CHAPTER 194 Acute Otitis Media and Otitis Media with Effusion
Otitis media is not only a disease of modern times. It has been a major health problem in many societies. In the 8th century, the most common approach to healing was water from wells named after various saints.1 Otitis media is in general a childhood disease, and in most children it resolves with anatomic and physiologic changes with growth. Until the condition has resolved, it may affect balance, hearing, and speech and language development and cause poor school performance. This disease affects not only the child but the whole family, from an economic as well as a societal standpoint. It is more than the cost for medicine or visits to the doctors; passing sleepless nights due to a crying child, having to take off from work to stay home with the child, and taking the child to the doctors can be very stressful for the family. The cost for medical and surgical therapy for otitis media in children 5 years of age or younger is estimated at $5 billion annually in the United States.2
Diagnosis
Immittance Testing (Tympanometry)
Immittance testing is an excellent adjunct to the assessment of middle ear status and the management of otitis media. When otoscopic evaluation is inconclusive or difficult to perform, tympanometry can be very useful in evaluating ear disease in children older than 6 months of age. It is easy to perform and most often is well accepted by the patient. It has been used for school screening as well as in pediatricians’ offices. It also is valuable for documentation of middle ear status over time with repeat testing. A small probe, which emits a tone, is placed in the ear canal with an airtight seal. The tympanogram is obtained by plotting the immittance (acoustic energy of the reflected tone) of the middle ear as a function of the pressure in the external ear canal, which varies, ranging from −400 to +200 daPa (decapascals). The instrument provides measures such as peak compensated (static) admittance, tympanometric peak pressure, acoustic reflex, and tympanometric width (TW) (a measure of gradient). The following set of criteria uses TW to categorize middle ear status3:
A flat or rounded pattern (TW greater than 350 daPa) with a small ear canal volume indicates MEE, whereas a flat pattern with a large ear canal volume suggests a perforation or a patent tympanostomy tube. In a normal air-filled middle ear with equal pressure on both sides of the TM, the peak pressure of the tympanogram is 0 daPa. Algorithms have been developed to determine the presence or absence of MEE, some combining pneumatic otoscopy and tympanometry. Other methods of ascertaining middle ear status have been investigated, including spectral gradient acoustic reflectometry4,5 and ultrasound evaluation,6 but they have been found to have limitations, although they may be useful for screening.
Pathophysiology and Pathogenesis
The pathophysiology of otitis media is multifactorial, with various overlapping factors (Fig. 194-1).
Eustachian Tube Function
The eustachian tube in the infant is shorter, wider, and more horizontal than in the adult, which accounts for the high rate of otitis media in infants and children. By the age of 7 years, when the tube has a more adult configuration, the prevalence of otitis media is low.7 The three physiologic functions of the eustachian tube are (1) pressure regulation (ventilation), (2) protection, and (3) clearance (drainage). Of these three functions, pressure regulation is the most important8 (Fig. 194-2). Middle ear pressure is equilibrated to atmospheric pressure through active intermittent openings of the eustachian tube caused by contraction of the tensor veli palatini muscle during swallowing, jaw movements, and yawning. If active function is impaired, negative pressure will develop in the middle ear. Pressure regulation can be impaired by failure of the opening mechanism (functional obstruction) or anatomic (mechanical) obstruction. Using a pressure chamber, Bylander and associates evaluated eustachian tube function in children and adults who were considered otologically normal with intact TMs.9 Although only 5% of the adults were not able to equilibrate negative middle ear pressure, as many as 35.8% of the children could not equilibrate negative pressure. Children 3 to 6 years of age performed worse than children 7 to 12 years of age. These studies indicate that in even apparently otologically normal children eustachian tube function is worse than in adults, but it does improve with age. The improvement in the eustachian tube function parallels the decrease in the incidence of otitis media. Stenström and colleagues tested eustachian tube function in 50 otitis-prone children and 49 children with no history of AOM (control group) using a pressure chamber.10 The otitis-prone children had significantly poorer active tubal function than the normal control subjects, suggesting that recurrent AOM is the result of functional obstruction of the eustachian tube.
Infection
Bacteriology
Before the year 2000, studies in the United States reported that Streptococcus pneumoniae was the single most common bacterial pathogen in AOM, followed by Haemophilus influenzae and Moraxella catarrhalis; Streptococcus pyogenes and other miscellaneous bacteria accounted for just a few cases. From ears with chronic OME, H. influenzae was the single most common pathogen, with S. pneumoniae, M. catarrhalis, and other bacteria accounting for a small percentage of cases each. Data obtained during the 1980s from a single institution is shown in Figure 194-3.11 There was great hope that the introduction and subsequent use of the heptavalent conjugated pneumococcal vaccine (PCV7) for infants and young children, licensed in the United States in 2000, would lead to a substantial decrease in AOM. However, data from vaccine trials conducted in California12 and Finland13 revealed only a 7.8% and a 6% relative risk reduction, respectively, in AOM. Casey and Pichichero examined the experience of children in a suburban Rochester, New York, practice in the years 2001 to 2003 (after the introduction of PCV7) compared with the years 1995 to 2000 and reported a 24% decrease in persistent AOM and AOM treatment failures.14 These investigators also reported a decrease in recovery of S. pneumoniae from the ears of such children and an increase in H. influenzae. McEllistrem and coworkers looked at middle ear cultures obtained between 1999 and 2002 at five institutions.15 They reported that the percentage of AOM episodes due to non-PCV7 serogroups increased after 2000, from 12% in 1999 to 32% in 2002, and the nonvaccine serotypes were more frequent in those who received more than one dose of vaccine. The frequency of penicillin nonsusceptible strains in the vaccine and nonvaccine serotypes remained unchanged.
Nasopharyngeal carriage also was examined. Pelton and associates enrolled 275 infants and children 2 to 24 months of age in a surveillance study in which nasopharyngeal swabs or nasal washes were obtained at all well-child visits and at visits for AOM.16 From 2000 to 2003, these workers noted a decrease in vaccine serotypes with an increase in nonvaccine serotypes. Tracking antimicrobial susceptibility, they found that the MIC50 and the MIC90 for amoxicillin were stable until the 6-month period from October 2002 to March 2003, when they both increased significantly. Jacobs and associates examined the nasopharyngeal flora in children undergoing tympanostomy tube placement in March 2004 to March 2005 and also found a decrease in vaccine serotypes with a concomitant increase in nonvaccine serotypes, many of which were antibiotic-resistant strains.17
Biofilms
Bacterial biofilms are sessile communities of interacting bacteria attached to a surface and encased in a protective matrix of exopolysaccharides rather than living in a motile “planktonic” or free-floating state. The exopolysaccharide matrix provides protection from phagocytosis and other host defense mechanisms due to lack of accessibility by immunoglobulins and complement. The reduced metabolic rate of bacteria in the biofilm renders them resistant to antimicrobial treatment. The bacterial community relies on a complex intracellular communication system that provides for organized growth characteristics known as “quorum sensing.” Biofilms have been known to exist on hard surfaces such as metal pipes or teeth. However, recent animal and human studies have suggested that biofilms can also be isolated from the middle ear. Post and coworkers, using polymerase chain reaction (PCR) methodology, found evidence of bacteria in 48% of culture-negative MEE specimens from children undergoing tympanostomy tube insertion for chronic OME.18 Hall-Stoodley and colleagues obtained confocal laser scanning microscopy images from the middle ear mucosa of 26 children scheduled for tympanostomy tube insertion for recurrent or persistent otitis media and 8 control subjects (3 children and 5 adults) who underwent cochlear implantation.19 Generic stains and species-specific probes for H. influenzae, S. pneumoniae, and M. catarrhalis were used to evaluate biofilm morphology. Mucosal biofilms were visualized in 92% of the mucosa from children with chronic and recurrent otitis media but were not observed in any of the 8 control subjects. The investigators suggested that the findings supported the hypothesis that chronic middle ear disease is biofilm-related. Biofilms also have been identified in the nasopharynx of children with otitis media, and it was suggested that the biofilm may act as a reservoir for bacterial pathogens resistant to antibiotics. The mechanical débridement of the nasopharyngeal biofilms may explain the observed clinical benefit associated with adenoidectomy in subsets of pediatric patients.20
Viruses
Until the introduction of PCR, viruses were not considered a major factor in the etiology of otitis media, probably because of the technical difficulties in isolating the viruses. Using PCR techniques, however, it has been possible to identify respiratory syncytial virus (RSV), influenzavirus, adenovirus, parainfluenza virus, and rhinoviruses in MEE.21,22 There is strong evidence that viruses have a crucial role in the development of AOM.22 In a majority of children, viral infection of the upper respiratory tract mucosa initiates the whole cascade of events that finally leads to the development of AOM, and AOM may be regarded as a complication of a preceding or concomitant viral infection. In a study of 60 children (in 24 families) who were followed from October 2003 through April 2004 by daily parental recording of illness signs, weekly pneumatic otoscopy and viral PCR analysis of nasal secretions collected during “colds” or when MEE was noted, or from enrolled siblings without these conditions, one or more viruses were identified from 73% of secretions collected during a cold but only 18% collected at a time without a cold.23 Of 93 episodes of otitis media, 70% occurred during a cold, and nasal secretions from 77% of children at the time of these episodes were positive for virus.
Allergy and Immunology
Even though allergy is considered in the pathogenesis of otitis media, the causal mechanism is not understood, and well-controlled studies to prove the efficacy of anti-allergic medication in the treatment of OM are lacking. Several mechanisms by which allergy may cause otitis media have been suggested: (1) the middle ear is a “shock organ” (target); (2) allergy may induce inflammatory swelling of the eustachian tube mucosa; (3) allergies produce inflammatory obstruction of the nose; and (4) bacteria-laden allergic nasopharyngeal secretions may be aspirated into the middle ear. The last three mechanisms involve an association between allergy and abnormal eustachian tube function. Prospective studies have reported a relationship between upper respiratory tract allergy and eustachian tube obstruction in a series of provocative, intranasal allergen–inhalation challenge studies.24
Recent studies focus on the analysis of the inflammatory markers in MEE to determine the role of allergy. Both TH1 and TH2 inflammatory patterns in helper T cells have been found in OME. TH1 cytokines antagonize allergic inflammation and play a key part in the defense against viruses and intracellular pathogens, whereas TH2 cytokines promote IgE production, eosinophil growth, and mucus production.25 The chemoattractant cytokine RANTES (regulated upon activation, normal T cell–expressed and secreted) and eosinophilic cationic protein (ECP), signs of a TH2 process, were assessed in MEE samples from 25 children with allergy and 20 nonallergic children who were 5 to 11 years old at the time of tympanostomy tube insertion.26 RANTES and ECP levels were significantly higher in the MEE samples from children with allergy than in children without allergy (P < 0.01 and < 0.05, respectively). A positive correlation also was found between RANTES and ECP in MEE (P < 0.01) in children with allergy, which suggests that allergy is a contributing factor in the pathogenesis of otitis media.
Children with major immune deficiencies may have recurrent otitis media as part of their overall clinical picture, but most otitis-prone children may have only a subtle immunologic abnormality that predisposes them to recurrent infections.27 The three major pathogens of otitis media—S. pneumoniae, H. influenzae, and M. catarrhalis—frequently colonize the nasopharynx. There are many strains of these organisms, and among the different strains are heterologous surface (strain specific) antigens and conserved antigens. Conserved antigens induce broadly protective antibodies while strain-specific antigens induce limited protection. Otitis-prone children may display strain-specific immunity but fail to develop a broadly protective antibody response, which makes them susceptible to recurrent and persistent otitis media.28
Recurrence of bilateral OME after tympanostomy tube insertion was more likely in children with poor eustachian tube function and low IgA or low IgG2 and decreased levels of mannose-binding lectin than in children with poor tubal function and high IgA and IgG2 measured at the time of tube insertion.29 This suggests that an interaction between various immune factors may increase the risk for the development of otitis media.
Gastroesophageal Reflux
In 2002 Tasker and associates reported finding pepsin/pepsinogen, using enzyme-linked immunosorbent assays, in 90.8% of 65 MEE samples obtained at the time of myringotomy in children.30 The pepsin-pepsinogen levels ranged from 0.8 to 213.9 µg/mL, which is 1000 times higher than serum levels. He and colleagues, using a sensitive and specific pepsin assay, detected pepsin activity in 14.4% of 152 subjects undergoing tympanostomy tube insertion.31 The levels of pepsin ranged from 13 to 687 ng/mL. Crapko and associates also identified pepsin activity in the middle ear fluid in 60% of 20 subjects also undergoing tympanostomy tube insertion.32 In this group of children the pepsin levels ranged from 80 to 1000 ng/mL. The latter studies have shown that pepsin is identified less frequently and at lower levels of pepsin, which may be due to differences in methodology. However, it may be concluded from these studies that gastroesophageal reflux may be a causative factor in otitis media, with a potential role for antireflux therapy in the treatment of otitis media in some children, but adequate controlled trials have not been done.
Epidemiology
Incidence and Prevalence
Otitis media is considered a worldwide pediatric health care problem. The Early Childhood Study—Birth Cohort (ECLS—B), a nationally representative longitudinal study of more than 8000 children born in 2001, showed that otitis media was diagnosed in 39% of children by 9 months and 62% of children by 2 years of age.33 Although the highest incidence of otitis media is in young children, it also occurs in older children, adolescents, and adults. The Oslo Birth Cohort enrolling infants born in a 15-month period in 1992 to 1993 found that 13% of 10-year-olds had at least one episode of otitis media in the previous 12 months.34 Approximately 3% to 15% of patients with otitis media referred to otolaryngology clinics are adults.35
Acute Otitis Media
Most children experience at least one episode of AOM during their childhood. The cumulative incidence of the first episode of AOM based on studies from various countries ranges from 19% to 62% by age 1 year and 50% to 84% by 3 years of age36–42 (Fig. 194-4). In a majority of these studies, the peak incidence of AOM was during the first 6 to 12 months of life.39 The incidence decreases with age, and by the age of 7 years, few children experience episodes of AOM. Recurrent episodes of AOM are common in young children. By 6 months of age, 20% have had two or more episodes.43 Occurrence of three or more episodes of AOM by the ages of 1, 3, 5, and 7 years has been documented in 10% to 19%, 50%, 65%, and 75% of children, respectively.39 Occurrence of six or more episodes of AOM was documented in 39% of children by the age of 7 years.
Otitis Media with Effusion
It may be difficult to determine the “true” incidence of OME because, by definition, OME is asymptomatic. Furthermore, most screening studies determine the presence of MEE without differentiating between AOM and OME. Also, a short time between observations is needed for precise assessment of the onset and time to resolution of each new episode of OME, because approximately 65% of OME episodes in children 2 to 7 years of age resolve within 1 month.44 Monthly examinations with otoscopy and tympanometry in 2- to 6-year-old children in a day care center in Pittsburgh revealed MEE at least once in 53% to 61% of the children. Lous and Fiellau-Nikolajsen found an incidence of MEE of 26% in 7-year-old children followed monthly for 1 year using tympanometry.45
The point prevalence of MEE from different countries shows a wide variation, depending on the age of the child, season of the year, and type of assessment. Thus, it must be emphasized that comparison of outcomes between studies will require careful evaluation of study methodology, as well as caution in drawing conclusions. However, nearly all children had experienced at least one episode by the age of 3 years.46
Trend over Time
Even though otitis media remains a major problem in the United States, several recent studies point to a trend toward reduction in the burden of otitis media. Grijalva and associates compared the national rates of outpatient visits for pneumonia and otitis media in children in 1994 to 1999 to those in 2002 to 2003 using the National Ambulatory Medical Care Survey and the National Hospital Ambulatory Medical Care Survey.47 The results showed that otitis media visit rates declined by 20% in children younger than 2 years of age while there was no significant decrease in outpatient visit rates for pneumonia or other respiratory infections in children in this same age group. The second study compared the risk of developing frequent otitis media or having tympanostomy tubes inserted in four birth cohorts (1998 to 1999, 1999 to 2000, 2000 to 2001, and 2001 to 2002).48 The study population included all children at birth in TennCare or selected upstate New York commercial insurance plans as of July 1998 and followed them for 5 years. When comparing the 2000-2001 cohort to the 1998-1999 cohort frequent otitis media declined by 17% and 28% and tympanostomy tube insertion declined by 16% and 23% for Tennessee and New York cohorts, respectively. The decline may be the result of infant immunization, catch-up doses in older children, and herd immunity after introduction of the heptavalent pneumococcal conjugate vaccine in the United States in 2000. In a comparison of the 2000 to 2001 and 2001 to 2002 birth cohorts, frequent otitis media and tympanostomy tubes remained stable in New York but increased in Tennessee. The investigators attributed this to several possible factors including increase in disease due to nonvaccine serotypes and change in medical care utilization.
A decline in the incidence of otitis media also has been reported from several European countries. Kvaerner conducted a study in Norway of hospital admissions for AOM from 1999 to 2005 among children aged 7 years or younger.49 Hospitalization for AOM was less frequent but the incidence of acute mastoiditis remained stable at approximately 6 per 100,000 children over time. The consultation rates to the general physicians in the Netherlands for otitis media during 1995 to 2003 in children from birth to 13 years of age were assessed using the research database of the Netherlands University Medical Center Utrecht Primary Care Network.50 The overall consultation rates decreased by 9% for AOM and 34% for OME. In children aged 2 to 6 years and those aged 6 to 13 years, the rate of AOM declined by 15% and 40%, respectively; the rates of OME declined by 41% and 48%, respectively. In children younger than 2 years of age, however, the rates of AOM and OME increased by 46% and 66%.
Data from the General Practitioner Database in the United Kingdom were analyzed for events classified as AOM or “glue ear” in the years 1991 to 2001.51 Total consultations for AOM and glue ear in children aged 2 to 13 years had changed from 105.3 to 34.7 and 15.2 to 16.7 per 1000 per year, respectively.
Risk Factors
Host-Related Factors
Age
The highest incidence of AOM is between 6 and 11 months of age,39 and onset of the first episode of AOM before 6 months39 or 12 months of age is a powerful predictor of recurrence.52 The risk of persistent MEE after an episode of AOM is inversely correlated with age,39 and children who experience their first episode of MEE before 2 months of age are at a higher risk for persistent fluid during their first year of life than are children who have their first episode later.53
Sex
Most investigators have reported no apparent sex-based difference in the incidence of OME.46 Paradise and coworkers, in a study involving more than 2000 infants, found males had more time with MEE, although the difference was not large.54 Some studies have found a significantly higher incidence of AOM in males as well as more recurrent episodes than in females, but others have not found this.46
Race
Earlier studies have suggested a lower incidence of otitis media in African-American children than in white children. More recent studies in which the children were followed prospectively from infancy to age 2 years with routine examinations of the ears every 6 weeks showed no difference between the African-American and white children in their experience with otitis media.54,55 A third study, part of an American population-based sample survey evaluating schoolchildren 6 to 10 years of age with tympanometry, reported no difference between African-American and white children. However, the prevalence of OME was significantly higher in Hispanic children compared with white children.56 In another study of more than 11,000 infants followed only from the ages of 1 through 6 months at two urban and one suburban location in the United States, non-Hispanic white children had significantly higher rates of otitis media after multivariant adjustment than non-Hispanic black or Asian children. There was no difference between Hispanic and non-Hispanic children.57
Prematurity
Some studies have shown a possible association between low birth weight and prematurity and otitis media, but others have not. In many of these studies, however, the sample sizes have been relatively small. A recent study analyzing data from more than 10,000 births showed some relationship between very low birth weight (less than 1500 g) and frequent otitis media, but no increased risk was found for recurrent otitis media and moderately low birth weight (1500 to 2499 g). In a prospective cohort study of 136 children treated with tympanostomy tubes, a positive but nonsignificant association was documented between low birth weight or low gestational age or history of incubator care and recurrent OME.58
Allergy
There is still controversy regarding the role of allergy in the pathogenesis of otitis media. Allergy is a common problem in children, occurring at a time when respiratory infections and otitis media are common. Most, but not all, of the epidemiologic studies have supported an association. Pukander and Karma followed 707 children with AOM and found that in children with atopic manifestations, persistence of MEE for 2 months or longer was more common than in children without allergy.59 In another study, however, atopic diathesis was not found to predispose affected children to the development of AOM.60 The Japanese investigators Tomonaga and associates determined that allergic rhinitis was present in 50% of 259 patients (mean age of 6 years) in whom OME had been diagnosed, and OME was present in 21% of 605 patients (mean age of 9 years) in whom allergic rhinitis had been diagnosed.61 Among 108 children 5 to 8 years of age (mean age of 6 years) in whom neither condition had been diagnosed, the occurrence rates for allergic rhinitis, OME, and both of these conditions were 17%, 6%, and 2%, respectively. Another investigative group (Kraemer and coworkers) compared risk factors for persistent MEE among 76 children hospitalized for bilateral myringotomy with tube insertion (M&T) and 76 control subjects matched by age, gender, and season of admission for a general surgical procedure.62 Results showed a nearly fourfold increase in the risk of persistent MEE in children who had atopic symptoms for more than 15 days per month. In another study, Bernstein and Reisman determined the allergy status of a group of 200 children who had undergone one or more M&T procedures.63 These workers diagnosed allergy in 23% of the entire group but found that the frequency was 35% among the 88 children with multiple myringotomies with tube insertions. The frequency of allergy in these patients was higher than that reported for children of similar age in the general population. The foregoing studies generally found a higher frequency of OME in allergic children than in age-matched nonallergic children, as well as higher frequencies of allergy in children with OME than in children without OME.
Immunocompetence
Children with recurrent otitis media as well as other recurrent infections may have a defect in the immune system such as a defect in phagocyte function, humoral immunity, local immunity, or other immune defects.8 Children infected with the human immunodeficiency virus (HIV) demonstrate a significantly higher recurrence rate than in normal children or children who have seroconverted.64
Cleft Palate/Craniofacial Abnormality
Otitis media is considered “universal” in infants younger than 2 years of age with unrepaired cleft palate.65 After surgical repair of the palate, the occurrence of otitis media is reduced, probably because of improvement in eustachian tube function. However, many children continue to have problems up into the teens. Otitis media also is common in children with other craniofacial abnormalities or Down syndrome, also due to anatomic or functional eustachian tube abnormalities.66 In children with Down syndrome, low resistance of the tube, in addition to poor active eustachian tube function, predisposes them to reflux of nasal secretions into the middle ear.67
Genetic Predisposition
Twin and triplet studies have been used to assess the heritability of otitis media. Two retrospective studies using questionnaires have been published. The first study, using 2750 Norwegian twin pairs, estimated the heritability at 0.74 in females and 0.45 in males.68 In the second study, the estimated heritability was, on average, 0.57 at the ages of 2, 3, and 4 years.69 In a prospective twin-triplet study from Pittsburgh with monthly assessments of middle ear status, the heritability estimate for otitis media at age 2 years was 0.79 in females and 0.64 in males.70 Heritability is a population statistic used to ascertain if a trait is heritable, and linkage and association studies may identify genetic regions or specific genes influencing the particular trait or disease. Daly and coworkers, using genome-wide linkage analysis, have suggested that chromosomal regions on 19q and 10q contain genes contributing to the susceptibility to chronic OME or recurrent AOM.71 Subsequent fine mapping of both regions further strengthened the evidence for this linkage.72 In addition, evidence for an association between polymorphisms in FBXO11, the human homolog of a mouse model gene for chronic otitis media, and chronic OME/recurrent AOM (COME/ROM) has been reported.73
Environmental Factors
A study was conducted to assess the variation in environmental risk factors for otitis media across Western countries, including European countries, the United States, Canada, and Australia.74 The main risk factors for otitis media were day care atendance, number of siblings, tobacco smoke exposure, breastfeeding, birth weight, socioeconomic status, and air pollution. However, the results indicated large variations in rates across the various countries: day care at ages 1 to 3 years: Sweden 75% versus Italy 6%; breastfed at 6 months: Norway 80% versus Poland 6%; and women smoking: Germany, France, and Norway 30% to 40% versus Portugal less than 10%.
Upper Respiratory Infection/Seasonality
Both epidemiologic evidence and clinical experience strongly suggest that otitis media frequently is a complication of an upper respiratory infection (URI). The incidence of AOM is highest during the fall and winter months and lowest during spring and summer months, which parallels the incidence of URI.44 This correlation supports the hypothesis that an episode of URI plays an important role in the etiology of otitis media. Rhinovirus, RSV, adenovirus, and coronavirus have been detected in MEE during episodes of AOM.21,75 Upper respiratory tract infections with RSV, influenzavirus, and adenovirus often precede an episode of AOM.22 In a prospective study of Finnish children, 54% of ensuing AOM episodes were diagnosed by day 4 after the onset of URI symptoms.76
Day Care/Home Care/Siblings
Almost universally, day care center attendance remains a very important risk factor for development of otitis media. For example, prevalence of high negative ME pressure and flat tympanograms (type B), indicative of MEE, have been shown to be highest in children cared for in day care centers with many children, intermediate in children in family day care with fewer children, and lowest in children cared for at home.77 Children attending day care centers also have been shown to be more likely to have tympanostomy tubes inserted than children cared for at home.78
Birth order has been shown to be associated with episodes of otitis media and percentage of time with MEE.55 First-born children had a lower rate of episodes of AOM and less time with MEE during the first 2 years of life than children with older siblings. Also, having more than one sibling was significantly related to early onset of otitis media.59
Tobacco Smoke Exposure
An association between otitis media and passive exposure to smoking has been reported by many investigators while others have not been able to demonstrate such an association. In most studies the information regarding smoke exposure has been obtained from the parents. A few recent studies have measured cotinine, a metabolite of nicotine, in blood, urine, or saliva of the child and have been able to more accurately determine the association between otitis media and smoke exposure.43,79 More information about the pathogenesis and duration and intensity of exposure is needed to clarify this association. Two recent studies have shown an association between middle ear status and parental smoking. In children exposed to parental smoking, tympanostomy tubes stayed in place for 59 weeks, compared with 86 weeks in children who were not exposed to smoking.80 Also, myringosclerosis was more prevalent in children with two smoking parents compared with those with no smoking parents (64% versus 20%) and maternal smoking was associated with a highly increased risk of recurrent otitis media after insertion of tympanostomy tubes.81
Breastfeeding/Bottle Feeding
Currently, most national and international authorities, including the American Academy of Pediatrics and the American Academy of Family Physicians, the World Health Organization (WHO), and the United Nations Children’s Fund, recommend 6 months of exclusive breastfeeding. In 2004, Kramer and Kakuma published a comprehensive review of the world literature to determine health benefits of exclusive breastfeeding for 6 months compared with 3 to 4 months.82 They reported a decrease in the risk for gastrointestinal infection even in developed areas. It had not been shown previously that exclusive breastfeeding for 6 months or longer compared with 4 to less than 6 months in the United States provided greater protection against respiratory infections. Therefore, a secondary analysis of data from the National Health and Nutrition Examination Survey III, a population-based cross-sectional home survey conducted from 1988 to 1994, was undertaken.83 After adjustment for demographic variables, child care, and smoke exposure, the data revealed statistically significant increased risk for both pneumonia (odds ratio [OR]: 4.27; 95% confidence interval [CI]: 1.27-14.35) and three or more episodes of otitis media (OR: 1.95; 95% CI: 1.06-3.59) in those children breastfed 4 to less than 6 months. These findings further support the current recommendations that infants receive breast milk exclusively during the first 6 months of life.
Socioeconomic Status
Socioeconomic status and access to health care are factors that may affect the incidence of otitis media. It has been generally thought that otitis media is more common among persons in the lower socioeconomic strata as a result of poor sanitary conditions and crowding. Paradise and colleagues followed 2253 infants for 2 years in the United States and found an inverse relationship between the cumulative proportion of days with MEE and socioeconomic status.84 However, many studies have not revealed any correlation between socioeconomic status of the child’s family and the incidence of MEE.
Pacifier Use
Niemelä and associates reported that pacifier use increased the annual incidence of AOM and calculated that pacifier use was responsible for 25% of AOM episodes in children younger than 3 years.85 They reported the results of an intervention trial in which parents in various well-baby clinics were taught that pacifier use was harmful and should be limited, and parents in other clinics were not provided with this information. This led to a 29% decrease in AOM in the group provided with pacifier information. Pacifier use has been theorized to contribute to the development of otitis media, possibly due to the sucking action of the child propelling nasopharyngeal secretions into the middle ear or by the pacifier acting as a fomite. However, Brook and Gober cultured the surface of pacifiers from children with AOM but found no pathogens typical of AOM.86 The role of pacifier use in AOM remains unclear.
Obesity
Recent studies have indicated a possible association between otitis media and body mass index (BMI). Tympanostomy tube insertion and overweight were studied in two cohorts of children followed from birth to 2 years of age.87 Weight-for-length and BMI-for-age were calculated using well-child visit data. In the cohort of predominantly white children, a history of tympanostomy tube insertion significantly predicted risk of BMI index beyond the 85th percentile at the age of 2 years. In the other cohort of American Indian children, a history of recurrent AOM increased the likelihood of weight-for-length index beyond the 95th percentile. Further studies are needed to clarify the association.
Prevention of Disease
Vaccines
Bacterial Vaccines
Streptococcus pneumoniae Vaccine
Pneumovax is a 23-valent polysaccharide vaccine that is not efficacious in children 2 years or younger because of poor antigen production in this age group. Prevnar is a conjugated vaccine in which pneumococcal polysaccharides are conjugated to a nontoxic mutant of diphtheria toxin. The 7-valent vaccine (PCV7), which includes serotypes 4, 6B, 9V, 14,18C, 19F, and 23F, was licensed for use in the United States in 2000 and recommended for use in children younger than 6 years of age by the American Academy of Pediatrics.88 Infants are immunized at the ages of 2, 4, 6, and 12 to 15 months. For previously unimmunized children 7 to 23 months of age, a reduced number of doses is recommended, and for children 24 to 59 months of age and children at high risk at any age, two doses of PCV7 and one dose of the 23-valent vaccine are recommended.
The efficacy for AOM of the conjugated vaccine has been evaluated in several randomized clinical trials. The trial from Northern California Kaiser Permanente enrolling 37,868 children randomly assigned to either pneumococcal conjugate vaccine or a meningococcal type C conjugate vaccine and followed prospectively, reported an overall reduction of episodes of otitis media of 7%.12 However, children with recurrent otitis media benefited from the vaccine with a reduction in otitis media episodes ranging from 9.3% to 22.8%, increasing as the frequency of episodes of AOM increased. In addition, immunized children were 20.1% less likely to require tube placement than control subjects. The PCV7 vaccine demonstrated a much higher efficacy (greater than 80%) in reducing invasive pneumococcal disease than in reducing the burden of otitis media. Additional follow-up of the study subjects continued to demonstrate a modest amount of protection against episodes of AOM and need for tympanostomy tube insertion.89 In a Finnish study of 1662 children randomized to receive either PCV7 vaccine or a hepatitis B vaccine, with follow-up to the age of 24 months, a bacterial diagnosis was made based on a culture of the middle ear fluid, in addition to the clinical assessment of episodes of AOM.13 The vaccine reduced the number of AOM episodes from any cause by 6%, the number of culture-confirmed pneumococcal episodes by 34%, and the number of episodes due to the vaccine serotypes by 57%. The number of episodes attributed to serotypes that were cross-reactive with those in the vaccine was reduced by 51%, whereas the number of episodes due to all other serotypes increased by 33%. The modest response in reducing number of episodes of AOM was due to a decrease in vaccine efficacy against serotype-specific infections as well as an increase in AOM episodes due to nonvaccine serotypes. A second trial was conducted in Finland using a 7-valent pneumococcal polysaccharide–meningococcal outer membrane protein complex conjugate vaccine (PncOMPC) in 1666 children who were followed for 24 months.90 This trial showed no reduction in overall number of AOM episodes. The reduction in number of AOM episodes due to any culture-confirmed pneumococci was 25% and to vaccine serotype pneumococci, 56%—results similar to those with the PCV7. This vaccine failed to show protection against cross-reactive serotypes. An increased rate of carriage of nonvaccine serotypes was noted among infants vaccinated with the pneumococcal conjugate vaccine.
With use of the vaccine, pneumococcal colonization in the nasopharynx has changed as the vaccine has caused a reduction in the carriage rate of vaccine serotypes and replacement with nonvaccine serotypes.17,91 It should be noted that a reduction in nasopharyngeal carriage of vaccine serotypes and an increase in nonvaccine serotypes has also been seen in children younger than 7 months of age.16
In the United States and Finland, the initial immunization with PCV7 is at age 2 months based on the results from the clinical trials. However, the question of whether or not the results obtained in the Finnish and U.S. trials could be extrapolated to older children who already had a history of recurrent episodes of AOM was addressed by Dutch and Belgian trials.92,93 Seventy-eight and 383 children with documented histories of recurrent AOM were enrolled in the Belgian and Dutch studies, respectively. In both trials children between 1 and 7 years old were immunized at entry with PCV7 followed by a booster immunization with a 23-valent pneumococcal polysaccharide vaccine 7 months later and followed up for a total of 18 months in the Dutch trial and 26 months in the Belgian trial. The results from these two studies did not lend any support to the use of pneumococcal conjugate vaccine to prevent AOM in previously unvaccinated toddlers and children with a history of recurrent AOM. The differences in outcome between the studies in older children and those in healthy infants may be caused by differences in age at immunization. In the infant studies the PCV7 immunization may prevent or delay nasopharyngeal colonization of the most frequent pneumococcal serotypes and consequently delay pneumococcal episodes of AOM until a later age. In older children, who already are carriers of pneumococci, there is rapid replacement of pneumococcal vaccine serotypes with nonvaccine serotypes.
An 11-valent polysaccharide pneumococcal serotype vaccine conjugated to H. influenzae–derived protein D was evaluated in a clinical trial conducted in the Czech Republic enrolling a total of 4968 children, who were followed to the end of the second year of life.94 Middle ear fluid was cultured when episodes of AOM were diagnosed. The vaccine reduced the overall incidence of AOM episodes by 33.6%. Vaccine efficacy was 57.6% for episodes of AOM caused by pneumococcal serotypes included in the vaccine and 66% for vaccine-related cross-reactive pneumococcal serotypes. There was no significant change in number of AOM episodes due to other nonvaccine serotypes. Of great interest is that there was 35.3% efficacy against AOM episodes caused by nontypeable H. influenzae. These results indicate that using the H. influenzae–derived protein D as a protein for conjugation of pneumococcal polysaccharides not only allowed for protection against pneumococcal otitis, but also against otitis media due to nontypeable H. influenzae. The use of the H. influenzae–derived protein D and the two additional pneumococcal serotypes probably explains the 33.6% reduction in numbers of all types of AOM episodes by the vaccine, which is significantly better than in the previously mentioned trials.12,13
