Management of acute otitis media in
an era of changing antimicrobial drug susceptibility
Jerome O Klein
Division of Pediatric Infectious Diseases, Boston City Hospital, Boston, Massachusetts, USA
Otitis media is the most frequent illness for which children
visit the physician. In 1990, approximately 24.5 million such visits were reported. For the majority of acute infections,
an antibiotic (most frequently amoxicillin) was prescribed. The emergence of multidrug-resistant strains of Streptococcus pneumoniae
and the increasing incidence of beta-lactamase-producing strains of Hemophilus
influenzae prompt a reexamination of the impact resistant
strains have on the selection of initial antibacterial therapy for acute otitis media (AOM). In response to emerging
drug resistance, some experts have suggested that mild to moderate cases of acute otitis media may not need antimicrobial
Epidemiology and Risk Features of Acute Otitis Media
Otitis media has been studied extensively in various populations in the United States, Canada, Great Britain and
Western Europe. Host, external and environmental factors have been defined in the groups most susceptible to recurrent
and severe disease. Alteration of these causative factors and promotion of protective factors may reduce the incidence
of AOM and thus the volume of antimicrobial agents used for this indication.
Acute otitis media is a disease of infancy and early childhood with peak incidence at 6 to 18 months. Although
episodes do occur throughout childhood and in adults, patients who have had few episodes of AOM before 3 years
of age are unlikely to experience recurrent AOM subsequently.
The incidence of AOM is significantly higher in males than in females.
The incidence of AOM is high among selected racial groups in developing areas or hostile environments, such as
Native Americans, Alaskan Eskimos and children in African villages. Within these groups, it may be difficult to
separate factors of race from those of poverty, inadequate hygiene and limited access to medical care.
Children with siblings with a history of middle ear infections are more likely to have recurrent episodes of AOM
than children whose siblings have had little or no disease.
Environmental and External Factors
Otitis media occurs throughout the year, but the highest incidence is during the fall and winter months, reflecting
the seasonal pattern of respiratory tract infections.
Breastfeeding has been found to be protective against respiratory infections in infancy, including otitis media.
Breastfeeding for as short a period as 3 months was associated with a decreased risk of AOM during the first year
Passive smoking and environmental pollutants have come under increased scrutiny as agents of structural and physiologic
changes in the respiratory tree. The concentration of cotinine in urine or serum was related to the number of smokers
in the household and an increased incidence of AOM.2
Large-group day care.
Children who attend large-group day care are exposed to more respiratory infections than children in home or family
care. Prospective studies have identified an increased incidence of acute and chronic otitis media and more surgical
procedures to resolve chronic disease in children in large-group day care than in those in family care.3
In an investigation of 14,000 infants in Bristol, England,4 more episodes of AOM were identified in children who slept prone (contrasted
with children who slept supine).
The microbiology of otitis media, based on cultures of middle ear effusions obtained by needle aspiration through
the tympanic membrane, has been remarkably consistent during the past 40 years in studies from the United States,
Europe and Japan. S. pneumoniae
is the leading bacterial pathogen, nontypable H. influenzae is the second most important and Moraxella
catarrhalis is usually the third most frequent pathogen;
less than 10% of all acute infections are due to group A streptococcus and Staphylococcus
aureus. The bacterial agent of AOM cannot be identified
by clinical signs, but disease caused by pneumococcus is more likely to be severe and less likely to remit spontaneously
than is disease caused by H. influenzae and M. catarrhalis.5
Only a few types of S. pneumoniae
are responsible for most cases of pneumococcal AOM. The most frequent types (in order of decreasing frequency)
are types 19, 23, 6, 14, 3, and 18.6-9 An effective vaccine that covers 7 or 8 types of S.
pneumoniae could prevent more than 80% of pneumococcal otitis
Most H. influenzae strains that cause otitis are nontypable. Before the introduction of the conjugate polysaccharide
type b vaccine, only about 10% of cases of hemophilus otitis were caused by type b strains; that percentage is
likely to be even lower in communities with effective immunization programs. Currently 20 to 50% of strains of
responsible for otitis media produce beta-lactamase, an increase seen during the past 10 years.
Management of AOM is based on the choice of an appropriate antimicrobial agent with activity against the 3 major
pathogens. Decongestants and antihistamines may provide some comfort for the patient who has mucosal inflammation
and congestion of the upper respiratory tract but provide no benefit for acute or chronic ear infections. The antimicrobial
agent should have documented clinical and microbiologic efficacy (determined by cultures of sequential aspirates
of middle ear fluids from patients with AOM), limited side effects, convenient dosages, palatability when provided
in oral form and reasonable cost. An appropriate response to therapy is substantial resolution of signs and symptoms
within 72 hours.
The In-Vivo Suceptibility Test
The US Food and Drug Administration (FDA) has approved 12 antimicrobial agents for the indication of AOM. Although
all are approximately equal in clinical efficacy, they differ in microbiologic efficacy as identified by the in
vivo susceptibility test.10
When fluid from the middle ear of children with AOM was aspirated before therapy and 2 to 7 days after therapy
pneumococci persisted in middle ear fluids if no treatment was given in 81% of cases; nontypable H. influenzae strains
persisted in only about 50% of cases without use of an antimicrobial agent (Table 1). Because many cases of bacterial
otitis media clear without an antimicrobial agent, the gold standard for evaluating microbiologic efficacy of the
drugs should be the rate of persistence of the organisms after administration of placebo. The microbiologic results
for selected drugs in Table 1 generally reflect data on in vitro activity and the concentrations of drug achieved
in the middle ear.
Antimicrobial Resistant S. pneumoniae
Resistance of S.
pneumoniae to multiple antibacterial agents has now been
recognized throughout the world. The resistance is based on alterations of penicillin-binding proteins and is incremental
rather than absolute. For penicillin G and V, intermediate resistance is defined at a minimum inhibitory concentrations
(MIC) of 0.1 to 1.0 µg/ml and high-level resistance at an MIC of 2.0 µg/ml or greater. The National
Committee for Clinical Laboratory Standards regularly reviews and publishes concentration guidelines for susceptibility
and intermediate or high-level resistance to antimicrobial agents.
Rates of antimicrobial resistance in pneumococci isolated from middle
ear fluids and the nasopharynx of children with AOM have been studied in the United States and Western Europe.
In a study from a private practice in Kentucky from 1992 to 1993,12 28% of isolates of S.
pneumoniae were resistant to penicillin; nearly half of
these strains had high-level resistance (MIC > 2.0 µg/ml). Current susceptibilities of nasopharyngeal
strains obtained from children in Boston are presented in Table 2.13 Approximately 20% of the pneumococcal strains were relatively resistant
to penicillin but only about 25% had high-level resistance. In Dallas, 40% of strains are penicillin resistant
and approximately 50% show high-level resistance (Personal communication, George H. McCracken, Jr., MD). Resistance
to trimethoprim-sulfamethoxazole is higher and resistance to cephalosporins lower than resistance to penicillin.
The macrolides, azithromycin, clarithromycin and clindamycin have low rates of resistance.
Antibiotic Resistant H. influenzae
Production of beta-lactamase by strains of H. influenzae was
first noted in the early 1970s. Beta-lactamase hydrolyzes ampicillin, amoxicillin and penicillins G and V. In resistant
organisms, the beta-lactam ring is broken, rendering the drugs inactive. Amoxicillin is no better than a placebo
against beta-lactamase-producing strains of H. influenzae (Table 1). Thus, the incremental increases in MIC required
to eradicate the intermediate or highly resistant pneumococci contrast with the more absolute activity of beta-lactamase-producing
organisms. Since the 1970s, approximately 15 to 40% of H.
Influenzae isolates in the United States have produced beta-lactamase,
though rates vary temporally and geographically. In Pittsburgh, the incidence of beta-lactamase-producing strains
of H. influenzae isolated
from middle ear fluids from children rose from less than 20% to more than 40% between 1980 and 1989.14 The incidence
of resistance of H. influenzae
to amoxicillin not associated with beta-lactamase production remains less than 5% in most centers in Europe and
Role of Antibacterial Resistance in Outcome of Therapy
The role of resistant organisms in treatment failures in AOM may be estimated by considering the number of infections
caused by a particular organism, the percentage of those organisms that can be expected to be resistant to the
drug and the percentage of infections caused by these organisms that might be expected to resolve spontaneously.
causes up to 40% of cases of AOM; approximately 20% of strains (in the Boston area) are resistant to penicillin
or amoxicillin; about 20% of these cases are likely to resolve spontaneously. Thus, 8% of cases of AOM might fail
clinically when treated with amoxicillin because of resistance. But about 20% clear spontaneously and clinical
failure would be expected only in the strains that have high-level resistance (about 25% of all resistant strains
in Boston). Thus, clinical failure would be expected in only 1.6% of cases (the 25% that are highly resistant of
the 8% with any resistant strain less the 20% that would spontaneously resolve).
Similar reasoning can be used to estimate the role of drug-resistant
in clinical failure. If H. influenzae
is responsible for about 25% of cases of AOM and about 30% involve beta-lactamase-producing strains, 8% of all
AOM would be due to a beta-lactamase-producing H. influenzae. But 50% of cases with these strains resolve spontaneously. Thus, clinical
failure due to a beta-lactamase-producing H. influenzae strain would be expected in only 4% of all cases treated with amoxicillin.
These data for antimicrobial resistance among the major pathogens of AOM indicate a need for concern and surveillance
of strains obtained from the respiratory tract of infants and young children. The incidence of resistant pneumococci
and H. influenzae
in Boston indicate that amoxicillin may still be considered the drug of choice. In other communities with higher
rates of resistance, alternative agents may need to be considered.
Management of Acute Otitis Media
Withholding Antimicrobial Agents for AOM
The microbiologic data suggest that many children with AOM need not be treated with antimicrobial agents. Approximately
33% of cases do not have a bacterial pathogen and presumably would not benefit from treatment with an antimicrobial
agent. Pneumococcal otitis media resolves without antimicrobial therapy in about 20% of children with clearance
of the organism from the middle ear and resolution of acute signs; approximately 50% of nontypable H. influenzae cases
clear similarly without use of an antimicrobial agent. Kaleida and colleagues16
demonstrated that approximately 90% of children with less
than severe disease (based on temperature and a system scoring otalgia) improved without antimicrobial therapy.
A meta-analysis of 5400 children from 33 randomized trials showed that resolution of clinical signs occurred without
antibiotics in 81% of cases, but the cure rate was 95% when antimicrobial agents were administered. The incidence
of suppurative complications, usually acute mastoiditis, was increased in 2 surveys in which children with AOM
were not treated with antimicrobial agents,17,18 and Kaleida and colleagues16 demonstrated that significantly fewer children who completed a 2-week course
of amoxicillin had effusion at that time (47%) when compared to those who received placebo (63%).
Because we cannot distinguish the small proportion of children who may benefit from antimicrobial therapy from
those who would respond without drug therapy, physicians should consider antimicrobial therapy for all children
with AOM. Current investigational studies may be able to define which children may be observed successfully rather
than treated immediately.
Is Garlic the Answer?
Although new antimicrobial agents are needed in response to the constant change in the susceptibility of important
microbial pathogens, alternative medications may also be of value. Farbman and colleagues19 demonstrated that garlic extract was effective in vitro against antibiotic-resistant
pneumococci, beta-lactamase-producing H. influenzae, methicillin-resistant staphylococci and other species of resistant bacteria.
Antibacterial activity was lost when the extract was boiled. Other common plant products could also be of possible
value because of their antibacterial activity.
Investigations of pneumococcal polysaccharide vaccines for prevention of AOM identified a notable protective effect
if the child responded with sufficient antibody titers.20 However, because the polysaccharide antigens were less immunogenic in infants
under 2 years of age (the group with the highest incidence of AOM), a more potent vaccine is needed. Current investigations
of a conjugated polysaccharide pneumococcal vaccine indicate that infants as young as 2 to 6 months of age respond
with protective titers. Clinical trials of the new vaccine are now underway in the United States; the results should
be available in the fall of 1997.
Epidemiologic data indicate that the incidence of AOM may be reduced by encouraging breastfeeding, small-group
day care and the supine sleep position and by discouraging smoking in the household. Counseling would be of particular
importance in those families in which parents or siblings or both have a history of severe and recurrent disease.
- Teele DW, Klein JO, Rosner B, the Greater Boston Otitis
Media Study Group. J Infect Dis 1989;160:83-94.
- Etzel RA, Pattishall EN, Haley NJ, et al. Pediatrics
- Wald ER, Dashefsky B, Byers C, et al. J Pediatr 1988;112:540-546.
- Gannon MM, Haggard MP, Golding J, Fleming P. In: Abstracts
of the Sixth International Symposium on Recent Advances in Otitis Media. Fort Lauderdale, FL, June 4-8, 1996. p.
- Bluestone CD, Klein JO. In: Otitis Media in Infants
and Children. 2nd ed. Philadelphia, PA: W.B. Saunders. 1995;55-72.
- Orange M, Gray BM. Pediatr Infect Dis J 1993;12:244-246.
- Austrian R, Howle VM, Ploussard JH. Johns Hopkins Med
- Kamme C, Ageberg M, Lundgren K. Scand J Infect Dis
- Spika JS, Facklam RR, Plikaytis BD, Oxtoby MJ. J Infect
- Howe VM, Ploussard JH, Sloyer JL, Hall HC. Pediatrics
- Klein JO. Pediatr Infect Dis 1993;12:973-975.
- Block S, Hedrick J, Wright P, et al. MMWR 1994;43:23-25,31.
- Serchuck L, Pelton S, Pires V, et al. (Abstract) Submitted
to 36th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC)
- Bluestone CD, Stephenson JS, Martin LM. Pediatr Infect
Dis J 1992;11:S7-S11
- Baquero F, Loza E. Pediatr Infect Dis J 1994;13:S9-S14.
- Kaleida PH, Casselbrant ML, Rockette HE, et al. Pediatrics
- Hoppe JE, Koster S, Bootz F, Niethammer D. Infecton
1994;22:180-182 (Medzin Verlag GmbH Munchen, Munchen).
- Slapak L, Hornik P, Slapakova MD. Proceedings of the
Sixth International Symposium in Recent Advances in Otitis Media. Fort Lauderdale, FL. June 4-8, 1995.
- Farbman KS, Barnett ED, Bolduc GR, Klein JO. Pediatr
Infect Dis J 1993;12:613-614.
- Teele DW, Klein JO, the Greater Boston Collaborative
Otitis Media Group. Rev Infect Dis 1981;3(Suppl):S113-S118.