Bacterial Infections of the Respiratory Tract

Pharyngitis, Retropharyngeal Cellulitis, and Retropharyngeal Abscess

Neonates with bacterial infection of the oropharynx may present with pharyngeal inflammation with or without exudate or with retropharyngeal cellulitis or abscess. Clinical signs and symptoms include respiratory distress, poor feeding, and irritability. Infants can have submandibular swelling, and some may have a weak or hoarse cry. Infection may extend to the surrounding structures, leading to formation of deep neck abscess. Microorganisms identified as the etiologic agents of these infections and their manifestations of disease include the following:

• Staphylococcus aureus: Although many children are colonized in the throat and nasopharynx with S. aureus, this organism is rarely a primary agent in the etiology of pharyngitis in infants (or adults). There have been reports, however, of localized abscesses in the oral cavity related to S. aureus. In 1936, Clark and Barysh reported a case of retropharyngeal abscess in a 6-week-old infant. The infant was critically ill but recovered after incision and drainage of the abscess. A 6-day-old infant described in a report from India presented with stridor, dysphagia, and lateral cervical swelling. The infant was found to have a retropharyngeal abscess caused by S. aureus. Steinhauer reported a case of cellulitis of the floor of the mouth (Ludwig angina) in a 12-day-old infant. The infant was febrile and toxic; examination of the mouth revealed swelling under the tongue. Purulent material was subsequently drained from this lesion, and S. aureus was isolated from the pus. A laceration was noted in the floor of the mouth, and the author considered this wound to be the portal of entry of the infection. Increased incidence of methicillin-resistant S. aureus (MRSA) has been reflected in three reports of retropharyngeal abscesses caused by MRSA in infants 3 weeks to 3 months of age. The first case involved a 2-month-old infant in Japan (who also had evidence of penicillin-resistant Pneumococcus ) ; the second was of a 3-week-old infant from Romania ; and the third report involved 3- and 4-month-old infants in the United States, one of whom had extension of the abscess into the mediastinum accompanied by venous thrombosis.

• Group A Streptococcus: Fever and pharyngeal inflammation may result from infection with group A Streptococcus (GAS) in the neonate.

• Streptococcus agalactiae: Retropharyngeal cellulitis has been associated with bacteremia caused by group B streptococci (GBS). Affected neonates presented with poor feeding, noisy breathing, persistent crying, irritability, and widening of the retropharyngeal space on radiographs of the lateral neck. Stridor also may be associated with retropharyngeal abscess, as reported in a 13-day-old infant in Hong Kong. A retropharyngeal abscess caused by GBS occurred in one of three neonates reported in a series of 31 cases of retropharyngeal abscess in Australian children during 1954 to 1990. This infant was found to have a third branchial arch pouch that was subject to recurrent infection until age 5 years.

• Listeria monocytogenes: Small focal granulomas on the mucous membrane of the posterior pharynx have been observed in neonates with L. monocytogenes infection. Necrosis of the granulomas can result in ulcers on the pharynx and tonsils.

• Treponema pallidum: Mucous patches occur on the lips, tongue, and palate of infants with congenital T. pallidum infection. Rhinitis may appear after the first week of life.

• Neisseria gonorrhoeae: A yellow mucoid exudate of the pharynx may be present simultaneously with ophthalmia with N. gonorrhoeae infection (A. Yu, personal communication, 1981). A case report of in utero gonococcal infection with involvement of multiple tissues included pharyngeal abscess.

• Enterococcus faecalis: In a case of retropharyngeal abscess, E. faecalis was isolated from a culture of aspirated pus, and two strains of coagulase-negative staphylococci occurred in a 2-week-old full-term infant from Australia. The infant was severely ill and had atlantoaxial dislocation resulting in paraplegia. At autopsy, the findings included bacterial endocarditis, diffuse bilateral pneumonia, and renal infarcts.

• Escherichia coli: E. coli can be a rare cause of infection of the pharyngeal cavity. Two strains of E. coli were isolated from pus from a retropharyngeal abscess in a 1-week-old infant. The infant was afebrile on presentation and had significant midline pharyngeal swelling.

Infants may have coryza and other signs of upper respiratory tract disease secondary to infection with respiratory viruses. Infections with respiratory viruses may damage the respiratory mucosa, increasing susceptibility to bacterial infection of the respiratory tract. Eichenwald described an apparent synergy of respiratory viruses and staphylococci that produced an upper respiratory tract infection called the “stuffy nose syndrome.” The syndrome occurred only when both organisms were present. Eichenwald and coworkers also documented increased dissemination of bacteria by newborns carrying staphylococci and echovirus 20 or adenovirus type 2 in the nasopharynx and coined the term “cloud babies” for these infants. Because these studies have not been repeated by other investigators, the significance of synergy of two or more microorganisms in neonatal respiratory infections remains uncertain.

Noma

Noma (cancrum oris) is a destructive gangrenous process that may affect the nose, lips, and mouth. It occurs almost exclusively in malnourished children in developing countries; nutrient deficiencies have been postulated to play a role in its pathogenesis. Although it is usually a chronic, destructive process in older children, in neonates it may be rapidly fatal. Affected neonates are usually premature and of low birth weight. In older children and adults, noma is caused by fusospirochetes such as Fusobacterium necrophorum. The disease in neonates is usually due to Pseudomonas aeruginosa . Ghosal and coworkers reported bacteriologic and histologic findings in 35 cases of noma in neonates in Calcutta. P. aeruginosa was isolated from blood or the gangrenous area in greater than 90% of the cases. In Israel, a full-term infant with bilateral choanal atresia who required an airway developed gangrenous lesions of the cheek on day 11 and palatal lesions that progressed to ulceration and development of an oronasal fistula. Cultures of material from the lesions grew P. aeruginosa . Freeman and associates reported the development of noma neonatorum in the third week of life in a 26-week-gestation premature infant; they suggested that this entity represents a neonatal form of ecthyma gangrenosum. Human immunodeficiency virus (HIV) testing has been recommended for infants in whom noma is diagnosed because of the difficulty distinguishing between early signs of noma and necrotizing diseases of the oral cavity associated with HIV infection.

Epiglottitis

Epiglottitis caused by S. aureus in an 8-day-old infant was reported by Baxter in a survey of experience with the disease at Montreal Children’s Hospital from 1951 to 1965. Rosenfeld and associates reported a case of epiglottitis caused by S. aureus in a 5-day-old infant. The infant presented with bradycardia, hoarseness, and inspiratory stridor and had diffuse inflammation of the arytenoids and epiglottis. S. aureus was cultured from pus on the epiglottic surface; blood culture was negative. Epiglottitis secondary to Streptococcus sanguinis in a newborn infant and secondary to GBS in an 11-week-old infant have also been reported.

Laryngitis

Laryngitis in the newborn is rare. A newborn infant with congenital syphilis may have laryngitis and an aphonic cry. Hazard and coworkers described a case of laryngitis caused by Streptococcus pneumoniae . A term infant was noted at 12 hours to have a hoarse cry, which progressed to aphonia during the next 3 days. Direct examination of the larynx revealed swelling and redness of the vocal cords. The infant was febrile (38.5° C), but the physical examination was unremarkable. S. pneumoniae was isolated from the amniotic fluid, the maternal cervix, and the larynx of the infant. The infant responded rapidly to treatment with parenteral penicillin G. Laryngitis may also be a sign of respiratory papillomatosis, usually associated with human papillomavirus. It has also been linked with acid reflux.

Infection of the Paranasal Sinuses

The paranasal sinuses of the fetus begin to differentiate at about the fourth month of gestation. The sinuses develop by local evagination of nasal mucosa and concurrent resorption of overlying bone. The maxillary and ethmoid sinuses are developed at birth and may be sites for suppurative infection. The sphenoid and frontal sinuses are rudimentary at birth and are not well defined until about 6 years of age.

Inflammation may occur simultaneously in the paranasal sinuses, the middle ears, and the lungs. Autopsy may reveal that purulent exudate and leukocytic infiltration of the mucosa are present at one or more of these sites. Infection of the ethmoid and maxillary sinuses may be severe and life threatening in the newborn. Clinical manifestations include general signs of infection, such as fever, lethargy, irritability, and poor feeding, and focal signs indicative of sinus involvement (i.e., nasal congestion, purulent drainage from the nostrils, and periorbital redness and swelling). Proptosis may occur in severely affected children. Although any of the organisms responsible for neonatal sepsis may cause sinusitis, S. aureus, GAS, and GBS are responsible for most infections. Suppurative infection of the maxillary sinus may progress to osteomyelitis of the superior maxilla (see Chapter 8 ).

Blood specimens, nasopharyngeal secretions, and purulent drainage material (if present) should be obtained for culture before treatment. Antibacterial therapy must include a penicillinase-resistant penicillin or cephalosporin for activity against S. aureus , GAS, and GBS. If no material is available for examination of Gram-stained pus or if results of the preparation are ambiguous, initial therapy should include an aminoglycoside or a third-generation cephalosporin to ensure activity against gram-negative enteric bacilli (see discussion of management in Chapter 37 ). Surgical drainage of the infected site should be considered. Drainage of the suppurative maxillary sinus should be performed through the nose to avoid scars on the face and damage to the developing teeth.

Diphtheria

Neonatal diphtheria, although now extremely rare in the United States, was common before widespread immunization with diphtheria toxoid. Outbreaks occurred in hospital nurseries. One of the most striking reports described three separate epidemics in a “foundling hospital” in Tipperary, Ireland, during 1937 to 1941; 36 infants younger than 1 month were affected, and 26 died. Goebel and Stroder described 109 infants younger than 1 year with diphtheria in Germany during the period from the fall of 1945 to the summer of 1947: 59 infants were younger than 1 month, and 26 died. In a report from the Communicable Disease Unit of the Los Angeles County Hospital covering a 10-year period ending June 1950, 1433 patients were admitted to the hospital with diphtheria; 19 patients were younger than 1 year, but only 2 patients were younger than 1 month. Diphtheria incidence continues to decline globally ; Mathur and associates identified only three cases of neonatal diphtheria in India from 1974 to 1984.

Respiratory diphtheria has been well controlled in the United States since the introduction of diphtheria toxoid in the 1920s, although it remained endemic in some states through the 1970s. The results of a survey of cases of diphtheria reported to the Centers for Disease Control and Prevention (CDC) of the U.S. Public Health Service, Atlanta, Georgia, for the period 1971 to October 1975 showed that no cases involved children younger than 1 month and that only six cases occurred in children younger than 1 year (the youngest was 5 months old) (G. Filice, personal communication, 1981). During the period 1980 to 1995, 41 cases of respiratory diphtheria were reported to the CDC; 4 (10%) were fatal, all of which occurred in unvaccinated children. A single case of diphtheria has been reported in the United States since 2003. Most cases of diphtheria in high-income countries are imported from countries where diphtheria remains endemic or where there is a resurgence of disease. Reemergence of diphtheria in the newly independent states of the former Soviet Union underscored the need to maintain control measures in the United States, including universal childhood immunization, adult boosters, and maintenance of surveillance activities. Maternal immunization may provide some protection to infants in the neonatal period before diphtheria vaccine is given.

A newborn receives antibodies to Corynebacterium diphtheriae from the mother if she is immune, and the titers of mother and infant at birth are approximately equivalent. Some degree of protection results in the neonate from this passively transferred antibody. Serologic surveys performed in the United States in the 1970s and 1980s suggested that 20% to 60% of adults older than 20 years may be susceptible to diphtheria. Additional data from Europe confirmed that many adults remain susceptible to diphtheria. As is the case in general with passively transferred immunity, protection depends on the level of maternal antibody at the time of the infant’s birth, and protection decreases during the months after birth unless the infant is immunized. Recommendations for one dose of tetanus toxoid-reduced diphtheria toxoid-acellular pertussis (Tdap) vaccine for all adults in the United States and for one dose of Tdap during every pregnancy should provide added protection to neonates against diphtheria.

Neonatal diphtheria usually is localized to the nares. Diphtheria of the fauces is less common. The skin and mucous membranes may be affected; diphtheritic conjunctivitis was reported in an 8-day-old neonate. Because isolation of C. diphtheriae requires inoculation of special culture media, notification of the laboratory about the possibility of diphtheria is important. Specimens of nasal and pharyngeal secretions may improve yield of positive cultures. Infants suspected to have diphtheria should be isolated and receive penicillin or erythromycin to eradicate the organism from the respiratory tract or other foci of infection to terminate toxin production and decrease likelihood of transmission. The mainstay of therapy is diphtheria antitoxin, which should be administered as soon as the diagnosis of diphtheria is considered. This product is available in the United States from the CDC.

Pertussis

The pathophysiology, clinical features, treatment, and prevention of Bordetella pertussis infection are provided in detail in Chapter 21 . Infants and young children in the United States are at the highest risk for pertussis and its complications. Although the incidence of pertussis has declined markedly since 1934, when more than 250,000 cases were recorded, resurgence of disease since the early 1980s underscores the need for continued awareness of this disease. Hospitalization rates for pertussis for infants were fairly constant from 1993 to 2004 in the United States, with rates highest for infants 1 to 2 months of age (293 hospitalizations per 100,000 live births). Substantial numbers of cases of pertussis continued into the 2010s, with 14 deaths of infants younger than 1 year in 2012.

Newborns may be exposed to pertussis through household or hospital contacts. From 1959 to 1977, pertussis was diagnosed in 400 children in Dallas hospitals; 69 patients (17%) were younger than 12 weeks. An adult in the household with undiagnosed mild disease was the usual source of infection for these neonates and young infants. Another risk factor for pertussis may be low birth weight. A study of cases of pertussis in Wisconsin infants and young children concluded that children of low birth weight were more likely than their normal birth weight counterparts to contract pertussis and to be hospitalized with the disease. Fatal pertussis was identified through a pediatric hospital–based active surveillance system in 16 infants in Canada from 1991 to 2001; 15 of 16 infants were 2 months of age or younger. When fatal cases were matched with 32 nonfatal cases by age, date, and geography, pneumonia and leukocytosis were identified as independent predictors of a fatal outcome in hospitalized infants. A more recent report described the histopathologic findings in the respiratory tracts of 15 infants younger than 4 months who died of pertussis. The findings suggested that the organism triggers a cascade of events, including pulmonary vasoconstriction and release of pertussis toxins leading to increased leukocyte mass and refractory pulmonary hypertension.

Clinical presentation of pertussis in newborns is similar to the presentation in older children, but may lack some features typical of disease in older children. The incubation period ranges from 5 to 10 days. The initial sign usually is mild coughing, which may progress over several days to severe paroxysms with regurgitation and vomiting of food. The characteristic “whoop” may be absent in infants. The clinical picture of the most severely affected infants may be dominated by marked respiratory distress, cyanosis, and apnea, rather than significant cough. Fever is usually absent. Lymphocyte counts are frequently greater than 30,000/mm ³ . Clinical signs of respiratory infection caused by Chlamydia trachomatis are similar to signs of pertussis (see Chapter 19 ).

Complications of pertussis in young infants include convulsions, bronchopneumonia, and hemorrhage. Bacterial and viral superinfection may occur. B. pertussis pneumonia may progress rapidly; pulmonary hypertension resulting from difficulty perfusing the congested lung may cause right-sided heart failure or fatal cardiac arrhythmias. Long-term sequelae of whooping cough in infancy and early childhood were studied by Johnston and coworkers ; there was a significant reduction in forced vital capacity in adulthood in individuals who had pertussis before age 7 years compared with individuals who did not have pertussis. Diagnostic methods for pertussis depend on the immunization status of the patient and the duration of cough. Culture of a nasopharyngeal specimen, collected with a Dacron or calcium alginate swab and inoculated onto specialized agar (Regan-Lowe or Bordet-Gengou) is the most specific technique, but the organism is fastidious and may be difficult to isolate in individuals who have been immunized and in those with cough duration of more than 3 weeks. Direct fluorescent antibody testing of nasopharyngeal secretions has low sensitivity and variable specificity and is no longer recommended. Polymerase chain reaction (PCR) assay shows promise as a diagnostic tool ; although it may lack sensitivity in immunized individuals, it may still be more sensitive than culture. A Dacron swab must be used to collect the nasopharyngeal specimen because calcium alginate swabs are inhibitory to PCR. Although no U.S. Food and Drug Administration (FDA)-licensed test is available, the CDC has released a “best practices” document to guide PCR assays. Serologic testing is available, but no commercial kit is licensed by the FDA, and diagnostic cutoff points have not been established. These tests also would be limited in the event of recent immunization with pertussis-containing vaccines.

Antimicrobial therapy may lessen severity of disease if it is given in the catarrhal stage, but it has no clinical effect after paroxysms occur. Antibiotic therapy eliminates carriage of organisms from the upper respiratory tract and is valuable in limiting communicability of infection, even if given late in the clinical course. Azithromycin, 10 mg/kg/day orally in one dose for 5 days, with a maximum daily dose of 600 mg, is the drug of choice for treatment or prevention of pertussis in infants younger than 1 month. An alternative is erythromycin estolate, 40 mg/kg/day orally in four divided doses for 14 days. Chemoprophylaxis is recommended for household and other close contacts, such as individuals in the hospital, including medical and surgical personnel.

Two Tdap vaccines were licensed in 2005 to enhance protection against pertussis in adolescents and adults. Currently, immunization with Tdap vaccine is recommended during every pregnancy. These vaccines have also been recommended for health care workers who have direct patient contact.

Pathogenesis and Pathology

During fetal life, amniotic fluid bathes the entire respiratory tree, including the lungs, paranasal sinuses, and middle ear cleft. Amniotic fluid and cellular debris usually are cleared from the middle ear in most infants within a few days after birth. In term infants, the middle ear usually is well aerated, with normal middle ear pressure and normal tympanic membrane compliance, within the first 24 hours. A study of 68 full-term infants examined by otoscopy, tympanometry, and acoustic reflectometry within the first 3 hours of life revealed the presence of middle ear effusion in all neonates; fluid was absent at 72 hours of life in almost all infants.

Studies of the middle ear at autopsy provide important information about the development of otitis media in the neonate. Inflammation in the lungs or paranasal sinuses usually was accompanied by inflammation in the middle ear. deSa examined 130 infants, including 36 stillborn infants, 74 neonates who died within 7 days of life, and 20 infants who died between 8 and 28 days. In 56 cases, the middle ear was aerated or contained a small amount of clear fluid. In 55 cases, amniotic debris was present; in 2 additional cases, cellular material was mixed with mucus. A purulent exudate was present in the middle ear of 17 infants; these exudates were cultured, and a bacterial pathogen was isolated from 13. Amniotic material was present in specimens obtained from most of the stillborn infants. Purulent exudate was not seen in the stillborn infants; the frequency of its presence increased with postnatal age at time of death. Of the 20 infants who lived for 7 or more days, 11 had purulent exudate in the middle ear. Each of the 17 infants with otitis media had one or more significant infections elsewhere; 12 had pneumonia, and 6 had meningitis. deSa subsequently identified mucosal metaplasia and chronic inflammation in the middle ears of newborns receiving ventilatory support.

Factors that may affect the development of otitis media in the neonate include the nature of the amniotic fluid, the presence of other infectious processes, the need for resuscitative efforts (especially positive-pressure ventilation), the presence of anatomic defects such as cleft palate, the immunologic status of the infant, and the general state of health of the infant. Aspiration of infected amniotic fluid through the eustachian tube may be one factor in the development of otitis media in the neonate; dysfunction of the eustachian tube, which is shorter, wider, and more horizontal than in the older child, and failure to clear aspirated material from the middle ear probably have etiologic roles as well. Piza and associates speculated that infants born through thick meconium fluid may be at greater risk for otitis media because of the inflammatory nature of this fluid. deSa noted that many infants in whom otitis media developed had required assistance in respiration and speculated that the pressure of ventilation efforts was responsible for propelling infected material into the middle ear. In infants, as in older children, middle ear effusion seems to be frequent in patients with nasotracheal tubes, and the effusion occurs first on the side of intubation. Berman and colleagues described an association between nasotracheal intubation for more than 7 days and the presence of middle ear effusion.

Infants with cleft palate are at high risk for recurrent otitis media and conductive hearing loss because of persistence of middle ear effusion. Attempts to reduce the incidence of permanent hearing impairment have included intensive monitoring of children with cleft palate for middle ear effusion and repair of these defects earlier in infancy. One study found, however, that early cleft palate repair did not reduce significantly the subsequent need for ventilating tubes in these children.

Breastfed infants are at lower risk than bottle-fed infants for acute otitis media. Results of studies of Canadian Eskimo infants and of infants in India, Finland, Denmark, and the United States indicate a significant decrease in the incidence of infection of the middle ear in breastfed infants compared with bottle-fed infants. A study from Cooperstown, New York, identified a significantly lower incidence of acute lower respiratory tract infection in infants who were breastfed, compared with infants who were bottle-fed; the incidence of otitis media was lower in the breastfed infants, but this difference was not statistically significant. Infants in Boston who were breastfed had a lower risk for either having had one or more episodes of acute otitis media or having had recurrent acute otitis media (three or more episodes) during the first year of life. The protective association of breastfeeding did not increase with increased duration of breastfeeding; infants who were breastfed for 3 months had an incidence of otitis media in the first year of life that was as low as infants who were breastfed for 12 months.

The beneficial effects of breastfeeding may be due to immunologic factors in breast milk or to development of musculature in the breastfed infant that may affect eustachian tube function and assist in promoting drainage of middle ear fluid. Alternatively, the findings could indicate harmful effects of bottle feeding, including the reclining or horizontal position of the bottle-fed infant that allows fluid to move readily into the middle ear, allergy to one or more components in cow’s milk or formula, or aspiration of fluids into the middle ear during feeding. The hypothesis that breast milk is protective is substantiated by results of studies of a special feeding bottle for infants with cleft palate. Among infants who were fed by this bottle containing breast milk, the number of days with middle ear effusion was less than in infants fed by this device containing formula, which suggests that protection was more likely to be due to a quality of the milk, rather than the mode of feeding. Adherence of S. pneumoniae and Haemophilus influenzae to buccal epithelial cells was inhibited by human breast milk.

Early onset of pneumococcal otitis media has been associated with low levels of cord blood pneumococcal immunoglobulin G (IgG) antibodies. Among a group of infants who had siblings with middle ear disease, low concentrations of cord blood antibody to pneumococcal serotype 14 or 19F were associated with earlier onset of otitis media. Low cord blood antibody concentrations to serotype 19F predicted more episodes of otitis media over the first year of life in a cohort of 415 infants whose mothers enrolled in the study during pregnancy. In these infants, early otitis media was associated significantly with type 14 IgG1 in the lowest quartile but not with type 19F IgG1 antibody or with either IgG2 antibody. These findings prompted study of maternal immunization to prevent pneumococcal disease in neonates. Immunization of pregnant chinchillas with heptavalent pneumococcal vaccine resulted in reduced incidence and severity of experimental otitis in their infants. Immunization of pregnant women in Bangladesh, The Gambia, the Philippines, and the United States with pneumococcal polysaccharide vaccine resulted in pneumococcal antibody concentrations that were higher at birth in infants of immunized mothers than in controls. In addition, pneumococcal IgG antibody acquired by infants of immunized mothers had greater opsonophagocytic activity than that in control infants.

Antibody to pneumococci in breast milk has been proposed to have a role in prevention of early otitis media. Early colonization with pneumococci or other bacteria is associated with early otitis media. The role of antibodies to pneumococci in human milk in prevention of nasopharyngeal colonization of infants with pneumococci is controversial. A study in Sweden involving 448 mother-infant pairs failed to show reduction in carriage of pneumococci in neonates fed milk with anticapsular and antiphosphorylcholine activity and showed an increase in colonization when infants were fed milk with anti–cell wall polysaccharide antibody activity. Maternal immunization with pneumococcal polysaccharide vaccine resulted in higher breast milk IgA antibodies to serotype 19F, but not type 6B.

Epidemiology

The incidence of acute otitis media or middle ear effusion in the newborn is uncertain because of the paucity of definitive studies. Warren and Stool examined 127 consecutive infants with birth weight less than 2300 g and found 3 with middle ear effusions (at 2, 7, and 26 days of life). Jaffe and coworkers examined 101 Navajo infants within 48 hours of birth and identified 18 with impaired mobility of the tympanic membrane. Berman and colleagues identified effusion in the middle ear of 30% of 125 consecutively examined infants who were admitted to a neonatal intensive care unit. The clinical diagnosis was corroborated by aspiration of middle ear fluid. The basis for the differences in incidence in the various studies is uncertain, but there may be an association with procedures used in the nurseries.

Acute otitis media is common in early infancy. In the prospective study of Boston children, 9% of children had an episode of middle ear infection by 3 months of age. Age at the time of first episode of acute otitis media seems to be an important predictor for recurrent otitis media. Children who experience a first episode during the first months of life are more likely to experience repeated infection than children whose first episode occurs after the first birthday. Additional risk factors include parental smoking and low socioeconomic status.

Some host factors that also are present in infants with neonatal sepsis have been identified in infants with middle ear infection. The incidence of infection is higher in premature infants than in infants delivered at term in some studies but not in the prospective study of Boston children. Male infants are more frequently infected than female infants. Otitis media also is associated with a prolonged interval after rupture of maternal membranes and with other obstetric difficulties. Middle ear infection is more severe in Native Americans and Canadian Eskimos than in the general population, and it is likely that this is true in neonates and older infants as well. Children with cleft palate have a high incidence of otitis media, which may begin soon after birth. Prenatal, innate, and early environmental exposures were assessed in relation to early otitis media in a cohort of 596 infants followed prospectively from birth. In multivariable analysis, prenatal factors were not associated with early onset of otitis media, but environmental (day care, upper respiratory infection, birth in the fall) and innate factors (parental and sibling history of otitis media) were associated with early or recurrent otitis media, or both.

Microbiology

The bacteriology of otitis media in infants has been studied by investigators in Honolulu, Dallas, Huntsville, Boston, Denver, Milwaukee, Tampere Hospital in Finland, and Beer-Sheva, Israel ( Table 7-1 ). S. pneumoniae and H. influenzae are isolated frequently from fluid aspirated from the middle ear in very young infants, as is the case in older infants and children. Although it has been suggested that otitis media in the youngest neonates (<2 weeks of age) is caused more frequently by organisms associated with neonatal sepsis, such as GBS, S. aureus, and gram-negative enteric bacilli, this pattern does not emerge consistently when multiple studies are examined. Pneumococci were isolated from middle ear fluid in the first 2 weeks of life, and otitis associated with gram-negative enteric organisms and GBS occurred in older infants. Microbiology of middle ear disease in infants who are in neonatal intensive care nurseries may be an exception to the pattern associated with otitis media in previously healthy infants and may reflect pathogens present in the neonatal intensive care unit. In a small series of 13 such infants, only gram-negative enteric organisms and staphylococcal species were identified in the 10 samples of middle ear fluid from which bacteria were identified. Table 7-1 shows the microbiology of middle ear isolates from eight studies of otitis media in infants; when possible, data from the youngest neonates have been separated from data from older infants.

Susceptibility patterns of organisms causing otitis media in newborns reflect local patterns. In general, trends toward increasing resistance of pneumococci to antibacterial agents and colonization and disease resulting from pneumococcal serotypes not present in the pneumococcal conjugate vaccine used routinely in the United States and other countries have been observed.

Gram-negative enteric bacilli have been the predominant organisms isolated at autopsy from purulent effusions of the middle ear. Of 17 infants studied by deSa, 7 were found to have E. coli, and 6 had P. aeruginosa . β-Hemolytic streptococci (not further identified) were isolated from one infant, and no organism was recovered from the remaining 3 infants. Because pneumonia and meningitis accompanied otitis in all of these cases, the predominance of gram-negative pathogens in this series is not unexpected.

Congenital tuberculosis of the ear and of the ear and parotid gland has been reported in preterm infants from Hong Kong and Turkey. Both cases were notable for significant regional lymphadenopathy, lack of response to antibacterial therapy, and presence of active pulmonary tuberculosis in the mother. Authors of both reports suggest that there is continued need for a high index of suspicion for this disease in appropriate circumstances. Otitis media and bacteremia resulting from P. aeruginosa occurring at 19 days of life was thought to occur after inoculation of the organism during a water birth. B. pertussis was isolated from middle ear fluid in a 1-month-old infant hospitalized with pertussis; intubation of the infant’s airway may have facilitated spread of the organism from the nasopharynx to the middle ear.