diet Diagnostic Test and sleeping position, weight reduction, and smoking cessation. In adults, empiric treatment with acid suppression, that Medications for use in GERD include agents to bufferis, without diagnostic testing, has been used for symp- gastric contents or suppress acid secretion. Agents affect-toms of heartburn (191), chronic cough (192,193), non- ing GI motility are discussed. Surgical therapy includescardiac chest pain (194), and dyspepsia (195). However, fundoplication and other procedures to eliminate reflux.empiric therapy has only modest sensitivity and speci-ficity as a diagnostic test for GERD, depending upon the 5.1. Lifestyle Changescomparative reference standard used (endoscopy, pHmonitoring, symptom questionnaires) (196), and the Parental education, guidance, and support are alwaysappropriate duration of a ‘‘diagnostic trial’’ of acid required and usually sufficient to manage healthy, thriv-suppression has not been determined. A meta-analysis ing infants with symptoms likely because of physiologicevaluating pooled data from 3 large treatment trials GER.among the adults with NERD showed that 85% of thepatients who had symptom resolution after 1 week of PPI 5.1.1. Feeding Changes in Infantstreatment remained well for the entire 4 weeks of PPItreatment, thus ‘‘confirming’’ the diagnosis of GERD About 50% of the healthy 3- to 4-month-old infants(197). However, 22% of the patients who had no regurgitate at least once per day (16,18) and up to 20% ofimprovement after 1 week of treatment did improve by caregivers in the United States seek medical help for thisthe fourth week of treatment. An uncontrolled trial of normal behavior (16). Breast-fed and formula-fed infantsesomeprazole therapy in adolescents with heartburn, have a similar frequency of physiologic GER, althoughepigastric pain, and acid regurgitation showed complete the duration of reflux episodes measured by pH proberesolution of symptoms in 30% to 43% by 1 week, but the may be shorter in breast-fed infants (201–203).responders increased to 65% following 8 weeks of treat- A subset of infants with allergy to cow’s milk proteinment (55). Another uncontrolled treatment trial of pan- experience regurgitation and vomiting indistinguishabletoprazole in children ages 5 to 11 years reported greater from that associated with physiologic GERsymptom improvement at 1 week with one 40-mg dose (9,69,142,204–206). In these infants, vomiting fre-compared with one 10-mg or 20-mg dose (64). After quency decreases significantly (usually within 2 weeks)8 weeks all of the treatment groups improved. Similar after the elimination of cow’s milk protein from the diet,improvement in symptoms over time has been observed and reintroduction causes recurrence of symptomsin adults with erosive esophagitis (198,199). One study of (206,207). Studies support the use of extensively hydro-infants with symptoms suggestive of GERD who were lyzed or amino acid formula in formula-fed infants withtreated empirically with a PPI showed no efficacy over bothersome regurgitation and vomiting for trials lastingplacebo (9). up to 4 weeks (206–208). Cow’s milk protein and other The treatment period required to achieve uniform proteins pass into human breast milk in small quantities.therapeutic responses with PPI therapy probably varies Breast-fed infants with regurgitation and vomiting maywith disease severity, treatment dose, and specific symp- therefore benefit from a trial of withdrawal of cow’s milktoms or complications (200). The 2-week ‘‘PPI test’’ and eggs from the maternal diet (209,210). The symp-lacks adequate specificity and sensitivity for use in toms of infant reflux are almost never so severe thatclinical practice. In an older child or adolescent with breast-feeding should be discontinued. There are nosymptoms suggesting GERD, an empiric PPI trial is studies specifically evaluating soy protein allergy injustified for up to 4 weeks. Improvement following treat- infants with regurgitation and vomiting, or the role ofment does not confirm a diagnosis of GERD because soy protein–based formula in the treatment of infantssymptoms may improve spontaneously or respond by a with regurgitation. Moreover, there are no data on allergyplacebo effect. There is no evidence to support an empiric to possible formula-thickening agents such as rice cereals.trial of pharmacologic treatment in infants and young One study in infants showed that large volume feed-children as a diagnostic test of GERD. ings promote regurgitation, probably by increasing the frequency of TLESR and reduced feeding volume and 5. TREATMENT decreased reflux frequency (211). Severe reduction in feeding volume during an extended period may deprive Management options for physiologic GER and for the infant of needed energy and adversely affect weightGERD discussed in this section include lifestyle changes, gain. Infants with inadequate weight gain because ofpharmacologic therapy, and surgery. Lifestyle changes in losses by regurgitation may benefit from increasing theinfants with physiologic GER include nutrition, feeding, energy density of formula when volume or frequency ofand positional modifications. In older children and ado- feedings is decreased as a part of therapy. J Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009
  15. 15. 512 NASPGHAN/ESPGHAN PEDIATRIC GER GUIDELINE COMMITTEE Adding thickening agents such as rice cereal to Although the actual number of esophageal reflux epi-formula does not decrease the time with pH <4 (reflux sodes may not decrease, the reduction in regurgitationindex) measured by esophageal pH studies, but it does may be a welcome improvement in quality of life fordecrease the frequency of overt regurgitation (211–215). caregivers. The impact of thickened formula on theStudies with combined pH/MII show that the height of natural history of physiologic GER or GERD has notreflux in the esophagus is decreased with thickened been studied. The allergenicity of commercial thickeningformula as well as the overt frequency of regurgitation, agents is uncertain, and the possible nutritional risks ofbut not the frequency of reflux episodes (114). One study long-term use require additional study.reported an improvement in esophageal pH parameters Infants with GERD who are unable to gain weightwith cornstarch-thickened formula (216). Another study despite conservative measures and in whom nasogastricshowed no change in esophageal impedance parameters or nasojejunal feeding may be beneficial are rare (231).of premature infants receiving cornstarch-thickened Similarly, nasojejunal feeding is occasionally usefulhuman milk (217). in infants with recurrent reflux-related pneumonia to In the United States, rice cereal is the most commonly prevent recurrent aspiration. Although these approachesused thickening agent for formula (214). Rice cereal– to therapy are widely used, there are no controlledthickened formula produces a decrease in the volume of studies comparing them to pharmacologic or surgicalregurgitation but may increase coughing during feedings treatments.(218). Formula with added rice cereal may require anipple with an enlarged hole to allow adequate flow. 5.1.2. Positioning Therapy for InfantsExcessive energy intake is a potential problem with long-term use of feedings thickened with rice cereal or corn- Several studies in infants have demonstrated signifi-starch (219). Thickening a 20-kcal/oz infant formula with cantly decreased acid reflux in the flat prone position1 tablespoon of rice cereal per ounce increases the energy compared with flat supine position (232–236). There isdensity to $34 kcal/oz ($1.1 kcal/mL). Thickening with conflicting evidence as to whether infants placed prone1 tablespoon per 2 oz of formula increases the energy with the head elevated have less reflux than those keptdensity to $27 kcal/oz ($0.95 kcal/mL). prone but flat (232–234,237). The amount of reflux in Commercial antiregurgitant (AR) formulae containing supine infants with head elevated is equal to or greaterprocessed rice, corn or potato starch, guar gum, or locust than in infants supine and flat (232,234,238,239). Thebean gum are available in Europe, Asia, and the United semisupine positioning as attained in an infant car seatStates. These formulae decrease overt regurgitation and exacerbates GER (240). Although the full upright pos-vomiting frequency and volume compared with unthick- ition appears to decrease measured reflux, 1 studyened formulae (1,220,221) or formulae thickened with suggested that using formula thickened with rice cerealrice cereal (216,222–226). However, a natural history is more effective in decreasing the frequency of regur-study showed only a nonsignificant decrease in episodes gitation than upright positioning after feeds (223).of regurgitation and no change in infant comfort among In the 1980s, prone positioning was recommendedinfants fed with a formula thickened with bean gum for the treatment of GERD in infants because studiesversus those fed with a formula thickened with rice showed less reflux in this position. Prone sleep position-cereal or regular formula (227). When ingested in normal ing is associated with longer uninterrupted sleep periodsvolumes, AR formulae contain an energy density, osmo- and supine sleep positioning with more frequent arousalslarity, protein, calcium, and fatty acid content appropriate and crying (241). However, concerns regarding theto an infant’s nutritional needs, whereas a formula with association between prone positioning and sudden infantadded thickener has a higher energy density, and in death syndrome (SIDS) required a reassessment of thenormal ingested volumes this may provide more energy benefits and risks of prone positioning for reflux man-than needed. A largely untested potential advantage of agement. The Nordic Epidemiological SIDS StudyAR formulae is that they do not require a substantially demonstrated that the odds ratio of mortality from SIDSincreased sucking effort, obviating the need for use of a was more than 10 times higher in prone-sleeping infantslarge-bore nipple hole. In vitro studies have shown a and 3 times higher in side-sleeping infants than indecrease in the absorption of minerals and micronutrients supine-sleeping infants (242–244). Therefore, pronefrom formulae commercially thickened with indigestible positioning is acceptable if the infant is observed andbut not digestible carbohydrates (228,229). The clinical awake, particularly in the postprandial period, but pronesignificance of these findings is unclear because a 3- positioning during sleep can only be considered inmonth follow-up study of children on formula thickened infants with certain upper airway disorders in whichwith indigestible carbohydrate showed normal growth the risk of death from GERD may outweigh the risk ofand nutritional parameters (230). SIDS. Prone positioning may be beneficial in children The use of AR formulae and formulae with added older than 1 year of age with GER or GERD whose risk ofthickener results in a decrease of observed regurgitation. SIDS is negligible.J Pediatr Gastroenterol Nutr, Vol. 49, No. 4, October 2009