SUPPLEMENT TO Otolaryngology– Head and Neck Surgery

MAY 2004 VOLUME 130 NUMBER 5

Clinical practice guideline: Otitis media with effusion

RICHARD M. ROSENFELD, MD, MPH, LARRY CULPEPPER, MD, MPH, KAREN J. DOYLE, MD, PHD, KENNETH M. GRUNDFAST, MD, ALEJANDRO HOBERMAN, MD, MARGARET A. KENNA, MD, ALLAN S. LIEBERTHAL, MD, MARTIN MAHONEY, MD, PHD, RICHARD A. WAHL, MD, CHARLES R. WOODS, JR, MD, MS, and BARBARA YAWN, MSC

The clinical practice guideline on otitis media with effusion, and presence and severity of associated effusion (OME) provides evidence-based recom- symptoms at each assessment of the child with OME; mendations on diagnosing and managing OME in (2) distinguish the child with OME who is at risk for children. This is an update of the 1994 clinical prac- speech, language, or learning problems from other tice guideline “Otitis Media With Effusion in Young children with OME and more promptly evaluate hear- Children,” which was developed by the Agency for ing, speech, language, and need for intervention in Healthcare Policy and Research (now the Agency children at risk; and (3) manage the child with OME for Healthcare Research and Quality). In contrast to who is not at risk with watchful waiting for 3 months the earlier guideline, which was limited to children from the date of effusion onset (if known), or from the aged 1 to 3 years with no craniofacial or neurologic date of diagnosis (if onset is unknown). abnormalities or sensory deficits, the updated The subcommittee also made recommendations guideline applies to children aged 2 months that (4) hearing testing be conducted when OME through 12 years with or without developmental persists for 3 months or longer, or at any time that disabilities or underlying conditions that predispose language delay, learning problems, or a significant to OME and its sequelae. The American Academy hearing loss is suspected in a child with OME; (5) of Pediatrics, American Academy of Family Physi- children with persistent OME who are not at risk cians, and American Academy of Otolaryngology– should be reexamined at 3- to 6-month intervals Head and Neck Surgery selected a subcommittee until the effusion is no longer present, significant composed of experts in the fields of primary care, hearing loss is identified, or structural abnormalities otolaryngology, infectious diseases, epidemiology, of the eardrum or middle ear are suspected; and hearing, speech and language, and advanced (6) when a child becomes a surgical candidate, practice nursing to revise the OME guideline. tympanostomy tube insertion is the preferred initial The subcommittee made a strong recommenda- procedure. Adenoidectomy should not be per- tion that clinicians use pneumatic otoscopy as the formed unless a distinct indication exists (nasal ob- primary diagnostic method and distinguish OME struction, chronic adenoiditis); repeat surgery con- from acute otitis media (AOM). sists of adenoidectomy plus myringotomy, with or The subcommittee made recommendations that cli- without tube insertion. Tonsillectomy alone or myr- nicians should (1) document the laterality, duration of ingotomy alone should not be used to treat OME. The subcommittee made negative recommenda- tions that (1) population-based screening programs Richard M. Rosenfeld, MD, SUNY-HSC Brooklyn, Depart- for OME not be performed in healthy, asymptom- ment of Pediatric Otolaryngology, 339 Hicks Street, atic children and (2) antihistamines and deconges- Brooklyn, NY 11201; e-mail, [email protected]. tants are ineffective for OME and should not be 0194-5998/$30.00 Copyright © 2004 by the American Academy of Otolaryn- used for treatment; antimicrobials and corticoste- gology–Head and Neck Surgery Foundation, Inc., and the roids do not have long-term efficacy and should American Academy of Pediatrics not be used for routine management. doi:10.1016/j.otohns.2004.02.002 The subcommittee gave as options that (1) tympa-

S95 Otolaryngology– Head and Neck Surgery S96 ROSENFELD et al May 2004 nometry can be used to confirm the diagnosis of the presence of middle-ear effusion, and signs and OME and (2) when children with OME are referred symptoms of middle-ear inflammation. Persistent by the primary clinician for evaluation by an oto- middle-ear fluid from OME results in decreased laryngologist, audiologist, or speech-language pa- mobility of the tympanic membrane and serves as thologist, the referring clinician should document a barrier to sound conduction.3 About 2.2 million the effusion duration and specific reason for referral diagnosed episodes of OME occur annually in the (evaluation, surgery), and provide additional rele- United States, yielding a combined direct and in- vant information such as history of AOM and devel- 2 opmental status of the child. The subcommittee direct annual cost estimate of $4.0 billion. made no recommendations for (1) complementary OME may occur spontaneously because of poor and alternative medicine as a treatment for OME eustachian tube function, or as an inflammatory based on a lack of scientific evidence document- response following AOM. About 90% of children ing efficacy and (2) allergy management as a (80% of individual ears) have OME at some time treatment for OME based on insufficient evidence before school age,4 most often between ages 6 of therapeutic efficacy or a causal relationship be- months and 4 years.5 In the first year of life, more tween allergy and OME. Last, the panel compiled a than 50% of children will experience OME, in- list of research needs based on limitations of the creasing to more than 60% by age 2 years.6 Many evidence reviewed. episodes resolve spontaneously within 3 months, The purpose of this guideline is to inform clinicians but about 30% to 40% of children have recurrent of evidence-based methods to identify, monitor, and manage OME in children aged 2 months OME and 5% to 10% of episodes last 1 year or 1,4,7 through 12 years. The guideline may not apply to longer. children older than 12 years because OME is un- The primary outcomes considered in the guide- common and the natural history is likely to differ line include hearing loss; effects on speech, lan- from younger children who experience rapid de- guage, and learning; physiologic sequelae; health velopmental change. The target population in- care utilization (medical, surgical); and quality of cludes children with or without developmental dis- life.1,2 The high prevalence of OME, difficulties in abilities or underlying conditions that predispose to diagnosis and assessing duration, increased risk of OME and its sequelae. The guideline is intended for conductive hearing loss, potential impact on lan- use by providers of health care to children, includ- guage and cognition, and significant practice vari- ing primary care and specialist physicians, nurses ations in management8 make OME an important and nurse practitioners, physician assistants, audi- ologists, speech-language pathologists, and child condition for the use of up-to-date evidence-based development specialists. The guideline is applica- practice guidelines. ble to any setting in which children with OME would be identified, monitored, or managed. METHODS This guideline is not intended as a sole source of General Methods and Literature guidance in evaluating children with OME. Rather, it Search is designed to assist primary care and other clini- In developing an evidence-based clinical prac- cians by providing an evidence-based framework tice guideline on managing OME, the American for decision-making strategies. It is not intended to Academy of Pediatrics (AAP), American Acad- replace clinical judgment or establish a protocol for all children with this condition, and may not provide emy of Family Physicians, and American Acad- the only appropriate approach to diagnosing and emy of OtolaryngologyÐHead and Neck Surgery managing this problem. (Otolaryngol Head Neck worked with the Agency for Healthcare Research Surg 2004;130:S95.) and Quality (AHRQ) and other organizations. This effort included representatives from each partner- Otitis media with effusion (OME) as discussed ing organization along with liaisons from audiol- in this guideline is defined as the presence of fluid ogy, speech-language pathology, informatics, and in the middle ear without signs or symptoms of advanced practice nursing. The most current liter- acute ear infection.1,2 OME is considered distinct ature on managing children with OME was re- from acute otitis media (AOM), which is defined viewed, and research questions were developed to as a history of acute onset of signs and symptoms, guide the evidence review process. Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S97

The AHRQ report on OME from the Southern piled and reviewed by the subcommittee California Evidence-Based Practice Center (EPC) cochairpersons. The recommendations contained focused on key questions of natural history; diag- in the practice guideline are based on the best nostic methods; and long-term speech, language, available published data through April 2003. and hearing outcomes.2 Searches were conducted Where data are lacking, a combination of clinical through January 2000 in MEDLINE, EMBASE, experience and expert consensus was used. A and the Cochrane Library. Additional articles were scheduled review process will occur at 5 years identified by review of reference listings in pro- from publication or sooner if new compelling ev- ceedings, reports, and other guidelines. The EPC idence warrants earlier consideration. accepted 970 articles for full review after screen- ing 3200 abstracts. The EPC reviewed articles Classification of Evidence-based using established quality criteria9,10 and included Statements randomized trials, prospective cohorts, and valida- Guidelines are intended to reduce inappropriate tions of diagnostic tests (validating cohort studies). variations in clinical care, produce optimal health The AAP subcommittee on OME updated the outcomes for patients, and minimize harm. The AHRQ review with articles identified by an elec- evidence-based approach to guideline develop- tronic MEDLINE search through April 2003 and ment requires that the evidence supporting a pol- with additional material identified manually by icy be identified, appraised, and summarized and subcommittee members. Copies of relevant arti- that an explicit link between evidence and state- cles were distributed to the subcommittee for con- ments be defined. Evidence-based statements re- sideration. A specific search for articles relevant to flect the quality of evidence and the balance of complementary and alternative medicine (CAM) benefit and harm that is anticipated when the state- was performed using MEDLINE and AMED ment is followed. The AAP definitions for evi- through April 2003. Articles relevant to allergy dence-based statements16 are listed in Tables 1 and OME were identified using MEDLINE and 2. through April 2003. The subcommittee met 3 Guidelines are never intended to overrule pro- times over a 1-year period, ending in May 2003, fessional judgment; rather, they may be viewed as with interval electronic review and feedback on a relative constraint on individual clinician discre- each guideline draft to ensure accuracy of content tion in a particular clinical circumstance. Less and consistency with standardized criteria for re- frequent variation in practice is expected for a porting clinical practice guidelines.11 strong recommendation than might be expected In May 2003 the Guidelines Review Group of with a recommendation. Options offer the most the Yale Center for Medical Informatics used the opportunity for practice variability.17 All clini- Guideline Elements Model12 to categorize content cians should always act and decide in a way that of the present draft guideline. Policy statements they believe will best serve their patients’ interests were parsed into component decision variables and needs, regardless of guideline recommenda- and actions, then assessed for decidability and tions. Guidelines represent the best judgment of a executability. Quality appraisal using established team of experienced clinicians and methodologists criteria13 was performed with Guideline Elements addressing the scientific evidence for a particular Model-Q Online.14,15 Implementation issues were topic.16 predicted using the Implementability Rating Pro- Making recommendations about health prac- file, an instrument under development by the Yale tices involves value judgments on the desirability Guidelines Review Group (R. Shiffman, MD, of various outcomes associated with management written communication, May 2003). OME sub- options. Values applied by the OME subcommit- committee members received summary results and tee sought to minimize harm and diminish unnec- modified an advanced draft of the guideline. essary therapy. Emphasis was placed on promptly The final draft practice guideline underwent ex- identifying and managing children at risk for tensive peer review by numerous entities identi- speech, language, or learning problems to maxi- fied by the subcommittee. Comments were com- mize opportunities for beneficial outcomes. Direct Otolaryngology– Head and Neck Surgery S98 ROSENFELD et al May 2004

Table 1. Guideline definitions for evidence-based statements

Statement Definition Implication

Strong recommendation A strong recommendation means the subcommittee Clinicians should follow a strong believes that the benefits of the recommended ap- recommendation unless a clear proach clearly exceed the harms (or that the harms and compelling rationale for an clearly exceed the benefits in the case of a strong alternative approach is present. negative recommendation) and that the quality of the supporting evidence is excellent (Grade A or B).* In some clearly identified circumstances, strong recommendations may be made based on lesser evidence when high-quality evidence is im- possible to obtain and the anticipated benefits strongly outweigh the harms. Recommendation A recommendation means the subcommittee believes Clinicians should also generally that the benefits exceed the harms (or that the follow a recommendation, but harms exceed the benefits in the case of a negative should remain alert to new infor- recommendation), but the quality of evidence is mation and sensitive to patient not as strong (Grade B or C).* In some clearly preferences. identified circumstances, recommendations may be made based on lesser evidence when high-quality evidence is impossible to obtain and the antici- pated benefits outweigh the harms. Option An option means that either the quality of evidence Clinicians should be flexible in their that exists is suspect (Grade D)* or that well-done decision making regarding appro- studies (Grade A, B, or C)* show little clear ad- priate practice, although they may vantage to one approach versus another. set bounds on alternatives; patient preference should have a substan- tial influencing role. No recommendation No recommendation means there is both a lack of Clinicians should feel little con- pertinent evidence (Grade D)* and an unclear bal- straint in their decision making ance between benefits and harms. and be alert to new published evi- dence that clarifies the balance of benefit versus harm; patient pref- erence should have a substantial influencing role.

*See Table 2 for definition of evidence grades.

Table 2. Evidence quality for grades of evidence costs were also considered in the statements con- Grade Evidence quality cerning diagnosis and screening, and to a lesser extent in other statements. A Well-designed randomized, controlled trials or diagnostic studies performed 1A. PNEUMATIC OTOSCOPY: Clinicians on a population similar to the guide- line’s target population should use pneumatic otoscopy as the primary B Randomized, controlled trials or diag- diagnostic method for OME. OME should be nostic studies with minor limitations; distinguished from AOM. Strong Recommenda- overwhelmingly consistent evidence tion based on systematic review of cohort studies from observational studies C Observational studies (case control and and preponderance of benefit over harm. cohort design) 1B. TYMPANOMETRY: Tympanometry D Expert opinion, case reports, reasoning can be used to confirm the diagnosis of OME. from first principles (bench research Option based on cohort studies and a balance of or animal studies) benefit and harm. Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S99

Diagnosing OME correctly is fundamental to dient analysis is a low-cost alternative to proper management. Moreover, OME must be dif- tympanometry that does not require an airtight ferentiated from AOM to avoid unnecessary anti- seal in the ear canal; however, validation studies microbial use.18,19 primarily have used children aged 2 years or older OME is defined as fluid in the middle ear with- with a high prevalence of OME.25-27 out signs or symptoms of acute ear infection.2 The While no research studies have examined tympanic membrane is often cloudy with dis- whether pneumatic otoscopy causes discomfort, 20 tinctly impaired mobility, and an air-fluid level expert consensus suggests that the procedure does or bubble may be visible in the middle ear. Con- not have to be painful, especially when symptoms versely, diagnosing AOM requires a history of of acute infection (AOM) are absent. A nontrau- acute onset of signs and symptoms, the presence matic examination is facilitated by using a gentle of middle-ear effusion, and signs and symptoms of touch, restraining the child properly when neces- middle-ear inflammation. The critical distinguish- sary, and inserting the speculum only into the ing feature is that only AOM has acute signs and outer one third (cartilaginous portion) of the ear symptoms. Distinct redness of the tympanic mem- canal.28 The pneumatic bulb should be slightly brane should not be a criterion for antibiotic pre- compressed before insertion because OME is of- scribing because it has poor predictive value for ten associated with a negative middle-ear pressure, AOM and is present in about 5% of ears with which can be more accurately assessed by releas- OME.20 ing the already compressed bulb. The otoscope The AHRQ evidence report2 systematically re- must be fully charged, the bulb (halogen or xenon) viewed the sensitivity, specificity, and predictive bright and luminescent,29 and the insufflator bulb values of 9 diagnostic methods for OME. Pneu- attached tightly to the head to avoid the loss of an matic otoscopy had the best balance of sensitivity and specificity, consistent with the 1994 guide- air seal. The window must also be sealed. line.1 Meta-analysis revealed a pooled sensitivity Evidence Profile: Pneumatic Otoscopy of 94% (95% CI, 91%-96%) and specificity of ● Aggregate evidence quality: A, diagnostic 80% (95% CI, 75%-86%) for validated observers studies in relevant populations using pneumatic otoscopy versus myringotomy as ● Benefit: improved diagnostic accuracy; inex- the gold standard. Pneumatic otoscopy should pensive equipment therefore remain the primary method of OME ● Harm: cost of training clinicians in pneumatic diagnosis because the instrument is readily avail- otoscopy able in practice settings, cost effective, and accu- ● Benefits-harms assessment: preponderance of rate in experienced hands. Nonpneumatic oto- benefit over harm scopy is not advised for primary diagnosis. ● Policy level: strong recommendation The accuracy of pneumatic otoscopy in routine Evidence Profile: Tympanometry clinical practice may be less than that shown in ● Aggregate evidence quality: B, diagnostic published results because clinicians have varying studies with minor limitations training and experience.21,22 When the diagnosis ● Benefit: increased diagnostic accuracy be- of OME is uncertain, tympanometry or acoustic yond pneumatic otoscopy; documentation reflectometry should be considered as an adjunct ● Harm: acquisition cost, administrative bur- to pneumatic otoscopy. Tympanometry with a den, recalibration standard 226-Hz probe tone is reliable for infants ● Benefits-harms assessment: balance of benefit aged 4 months or older and has good interobserver and harm agreement of curve patterns in routine clinical ● Policy level: option practice.23,24 Younger infants require specialized equipment with a higher probe tone frequency. 1C. SCREENING: Population-based screen- Tympanometry generates costs related to instru- ing programs for OME are not recommended ment purchase, annual calibration, and test admin- in healthy, asymptomatic children. Recommen- istration. Acoustic reflectometry with spectral gra- dation based on randomized, controlled trials, and Otolaryngology– Head and Neck Surgery S100 ROSENFELD et al May 2004 cohort studies with a preponderance of harm over guage, or expressive language.2,39,41,42 Therefore, benefit. population-based screening for early detection of This recommendation concerns population- OME in asymptomatic children has not been based screening programs of all children in a com- shown to improve outcomes and is not recom- munity or a school without regard to any preex- mended. isting symptoms or history of disease. This Evidence Profile: Screening recommendation does not address hearing screen- ● Aggregate evidence quality: B, randomized, ing or monitoring of specific children with previ- controlled trials with minor limitations and ous or recurrent OME. consistent evidence from observational stud- OME is highly prevalent in young children. ies Screening surveys of healthy children ranging in ● Benefit: potentially improved developmental age from infants to age 5 years show a 15% to outcomes, which have not been demonstrated 40% point prevalence of middle-ear effu- 5,7,30-36 in the best current evidence sion. Among children examined at regular ● Harm: inaccurate diagnosis (false positive, intervals for a year, about 50% to 60% of child 32 false negative), overtreating self-limited dis- care center attendees and 25% of school-aged ease, parental anxiety, cost of screening and children37 were found to have a middle-ear effu- unnecessary treatment sion at some time during the examination period, ● Benefits-harms assessment: preponderance of with peak incidence during the winter months. harm over benefit Population-based screening has not been found ● Policy level: recommendation against to influence short-term language outcomes,33 and its long-term effects have not been evaluated in a 2. DOCUMENTATION: Clinicians should randomized clinical trial. Therefore, the recom- document the laterality, duration of effusion, mendation against screening is based not only on and presence and severity of associated symp- the ability to identify OME, but more importantly toms at each assessment of the child with OME. on a lack of demonstrable benefits from treating Recommendation based on observational studies children so identified that exceed the favorable and strong preponderance of benefit over harm. natural history of the disease. The New Zealand Documentation in the medical record facilitates Health Technology Assessment38 could not deter- diagnosis and treatment, and communicates perti- mine whether preschool screening for OME was nent information to other clinicians to ensure pa- effective. More recently, the Canadian Task Force tient safety and reduce medical errors.43 Manage- on Preventive Health Care39 reported that insuffi- ment decisions in children with OME depend on cient evidence was available to recommend in- cluding or excluding routine early screening for effusion duration and laterality plus the nature and OME. Although screening for OME is not inher- severity of associated symptoms. Therefore, these ently harmful, potential risks include inaccurate features should be documented at every medical diagnoses, overtreating self-limited disease, pa- encounter for OME. Although no studies have rental anxiety, and the costs of screening and specifically addressed documentation for OME, unnecessary treatment. there is room for improvement in documentation 44 Population-based screening is appropriate for of ambulatory care medical records. conditions that are common, can be detected by a Ideally, the time of onset and laterality of OME sensitive and specific test, and benefit from early can be defined through diagnosis of an antecedent detection and treatment.40 The first 2 requirements AOM, a history of acute onset of signs or symp- are fulfilled by OME, which affects up to 80% of toms directly referable to fluid in the middle ear, children by school entry2,5,7 and can be easily or the presence of an abnormal audiogram or tym- screened with tympanometry (see Recommenda- panogram closely following a previously normal tion 1B). Early detection and treatment of OME test. Unfortunately, these conditions are often identified by screening, however, has not been lacking, and the clinician is forced to speculate on shown to improve intelligence, receptive lan- the onset and duration of fluid in the middle ear(s) Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S101 in a child found to have OME at a routine office Table 3. Risk factors for developmental difficulties* visit or school screening audiometry. Permanent hearing loss independent of otitis media In about 40% to 50% of cases of OME, neither with effusion the affected children nor their parents or caregiv- Suspected or diagnosed speech and language delay ers describe significant complaints referable to a or disorder 45,46 Autism-spectrum disorder and other pervasive devel- middle-ear effusion. In some children, how- opmental disorders ever, OME may have associated signs and symp- Syndromes (eg, Down) or craniofacial disorders that toms caused by inflammation or the presence of include cognitive, speech, and language delays effusion (not acute infection) that should be doc- Blindness or uncorrectable visual impairment umented, such as Cleft palate, with or without associated syndrome Developmental delay ● Mild intermittent ear pain, fullness or “pop- ping” *Sensory, physical cognitive, or behavioral factors that place children who have otitis media with effusion at increased risk for develop- ● Secondary manifestations of ear pain in in- mental difficulties (delay or disorder). fants, which may include ear rubbing, exces- sive irritability, and sleep disturbances ● Failure of infants to respond appropriately to ● Benefits-harms assessment: preponderance of voices or environmental sounds, such as not benefit over harm turning accurately toward the sound source ● Policy level: recommendation ● Hearing loss, even when not specifically de- scribed by the child, suggested by seeming 3. CHILD AT RISK: Clinicians should dis- lack of attentiveness, behavioral changes, tinguish the child with OME who is at risk for failure to respond to normal conversational speech, language, or learning problems from level speech, or the need for excessively high other children with OME, and should more sound levels when using audio equipment or promptly evaluate hearing, speech, language, viewing television and need for intervention. Recommendation ● Recurrent episodes of AOM with persistent based on case series, preponderance of benefit OME between episodes over harm, and ethical limitations in studying ● Problems with school performance children with OME who are at risk. ● Balance problems, unexplained clumsiness, The panel defines the child at risk as one who is 47-50 or delayed gross motor development at increased risk for developmental difficulties ● Delayed speech or language development (delay or disorder) because of sensory, physical, The laterality (unilateral vs bilateral), duration cognitive, or behavioral factors listed in Table 3. of effusion, and the presence and severity of as- These factors are not caused by OME but can sociated symptoms should be documented in the make the child less tolerant of hearing loss or medical record at each assessment of the child vestibular problems secondary to middle-ear effu- with OME. When OME duration is uncertain, the sion. In contrast the child with OME who is not at clinician must take whatever evidence is at hand risk is otherwise healthy and does not have any of and make a reasonable estimate. the factors in Table 3. Evidence Profile: Documentation Earlier guidelines for managing OME have ap- ● Aggregate evidence quality: C, observational plied only to young children who are healthy and studies exhibit no developmental delays.1 Studies of the ● Benefits: defines severity, duration has prog- relationship between OME and hearing loss or nostic value, facilitates future communication speech/language development typically exclude with other clinicians, supports appropriate children with craniofacial anomalies, genetic syn- timing of intervention, and if consistently uni- dromes, and other developmental disorders. lateral may identify a problem with specific Therefore, the available literature mainly applies ear other than OME (eg, retraction pocket or to otherwise healthy children who meet inclusion cholesteatoma) criteria for randomized, controlled trials. Few, if ● Harm: administrative burden any, existing studies dealing with developmental Otolaryngology– Head and Neck Surgery S102 ROSENFELD et al May 2004 sequelae caused by hearing loss from OME can be areas of balance, sound localization, and commu- generalized to children who are at risk. nication. Children who are at risk for speech or language Management of the child with OME who is at delay would likely be further affected by hearing increased risk for developmental delays should problems from OME,51 even though definitive include hearing testing and speech and language studies are lacking. For example, small compara- evaluation, and may include speech and language tive studies of children or adolescents with Down therapy concurrent with managing OME, hearing syndrome52 or cerebral palsy53 show poorer artic- aids or other amplification devices for hearing loss ulation and receptive language associated with a independent of OME, tympanostomy tube inser- history of early otitis media. Large studies are tion,54,63,66,67 and hearing testing after OME re- unlikely to be forthcoming because of method- solves to document improvement, because OME ologic and ethical difficulties inherent in studying can mask a permanent underlying hearing loss and children who are delayed or at risk for further delay detection.59,68,69 delays. Therefore, clinicians who manage children Evidence Profile: Child At Risk with OME should determine whether other condi- ● Aggregate evidence quality: C, observational tions coexist that put a child at risk for develop- studies of children at risk; D, expert opinion mental delay (Table 3), and then take these con- on the ability of prompt assessment and man- ditions into consideration when planning agement to alter outcomes assessment and management. ● Benefits: optimizing conditions for hearing, Children with craniofacial anomalies (eg, cleft speech, and language; enabling children with palate, Down syndrome, Robin sequence, special needs to reach their potential; avoid- CHARGE association) have a higher prevalence ing limitations on the benefits of educational of chronic OME, hearing loss (conductive and interventions because of hearing problems sensorineural), and speech or language delay than from OME 54-57 do children without these anomalies. Other ● Harm: cost, time, and specific risks of medi- children may not be more prone to OME but are cations or surgery likely to have speech and language disorders, such ● Benefits-harms assessment: exceptional pre- as those children with permanent hearing loss in- ponderance of benefits over harm based on dependent of OME,58,59 specific language impair- subcommittee consensus because of circum- 60 61 ment, autism-spectrum disorders, or syn- stances to date precluding randomized trials dromes that adversely affect cognitive and ● Policy level: recommendation linguistic development. Some retrospective stud- ies52,62,63 have found that hearing loss caused by 4. WATCHFUL WAITING: Clinicians OME in children with cognitive delays, such as should manage the child with OME who is not Down syndrome, has been associated with lower at risk with watchful waiting for 3 months from language levels. Children with language delays or the date of effusion onset (if known) or from the disorders with OME histories perform poorer on date of diagnosis (if onset is unknown). Recom- speech perception tasks than do children with mendation based on systematic review of cohort OME histories alone.64,65 studies and preponderance of benefit over harm. Children with severe visual impairments may This recommendation is based on the self-lim- be more susceptible to the effects of OME because ited nature of most OME, which has been well they depend on hearing more than children with documented in cohort studies and in control normal vision.51 Any decrease in their most im- groups of randomized trials.2,70 portant remaining sensory input for language The likelihood of spontaneous resolution of (hearing) may significantly compromise language OME is determined by the cause and duration of development and their ability to interact and com- effusion.70 For example, about 75% to 90% of municate with others. All children with severe residual OME after an AOM episode resolves visual impairments should be considered more spontaneously by 3 months.71-73 Similar outcomes vulnerable to OME sequelae, especially in the of defined onset during a period of surveillance in Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S103 a cohort study are observed for OME.32,37 An- both. Factors to consider in determining the opti- other favorable situation involves improvement mal interval(s) for follow-up include clinical judg- (not resolution) of newly detected OME defined as ment, parental comfort level, unique characteris- change in tympanogram from type B (flat curve) tics of the child and/or his environment, access to to non-B (anything other than a flat curve). About a health care system, and hearing levels if known. 55% of children so defined improve by 3 After documented resolution of OME in all months,70 but one third will have OME relapse affected ears, further follow-up is unnecessary. within the next 3 months.4 Although a type B Evidence Profile: Watchful Waiting tympanogram is an imperfect measure of OME ● Aggregate evidence quality: B, systematic re- (81% sensitivity and 74% specificity vs myringot- view of cohort studies omy), it is the most widely reported measure suit- ● Benefit: avoid unnecessary interventions, take able for deriving pooled resolution rates.2,70 advantage of favorable natural history, avoid About 25% of newly detected OME of un- unnecessary referrals and evaluations known prior duration in children aged 2 to 4 years ● Harm: delays in therapy for OME that will resolves by 3 months when resolution is defined as not resolve with observation; prolongation of a change in tympanogram from type B to type hearing loss 2,70,74-77 A/C1 (peak pressure Ͼ200 daPa). Reso- ● Benefits-harms assessment: preponderance of lution rates may be higher for infants and young benefit over harm children in whom the preexisting duration of ef- ● Policy level: recommendation fusion is generally shorter, and particularly for those observed prospectively in studies or in the 5. MEDICATION: Antihistamines and de- course of well-child care. Documented bilateral congestants are ineffective for OME and are OME of 3 months’ duration or longer resolves not recommended for treatment. Antimicrobi- spontaneously after 6 to 12 months in about 30% als and corticosteroids do not have long-term of children aged primarily 2 years or older, with efficacy and are not recommended for routine only marginal benefits if observed longer.70 management. Recommendation based on system- Any intervention for OME (medical or surgical) atic review of randomized, controlled trials and other than observation carries some inherent harm. preponderance of harm over benefit. There is little harm associated with a specified Therapy for OME is appropriate only if persis- period of observation in the child who is not at risk tent and clinically significant benefits can be for speech, language, or learning problems. When achieved beyond spontaneous resolution. Al- observing children with OME, clinicians should though statistically significant benefits have been inform the parent or caregiver that the child may demonstrated for some medications, they are short experience reduced hearing until the effusion re- term and relatively small in magnitude. Moreover, solves, especially if bilateral. Clinicians may dis- significant adverse events may occur with all med- cuss strategies for optimizing the listening and ical therapies. learning environment until the effusion resolves. The prior OME guideline1 found no data sup- These strategies include speaking in close prox- porting antihistamine-decongestant combinations imity to the child, facing the child and speaking in treating OME. Meta-analysis of 4 randomized clearly, repeating phrases when misunderstood, trials showed no significant benefit for antihista- and providing preferential classroom seating.78,79 mines or decongestants versus placebo. No addi- The recommendation for a 3-month period of tional studies have been published since 1994 to observation is based on a clear preponderance of change this recommendation. Adverse effects of benefit over harm and is consistent with the orig- antihistamines and decongestants include insom- inal OME guideline intent of avoiding unneces- nia, hyperactivity, drowsiness, behavioral change, sary surgery.1 At the discretion of the clinician, and blood pressure variability. this 3-month period of watchful waiting may in- Long-term benefits of antimicrobial therapy for clude interval visits at which OME is monitored OME are unproved despite a modest short-term using pneumatic otoscopy, tympanometry, or benefit for 2 to 8 weeks in randomized trials.1,80,81 Otolaryngology– Head and Neck Surgery S104 ROSENFELD et al May 2004

Initial benefits, however, can become nonsignifi- Evidence Profile: Medication cant within 2 weeks of stopping the medication.82 ● Aggregate evidence quality: A, systematic re- Moreover, about 7 children would need to be view of well-designed randomized, controlled treated with antimicrobials to achieve one short- trials term response.1 Adverse effects of antimicrobials ● Benefit: avoid side effects and reduce cost by are significant and may include rashes, vomiting, not administering medications; avoid delays diarrhea, allergic reactions, alteration of the in definitive therapy caused by short-term im- child’s nasopharyngeal flora, development of bac- provement then relapse terial resistance,83 and cost. Societal consequences ● Harm: adverse effects of specific medications include direct transmission of resistant bacterial as listed previously; societal impact of anti- pathogens in homes and child care centers.84 microbial therapy on bacterial resistance and The prior OME guideline1 did not recommend transmission of resistant pathogens oral steroids for treating OME in children. A later ● Benefits-harms assessment: preponderance of meta-analysis85 showed no benefit for oral steroid harm over benefit versus placebo within 2 weeks, but did show a ● Policy level: recommendation against short-term benefit for oral steroid plus antimicro- bial versus antimicrobial alone in 1 out of 3 chil- 6. HEARING AND LANGUAGE: Hearing dren treated. This benefit became nonsignificant testing is recommended when OME persists for after several weeks in a prior meta-analysis1 and 3 months or longer, or at any time that lan- in a large randomized trial.86 Oral steroids can guage delay, learning problems, or a significant produce behavioral changes, increased appetite, hearing loss is suspected in a child with OME. and weight gain.1 Additional adverse effects may Language testing should be conducted for chil- include adrenal suppression, fatal varicella infec- dren with hearing loss. Recommendation based tion, and avascular necrosis of the femoral head.3 on cohort studies and preponderance of benefit Although intranasal steroids have fewer adverse over risk. effects, one randomized trial87 showed statistically equivalent outcomes at 12 weeks for intranasal Hearing Testing beclomethasone plus antimicrobials versus antimi- Hearing testing is recommended when OME crobials alone for OME. persists for 3 months or longer, or at any time that Antimicrobial therapy, with or without steroids, language delay, learning problems, or a significant has not been demonstrated to be effective in long- hearing loss is suspected. Conductive hearing loss term resolution of OME, but in some cases this often accompanies OME1,88 and may adversely therapy can be considered an option because of affect binaural processing,89 sound localization,90 short-term benefit in randomized trials, when the and speech perception in noise.91-94 Hearing loss parent or caregiver expresses a strong aversion to caused by OME may impair early language acqui- impending surgery. In this circumstance a single sition,95-97 but the child’s home environment has a course of therapy for 10 to 14 days may be used. greater impact on outcomes98; recent randomized The likelihood that the OME will resolve long trials41,99,100 suggest no impact on children with term with these regimens is small, and prolonged OME who are not at risk identified by screening or or repetitive courses of antimicrobials or steroids surveillance. are strongly not recommended. Studies examining hearing sensitivity in chil- Other nonsurgical therapies that are discussed dren with OME report that average pure tone in the OME literature include autoinflation of the hearing loss at 4 frequencies (500, 1000, 2000, and eustachian tube, oral or intratympanic use of mu- 4000 Hz) ranges from normal hearing to moderate colytics, and systemic use of pharmacologic hearing loss (0-55 dB). The 50th percentile is agents other than antimicrobials, steroids and an- about 25 dB hearing level (HL) and about 20% of tihistamine-decongestants. Insufficient data exist ears exceed 35 dB HL.101,102 Unilateral OME with for any of these therapies to be recommended in hearing loss results in overall poorer binaural treating OME.3 hearing than in infants with normal middle-ear Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S105 function bilaterally.103,104 Although based on lim- problems and later delays in reading and writ- ited research, there is evidence that children expe- ing.110-112 In one study, 6% to 8% of children riencing the greatest conductive hearing loss for aged 3 years and 2% to 13% of kindergartners had the longest periods may be more likely to exhibit language impairment.113 Language intervention developmental and academic sequelae.1,95,105 can improve communication and other functional Initial hearing testing for children aged 4 years outcomes for children with histories of OME.114 or older can be done in the primary care setting.106 Children who experience repeated and persis- Testing should be performed in a quiet environ- tent episodes of OME and associated hearing loss ment, preferably in a separate closed or sound- during early childhood may be at a disadvantage proofed area set aside specifically for that purpose. for learning speech and language.79,115 Although Conventional audiometry with earphones is per- Shekelle et al2 concluded there was no evidence to formed with a fail criterion of Ͼ20 dB HL at 1 or support the concern that OME during the first 3 more frequencies (500, 1000, 2000, 4000 Hz) in years of life was related to later receptive or ex- either ear.106,107 Methods not recommended as pressive language, this meta-analysis should be substitutes for primary care hearing testing include interpreted cautiously because it did not examine tympanometry and pneumatic otoscopy102; care- specific language domains, such as vocabulary, giver judgment regarding hearing loss108,109; and because the independent variable was OME speech audiometry; and tuning forks, acoustic re- and not hearing loss. Other meta-analyses79,115 flectometry, and behavioral observation.1 have suggested at most a small negative associa- Comprehensive audiologic evaluation is recom- tion of OME and hearing loss on children’s recep- mended for children who fail primary care testing, tive and expressive language through the elemen- are younger than 4 years, or cannot be tested in the tary school years. The clinical significance of primary care setting. Audiologic assessment in- these effects for language and learning is unclear cludes evaluating air-conduction and bone-con- for the child not at risk. For example, in one duction thresholds for pure tones, speech detection randomized trial,100 prompt insertion of tympa- or speech recognition thresholds,102 and measur- nostomy tubes for OME did not improve develop- ing speech understanding if possible.94 The mental outcomes at age 3 years, regardless of method of assessment depends on the develop- baseline hearing levels. In another randomized mental age of the child and might include visual trial,116 however, prompt tube insertion achieved reinforcement or conditioned orienting response small benefits for children with bilateral OME and audiometry for infants aged 6 to 24 months, play hearing loss. audiometry for children aged 24 to 48 months, or Clinicians should ask the parent or caregiver conventional screening audiometry for children about specific concerns regarding their child’s lan- 106 aged 4 years and older. The auditory brain stem guage development. Children’s speech and lan- response and otoacoustic emission are tests of guage can be tested at ages 6 to 36 months by auditory pathway structural integrity, not hearing, direct engagement of a child and interviewing the and should not substitute for behavioral pure tone parent using the Early Language Milestone 106 audiometry. Scale.117 Other approaches require interviewing only the child’s parent or caregiver, such as the Language Testing MacArthur Communicative Development Inven- Language testing should be conducted for chil- tory118 and the Language Development Survey.119 dren with hearing loss (pure tone average greater For older children the Denver Developmental than 20 dB HL on comprehensive audiometric Screening Test II120 can be used to screen general evaluation). Testing for language delays is impor- development, including speech and language. tant because communication is integral to all as- Comprehensive speech and language evaluation is pects of human functioning. Young children with recommended for children who fail testing or speech and language delays during the preschool whenever the child’s parent or caregiver expresses years are at risk for continued communication concern.121 Otolaryngology– Head and Neck Surgery S106 ROSENFELD et al May 2004

Evidence Profile: Hearing and Language ● Hearing loss greater than 30 dB HL in the ● Aggregate evidence quality: B, diagnostic better-hearing ear studies with minor limitations; C, observa- ● History of prior tympanostomy tubes tional studies ● Not having had an adenoidectomy ● Benefit: to detect hearing loss and language Children with chronic OME are at risk for struc- delay and identify strategies or interventions tural damage of the tympanic membrane126 be- to improve developmental outcomes cause the effusion contains leukotrienes, prosta- ● Harm: parental anxiety, direct and indirect glandins, and arachidonic acid metabolites that costs of assessment, false-positive results invoke a local inflammatory response.127 Reactive ● Balance of benefit and harm: preponderance changes may occur in the adjacent tympanic mem- of benefit over harm brane and mucosal linings. A relative underventi- ● Policy level: recommendation lation of the middle ear produces a negative pres- sure that predisposes to focal retraction pockets, 7. SURVEILLANCE: Children with persis- generalized atelectasis of the tympanic membrane, tent OME who are not at risk should be reex- and cholesteatoma. amined at 3- to 6-month intervals until the Structural integrity is assessed by carefully ex- effusion is no longer present, significant hear- amining the entire tympanic membrane, which, in ing loss is identified, or structural abnormali- many cases, can be accomplished by the primary ties of the eardrum or middle ear are suspected. care clinician using a handheld pneumatic oto- Recommendation based on randomized, con- scope. A search should be made for retraction trolled trials and observational studies with a pre- pockets, ossicular erosion, and areas of atelectasis ponderance of benefit over harm. If OME is asymptomatic and is likely to resolve or atrophy. If there is any uncertainty that all spontaneously, intervention is unnecessary even if observed structures are normal, the patient should OME persists for more than 3 months. The clini- be examined using an otomicroscope. All children cian should determine if risk factors exist that with these tympanic membrane conditions, regard- would predispose to undesirable sequelae or pre- less of OME duration, should have a comprehen- dict nonresolution of the effusion. As long as sive audiologic evaluation. OME persists, the child is at risk for sequelae and Conditions of the tympanic membrane that gen- must be periodically reevaluated for factors that erally mandate inserting a tympanostomy tube are would prompt intervention. posterosuperior retraction pockets, ossicular ero- The 1994 OME guideline1 recommended sur- sion, adhesive atelectasis, and retraction pockets gery for OME persisting 4 to 6 months with hear- that accumulate keratin debris. Ongoing surveil- ing loss, but requires reconsideration because of lance is mandatory because the incidence of struc- 128 later data on tubes and developmental sequelae.122 tural damage increases with effusion duration. For example, selecting surgical candidates using As noted in Recommendation 6, children with duration-based criteria (eg, OME more than 3 persistent OME for 3 months or longer should months or exceeding a cumulative threshold) does have their hearing tested. Based on these results, not improve developmental outcomes in infants clinicians can identify 3 levels of action based on and toddlers who are not at risk.41,42,99,100 Further, hearing levels obtained for the better-hearing ear the 1994 OME guideline did not specifically ad- using earphones, or in sound field using speakers dress managing effusion without significant hear- if the child is too young for ear-specific testing. ing loss persisting more than 6 months. 1. Hearing levels Ն40 dB (at least a moderate Asymptomatic OME usually resolves spontane- hearing loss). Comprehensive audiologic ously, but resolution rates decrease the longer the evaluation is indicated if not previously per- effusion has been present,36,76,77 and relapse is formed. If moderate hearing loss is docu- common.123 Risk factors that make spontaneous mented, and persists at this level, surgery is resolution less likely include124,125 recommended because persistent hearing ● Onset of OME in the summer or fall season loss of this magnitude that is permanent in Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S107

Table 4. Strategies for optimizing the listening- nal educational level, unfavorable child care envi- learning environment for children with OME and ronment, and low socioeconomic status. In such hearing loss* cases, these factors may be additive to the hearing Get within 3 feet of the child before speaking. loss in affecting lower school performance and Turn off competing audio signals, such as unneces- classroom behavior problems. sary music and television in the background. Persistent OME may be associated with physi- Face the child and speak clearly, using visual clues (hands, pictures) in addition to speech. cal or behavioral symptoms, including hyperactiv- Slow the rate, raise the level, and enunciate speech ity, poor attention, and behavioral problems in directed at the child. some studies134-136 and reduced child quality of Read to or with the child, explaining pictures and life.46 Conversely, young children randomized to asking questions. Repeat words, phrases, and questions when misun- early versus late tube insertion for persistent OME derstood. showed no behavioral benefits from early sur- Assign preferential seating in the classroom near the gery.41,100 Children with chronic OME also have teacher. significantly poorer vestibular function and gross Use a frequency modulated personal or sound field amplification system in the classroom. motor proficiency when compared with non-OME controls.48-50 Moreover, vestibular function, be- 78-79 *Modified with permission from Roberts et al. havior, and quality of life can improve after tym- panostomy tube insertion.47,137,138 Other physical nature has been shown to impact speech, symptoms of OME that, if present and persistent, 129-131 language, and academic performance. may warrant surgery include otalgia, unexplained 2. Hearing levels 21 to 39 dB (mild hearing sleep disturbance, and coexisting recurrent AOM. loss). Comprehensive audiologic evaluation Tubes reduce the absolute incidence of recurrent is indicated if not previously performed. AOM by about 1 episode per child per year, but Mild sensorineural hearing loss has been as- the relative risk reduction is 56%.139 sociated with difficulties in speech, lan- The risks of continued observation of children guage, and academic performance in 129,132 with OME must be balanced against the risks of school, and persistent mild conductive surgery. Children with persistent OME examined hearing loss from OME may have similar regularly at 3- to 6-month intervals, or sooner if impact. Further management should be indi- OME-related symptoms develop, are most likely vidualized based on effusion duration, sever- at low risk for physical, behavioral, or develop- ity of hearing loss, and parent or caregiver mental sequelae of OME. Conversely, prolonged preference, and may include strategies to op- watchful waiting of OME is not appropriate when timize the listening and learning environ- regular surveillance is impossible or when the ment (Table 4) or surgery. Repeat hearing testing should be performed in 3 to 6 months child is at risk for developmental sequelae of if OME persists at follow-up evaluation or OME because of comorbidities (Table 3). For tympanostomy tubes have not been placed. these children, the risks of anesthesia and surgery 3. Hearing levels Յ20 dB (normal hearing). (see Recommendation 9) may be less than contin- Repeat hearing test should be performed in 3 ued observation. to 6 months if OME persists at follow-up Evidence Profile: Surveillance evaluation. ● Aggregate evidence quality: C, observational In addition to hearing loss and speech or lan- studies and some randomized trials guage delay, other factors may influence the deci- ● Benefit: avoiding interventions that do not sion to intervene for persistent OME. Roberts et improve outcomes al98,133 showed that the caregiving environment is ● Harm: allowing structural abnormalities to more strongly related to school outcome than was develop in the tympanic membrane, underes- OME or hearing loss. Risk factors for delays in timating the impact of hearing loss on a child, speech and language development caused by a failing to detect significant signs or symptoms poor caregiving environment included low mater- that require intervention Otolaryngology– Head and Neck Surgery S108 ROSENFELD et al May 2004

● Balance of benefit and harm: preponderance or caregiver should be encouraged to express of benefit over harm to the surgeon any concerns they may have ● Policy level: recommendation about recommendations made. When referring a child to an otolaryngologist, 8. REFERRAL: When children with OME audiologist, or speech-language pathologist, the are referred by the primary care clinician for minimum information that should be conveyed in evaluation by an otolaryngologist, audiologist, writing includes the following: or speech-language pathologist, the referring ● Duration of OME—State how long fluid has clinician should document the effusion dura- been present. tion and specific reason for referral (evaluation, ● Laterality of OME—State whether 1 or both surgery), and provide additional relevant infor- ears have been affected. mation such as history of AOM and develop- ● Results of prior hearing testing or tympanom- mental status of the child. Option based on panel etry. consensus and a preponderance of benefit over ● Suspected speech or language problems— harm. State if there had been a delay in speech and This recommendation emphasizes the impor- language development or if the parent or a tance of communication between the referring pri- caregiver has expressed concerns about the mary care clinician and the otolaryngologist, au- child’s communication abilities, school diologist, and speech-language pathologist. achievement, or attentiveness. Parents and caregivers may be confused and frus- ● Conditions that might exacerbate the delete- trated when a recommendation for surgery is made rious effects of OME—State if the child has for their child because of conflicting information conditions such as permanent hearing loss, about alternative management strategies. Choos- impaired cognition, developmental delays, ing among management options is facilitated cleft lip or palate, or unstable or nonsupport- when primary care physicians and advanced prac- ive family or home environment. tice nurses who best know the patient’s history of ● AOM history—State if the child has a history ear problems and general medical status provide of recurrent AOM. the specialist with accurate information. Although Additional medical information that should be there are no studies showing improved outcomes provided to the otolaryngologist by the primary from better documentation of OME histories, there care clinician includes: is a clear need for better mechanisms to convey ● Parental attitude toward surgery—State if the information and expectations from primary care parents have expressed a strong preference clinicians to consultants and subspecialists.140-142 for or against surgery as a management op- When referring a child for evaluation to an tion. otolaryngologist, the primary care physician ● Related conditions that might require con- should explain the following to the parent or care- comitant surgery—State if there have been giver of the patient: other conditions that might warrant surgery if ● Reason for referral—Explain that the child is the child is going to have general anesthesia seeing an otolaryngologist for evaluation, (eg, nasal obstruction and snoring that might which is likely to include ear examination and be an indication for adenoidectomy, or ob- audiologic testing, and not necessarily simply structive breathing during sleep that might to be scheduled for surgery. mean tonsillectomy is indicated). ● What to expect—Explain that surgery may be ● General health status—State if there are any recommended and let the parent know that the conditions that might present problems for otolaryngologist will further explain the op- surgery or administering general anesthesia tions, benefits, and risks. such as congenital heart abnormality, bleed- ● Decision-making process—Explain that there ing disorder, asthma or reactive airway dis- are many alternatives for management and ease, or family history of malignant that surgical decisions are elective; the parent hyperthermia. Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S109

After evaluating the child, the otolaryngologist, and an absolute decrease of 128 effusion days per audiologist, or speech-language pathologist should child during the next year.139,143-145 Hearing lev- inform the referring physician regarding their di- els improve by a mean of 6 to 12 dB while the agnostic impression, plans for further assessment, tubes remain patent.146,147 Adenoidectomy plus and recommendations for ongoing monitoring and myringotomy (without tube insertion) has compa- management. rable efficacy in children aged 4 years or older,143 Evidence Profile: Referral but is more invasive with additional surgical and ● Aggregate evidence quality: C, observational anesthetic risks. Similarly, the added risk of ade- studies noidectomy outweighs the limited, short-term ben- ● Benefit: better communication, improved de- efit for children aged 3 years or older without prior cision making tubes.148 Consequently, adenoidectomy is not rec- ● Harm: confidentiality concerns, administra- ommended for initial OME surgery unless a dis- tive burden, increased parent or caregiver tinct indication exists, such as adenoiditis, postna- anxiety sal obstruction, or chronic sinusitis. ● Benefits-harms assessment: balance of benefit About 20% to 50% of children who have had and harm tympanostomy tubes have OME relapse after tube ● Policy level: option extrusion that may require additional sur- gery.144,145,149 When a child needs repeat surgery 9. SURGERY: When a child becomes a sur- for OME, adenoidectomy is recommended (unless gical candidate, tympanostomy tube insertion is the child has an overt or submucous cleft palate) the preferred initial procedure; adenoidectomy because it confers a 50% reduction in the need for should not be performed unless a distinct indi- future operations.143,150,151 The benefitofade- cation exists (nasal obstruction, chronic ade- noidectomy is apparent at age 2 years,150 greatest noiditis). Repeat surgery consists of adenoidec- for children aged 3 years or older, and independent tomy plus myringotomy, with or without tube of adenoid size.143,151,152 Myringotomy is per- insertion. Tonsillectomy alone or myringotomy formed concurrent with adenoidectomy. Myrin- alone should not be used to treat OME. Recom- gotomy plus adenoidectomy is effective for chil- mendation based on randomized, controlled trials dren aged 4 years or older,143 but tube insertion is with a preponderance of benefit over harm. advised for younger children, when potential re- Surgical candidacy for OME depends largely on lapse of effusion must be minimized (eg, children hearing status, associated symptoms, the child’s at risk), or when pronounced inflammation of the developmental risk (Table 3), and the anticipated tympanic membrane and middle-ear mucosa is chance of timely spontaneous resolution of the present. effusion. Candidates for surgery include children Tonsillectomy or myringotomy alone (without with OME lasting 4 months or longer with persis- adenoidectomy) is not recommended to treat tent hearing loss or other signs and symptoms, OME. Although tonsillectomy is either ineffec- recurrent or persistent OME in children at risk tive152 or of limited efficacy,148,150 the risks of regardless of hearing status, and OME and struc- hemorrhage (about 2%) and additional hospital- tural damage to the tympanic membrane or middle ization outweigh any potential benefits unless a ear. Ultimately the recommendation for surgery distinct indication for tonsillectomy exists. Myr- must be individualized, based on consensus be- ingotomy alone, without tube placement or ade- tween the primary care physician, otolaryngolo- noidectomy, is ineffective for chronic OME144,145 gist, and parent or caregiver that a particular child because the incision closes within several days. would benefit from intervention. Children with Laser-assisted myringotomy extends the ventila- OME of any duration who are at risk are candi- tion period several weeks,153 but randomized trials dates for earlier surgery. with concurrent controls have not been conducted Tympanostomy tubes are recommended for ini- to establish efficacy. In contrast, tympanostomy tial surgery because randomized trials show a tubes ventilate the middle ear for an average of 12 mean 62% relative decrease in effusion prevalence to 14 months.144,145 Otolaryngology– Head and Neck Surgery S110 ROSENFELD et al May 2004

Anesthesia mortality has been reported to be 10. COMPLEMENTARY AND ALTERNA- about 1:50,000 for ambulatory surgery,154 but the TIVE MEDICINE: No recommendation is current fatality rate may be lower.155 Laryngo- made regarding CAM as a treatment for OME. spasm and bronchospasm occur more often in No recommendation based on lack of scientific children receiving anesthesia than adults. Tympa- evidence documenting efficacy and an uncertain nostomy tube sequelae are common156 but are balance of harm and benefit. generally transient (otorrhea) or do not affect The 1994 OME guideline1 made no recommen- function (tympanosclerosis, focal atrophy, or shal- dation regarding CAM as a treatment for OME, low retraction pocket). Tympanic membrane per- and no subsequent controlled studies have been forations, which may require repair, are seen in published to change this conclusion. The current 2% of children after placement of short-term statement of “no recommendation” is based on (grommet-type) tubes and 17% after long-term lack of scientific evidence documenting efficacy tubes.156 Adenoidectomy has a 0.2% to 0.5% in- plus a balance of benefit and harm. cidence of hemorrhage150,157 and 2% incidence of Evidence concerning CAM is insufficient to transient velopharyngeal insufficiency.148 Other determine if the outcomes achieved for OME dif- potential risks of adenoidectomy, such as nasopha- fer from those achieved by watchful waiting and ryngeal stenosis and persistent velopharyngeal in- spontaneous resolution. There are no randomized, sufficiency, can be minimized with appropriate controlled trials with adequate sample size on the patient selection and surgical technique. efficacy of CAM for OME. While many case There is a clear preponderance of benefit over reports and subjective reviews on CAM treatment of AOM were found, little is published on OME harm when considering the impact of surgery for treatment or prevention. Homeopathy158 and chi- OME on effusion prevalence, hearing levels, sub- ropractic treatments159 were assessed in pilot stud- sequent incidence of AOM, and the need for re- ies with small numbers of patients that failed to operation after adenoidectomy. Information about show clinically or statistically significant benefits. adenoidectomy in children younger than 4 years, Consequently, there is no research base on which however, remains limited. Although the cost of to develop a recommendation concerning CAM surgery and anesthesia is nontrivial, it is offset by for OME. reduced OME and AOM after tube placement and The natural history of OME in childhood (dis- by reduced need for reoperation after adenoidec- cussed previously) is such that almost any inter- tomy. About 8 adenoidectomies are needed to vention can be “shown” to have helped in an avoid a single instance of tube reinsertion; how- anecdotal, uncontrolled report or case series. The ever, each avoided surgery probably represents a efficacy of CAM, or any other intervention for larger reduction in the number of AOM and OME OME, can only be shown with parallel group episodes, including those in children who did not randomized, controlled trials with valid diagnostic 150 require additional surgery. methods and adequate sample size. Unproved mo- Evidence Profile: Surgery dalities that have been claimed to provide benefit ● Aggregate evidence quality: B, randomized, in middle-ear disease include osteopathic and chi- controlled trials with minor limitations ropractic manipulation, dietary exclusions (such as ● Benefit: improved hearing, reduced preva- dairy), herbal and other dietary supplements, acu- lence of OME, reduced incidence of AOM, puncture, traditional Chinese medicine, and home- and less need for additional tube insertion opathy. None of these modalities, however, have (after adenoidectomy) yet been subjected to a published, peer-reviewed ● Harm: risks of anesthesia and specific surgi- clinical trial. cal procedures, sequelae of tympanostomy The absence of any published clinical trials also tubes means that all reports of CAM adverse effects are ● Benefits-harms assessment: preponderance of anecdotal. A systematic review of recent evi- benefit over harm dence160 found significant serious adverse effects ● Policy level: recommendation of unconventional therapies for children, most of Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S111 which were associated with inadequately regu- has been reported to range from less than 10% to lated herbal medicines. One report on malpractice more than 80%.167 Allergy has long been postu- liability associated with CAM therapies161 did not lated to cause OME through its contribution to specifically address childhood issues. Allergic re- eustachian tube dysfunction.168 The cellular re- actions to echinacea occur but seem to be rare in sponse of respiratory mucosa to allergens has been children.162 A general concern about herbal prod- well studied. Therefore, like other parts of respi- ucts is the lack of any governmental oversight into ratory mucosa, the mucosa lining the middle-ear product quality or purity.160,163,164 Further, herbal cleft is capable of an allergic response.169,170 Sen- products may alter blood levels of allopathic med- sitivity to allergens varies among individuals, and ications, including anticoagulants. A possible con- atopy may involve neutrophils in type I allergic cern with homeopathy is the worsening of symp- reactions that enhance the inflammatory re- toms, which is viewed as a positive, early sign of sponse.171 homeopathic efficacy. The adverse effects of ma- The correlation between OME and allergy has nipulative therapies (such as chiropractic treat- been widely reported, but no prospective studies ments and osteopathy) in children are difficult to have examined the effects of immunotherapy assess because of scant evidence, but a case series compared with observation alone or other man- of 332 children treated for AOM or OME with agement options. Reports of OME cure after im- chiropractic manipulation did not mention any munotherapy or food elimination diets172 are im- 165 side effects. Quadriplegia has been reported, possible to interpret without concurrent control however, following spinal manipulation in an in- groups because of the favorable natural history of 166 fant with torticollis. most untreated OME. The documentation of al- Evidence Profile: Complementary and Alter- lergy in published reports has been defined incon- native Medicine sistently (medical history, physical examination, ● Aggregate evidence quality: D, case series skin-prick testing, nasal smears, serum IgE and without controls eosinophil counts, inflammatory mediators in ef- ● Benefit: not established fusions). Study groups have been drawn primarily ● Harm: potentially significant, depending on from specialist offices, likely lack heterogeneity, the intervention and are not representative of general medical prac- ● Benefits-harms assessment: uncertain balance tice. of benefit and harm Evidence Profile: Allergy Management ● Policy level: no recommendation ● Aggregate evidence quality: D, case series 11. ALLERGY MANAGEMENT: No recom- without controls ● mendation is made regarding allergy manage- Benefit: not established ● ment as a treatment for OME. No recommen- Harm: adverse effects and cost of medication, dation based on insufficient evidence of physician evaluation, elimination diets, and therapeutic efficacy or a causal relationship be- desensitization tween allergy and OME. ● Benefits-harms assessment: balance of benefit The 1994 OME guideline1 made no recommen- and harm dation regarding allergy management as a treat- ● Policy level: no recommendation ment for OME and no subsequent controlled stud- ies have been published to change this conclusion. RESEARCH NEEDS The current statement of “no recommendation” is Diagnosis based on insufficient evidence of therapeutic effi- ● Further standardize the definition of OME. cacy or a causal relationship between allergy and ● Assess the performance characteristics of OME, plus a balance of benefit and harm. pneumatic otoscopy as a diagnostic test for A linkage between allergy and OME has long OME when performed by primary care phy- been speculated but to date remains unquantified. sicians and advanced practice nurses in the The prevalence of allergy among OME patients routine office setting. Otolaryngology– Head and Neck Surgery S112 ROSENFELD et al May 2004

● Determine the optimal methods for teaching symptomatic children with OME referred for pneumatic otoscopy to residents and clini- evaluation. cians. ● Develop a brief, reliable, objective method Medication for diagnosing OME. ● Clarify which children, if any, should receive ● Develop a classification method for identify- antimicrobials, steroids, or both for OME. ing the presence of OME for practical use by ● Conduct a randomized, placebo-controlled clinicians that is based on quantifiable tym- trial on the efficacy of antimicrobial therapy, panometric characteristics. with or without concurrent oral steroid, in ● Assess the usefulness of algorithms combin- avoiding surgery in children with OME who ing pneumatic otoscopy and tympanometry are surgical candidates and have not received for detecting OME in clinical practice. recent antimicrobials. ● Conduct additional validating cohort studies ● Investigate the role of mucosal surface bio- of acoustic reflectometry as a diagnostic films in refractory or recurrent OME and de- method for OME, particularly in children velop targeted interventions. younger than 2 years. Hearing and Language Child At Risk ● Conduct longitudinal studies on the natural history of hearing loss accompanying OME. ● Better define the child with OME who is at ● risk for speech, language, and learning prob- Develop improved methods for describing and quantifying the fluctuations in hearing of lems. children with OME over time. ● Conduct large, multicenter observational co- ● Conduct prospective controlled studies on the hort studies to identify the child at risk who is relation of hearing loss associated with OME most susceptible to potential adverse sequelae to later auditory, speech, language, behav- of OME. ioral, and academic sequelae. ● Conduct large, multicenter observational co- ● Develop reliable, brief, objective methods for hort studies to analyze outcomes achieved estimating hearing loss associated with OME. with alternative management strategies for ● Develop reliable, brief, objective methods for OME in children at risk. estimating speech or language delay associ- ated with OME. Watchful Waiting ● Evaluate the benefits and administrative bur- ● Define the spontaneous resolution of OME in den of language testing by primary care cli- infants and young children (existing data are nicians. limited primarily to children aged 2 years or ● Agree on the aspects of language that are older). vulnerable to, or affected by, hearing loss ● Conduct large-scale, prospective cohort stud- caused by OME, and reach a consensus on the ies to obtain current data on the spontaneous best tools for measurement. resolution of newly diagnosed OME of un- ● Determine if OME and associated hearing known prior duration (existing data are pri- loss place children from special populations marily from the late 1970s and early 1980s). at greater risk for speech and language delays. ● Develop prognostic indicators to identify the best candidates for watchful waiting. Surveillance ● Determine if the lack of impact from prompt ● Develop better tools for monitoring children insertion of tympanostomy tubes on speech with OME, suitable for routine clinical care. and language outcomes seen in asymptomatic ● Assess the value of new strategies for moni- young children with OME identified by toring OME, such as acoustic reflectometry screening or intense surveillance can be gen- performed at home by the parent or caregiver, eralized to older children with OME or to in optimizing surveillance. Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S113

● Improve our ability to identify children who CONCLUSION would benefit from early surgery instead of This evidence-based practice guideline offers prolonged surveillance. recommendations for identifying, monitoring, and ● Promote early detection of structural abnor- managing the child with OME. The guideline em- malities in the tympanic membrane associated phasizes appropriate diagnosis and provides op- with OME that may require surgery to pre- tions for various management strategies including vent complications. observation, medical intervention, and referral for ● Clarify and quantify the role of parent or surgical intervention. These recommendations caregiver education, socioeconomic status, should provide primary care physicians and other and quality of the caregiving environment as health care providers with assistance in managing modifiers of OME developmental outcomes. children with OME. ● Develop methods for minimizing loss to fol- low-up during OME surveillance. SUBCOMMITTEE ON OTITIS MEDIA WITH EFFUSION Surgery Richard M. Rosenfeld, MD, MPH, Cochairperson, AAP, AAO-HNS ● Define the role of adenoidectomy in children Larry Culpepper, MD, MPH, Cochairperson, aged 3 years or younger as a specific OME AAFP therapy. Karen J. Doyle, MD, PhD, AAO-HNS ● Conduct controlled trials on the efficacy of Kenneth M. Grundfast, MD, AAO-HNS tympanostomy tubes for developmental Alejandro Hoberman, MD, AAP outcomes in children with hearing loss, Margaret A. Kenna, MD, AAO-HNS other symptoms, or speech and language Allan S. Lieberthal, MD, AAP delay. Martin Mahoney, MD, PhD, AAFP ● Conduct randomized, controlled trials of sur- Richard A. Wahl, MD, AAP gery versus no surgery that emphasize pa- Charles R. Woods, Jr, MD, MS, AAP tient-based outcome measures (quality of life, Barbara Yawn MD, MSc, AAFP functional health status) in addition to objec- tive measures (effusion prevalence, hearing CONSULTANTS levels, AOM incidence, reoperation). S. Michael Marcy, MD ● Identify the optimal ways to incorporate par- Richard N. Shiffman, MD ent or caregiver preference into surgical de- cision making. LIAISONS Linda Carlson, MS, CPNP, National Association Complementary and Alternative of Pediatric Nurse Practitioners Medicine Judith Gravel, PhD, American Academy of Audi- ology ● Conduct randomized, controlled trials on the Joanne Roberts, PhD, American Speech-Lan- efficacy of CAM modalities for OME. guage-Hearing Association ● Develop strategies to identify parents or care- givers who use CAM therapies for their STAFF child’s OME, and encourage surveillance by Maureen Hannley, PhD, AAO-HNS the primary care clinician. Carla T. Herrerias, MPH, AAP Bellinda K. Schoof, MHA, CPHQ, AAFP Allergy Management ● Evaluate the causal role of atopy in OME. CONFLICTS OF INTEREST ● Conduct randomized, controlled trials on the S. Michael Marcy, MD: consultant to Abbott Lab- efficacy of allergy therapy for OME that are oratories; consultant to GlaxoSmithKline generalizable to the primary care setting. (vaccines). Otolaryngology– Head and Neck Surgery S114 ROSENFELD et al May 2004

REFERENCES 17. Eddy DM. A manual for assessing health practices and 1. Stool SE, Berg AO, Berman S, et al. Otitis media with designing practice policies: the explicit approach. Phil- effusion in young children. Clinical Practice Guideline, adelphia, PA: American College of Physicians; 1992. Number 12. Rockville, MD: Agency for Health Care 18. Dowell SF, Marcy MS, Phillips WR, et al. Otitis media: Policy and Research, Public Health Service, US Depart- principles of judicious use of antimicrobial agents. Pe- ment of Health and Human Services; 1994; AHCPR diatrics 1998;101:165-71. Publication No. 94-0622. 19. Dowell SF, Butler JC, Giebink GS, et al. Acute otitis 2. Shekelle P, Takata G, Chan LS, et al. Diagnosis, natural media: management and surveillance in an era of pneu- history and late effects of otitis media with effusion. mococcal resistance: a report from the Drug-resistant Streptococcus Pneumoniae Therapeutic Working Evidence report/technology assessment no. 55. Rock- Group. Pediatr Infect Dis J 1999;18:1-9. ville, MD: Agency for Healthcare Research and Qual- 20. Karma PH, Penttila MA, Sipila MM, et al. Otoscopic ity; 2003. AHRQ Publication No. 03-E023. diagnosis of middle ear effusion in acute and non-acute 3. Williamson I. Otitis media with effusion. Clin Evid otitis media. I. the value of different otoscopic findings. 2002;7:469-76. Int J Pediatr Otorhinolaryngol 1989;17:37-49. 4. Tos M. Epidemiology and natural history of secretory 21. Pichichero ME, Poole MD. Assessing diagnostic accu- otitis. Am J Otol 1984;5:459-62. racy and tympanocentesis skills in the management of 5. Paradise JL, Rockette HE, Colborn DK, et al. Otitis otitis media. Arch Pediatr Adolesc Med 2001;155:1137- media in 2253 Pittsburgh area infants: prevalence and 42. risk factors during the first two years of life. Pediatrics 22. Steinbach WJ, Sectish TC. Pediatric resident training in 1997;99:318-33. the diagnosis and treatment of acute otitis media. Pedi- 6. Casselbrant ML, Mandel EM. Epidemiology. In: Rosen- atrics 2002;109:404-08. feld RM, Bluestone CD, editors. Evidence-based otitis 23. Palmu A, Puhakka H, Rahko T, et al. Diagnostic value media, 2nd ed. Hamilton, Ontario: BC Decker Inc; of tympanometry in infants in clinical practice. Int J Pe- 2003. pp. 147-62. diatr Otorhinolaryngol 1999;49:207-13. 7. Williamson IG, Dunleavy J, Baine J, et al. The natural 24. van Balen FA, Aarts AM, De Melker RA. Tympanom- history of otitis media with effusionz: a three-year study etry by general practitioners: reliable? Int J Pediatr of the incidence and prevalence of abnormal tympano- Otorhinolaryngol 1999;48:117-23. grams in four South West Hampshire infant and first 25. Block SL, Mandel E, McLinn S, et al. Spectral gradient schools. J Laryngol Otol 1994;108:930-4. acoustic reflectometry for the detection of middle ear 8. Coyte PC, Croxford R, Asche CV, et al. Physician and effusion by pediatricians and parents. Pediatr Infect Dis population determinants of rates of middle-ear surgery J 1998;17:560-64, 580. in Ontario. JAMA 2001;286:2128-35. 26. Barnett ED, Klein JO, Hawkins KA, et al. Comparison 9. Tugwell P. How to read clinical journals: III. To learn of spectral gradient acoustic reflectometry and other the clinical course and prognosis of disease. Can Med diagnostic techniques for detection of middle ear effu- Assoc J 1981;124:869-72. sion in children with middle ear disease. Pediatr Infect 10. Jaeschke R, Guyatt G, Sackett DL. Users’ guides to the Dis J 1998;17:556-9, 580. medical literature: III. How to use an article about a 27. Block SL, Pichichero ME, McLinn S, et al. Spectral diagnostic test. A. Are the results of the study valid? gradient acoustic reflectometry: detection of middle ear Evidence-Based Medicine Working Group. JAMA effusion by pediatricians in suppurative acute otitis me- 1994;271:389-91. dia. Pediatr Infect Dis J 1999;18:741-4. 11. Shiffman RN, Shekelle P, Overhage JM, et al. Standard- 28. Schwartz RH. A practical approach to the otitis prone ized reporting of clinical practice guidelines: a proposal child. Contemp Pediatr 1987;4:30-54. 29. Barriga F, Schwartz RH, Hayden GF. Adequate illumi- from the Conference on Guideline Standardization. Ann nation for otoscopy: variations due to power source, Intern Med 2003;139:493-8. bulb, and head and speculum design. Am J Dis Child 12. Shiffman RN, Karras BT, Agrawal A, et al. GEM: a 1986;140:1237-40. proposal for a more comprehensive guideline document 30. Sorenson CH, Jensen SH, Tos M. The post-winter prev- model using XML. J Am Med Informatics Assoc 2000; alence of middle-ear effusion in four-year-old children, 7:488-98. judged by tympanometry. Int J Pediatr Otorhinolaryngol 13. Shaneyfelt TM, Mayo-Smith MF, Rothwangl J. Are 1981;3:119-28. guidelines following guidelines? the methodological 31. Fiellau-Nikolajsen M. Epidemiology of secretory otitis quality of clinical practice guidelines in the peer-re- media: a descriptive cohort study. Ann Otol Rhinol viewed medical literature. JAMA 1999;281:1900-5. Laryngol 1983;92:172-7. 14. Agrawal A, Shiffman RN. Evaluation of guideline qual- 32. Casselbrant ML, Brostoff LM, Cantekin EI, et al. Otitis ity using GEM-Q. Medinfo 2001;10:1097-101. media with effusion in preschool children. Laryngo- 15. Yale Center for Medical Informatics. GEM: The Guide- scope 1985;95:428-36. line Elements Model. Available at: http://yemi.med. 33. Zielhuis GA, Rach GH, van den Broek P. Screening for yale.edu/GEM/. Accessed December 8, 2003. otitis media with effusion in preschool children. Lancet 16. American Academy of Pediatrics, Steering Committee 1989;1:311-14. on Quality Improvement and Management. A taxonomy 34. Poulsen G, Tos M. Repetitive tympanometric screen- of recommendations for clinical practice guidelines. Pe- ings of two-year-old children. Scand Audiol 1980;9: diatrics. (In press.) 21-8. Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S115

35. Tos M, Holm-Jensen S, Sorensen CH. Changes in prev- lescents with Down syndrome. Ment Retard 1986;24: alence of secretory otitis from summer to winter in 353-6. four-year-old children. Am J Otol 1981;2:324-7. 53. van der Vyver M, van der Merwe A, Tesner HE. The 36. Thomsen J, Tos M. Spontaneous improvement of secre- effects of otitis media on articulation in children with tory otitis: a long-term study. Acta Otolaryngol 1981; cerebral palsy. Int J Rehab Res 1988;11:386-9. 92:493-9. 54. Paradise JL, Bluestone CD. Early treatment of the uni- 37. Lous J, Fiellau-Nikolajsen M. Epidemiology of middle versal otitis media of infants with cleft palate. Pediatrics ear effusion and tubal dysfunction: a one-year prospec- 1974;53:48-54. tive study comprising monthly tympanometry in 387 55. Schwartz DM, Schwartz RH. Acoustic impedance and non-selected seven-year-old children. Int J Pediatr Oto- otoscopic findings in young children with Down’s syn- rhinolaryngol 1981;3:303-17. drome. Arch Otolaryngol 1978;104:652-6. 38. New Zealand Health Technology Assessment. Screen- 56. Corey JP, Caldarelli DD, Gould HJ. Otopathology in ing programmes for the detection of otitis media with cranial facial dysostosis. Am J Otol 1987;8:14-7. effusion and conductive hearing loss in pre-school and 57. Schonweiler R, Schonweiler B, Schmelzeisen R. Hear- new entrant school children: a critical appraisal of the ing capacity and speech production in 417 children with literature. New Zealand, Christchurch: New Zealand facial cleft abnormalities [in German]. HNO 1994;42: Health Technology Assessment; 1998, pp. 61. 691-6. 39. Canadian Task Force on Preventive Health Care. 58. Ruben RJ, Math R. Serous otitis media associated with Screening for otitis media with effusion: recommenda- sensorineural hearing loss in children. Laryngoscope tion statement from the Canadian Task Force on Pre- 1978;88:1139-54. ventive Health Care. CMAJ 2001;165:1092-1093. 59. Brookhouser PE, Worthington DW, Kelly WJ. Middle 40. US Preventive Services Task Force. Guide to clinical ear disease in young children with sensorineural hearing preventive services, 2nd ed. Baltimore, MD: Williams loss. Laryngoscope 1993;103:371-8. & Wilkins; 1995. 60. Rice ML. Specific language impairments: in search of 41. Paradise JL, Feldman HM, Campbell TF, et al. Effect of diagnostic markers and genetic contributions. Mental early or delayed insertion of tympanostomy tubes for Retard Dev Disabil Res Rev 1997;3:350-7. persistent otitis media on developmental outcomes at 61. Rosenhall U, Nordin V, Sandstrom M, et al. Autism and the age of three years. N Engl J Med 2001;344:1179-87. hearing loss. J Autism Dev Disord 1999;29:349-57. 62. Cunningham C, McArthur K. Hearing loss and treat- 42. Rovers MM, Krabble PF, Straatman H, et al. Random- ment in young Down’s syndrome children. Childcare ized controlled trial of the effect of ventilation tubes Health Dev 1981;7:357-74. (grommets) on quality of life at age 1-2 years. Arch Dis 63. Shott SR, Joseph A, Heithaus D. Hearing loss in chil- Child 2001;84:45-9. dren with Down syndrome. Int J Pediatr Otorhinolaryn- 43. Wood DL. Documentation guidelines: evolution, future gol 2001;61:199-205. direction, and compliance. Am J Med 2001;110:332-4. 64. Clarkson RL, Eimas PD, Marean GC. Speech percep- 44. Soto CM, Kleinman KP, Simon SR. Quality and corre- tion in children with histories of recurrent otitis media. lates of medical record documentation in the ambulatory J Acoust Soc Am 1989;85:926-33. care setting. BMC Health Serv Res 2002;2:22-35. 65. Groenen P, Crul T, Maassen B, et al. Perception of 45. Marchant CD, Shurin PA, Turczyk VA, et al. Course voicing cues by children with early otitis media with and outcome of otitis media in early infancy: a prospec- and without language impairment. J Speech Hear Res tive study. J Pediatr 1984;104:826-31. 1996;39:43-54. 46. Rosenfeld RM, Goldsmith AJ, Tetlus L, et al. Quality of 66. Hubbard TW, Paradise JL, McWilliams BJ, et al. Con- life for children with otitis media. Arch Otolaryngol sequences of unremitting middle-ear disease in early Head Neck Surg 1997;123:1049-54. life: otologic, audiologic, and developmental findings in 47. Casselbrant ML, Furman JM, Rubenstein E, et al. Effect children with cleft palate. N Engl J Med 1985;312: of otitis media on the vestibular system in children. Ann 1529-34. Otol Rhinol Laryngol 1995;104:620-4. 67. Nunn DR, Derkay CS, Darrow DH, et al. The effect of 48. Orlin MN, Effgen SK, Handler SD. Effect of otitis very early cleft palate closure on the need for ventilation media with effusion on gross motor ability in preschool- tubes in the first years of life . Laryngoscope 1995;105: aged children: preliminary findings. Pediatrics 1997;99: 905-8. 334-37. 68. Pappas DG, Flexer C, Shackelford L. Otological and 49. Golz A, Angel-Yeger B, Parush S. Evaluation of bal- habilitative management of children with Down syn- ance disturbances in children with middle ear effusion. drome. Laryngoscope 1994;104:1065-70. Int J Pediatr Otorhinolaryngol 1998;43:21-6. 69. Vartiainen E. Otitis media with effusion in children with 50. Casselbrant ML, Redfern MS, Furman JM, et al. Visual- congenital or early-onset hearing impairment. J Otolar- induced postural sway in children with and without yngol 2000;29:221-3. otitis media. Ann Otol Rhinol Laryngol 1998;107: 70. Rosenfeld RM, Kay D. Natural history of untreated 401-5. otitis media. Laryngoscope 2003;113:1645-57. 51. Ruben R. Host susceptibility to otitis media sequelae. 71. Teele DW, Klein JO, Rosner BA. Epidemiology of In: Rosenfeld RM, Bluestone CD, editors. Evidence- otitis media in children. Ann Otol Rhinol Laryngol based otitis media, 2nd ed. Hamilton, Ontario: BC Suppl 1980;89:5-6. Decker Inc; 2003. 505-14. 72. Mygind N, Meistrup-Larsen KI, Thomsen J, et al. Pen- 52. Whiteman BC, Simpson GB, Compton WC. Relation- icillin in acute otitis media: a double-blind, placebo- ship of otitis media and language impairment on ado- controlled trial. Clin Otolaryngol 1981;6:5-13. Otolaryngology– Head and Neck Surgery S116 ROSENFELD et al May 2004

73. Burke P, Bain J, Robinson D, et al. Acute red ear in 90. Besing J, Koehnke J. A test of virtual auditory local- children: controlled trial of nonantibiotic treatment in ization. Ear Hear 1995;16:220-9. general practice. BMJ 1991;303:558-62. 91. Jerger S, Jerger J, Alford BR, et al. Development of 74. Fiellau-Nikolajsen M, Lous J. Prospective tympanom- speech intelligibility in children with recurrent otitis etry in 3-year-old children: a study of the spontaneous media. Ear Hear 1983;4:138-45. course of tympanometry types in a nonselected popula- 92. Gravel JS, Wallace IF. Listening and language at 4 tion. Arch Otolaryngol 1979;105:461-6. years of age: effects of early otitis media. J Speech Hear 75. Fiellau-Nikolajsen M. Tympanometry in 3-year-old Res 1992;35:588-95. children: type of care as an epidemiological factor in 93. Schilder AG, Snik AF, Straatman H, et al. The effect of secretory otitis media and tubal dysfunction in uns- otitis media with effusion at preschool age on some elected populations of 3-year-old children. ORL J Oto- aspects of auditory perception at school age. Ear Hear rhinolaryngol Relat Spec 1979;41:193-205. 1994;15:224-31. 76. Tos M. Spontaneous improvement of secretory otitis 94. Rosenfeld RM, Madell JR, McMahon A. Auditory func- and impedance screening. Arch Otolaryngol 1980;106: tion in normal-hearing children with middle ear effu- 345-9. sion. In: Lim DJ, Bluestone CD, Casselbrant M, et al, 77. Tos M, Holm-Jensen S, Sorensen CH, Mogensen C. editors. Recent advances in otitis media: proceedings of Spontaneous course and frequency of secretory otitis in the 6th international symposium. Hamilton, Ontario: 4-year-old children. Arch Otolaryngol 1982;108:4-10. BC Decker Inc; 1996. pp. 354-356. 78. Roberts JE, Zeisel SA. Ear infections and language 95. Friel-Patti S, Finitzo T. Language learning in a prospec- development. Rockville, MD: American Speech-Lan- tive study of otitis media with effusion in the first two guage-Hearing Association and the National Center for years of life. J Speech Hear Res 1990;33:188-94. Early Development and Learning; 2000. 96. Wallace IF, Gravel JS, McCarton CM, et al. Otitis 79. Roberts JE, Rosenfeld RM, Zeisel SA. Otitis Media and media, auditory sensitivity, and language outcomes at Speech and Language. Pediatrics 2004;113:e238-48. one year. Laryngoscope 1988;98:64-70. 80. Williams RL, Chalmers TC, Stange KC, et al. Use of 97. Roberts JE, Burchinal MR, Medley LP, et al. Otitis antibiotics in preventing recurrent otitis media and in media, hearing sensitivity, and maternal responsiveness treating otitis media with effusion: a meta-analytic at- in relation to language during infancy. J Pediatr 1995; tempt to resolve the brouhaha. JAMA 1993;270:1344- 126:481-9. 98. Roberts JE, Burchinal MR, Zeisel SA. Otitis media in 51. early childhood in relation to children’s school-age lan- 81. Rosenfeld RM, Post JC. Meta-analysis of antibiotics for guage and academic skills. Pediatrics 2002;110:696- the treatment of otitis media with effusion. Otolaryngol 706. Head Neck Surg 1992;106:378-86. 99. Rovers MM, Straatman H, Ingels K, et al. The effect of 82. Mandel EM, Rockette HE, Bluestone CD, et al. Efficacy ventilation tubes on language development in infants of amoxicillin with and without decongestant-antihista- with otitis media with effusion: a randomized trial. mine for otitis media with effusion in children: results of Pediatrics 2000;106:E42. a double-blind, randomized trial. N Engl J Med 1987; 100. Paradise JL, Feldman HM, Campbell TF, et al. Early 316:432-7. versus delayed insertion of tympanostomy tubes for 83. McCormick AW, Whitney CG, Farley MM, et al. Geo- persistent otitis media: developmental outcomes at the graphic diversity and temporal trends of antimicrobial age of three years in relation to prerandomization illness resistance in Streptococcus pneumoniae in the United patterns and hearing levels. Pediatr Infect Dis J 2003; States. Nat Med 2003;9:424-30. 22:309-14. 84. Levy SB. The antibiotic paradox: how the misuse of 101. Kokko E. Chronic secretory otitis media in children: a antibiotic destroys their curative powers. Cambridge, clinical study. Acta Otolaryngol Suppl 1974;327:1-44. MA: Perseus Publishing; 2002. 102. Fria TJ, Cantekin EI, Eichler JA. Hearing acuity of 85. Butler CC, van der Voort JH. Oral or topical nasal children with otitis media with effusion. Arch Otolar- steroids for hearing loss associated with otitis media yngol 1985;111:10-6. with effusion in children. Cochrane Database Syst Rev 103. Gravel JS, Wallace IF. Effects of otitis media with 2002;4:CD001935. effusion on hearing in the first three years of life. J 86. Mandel EM, Casselbrant ML, Rockette HE, et al. Sys- Speech Lang Hear Res 2000;43:631-44. temic steroid for chronic otitis media with effusion in 104. Roberts JE, Burchinal MR, Zeisel S, et al. Otitis media, children. Pediatrics 2002;110:1071-80. the caregiving environment, and language and cognitive 87. Tracy JM, Demain JG, Hoffman KM, et al. Intranasal outcomes at 2 years. Pediatrics 1998;102:346-54. beclomethasone as an adjunct to treatment of chronic 105. Gravel JS, Wallace IF, Ruben RJ. Early otitis media and middle ear effusion. Ann Allergy Asthma Immunol later educational risk. Acta Otolaryngol 1995;115:279- 1998;80:198-206. 81. 88. Joint Committee on Infant Hearing. Year 2000 position 106. Cunningham M, Cox EO, American Academy of Pedi- statement: principles and guidelines for early hearing atrics, Committee on Practice and Ambulatory Medi- detection and intervention programs. Am J Audiol cine, Section on Otolaryngology and 2000;9:9-29. Bronchoesophagology. Clinical report: hearing assess- 89. Pillsbury HC, Grose JH, Hall JW III. Otitis media with ment in infants and children: recommendations beyond effusion in children. Binaural hearing before and after neonatal screening. Pediatrics 2003;111:436-40. corrective surgery. Arch Otolaryngol Head Neck Surg 107. American Speech-Language-Hearing Association Panel 1991;117:718-23. on Audiologic Assessment. Guidelines for Audiologic Otolaryngology– Head and Neck Surgery Volume 130 Number 5 ROSENFELD et al S117

Screening. Rockville, MD: American Speech-Language- follow-up study in children aged 6 months to 6 years. J Hearing Association; 1996. Fam Pract 2000;49:605-11. 108. Rosenfeld RM, Goldsmith AJ, Madell JR. How accurate 126. Sano S, Kamide Y, Schachern PA, et al. Micropatho- is parent rating of hearing for children with otitis media? logic changes of pars tensa in children with otitis media Arch Otolaryngol Head Neck Surg 1998;124:989-92. with effusion. Arch Otolaryngol Head Neck Surg 1994; 109. Brody R, Rosenfeld RM, Goldsmith AJ, Madell JR. 120:815-9. Parents cannot detect mild hearing loss in children. 127. Yellon RF, Doyle WJ, Whiteside TL, et al. Cytokines, Otolaryngol Head Neck Surg 1999;121:681-6. immunoglobulins, and bacterial pathogens in middle ear 110. Catts HW, Fey ME, Zhang X, et al. Language basis of effusions. Arch Otolaryngol Head Neck Surg 1995;121: reading and reading disabilities: evidence from a longi- 865-9. tudinal investigation. Sci Studies Reading 1999;3:331- 128. Maw RA, Bawden R. Tympanic membrane atrophy, 62. scarring, atelectasis and attic retraction in persistent, 111. Johnson CJ, Beitchman JH, Young A, et al. Fourteen- untreated otitis media with effusion and following ven- year follow-up of children with and without speech/ tilation tube insertion. Int J Pediatr Otorhinolaryngol language impairments: speech/language stability and 1994;30:189-204. outcomes. J Speech Lang Hear Res 1999;42:744-60. 129. Davis JM, Elfenbein J, Schum R, et al. Effects of mild 112. Scarborough H, Dobrich W. Development of children and moderate hearing impairment on language, educa- with early language delay. J Speech Hear Res 1990;33: tional, and psychosocial behavior of children. J Speech 70-83. Hear Disord 1986;51:53-62. 113. Tomblin JB, Records NL, Buckwalter P, et al. Preva- 130. Carney AE, Moeller MP. Treatment efficacy: hearing lence of specific language impairment in kindergarten loss in children. J Speech Lang Hear Res 1998;41:S61- children. J Speech Lang Hear Res 1997;40:1245-60. S84. 114. Glade MJ. Diagnostic and therapeutic technology as- 131. Karchmer MA, Allen TE. The functional assessment of sessment: speech therapy in patients with a prior history deaf and hard of hearing students. Am Ann Deaf 1999; of recurrent acute or chronic otitis media with effusion. 144:68-77. Chicago, IL: American Medical Association; 1996, pp. 132. Bess FH, Dodd-Murphy J, Parker RA. Children with 1-14. minimal sensorineural hearing loss: prevalence, educa- 115. Casby MW. Otitis media and language development: a tional performance, and functional status. Ear Hear 1998;19:339-54. meta-analysis. Am J Speech Lang Pathol 2001;10:65- 133. Roberts JE, Burchinal MR, Jackson SC, et al. Otitis 80. media in early childhood in relation to preschool lan- 116. Maw R, Wilks J, Harvey I, et al. Early surgery com- guage and school readiness skills among black children. pared with watchful waiting for glue ear and effect on Pediatrics 2000;106:725-35. language development in preschool children: a random- 134. Haggard MP, Birkin JA, Browning GG, et al. Behavior ised trial. Lancet 1999;353:960-3. problems in otitis media. Pediatr Infect Dis J 1994;13: 117. Coplan J. Early language milestone scale, 2nd ed. Aus- S43-S50. tin, TX: PRO-ED; 1983. 135. Bennett KE, Haggard MP. Behaviour and cognitive 118. Fenson L, Dale PS, Reznick JS, et al. MacArthur com- outcomes from middle ear disease. Arch Dis Child municative development inventories: user’s guide and 1999;80:28-35. technical manual. San Diego, CA: Singular Publishing 136. Bennett KE, Haggard MP, Silva PA, et al. Behaviour Group; 1993. and developmental effects of otitis media with effusion 119. Rescoria L. The language development survey: a into the teens. Arch Dis Child 2001;85:91-5. screening tool for delayed language in toddlers. J 137. Wilks J, Maw R, Peters TJ, et al. Randomised controlled Speech Hear Dis 1989;54:587-99. trial of early surgery versus watchful waiting for glue 120. Frankenburg WK, Dodds JA, Faucal A, et al. Denver ear: the effect on behavioural problems in pre-school developmental screening test II. Denver, CO: University children. Clin Otolaryngol 2000;25:209-14. of Colorado Press; 1990. 138. Rosenfeld RM, Bhaya MH, Bower CM, et al. Impact of 121. Klee T, Pearce K, Carson DK. Improving the positive tympanostomy tubes on child quality of life. Arch Oto- predictive value of screening for developmental lan- laryngol Head Neck Surg 2000;126:585-92. guage disorder. J Speech Lang Hear Res 2000;43:821- 139. Rosenfeld RM, Bluestone CD. Clinical efficacy of sur- 33. gical therapy. In: Rosenfeld RM, Bluestone CD, editors. 122. Shekelle PG, Ortiz E, Rhodes S, et al. Validity of the Evidence-based otitis media, 2nd ed. Hamilton, Ontario: Agency for Healthcare Research and Quality clinical BC Decker Inc; 2003. pp. 227-40. practice guidelines: how quickly do guidelines become 140. Kuyvenhoven MM, De Melker RA. Referrals to spe- outdated? JAMA 2001;286:1461-7. cialists: an exploratory investigation of referrals by 13 123. Zielhuis GA, Straatman H, Rach GH, et al. Analysis and general practitioners to medical and surgical depart- presentation of data on the natural course of otitis media ments. Scand J Prim Health Care 1990;8:53-7. with effusion in children. Int J Epidemiol 1990;19: 141. Haldis TA, Blankenship JC. Telephone reporting in the 1037-44. consultant-generalist relationship. J Eval Clin Pract 124. MRC Multi-centre Otitis Media Study Group. Risk fac- 2002;8:31-5. tors for persistence of bilateral otitis media with effu- 142. Reichman S. The generalist’s patient and the subspe- sion. Clin Otolaryngol 2001;26:147-56. cialist. Am J Manage Care 2002;8:79-82. 125. van Balen FA, De Melker RA. Persistent otitis media 143. Gates GA, Avery CA, Prihoda TJ, et al. Effectiveness of with effusion: can it be predicted? a family practice adenoidectomy and tympanostomy tubes in the treat- Otolaryngology– Head and Neck Surgery S118 ROSENFELD et al May 2004

ment of chronic otitis media with effusion. N Engl 158. Harrison H, Fixsen A, Vickers A. A randomized com- J Med 1987;317:1444-51. parison of homeopathic and standard care for the treat- 144. Mandel EM, Rockette HE, Bluestone CD, et al. Myrin- ment of glue ear in children. Complement Ther Med gotomy with and without tympanostomy tubes for 1999;7:132-5. chronic otitis media with effusion. Arch Otolaryngol 159. Sawyer CE, Evans RL, Boline PD, et al. A feasibility Head Neck Surg 1989;115:1217-24. study of chiropractic spinal manipulation versus sham spi- 145. Mandel EM, Rockette HE, Bluestone CD, et al. Efficacy nal manipulation for chronic otitis media with effusion in of myringotomy with and without tympanostomy tubes children. J Manipulative Physiol Ther 1999;22:292-8. for chronic otitis media with effusion. Pediatr Infect Dis 160. Ernst E. Serious adverse effects of unconventional ther- J 1992;11:270-7. apies for children and adolescents: a systematic review 146. University of York Centre for Reviews and of recent evidence. Eur J Pediatr 2003;162:72-80. Dissemination. The treatment of persistent glue ear in 161. Cohen MH, Eisenberg DM. Potential physician mal- children. Effect Health Care 1992;4:1-16. practice liability associated with complementary and 147. Rovers MM, Straatman H, Ingels K, et al. The effect of integrative medical therapies. Ann Int Med 2002;136: short-term ventilation tubes versus watchful waiting on hearing in young children with persistent otitis media with 596-603. effusion: a randomized trial. Ear Hear 2001;22:191-9. 162. Mullins RJ, Heddle R. Adverse reactions associated 148. Paradise JL, Bluestone CD, Colborn DK, et al. Ade- with echinacea: the Australian experience. Ann Allergy noidectomy and adenotonsillectomy for recurrent acute Asthma Immunol 2002;88:42-51. otitis media: parallel randomized clinical trials in chil- 163. Miller LG, Hume A, Harris IM, et al. White paper on dren not previously treated with tympanostomy tubes. herbal products, American College of Clinical Phar- JAMA 1999;282:945-53. macy. Pharmacotherapy 2000;20:877-91. 149. Boston M, McCook J, Burke B, et al. Incidence of and 164. Angell M, Kassirer JP. Alternative medicine: the risks risk factors for additional tympanostomy tube insertion of untested and unregulated remedies. N Engl J Med in children. Arch Otolaryngol Head Neck Surg 2003; 1998;339:839-41. 129:293-6. 165. Fallon JM. The role of chiropractic adjustment in the 150. Coyte PC, Croxford R, McIsaac W, et al. The role of care and treatment of 332 children with otitis media. adjuvant adenoidectomy and tonsillectomy in the out- J Clin Chiropractic Pediatr 1997;2:167-83. come of insertion of tympanostomy tubes. N Engl 166. Shafrir Y, Kaufman BA. Quadriplegia after chiropractic J Med 2001;344:1188-95. manipulation in an infant with congenital torticollis 151. Paradise JL, Bluestone CD, Rogers KD, et al. Efficacy caused by a spinal cord astrocytoma. J Pediatr 1992; of adenoidectomy for recurrent otitis media in children 120:266-9. previously treated with tympanostomy-tube placement: 167. Corey JP, Adham RE, Abbass AH, et al. The role of results of parallel randomized and nonrandomized trials. IgE-mediated hypersensitivity in otitis media with effu- JAMA 1990;263:2066-73. sion. Am J Otolaryngol 1994;15:138-44. 152. Maw AR. Chronic otitis media with effusion (glue ear) 168. Bernstein JM. Role of allergy in eustachian tube block- and adenotonsillectomy: prospective randomised con- age and otitis media with effusion: a review. Otolaryn- trolled study. Br Med J (Clin Res Ed) 1983;287:1586-8. gol Head Neck Surg 1996;114:562-8. 153. Cohen D, Schechter Y, Slatkine M, et al. Laser myrin- 169. Ishii TM, Toriyama M, Suzuki JI. Histopathological gotomy in different age groups. Arch Otolaryngol Head Neck Surg 2001;127:260-4. study of otitis media with effusion. Ann Otol Rhinol 154. Holzman RS. Morbidity and mortality in pediatric an- Laryngol 1980;89(suppl):83-6. esthesia. Pediatr Clin North Am 1994;41:239-56. 170. Hurst DS, Venge P. Evidence of eosinophil, neutrophil, 155. Cottrell JE, Golden S. Under the mask: a guide to feeling and mast-cell mediators in the effusion of OME patients secure and comfortable during anesthesia and surgery. with and without atopy. Allergy 2000;55:435-41. New Brunswick, NJ: Rutgers University Press; 2001. 171. Hurst DS, Venge P. The impact of atopy on neutrophil 156. Kay DJ, Nelson M, Rosenfeld RM. Meta-analysis of activity in middle ear effusion from children and adults tympanostomy tube sequelae. Otolaryngol Head Neck with chronic otitis media. Arch Otolaryngol Head Neck Surg 2001;124:374-80. Surg 2002;128:561-6. 157. Crysdale WS, Russel D. Complications of tonsillectomy 172. Hurst DS. Allergy management of refractory serous and adenoidectomy in 9409 children observed over- otitis media. Otolaryngol Head Neck Surg 1990;102: night. CMAJ 1986;135:1139-42. 664-9.