Acute otitis media (AOM) remains one of the most common clinical problems in childhood. An estimated 30 million cases occurred in 1996. Its reported frequency has increased substantially in the past two decades, as have its associated costs. The annual costs of treating otitis media in children younger than 5 years have been estimated at more man $5 billion.1 Of further concern are possible indirect effects of otitis media on children's cognitive, language, speech, and emotional development. However, these relationships remain to be clarified.
Age, sex, race, genetics, feeding practices, and environmental conditions need to be considered in the epidemiology of AOM. The peak incidence is in the first 2 years of life, particularly between the ages of 6 and 12 months.2 The increased susceptibility among infants may be explained, at least in part, by characteristics of the eustachian tube, such as short length, horizontal position, and high compliance, plus immunologie factors, such as a limited response to antigens and lack of previous exposure to common bacterial and viral pathogens. Age at the first episode of AOM has been inversely associated with risk for recurrent episodes: the younger the child, the greater the risk. Children who have Down syndrome or congenital or acquired immunologie deficiencies have a high incidence of AOM, as do Native American and Alaskan and Canadian Inuit children.3
A genetic predisposition to otitis media was suggested originally through a study of Apache children,4 and more recently documented in twins and triplets in Pittsburgh.5 The former compared Apache children who lived on a reservation with those who were adopted and reared outside the reservation. The Apache children who were adopted had more episodes of AOM than did their nonApache siblings, and their illness rate was similar to that of Apache children living on a reservation. The twin and triplet model permitted control for environmental factors and suggested that a strong genetic component determines the number of episodes of middle ear effusion (MEE) and AOM in children and the amount of time a child will have MEE. AOM is also remarkably seasonal, occurring predominantly during the winter but also in the fall and spring.
A recent study at Children's Hospital of Pittsburgh delineated the occurrence and course of otitis media during the first 2 years of life in a sociodemographically diverse population of 2,253 infants.2 It affirmed the importance of certain risk factors for the development and persistence of otitis media, particularly low sodoeconomic status and exposure to a large number of other children (eg, in day care). Other less important independent variables were sex (boys were affected more than girls), duration of breastfeeding (longer was associated with less otitis), and exposure to tobacco smoke. Of the infants observed in this study until the age of 2 years, the proportions having one or more episodes of MEE between 2 and 6, 12, or 24 months were 47.8%, 78.9%, and 91.1%, respectively. Overall the mean cumulative proportion of days that MEE was present was 20.4% in the first year of life and 16.6% in the second year. In contrast to other studies, this study found that the prevalence of otitis media in black infants of lower sodoeconomic status was as high as, if not higher than, that in white infants of lower sodoeconomic status, and certainly higher than that in white middle-dass infants.
Three organisms are the predominant pathogens in AOM: Streptococcus pneumonías, non-typeable Haemophilus influenzas, and Moraxella catarrhalis. S. pneumoniae causes up to 40% of AOM cases, H. influenzas approximately 25% to 30%, and M. catarrhalis approximately 10% to 20%.6 Other less frequently involved pathogens indude group A streptococd (3%), Staphylococcus aureus (2%), and gramnegative organisms such as Pseudomonas aeruginosa (1% to 2%). Respiratory viruses, generally in combination with bacterial pathogens, have been identified in most patients with AOM. A recent study found that respiratory syncytial virus was most common, but that other viruses were also present.7
Bacterial resistance occurs by a variety of adaptive mechanisms and is an increasing problem in AOM. Although certain strains of H. influenzas and most strains of M. catarrhalis are resistant to amoxicillin by virtue of ß-lactamase production, in most cases such resistance can be overcome by a combination of a ß-Iactam antibiotic and a ß-lactamase inhibitor (eg, amoxicillin-davulanate), or by using ß-lactamase-stable antibiotics (eg, cefixime). The primary mechanism of resistance among strains of S. pneumoniae involves alterations in penicillin-binding proteins. There are at least six known penicillinbinding proteins, and a higher number of alterations in multiple penicillin-binding proteins correlates with higher resistance. In other words, resistance occurs in a graded fashion.8 This mechanism can be overcome by higher concentrations of antibiotic at the site of infection. Two mechanisms of resistance to macrolides have been identified among strains of S. pneumoniae. One involves an efflux pump and results in low-level resistance, whereas the other involves alterations in ribosomes and results in high-level resistance.9
The prevalence of drug-resistant S. pnsumoniae has increased significantly worldwide during the past two decades, and in the United States more recently.10 The prevalence is relatively high in (1) children younger than 24 months, (2) those who have recently received ß-lactam drugs, and (3) children exposed to large numbers of other children, as in day care or large families.11 It is antidpated that the proportion of penicillin-resistant S. pneumoniae strains may reach 40% to 50% and that half of the resistant strains will be highly resistant. Further, 5%, 37%, and 66% of susceptible, intermediate, and highly penicillin-resistant strains of S. pnsumoniae, respectively, are resistant to macrolide antimicrobials.12 Because the prevalence of AOM and resistance among pathogens causing AOM are increasing and because antibiotic use is an important contributor to bacterial resistance, antibiotics need to be used judidously. At the same time, timely treatment of AOM may avoid potential problems such as mastoiditis, meningitis, and hearing impairment.
CLINICAL CLASSIFICATION OF OTITIS MEDIA
Otitis media is a general term for middle ear inflammation. It ranges from acute to chronic and from with symptoms to without symptoms. Most management strategies require that it be dassified clinically as either AOM or otitis media with effusion (OME). This distinction is, to some extent, arbitrary and often varies among clirddans.10 OME is diagnosed if there is evidence of MEE on pneumatic otoscopy, but no signs of acute inflammation that would constitute evidence of middle ear infection. Evidence of MEE consists of the presence of either (1) at least two of three tympanic membrane (TM) abnormalities (abnormal color, including white, yellow, amber, or blue; opadfication other than due to scarring; and decreased or absent mobility), or (2) visible bubbles or air-fluid interfaces behind the TM. The main symptom associated with OME is hearing loss; however, this is often not apparent in infants and young children.
A diagnosis of AOM is justified when, in addition to evidence of MEE, there is recent ear pain (including unaccustomed tugging or rubbing of the ear) and /or marked redness or distinct fullness or bulging of the TM.10 Fever or irritability may or may not be present. OME may occur as either the aftermath of AOM or a consequence of eustachian tube dysfunction due to another cause, such as an upper respiratory tract infection.10 However, OME may also precede and apparently predispose to AOM. Accordingly, these two forms of otitis media may be considered segments of a disease continuum.
ISSUES COMPUCATING THE DIAGNOSIS OF OTITIS MEDIA
Accurate diagnosis of acute otitis media is often difficult in infants and young children. Symptoms may be absent or inapparent, and frequently overlap with those of an upper respiratory illness. Ear pain is the most specific symptom, but often seems absent in children with AOM. Moreover, fever and irritability may be present in acute respiratory infections with or without OME. The TM may be obscured by cerumen, and subtle changes in the TM may be difficult to discern. Additional challenges include lack of cooperation on the part of the patient, less-thanoptimal diagnostic equipment or instruments for clearing cerumen from the external auditory canal, and inadequate assistance or lack of experience in removing cerumen or performing pneumatic otoscopy
Several instruments help optimize the accuracy of the clinician's diagnosis of otitis media. The pneumatic otoscope (model 20200, Welch Allyn, Inc., Skaneateles Falls, NY) is the standard tool. Also valuable is a surgical head (model 21700, Welch Allyn), which facilitates cleaning cerumen from an infant's external auditory canal and performing diagnostic tympanocentesis. If the TM is initially obscured by cerumen, a Buck curette (N-400-0, Storz Instrument Co., St. Louis, MO) may be used to carefully remove the cerumen from the canal under direct vision through the otoscope. Any remaining bits can then be wiped away using a Farrell applicator (N-2001A, Storz) with its tip (triangular in crosssection) wrapped with a bit of dry or alcohol-moistened cotton to create a dry or wet "mop." Alternatively, gentle suction may be applied using a no. 7 French ear suction tube.
We prefer to restrain the uncooperative patient in the prone position throughout the procedure, turning his or her head to the left or right as each ear is cleared. One adult usually a parent, can place one hand on each of the patient's buttocks and brace the patient's hips against the examining table, using his or her own weight if necessary. Another adult can restrain the patient's head with one hand and the patient's free arm with the other, changing hands for the opposite ear.
The pneumatic otoscope should have a light source of sufficient brightness and include an airtight system that permits application of positive and negative pressure. Pneumatic otoscopy permits assessment of the contour of the TM (normal, retracted, full, or bulging), its color (gray, yellow, pink, amber, white, red, or blue), its translucency (translucent, semiopaque, or opaque), and its mobility (normal, increased, decreased, or absent) in assessing middle ear status. The normal TM is translucent, is pearly gray, and has a ground-glass appearance (Fig. 1). Specific landmarks can be easily visualized, including the short process and the manubrium of the malleus and the pars flaccida, located superiorly. These help identify the position of the TM. Further, inward movement of the TM should be induced by positive pressure in the external canal and outward movement by negative pressure, especially in the superior posterior quadrant.
When the TM is retracted, the short process of the malleus becomes more prominent and the manubrium appears shortened because of its change in position within the middle ear. Inward motion with positive pressure is restricted or absent because the TM is frequently as far inward as it can go. However, outward mobility can be visualized when negative pressure is applied. If the TM does not move perceptibly with applications of gentle positive or negative pressure, MEE is likely. Sometimes the application of pressure will make an air-fluid interface behind the TM more evident. As stated earlier, AOM is diagnosed if, in addition to MEE7 there is recent ear pain or marked redness, or distinct fullness or bulging of the TM. In contrast, OME is diagnosed if MEE is present, but there are no symptoms or signs of acute inflammation (Fig. 2).
Figure 1. Normal tympanic membrane
Figure 2. Otitis medía with effusion.
Figure 3. Acute otitis media.
Figure 4. Acute otitis media (bultous myringitis).
In patients with AOM, the TM is frequently bulging. The malleus may be obscured and the TM red, pale yellow, or white, and opaque. The TM may (1) resemble a bagel with a central depression, but without a hole (Fig. 3); (2) present with bullae filled with purulent, serous, or bloody effusion (Fig. 4); or (3) have a cobblestoned appearance (Fig. 5). The algorithm in Figure 6 may be used as a guide to (1) determine the presence or absence of MEE on the basis of TM findings on pneumatic otoscopy, and (2) classify otitis media as either AOM or OME on the basis of the clinical history and otoscopie findings.
Tympanometry is a simple, rapid, atraumatic test that offers objective evidence of the presence or absence of MEE.13 It is easily administered by paraprofessional personnel. The tympanogram provides information about TM compliance in electroacoustic terms that is roughly equivalent to TM mobility discerned visually by pneumatic otoscopy. Compliance is measured by presenting a probe tone to the sealed ear canal, and by measuring the sound reflection as a result of controlled changes of ear canal pressure. Compliance is measured in milliliters and is rated as high when it is 0.5 mL or greater, intermediate when it is less than 0.5 mL but greater than 0.2 mL, and low when it is 0.2 mL or less.
Figure 5. Acute otitis media (cobblestoned tympanic membrane).
In addition to compliance, the tympanogram also provides information about middle ear pressure (measured in mm H2O and categorized as normal, negative, or positive) and the peak, or shape, of the curve. A peak falling between -100 and +50 mm H2O signifies normal middle ear pressure, a peak at less than -100 mm H2O signifies high negative pressure, and a peak at greater than +50 mm H2O signifies high positive pressure. The tympanometric peak will be sharp, suggesting a low likelihood of MEE; rounded, suggesting a greater likelihood of MEE; or flat, suggesting a high likelihood of MEE (Fig. 7).13
Although tympanometry is sensitive in detecting MEE, its positive predictive value for MEE is limited. Accordingly, abnormal or questionable tympanograms are not uncommon with normal middle ear status, especially in infants. On the other hand, tympanograms are sometimes normal early in AOM. Tympanometry may be helpful in office screening by obviating the need for routine otoscopie examination in difficult-to-examine patients older than 6 months whose TMs have been visualized previously, who are asymptomatic, and whose tympanograms are classified as normal, and by identifying patients who require further attention because their tympanograms are abnormal, Tympanometry also may confirm, refine, or clarify questionable otoscopie findings; objectify tibe f oUow-up evaluation of patients with known middle ear disease; and serve as a validator (or invalidate*) of otoscopie diagnoses of MEE. Although tympanometry can predict the probability of MEE, it cannot distinguish the effusion of OME from that of AOM.
Figure 6. Algorithm for distinguishing acute otitis media from otitis media with effusion.
Figure 7. The tympanographtc peak will be (left) sharp, suggesting a low likelihood of middle ear effusion (MEE); (center) rounded, suggesting a greater likelihood of MEE; or (right) flat suggesting a high likelihood of MEE.
SPECTRAL GRADIENT ACOUSTIC REFLECTOMETRY (SOAR)
Reflectometry is a relatively new adjunctive technique that measures the condition of the middle ear by assessing the response of the TM to a sound stimulus. New versions analyze the reflectance of sound waves with respect to frequency and present results as spectral grathent angles that correspond to the likelihood of MEE.14 A spectral grathent angle of greater than 95° indicates a low risk of MEE. The instrument is small and portable, and gives readings rapidly. Small amounts of cerumen partially occluding the ear canal do not affect findings. SGAR has been reported to be equivalent to tympanometry for the diagnosis of MEE.14 However, recent evaluation of SGAR in children with AOM identified a relatively high proportion of errors at our institution, particularly in young children.15 Although SGAR is promising, further research is warranted. As in the case of tympanometry, SGAR cannot distinguish the emisión of OME from mat of AOM.
DIAGNOSTIC TYMPANOCENTESIS AND MYRINGOTOMY
Diagnostic tympanocentesis involves puncturing the TM and aspirating middle ear fluid to relieve pressure and permit the identification of infecting organisms, and myringotomy involves incising the TMf with or without aspiration. These procedures have become lost arts for most pediatricians trained during the antimicrobial era. Today, with the increasing incidence of antimicrobial resistance among organisms causing otitis media, the use of diagnostic tympanocentesis has taken on greater importance in selecting optimal antimicrobial therapy. When symptoms or signs fail to improve after treatment with amoxicillin (at standard or higher dosages), a second-line antimicrobial agent effective against drug-resistant S. pneumoniae and strains of H. influenzae and M. catarrhalis mat produce ß-lactamase is usually used empirically. If the patient then fails to improve after 48 to 72 hours of the secondline drug, diagnostic tympanocentesis is indicated, particularly in light of the increasing incidence of drug-resistant S. pnewnaniae during the past few years.16
Other indications for tympanocentesis or myringotomy include severe otalgia before or during therapy; refractory AOM in children who have recently received multiple courses of antimicrobials, particularly those attending day care; and suppurative complications (mastoiditis, meningitis, labyrinmitis, and facial nerve paralysis).17
Finally, regional and temporal variability in causative pathogens underscores the desirability of knowing the local flora before prescribing therapy. The Drug-resistant Streptococcus pneumoniae Therapeutic Working Group of the Centers for Disease Control and Prevention (CDC) recently recommended that "in an era of increasing antimicrobial resistance, clinicians treating children with AOM should consider developing the capacity to perform tympanocentesis themselves or establish ready referral mechanisms to a clinician with this capacity."11 Potential complications of tympanocentesis include chronic perforation of the TM, bleeding from an exposed jugular bulb, dislocation of the incudostapedial joint, and facial nerve paralysis. To our knowledge, no systematic study of me frequency of these has been conducted, but they must be rare. Whether tympanocentesis or myringotomy may themselves favorably influence outcome remains an unsettled question.
Also to our knowledge, no method of inducing analgesia before performing tympanocentesis has been formally evaluated. For several years we have used acetaminophen with codeine (1 mg/kg of codeine plus 10 mg/kg of aeetaminophen, supplemented with 5 mg/kg of acetaminophen to reach a total acetaminophen dose of 15 mg/kg). Alternatively, we have used oral midazolam (0.7 mg/kg, maximum dose 18 mg) combined with ibuprofen (10 mg/kg). Importantly, continuous pulse oximeter monitoring and dose observation are necessary when using the latter combination, and flumazenil (10 pg/kg every 30 to 60 seconds, maximum dose 1 mg), which reverses the effect of midazolam, should be available in the unlikely event that hypoventilation, cardiac suppression, or oversedation occurs. Midazolam's anxiolytic and amnesic effects are particularly valuable for future rapport with children,17 We are currently conducting a randomized clinical trial comparing the effectiveness of (1) acetaminophen with codeine plus acetaminophen, (2) oral midazolam plus ibuprofen, and (3) acetaminophen alone (control group).
Some authors have recommended 8% tetracaine with a Pope Otowick.17 However, concern exists about topical anesthetics because they are bacteriostatic and can affect culture results. Some otolaryngologists have reported that touching the TM with the tip of a phenol applicator produces excellent anesthetic results. In our experience, however, application of phenol often results in pain comparable to that of tympanocentesis. EMLA cream has also been useful in adults. Uniform contact of the cream with the TM and complete suction prior to tympanocentesis may not be easy to achieve, and safety concerns arise from animal studies describing ototoxicity when EMLA was instilled into the middle ear. Additional pain management methods include (1) 1% lidocaine injection in the posterior, inferior bony canal, which requires an otomicroscope and subjects patients to another injection; and (2) iontophoresis with lidocaine, which gives inconsistent results. A technique for performing diagnostic tympanocentesis is detailed in die table.
TREATMENT OF AOM
Some authors have recommended withholding antimicrobial treatment entirely in some or all cases of AOM unless symptoms persist or worsen.18 As noted previously,10 this policy seems questionable because of the increased risk of delayed improvement19 or suppurative complications. A report from Germany described an increase in the occurrence of mastoiditis, mostly in children younger than 2 years: many of these had received (for immediately antecedent AOM) either no antibiotic treatment, a drug with limited efficacy, or treatment for only 5 days.20 In contrast, a widely accepted strategy to reduce antimicrobial pressure has been to drastically curtail the use of antimicrobial prophylaxis for recurrent AOM.10 Another proposed strategy has been to limit the duration of antimicrobial treatment, in presumably low-risk episodes of AOM, to 5 days rather than the usual 10 days (see below).21
PHARMACOKINETIG AND PHARMACODYNAMIC PRINCIPLES OF ANTIMICROBIAL ACTIVITY
The primary goal of antimicrobial therapy is to provide an adequate concentration at the site of infection to eradicate the pathogens. Historically, the variable used to predict effectiveness has been the minimum inhibitory concentration (MIC) breakpoint. It has been assumed that if in vitro testing found a pathogen to be susceptible, treatment with that agent could eradicate the organism, However, the classification of pathogens as susceptible, intermediate, or resistant based on breakpoints can be difficult to interpret, because MIC breakpoints are merely one factor in the equation. For example, although MIC data may indicate that a pathogen is susceptible to a particular antimicrobial agent, the fact that the agent does not achieve sufficient concentrations in the middle ear may predude its use for AOM. Also important are pharmacokinetic and pharmacodynamic parameters such as the ability to achieve adequate middle ear concentrations for a sufficient period of time, the mechanism of antimicrobial effect (ie, concentrationdependent or time-dependent killing), and whether there are persistent post-antibiotic effects. The different antimicrobials used in AOM treatment vary significantly in their pharmacokinetic and pharmacodynamic properties,
Technique for Performing Tympanocentesls
However, for each class of antimicrobials, certain parameters are most important in predicting activity. The efficacy of penicillins, cephalosporins, dindamycin, and macrolides is determined primarily by the length of time they are above the MIC of the pathogen in question, because of their time-dependent killing mechanism and minimal to moderate post-antibiotic effects. Achieving serum concentrations above the MIC for greater than 40% to 50% of the dosing interval has resulted in bactériologie cure rates in 80% to 85% of patients.22
On the other hand, the efficacy of fluoroquinolones, ketolides, and azithromycin correlates better with the ratio of the area under the concentration-time curve (AUC) to the MIC because of the concentration-dependent mechanism of killing and the prolonged post-antibiotic effect associated with these drugs. The AUC to the MIC ratio that predicts efficacy is approximately 25 to 30.
In the case of ß-lactam treatment of S. pneumoniae, inability to reach the necessary concentration for the necessary length of time is associated with an increased risk of bactériologie and clinical failure. Similar considerations may apply to macrolides and azithromycin at respiratory sites where serum concentrations are predictive of efficacy (ie, sinus or middle ear).
SELECTING FIRST-LINE AND SECOND-UNE ANTIMICROBUL THERAPY
Choice of an antimicrobial drug is guided by knowledge of overall bacterial etiology, bacterial susceptibility, and expected spontaneous cure rates. As mentioned previously, as many as 40% to 50% of S. pneumoniae strains are resistant to penicillin. On the other hand, approximately 20% of AOM episodes caused by S. pneumoniae can resolve spontaneously.23 OfK influenzae and M. catarrhalis strains causing AOM, approximately 40% and 100%, respectively, produce ß-lactamase, but approximately 50% and 80% of AOM episodes caused by these pathogens can be expected to resolve spontaneously.23 Although several antibiotics are currently available for the treatment of AOM, a select few remain effective against all predominant pathogens. As a result, selection of appropriate empiric therapy can be difficult. The recent report of the CDC Working Group was developed to help guide decisions regarding this empiric therapy. It states that an appropriate first-line agent should have sufficient activity against resistant S. pneumoniae. The report also states that, for clinically defined initial treatment failures, an appropriate second-line drug should be effective against S. pneumoniae, including most drugresistant S. pneumoniae, and strains of H. influenzae and M, catarrhalis that produce β-lactamase.11
Amoxicillin, a relatively inexpensive drug with good tolerability and an excellent safety record, continues to be the most efficacious antimicrobial for susceptible S. pneumoniae strains. Increasing the dose to 80 to 90 mg/kg/d (to a maximum of 2.0 g/d) in two divided doses provides increased concentrations in middle ear fluid (3 to 8 pg/mL for at least 3 hours after the dose) and thus will generally be effective against intermediately resistant (0.1 to 1 ^g/ mL) and some highly resistant strains (s* 2 /ig/mL).24 The higher dose is appropriate for children in whom the prevalence of resistant strains will be relatively high (ie, children younger than 24 months, children who have recently received treatment with ß-lactam drugs, and children who are exposed to large numbers of other children). These are also the children who are most likely to respond unsatisfactorily to treatment. For those allergic to penicillin, or when palatability or convenience is an important issue, azithromycin is an attractive alternative. It also has an excellent safety record.
Oral cephalosporins (except cefuroxime axetil) and macrolides do not reliably provide activity against resistant strains of S. pneumoniae. Drugs with inadequate activity against susceptible S. pneumoniae strains, such as cefaclor, ceftibuten, and cefixime, are, of course, also ineffective against intermediately or highly resistant strains. Of the currently available regimens for AOM, only amoxicillin, amoxicillinclavulanate (because of the amoxicillin component), cefuroxime axetil, and intramuscular ceftriaxone (in two or three daily doses) provide concentrations above the MIC90 of most drug-resistant S. pneumoniae strains for at least 40% of the dosing interval. Clindamycin was recommended by the CDC Working Group as an option for patients with pneumococcal infection who fail initial therapy and had received antibiotics within the prior month.11 The rate of pneumococcal resistance to clindamyän is lower than that to other oral antimicrobials, but clindamyän has limited activity against H. influenzae and M. catarrhalis. Accordingly, clindamyan should be reserved for patiente known to have infection caused by resistant S. pneumoniae.
Activity against β-lactamase-positive H. influenzae strains varies significantly among the second-line antimicrobial agents. Amoxicillin-davulanate and the third-generation cephalosporins (cefixime, cefpodoxime, ceftibuten, cefdinir, and ceftriaxone) have the greatest activity. Cefuroxime axetil is active, whereas the activity of cefador, loracarbef, clarithromycin, cefprozil, and trimethoprim-sulfamethoxazole (TMP-SMX) is limited.25
Using a repeat tympanocentesis design in children treated with either azithromycin for 5 days or amoxicillin-clavulanate for 10 days, Dagan et al. found lower eradication rates for H. infiuenzae after 4 to 6 days of azithromycin than with amoxicillindavulanate (39% vs 87%, respectively).26 Among children with H. influenzae isolated at study entry, significant differences in clinical outcomes (symptomatic response and the presence of TM findings indicative of acute middle ear inflammation) were also noted between these two antimicrobial agents (clinical cure rates of 65% and 91% at the end of therapy for azithromycin and amoxicillin-clavulanate, respectively).26 However, no difference in efficacy was found between the two agents regarding pneumococcal infection, and results were not reported in approximately 40% of patients who either had initially negative results on culture or, for one reason or another, were not able to be evaluated. All currently available agents except TMP-SMX amoxicillin, and cefador are active against β-lactamase-positive strains of M. catarrhalis.
Second-line treatment options for the patient who is allergic to penicillin include oral cephalosporins (if he or she is not also allergic to cephalosporins), macrolides, and TMP-SMX. Rates of resistance by pneumococci to TMP-SMX are high, and in some areas are higher than rates of resistance to penicillin,11 Approximately 25% of H. infiuenzae strains are also resistant to TMP-SMX, and TMP-SMX is not effective against group A streptococci, which may cause up to 10% of AOM episodes in older children. Among the macrolides, some children refuse the combination of erythromycin plus sulfisoxazole because of taste or gastrointestinal upset. Azithromycin was not selected by the CDC Working Group because pneumococcal resistance, when it occurs, is absolute (unaffected by increasing the dose), and because of the previously mentioned limited efficacy against H. influenzae,11 However, it may be the best available alternative for children allergic to ß-lactam. Newer antimicrobials likely to play a role in the future management of AOM indude streptogramins, ketolides, fluoroquinolones, and oxazolidinones.
DURATION OF ANTIMICROBIAL THERAPY
A recent editorial by one of us (JLP)21 discussed some of the issues surrounding clinical trials that compare shorter-duration therapy with longerduration therapy. Some trials found no differences in outcome, whereas others found differences favoring longer treatment. Most studies were characterized by one or more limitations: small sample size and, accordingly, limited statistical power; few or no children younger than 2 years; lack of analysis of outcome by age; exdusion of children prone to otitis media; lack of standard criteria for the diagnosis of AOM; and use of a lower than recommended dosage of a drug or use of a drug that has less than optimal efficacy against common middle ear pathogens.21
Four recent randomized clinical trials compared longer-duration therapy with shorter-duration therapy for AOM in infante and young children.27"^ Three27"29 used stringent criteria for the diagnosis at entry and determined outcome based on both symptomatic response and TM findings indicative of acute middle ear inflammation. The fourth trial30 used nonstringent criteria for diagnosis and outcome. In the first three,27'29 success rates consistently favored longer courses of therapy, reaching statistical and clinical significance at the 12- to 14-day visit. In two of the trials,28'29 children who spent time outside the home (in day care centers or with babysitters) had significantly lower success rates than did those cared for entirely at home.
Results were discrepant regarding outcomes at the late follow-up visit (4 to 6 weeks following diagnosis). One trial27 showed significant differences favoring longer treatment in children younger than 2 years, whereas two trials28*29 showed no differences between outcomes of shorter and longer treatment. In the fourth trial,30 no differences in outcome at any time were noted between shorter and longer treatments. However, the validity and generalizability of the latter trial's findings are rendered questionable by the large number of observers (50 sites in 8 countries), use of nonstringent diagnostic and outcome criteria, routine performance of tympanocentesis at outset, and lack of blinding of parents and investigators.
Three additional trials31'33 compared the efficacy of a single dosé of intramuscular ceftriaxone, which also may be viewed as short-duration therapy, with 10-day courses of oral antimicrobial treatment. Again, the validity and generalizability of mese trials' findings are rendered questionable by several limitations in design. Caution is further occasioned by a recently published study in children with refractory AOM34 that showed significantly better outcomes with 3 doses compared with a single dose of intramuscular ceftriaxone. Finally, intramuscular ceftriaxone has a potential impact, as a potent third-generation parenteral antibiotic, on the development of bacterial resistance. The pain associated with injection and its cost compared with that of oral treatment are additional problems.
In summary, most recent studies involving large numbers of children and using stringent diagnostic and outcome criteria confirm previous studies of AOM: shorter-course antimicrobial treatment seems less effective than longer-course treatment in infants and young children. Accordingly, short-course treatment seems injudicious as routine for AOM in those younger than 2 years or those cared for outside the home. Further study is warranted in children 2 to 5 years old to determine appropriate length of therapy, although individualizing management may prevent overtreatment or undertreatment. For example, children in day care with bilateral disease, a history of recurrent AOM, or AOM during the winter months may benefit from a longer course of therapy. In contrast, older children who have been relatively free of AOM, or who have a mild episode during spring or summer, or who have a prompt symptomatic response might do well with a shorter course of therapy. Use of intramuscular ceftriaxone for AOM is, in general, not recommended. Exceptions to this include when oral treatment is not feasible, in highly selected cases of refractory AOM following failure of orally administered second-line antimicrobials, or when highly resistant S. pneumoniae is found by diagnostic tympanocentesis. If ceftriaxone is used, consideration should be given to administering 2 or 3 doses at intervals of 2 to 3 days.
OTHER THERAPEUTIC CONSIDERATIONS
Considerations other than the "drug versus bug" interaction are important. Additional factors that may contribute to persistent disease despite appropriate antimicrobial therapy include concomitant viral infection, persistent inflammation and effusion despite bacterial eradication, impaired host defenses, and the adenoidal or nasopharyngeal reservoir of pathogens.
Compliance with treatment regimens and cost also need to be considered. In estimating compliance with treatment regimens, antimicrobials that require dosing more than once or twice a day are not likely to constitute favorable options for most pediatrie patients. Also, orally administered antimicrobials vary greatly in palatability. For example, cefuroxime axetil has excellent activity against S. pneumoniae and H. influenzae, but the current preparation is not well accepted. A recent evaluation judged cefuroxime, cefpodoxime proxetil, and erythromycin plus sulfisoxazole unpalatable.35 Cost is also important: insurance coverage may limit choice.
Clinicians must also take the time to discuss with parents (1) the reasons for using or not using an antimicrobial; (2) why the specific antimicrobial was chosen; (3) the importance of adherence to the prescribed regimen, particularly in young children; and (4) that low or inconsistent antimicrobial concentrations in the middle ear, by eradicating susceptible strains, may result in selection for resistant strains.36
Measures that may reduce the incidence of AOM include breastfeeding, eliminating smoking in the household, and avoidance of large group day care. Additional interventions include the recently licensed septavalent conjugate pneumococcal vaccine, influenza vaccine, tympanostomy tube placement, and adenoidectomy. Based on available data, conjugate pneumococcal vaccine appears likely to reduce the overall incidence of AOM only approximately 6% to 8%. On the other hand, the seven serotypes in the vaccine have been the ones most likely to exhibit resistance.37 A more substantial impact may be anticipated in recurrent AOM, particularly in reducing the likelihood of placement of a tympanostomy tube.37 Inactivated trivalent influenza vaccine and the newer live attenuated intranasal vaccines have been reported to reduce the incidence of AOM by approximately 30% during the respiratory season.38"40
When recurrences become intolerable, placement of a tympanostomy tube should be considered as the first surgical option.41 Adenoidectomy should be reserved for otitis media that continues after tympanostomy tubes.42,43 Because of the potential for bacterial resistance, prolonged low doses of antimicrobials should no longer be recommended for routine prophylaxis of recurrent AOM, but may be appropriate in selected cases (eg, when exposure to other young children is limited).10
In an era of increased bacterial resistance, use of stringent criteria for distinguishing AOM from OME and use of ancillary diagnostic tools (tympanometry, acoustic reflectometry, or both) may result in more judicious use of antimicrobials. Amoxicillin (at higher dosages in young and otherwise high-risk children) continues to be the standard regimen for treating AOM. When amoxicillin fails, an appropriate second-line antimicrobial should be effective against S. pneumonias, including most drug-resistant S. pneumoniae, and also against strains of H. influenze^ and M. catarrhalis that produce β-lactamase. Tympanocentesis should be considered in treatment failure despite use of a second-line drug with the appropriate coverage. In infants and young children, shortercourse antimicrobial treatment appears less effective than longer-course treatment. It is to be hoped that widespread use of the recently approved pneumococcal conjugate vaccine and the influenza vaccine will reduce the burden of this common condition.
1. BondyJ, Berman S, Glazner J, Lezotte D. Direct expenditures related to otitis media diagnoses: extrapolations from a pediatrie Medicaid cohort. Pediatrics. 2000;105:E72. Available at: www.pediatrics.org / cgj/ content / full / 105 / 6 / e72.
2. Paradise JL, Rockette HE, Colborn DK, et aL Otitis media in 2253 Pittsburgh-area infants: prevalence and risk factors during the first two years of life. Pediatrics. 1997;99:318-333.
3. Reed D, Struve S, Maynard JE. Otitis media and hearing deficiency among Eskimo children: a cohort study. Am J Public Health. 1967;57:1657-1662.
4. Spivey GH, Hirschhorn N. A migrant study of adopted Apache children. Johns Hopkms Mediad journal. 1977;140:4346.
5. Casselbrant ML, Mandel EM, Fall PA, et al. The heritability of otitis media: a twin and triplet study. ]AMA. 1999282:21252130.
6. Barnett ED, Klein JO, The problem of resistant bacteria for the management of acute otitis media. Pediatr Om North Am, 1995;42:509-517.
7. Heikkinen T, Thint M, Chonmaitree T. Prevalence of various respiratory viruses in the middle ear during acute otitis media. N Engi } Med. 1999^40260-264.
8. Dowson C Coffey TJ, Spratt BG. Origin and molecular epidemiology of penicillin-binding-protein-mediated resistance to beta-lactam antibiotics. Trends Microbiol. 19942361-366.
9. Shortridge VD, Flamm RK, Ramer N, Beyer J, Tanaka SK, Novel mechanism of macrolide resistance in Streptococcus pneumoniae. Diagn Microbiol Inject Dis. 199626:73-78.
10. Paradise JL. Managing otitis media: a time for change. Pediatrics. 1995,-%:712-715.
11. Dowell SF, Butler JC, Giebink GS, et al. Acute otitis media: management and surveillance in an era of pneumococcal resistance: a report from the Drug-resistant Streptococcus pneumoniae Therapeutic Working Group. Pediatr Inject Dis ]. 1999;18:l-9.
12. Jacobs MR, Bajaksouzian S, Zilles A, Lin G, Pankuch GA, Appelbaum PC. Susceptibilities of Streptococcus pneumoniae and Haemophilus influenzae to 10 oral antimicrobial agents based on pharmacodynamic parameters: 1997 U.S. Surveillance study. Antimicrob Agents Chemother. 1999;43: 1901-1908.
13. Paradise JL, Smith CG, Bluestone CD. Tympanometric detection of middle ear effusion in infants and young children. Pediatrics. 1976^8:198-210.
14. Combs J, Schuman AJ. Three technologies for taming otitis media. Contemp Pediatr. 1999;16:78-101.
15. Block SL, Pichichero ME, McLinn S, Aronovitz G, Kimball S. Spectral grathent acoustic reflectometry: detection of middle ear effusion in suppurative acute otitis media. Pediatr Infect Dis ]. 1999;18:741-744.
16. Hoberman A, Paradise JL, WaId ER. Tympanocentesis technique revisited. Pediatr Infect Dis J. 1997;16(suppl 2): S25-S26.
17. Block S. Tympanocentesis: why, when, how. Contemporary Pediatrics. 1999;16:103-127.
18. Cunningham A. Antibiotics for otitis media: restraint, not routine. Contemporary Pediatrics. 1994;11:17-30.
19. Kaleida PH, Casselbrant ML, Rockette HE, et al. Amoxicillin or myringotomy or bom for acute otitis media: results of a randomized clinical trial. Pediatrics. 1991^7:466-474.
20. Hoppe JE, Koster S, Bootz F, Niethammer D. Acute mastoiditis: relevant once again. Infection. 199422:178-182.
21. Paradise JL. Short-course antimicrobial treatment for acute otitis media: not best for infants and young children. JAMA. 1997278:1640-1642.
22. Craig WA, Andes D. Pharmacokinetics and pharmacodynamics of antibiotics in otitis media. Pediatr Infect Ois J. 1996;15:255-259.
23. Barnett ED, Klein JO. The problem of resistant bacteria for the management of acute otitis media. Pediatr Clin North Am. 1995^2^09-517.
24. Seikel K, Shelton S, McCracken GH Jr. Middle ear fluid concentrations of amoxicillin after large dosages in children with acute otitis media. Pediatr Infect Dis J. 1997;16:710-711.
25. Blumer J. Fundamental basis for rational therapeutics in acute otitis media. Pediatr Infect Dis J. 1999;18:11304140.
26. Dagan R, Johnson C McLinn S, et al. Bactériologie and clinical efficacy of amoxicillin/davulanate vs. azithromyan in acute otitis media. Pediatr Infect Dis J. 2000;19:95-104.
27. Hoberman A, Paradise JL, Burch DJ, et al. Equivalent efficacy and reduced occurrence of diarrhea from a new formulation of amoxiällin/davulanate potassium (Augmentin) for treatment of acute otitis media in children. Pediatr Infect Dis J. 1997;16:463470.
28. Cohen R, Levy C, Boucherai M, Langue J, de La Rocque F. A multicenter, randomized, double-blind trial of 5 versus 10 days of antibiotic therapy for acute otitis media in young children. / Pediatr. 1998;133:634-639.
29. Cohen R, Levy C, Boucherat M, et al. Five versus ten days of antibiotic therapy for acute otitis media in young children. Pediatr Infect Dis J. 2000;19:458463.
30. Pessey JJ, Gehanno P, Thoroddsen E, et al. Short course therapy with cefuroxime axetil for acute otitis media: results of a randomized multicenter comparison with amoxicillin/ davulanate. Pediatr infect Dis }. 1999;18:854-859.
31. Green SM, Rofhrock SG. Single-dose intramuscular ceftriaxone for acute otitis media in children. Pediatrics. 1993,391:23-30.
32. Barnett ED, Teele DW, Klein JO, Cabrai HJ, Kharasch SJ. Comparison of ceftriaxone and trimethoprim-sulfamethoxazole for acute otitis media. Pediatrics. 1997&92Z-28.
33. Chamberlain JM, Boenning DA, Waisman Y, Ochsenschlager DW, Klein BL. Single-dose ceftriaxone versus 10 days of cefador for otitis media. Clin Pediatr. 1994^33:642-646.
34. Leibovitz E, Piglansky L, Raiz S, et al. Bactériologie efficacy of a three-day intramuscular ceftriaxone regimen in nonresponsive acute otitis media. Pediatr Inject Dis }. 1998;17:1126-1131.
35. Steele R, Thomas M, Begue R, Despinasse B. Selection of pediatrie antibiotic suspensions: taste and cost factors. Inject Med. 1999;16:197-200.
36. Klein JO. Protecting the therapeutic advantage of antimicrobial agents used for otitis media. Pediatr Inject Dis J. 1998; 17:571-575.
37. Black S, Shinefield H, Fireman B, et al. Efficacy, safety and immunogenicity of heptavalent pneumococcal conjugate vaccine in children. Pediatr Inject Dis }. 2000;19:187-195.
38. Heikkinen T, Ruuskanen O, Waris M, Ziegler T, Aróla M, Halonen P. Influenza- vaccination in aie prevention of acute otitis media in children. Am J Dis (Md. 1991;145:445-448.
39. dements DA, Langdon L, Bland C, Walter E. Influenza A vaccine decreases the incidence of otitis media in 6- to 30month-old children in day care. Arch Pediatr Malese Med. 1995;149:1113-1U7.
40. Belshe RB, Mendelman PM, Treanor J, et al. The efficacy of live attenuated, cold-adapted, bivalent, intranasal influenza virus vaccine in children. N Engl } Med. 1998^338:1405-1412.
41. Casselbrant ML, Kaleida PH, Rockette HE, et al. Efficacy of antimicrobial prophylaxis and of tympanostomy tube insertion for prevention of recurrent acute otitis media: results of a randomized clinical trial. Pediatr Infect Dis J. 1992;11:278286.
42. Paradise JL, Bluestone CD, Rogers KD, et al. Efficacy of adenoidectomy for recurrent otitis media in children previously treated with tympanostomy-tube placement: results of parallel randomized and nonrandomized trials. JAMA. 1990^63:2066-2073.
43. Paradise JL, Bluestone CD, Colborn DK, et al. Adenoidectomy and adenotonsiHectomy for recurrent acute otitis media: parallel randomized clinical trials in children not previously treated with tympanostomy tubes. JAMA. 1999^82:945-953.
Technique for Performing Tympanocentesls