BACKGROUND

Epidemiology and Pathophysiology

Pneumonia is the leading killer of children worldwide, with nearly 2 million deaths annually among children less than 5 years of age.^1,2 Mortality is exceedingly rare in the developed world, occurring in less than 1% of hospitalized pneumonia cases;^3-5 however, morbidity remains substantial with hospitalization rates estimated at 74 to 790 per 100,000 children in the United States, with the highest rates occurring among those less than five years of age.⁵ This translates to over 100,000 hospitalizations per year in the United States, making childhood pneumonia one of the most common conditions encountered in the hospital setting and also one of the most costly, with aggregate yearly hospitalization costs approaching $1 billion.^6,7

ETIOLOGY

Viruses

Respiratory viruses lead to pneumonia by direct extension from the upper respiratory tract and progressive invasion of the respiratory epithelium, which may lead to diffuse lymphocytic proliferation and interstitial pneumonitis and/or alveolar destruction depending on the specific etiology.⁸ Viruses are detected in 60% to 80% of children with pneumonia in both ambulatory and hospitalized settings (Table 102-1).^9-15 Viral pneumonias are most commonly encountered among young children, especially those less than 5 years of age in whom viral pathogens represent the predominant cause of pneumonia, either alone or in combination with bacteria. Respiratory syncytial virus, parainfluenza viruses, influenza viruses, and adenovirus have long been considered causative pneumonia pathogens.^16-22 More recently, human metapneumovirus, human bocavirus, novel species of coronaviruses, and rhinoviruses have also been implicated; however, with the possible exception of human metapneumovirus, the precise contribution of these microorganisms is still debated.^23-25

Bacteria

Bacteria causing pneumonia often colonize the nasopharynx and gain access to the lower respiratory tract directly by inhalation, although this alone is not sufficient to cause disease as it likely happens many times per day. Only when bacteria overcome the typical pulmonary host defenses does pneumonia develop. Often this occurs in the setting of a viral upper respiratory tract infection or other period of impaired immunity.^26,27 Bacteria replication in the lower airways leads to alveolar fluid accumulation and neutrophilic inflammation with rapid involvement of contiguous alveoli leading to lobar consolidation.²⁸ Occasionally, pneumonia develops by hematogenous dissemination, either from a distant site of infection (e.g. osteomyelitis) or from primary bacteremia.

A number of different bacteria may cause pneumonia (Table 102-1); however, S. pneumoniae has long been considered the major pathogen, implicated in up to 50% of pneumonia cases in clinical and epidemiologic studies.^16-22 Following the introduction of the 7-valent pneumococcal conjugate vaccine, rates of all-cause and pneumococcal pneumonia declined substantially, as did rates of pneumococcal penicillin resistance.^3,5,29-31 However, non-vaccine serotypes quickly emerged,^31-34 and despite reductions in pneumonia hospitalizations and systemic complications, disease complicated by parapneumonic effusion and empyema increased substantially.^3,35-39

Non-vaccine serotype 19A, often multidrug-resistant and associated with severe disease, was responsible for the majority of culture-positive pneumococcal cases reported following the introduction of the heptavalent pneumococcal conjugate vaccine and is likely partially responsible for recent increases in rates of complicated pneumonia.^32-34,40-42 Overall disease rates, however, have yet to approach preconjugate vaccine era levels, and the licensure of the 13-valent pneumococcal conjugate vaccine in 2010,⁴³ targeting six additional clinically important serotypes including 19A, has led to further declines.

Other bacteria-causing pneumonia include the atypical pathogens Mycoplasma pneumoniae and Chlamydophila pneumoniae, as well as Staphylococcus aureus, Streptococcus pyogenes, Haemophilus influenzae, and Moraxella catarrhalis. Atypical pathogens, particularly M. pneumoniae, are common causes of pneumonia, especially in school-age children and outpatients;^16-22,44 however, high rates of M. pneumoniae have also been reported among younger children in several studies.^18,45,46 Both S. aureus and S. pyogenes are notable for causing rapidly progressive disease that is often severe and associated with frequent complications. S. aureus, especially community-associated methicillin-resistant S. aureus (CA-MRSA), has become an increasingly important consideration in the United States.^47-50 The CA-MRSA epidemic also parallels the rise of local pneumonia complications.^35,47,51,52 In contrast, H. influenzae is no longer a major pneumonia pathogen due to widespread use of conjugate vaccines targeting H. influenzae type B. Non-type B H. influenzae, along with M. catarrhalis, are frequent colonizers of the upper respiratory tract, but only occasionally cause pneumonia.^53,57,58

Viral-Bacterial Coinfection

The influenza pandemic of 1918 killed more than 600,000 US children and adults, but it is believed the vast majority of deaths resulted from bacterial coinfection.⁵³ Thus, viral-bacterial coinfections are well recognized, particularly the association of influenza with bacterial pneumonia caused by either S. pneumoniae or S. aureus. Other viruses have also been implicated as co-pathogens in bacterial pneumonia (20%–35%).^{9,10,12,16,18,21,22,54} Viral infections often precede the development of bacterial pneumonia,^26,27,55 although it is not always clear if the infection is purely a “secondary” bacterial pneumonia or if the virus and bacterium actually coinfect the lower airways simultaneously. However, in vitro and animal studies demonstrate that the virulence of both bacteria and viruses is enhanced in coinfections,⁵⁶ and clinical studies of pediatric pneumonia indicate worse outcomes among coinfected children compared with those without viral coinfection.^53,57-60

Occasional Pathogens and Disease Occuring in Specific Populations

A number of other pathogens, including viruses, bacteria, and fungi, may occasionally cause pneumonia. Often these occur in neonates, children with congenital or acquired immunodeficiency, those with specific animal or environmental exposures, and in children at risk for aspiration pneumonia. Key animal exposures for bacterial and viral pneumonia include wild and domesticated animals such as cattle, sheep, and goats (Coxiella burnetii, Bacillus anthracis, Burcella abortus), rodents (Yersinia pestis, lymphocytis choriomeningitis virus), pigeons or pet birds (Chlamydia psittaci), and rabbits (Francisella tularensis). Key exposures for endemic fungi include bird droppings (Histoplasma capsulaturm, Cryptococcus neoformans) and geographic location: Midwestern United States (H. capsulatum, Blastomyces dermatiditis), Southeast (B. dermatiditis), Southwest (Coccidioides immitis), and Eastern and Central (H. capsulatum).

CLINICAL PRESENTATION

Signs and Symptoms

Children with pneumonia typically present with one or more of the following clinical signs and symptoms: fever, cough, tachypnea, increased respiratory effort (retractions, nasal flaring, grunting), hypoxemia, and asymmetric or decreased breath sounds and crackles on auscultation.^61-69 Fever and abdominal pain suggestive of an acute abdomen may occur, particularly in the presence of a parapneumonic effusion or lobar consolidation adjacent to the diaphragm. Tachypnea is considered the most sensitive marker for pneumonia, present in up to 80% of radiographic-confirmed cases.^70-73 However, tachypnea alone is relatively nonspecific and most children with this finding do not have pneumonia. Further, defining appropriate age-based reference ranges for respiratory rate and other vital signs has proven difficult, especially among hospitalized children.⁷⁴ The World Health Organization (WHO) case definition of pneumonia, applicable to children less than 5 years of age, includes cough or difficulty breathing and tachypnea at rest, and is widely used in the developing world.⁷⁵ A Brazilian study demonstrated sensitivity of 84% for the WHO case definition but specificity of only 19%, although the addition of fever improved specificity to 42% without reducing sensitivity.⁷⁶ In contrast, a study conducted among US children with clinical suspicion of pneumonia demonstrated poor sensitivity for the WHO case definition (34%) compared with the reference standard of radiograph-confirmed pneumonia.⁷⁷

Presentation According to Microbiologic Etiology

The clinical presentation may also provide etiologic clues. Viral pneumonia typically evolves gradually over several days and includes both upper and lower respiratory tract symptoms, such as cough, coryza, post-tussive emesis, and generalized malaise. Fever is typical but may not be present. Upper tract rhonchi and bilateral wheezes or rales may be appreciated on auscultation. In contrast, bacterial pneumonia often presents rapidly with high fever and ill or toxic appearance. In a child with preceding viral illness, this typically occurs several days after the initial illness and represents an abrupt and significant change in clinical status. Auscultation may reveal focal abnormality (e.g. rales, whispered pectoriloquy, egophony) indicative of parenchymal consolidation; such focal findings are more challenging to appreciate in young children. The presentation of atypical pneumonia is more insidious and is difficult to distinguish from viral pneumonia;⁷⁸ persistent cough is typical. Respiratory symptoms may be preceded by pharyngitis. Wheezing is also frequently present in children with atypical pneumonia, and like some viral pathogens, has been implicated as a trigger of recurrent wheezing and asthma.⁷⁹

COMPLICATIONS AND SPECIAL CONSIDERATIONS

Parapneumonic Effusion and Empyema

Parapneumonic effusion and empyema represent the most common complications of pediatric pneumonia, occurring in up to 25% of children hospitalized with pneumonia.¹⁸ Parapneumonic effusion is not always apparent at presentation and may develop and evolve despite adequate antimicrobial therapy. Persistent fever, worsening respiratory distress, painful respirations, or pain referred to the back, shoulder, or abdomen may herald the development of a significant parapneumonic effusion. Physical examination may reveal dullness to percussion, decreased breath sounds, or egophany; tracheal deviation is also possible. Most effusions are small and resolve following treatment of the underlying pneumonia; however, about 5% of hospitalized children develop large and/or persistent complicated effusions that necessitate intervention and are associated with significant morbidity.³ Large effusions are most indicative of typical bacterial disease; however, small effusions, either unilateral or bilateral, are occasionally encountered in children with M. pneumoniae and viral pneumonia.⁸⁰

Necrotizing Pneumonia and Lung Abscess

Other local complications include necrotizing pneumonia (Figure 102-1) and lung abscess. These complications are rare as a result of routine and prompt antimicrobial therapy. The same organisms that cause complicated parapneumonic effusions are also implicated in necrotizing pneumonia, specifically S. pneumoniae, S. aureus, or S. pyogenes, and may also lead to abscess formation as well as bronchopleural fistula. Lung abscess may also be the primary source of infection and is often associated with aspiration; anaerobes and polymicrobial infections are frequently implicated. Prolonged fever and poor response to therapy are usually cardinal features of both necrotizing pneumonia and lung abscess; cough productive of purulent sputum and/or blood may also be noted.

Systemic and Metastatic Complications

Systemic complications, including respiratory failure and severe sepsis/shock, occur in approximately 5% to 10% of hospitalized children, and are highest in very young children.³ Respiratory failure is the most common systemic complication, and may progress to acute respiratory distress syndrome (ARDS) in a small minority of children. For these children, extracorporeal membrane oxygenation (ECMO) can be life saving. S. pneumoniae-associated hemolytic uremic syndrome is also a rare but severe complication of pneumococcal disease that is most often associated with empyema.^81-83

Metastatic complications include distant sites of infection (e.g. meningitis, endocarditis, osteomyelitis) and occur as a result of hematogenous seeding; however, these complications are exceedingly rare.

Aspiration Pneumonia

Children with lack of adequate airway protection (e.g. seizure disorders, neuromuscular disease, or other neurologic impairment) may aspirate oral or gastric contents, leading to chemical pneumonitis and/or infectious pneumonia. Aspiration is a frequent reason for hospitalization among these children, although distinguishing between pneumonitis and pneumonia is difficult. Sterile chemical pneumonitis results from the aspiration of sterile stomach contents with symptoms beginning immediately following or within hours of a witnessed aspiration event, whereas infectious pneumonia typically follows aspiration of oropharyngeal contents and becomes symptomatic many hours to days following an aspiration event.⁸⁴ In the case of pneumonia, although typical community-acquired pneumonia pathogens are a consideration, anaerobes and gram-negative bacilli may be more prevalent. Typically, radiographic infiltrates develop in dependent areas of the lungs, such as the posterior upper lobes or apical lower lobes, in children who aspirate while lying down, or basilar lower lobes in those who aspirate while upright; the right lung is also more often implicated in aspiration.

DIFFERENTIAL DIAGNOSIS

The clinical presentation of pneumonia may mimic a number of other infectious and non-infectious conditions. Lack of fever, historical features or associated signs or symptoms, and poor response to therapy should prompt consideration of alternative diagnoses. Children with asthma or bronchiolitis may have opacities on chest radiograph that represent alveolar consolidation of viral origin or atelectasis.

DIAGNOSTIC EVALUATION

Imaging

Chest Radiographs

Chest radiographs are typically recommended for children suspected of having pneumonia who require hospitalization, as they help to distinguish among other possible diagnoses and can reveal complications and provide valuable prognostic and diagnostic clues.⁸⁵ Both frontal and lateral views are recommended.⁸⁶ Infiltrate patterns may be classified as alveolar (e.g. lobar consolidation) (Figure 102-2), interstitial (bilateral patchy or linear densities), or mixed. Chest radiographs are subject to variability in interpretation, although recent efforts have focused on standardizing radiographic assessments for pneumonia in children.⁸⁷ Alveolar infiltrate has high inter-observer agreement (κ = 0.58–0.73 across studies), indicating good reliability across readers, whereas inter-observer agreement for interstitial infiltrate is poor, contributing to diagnostic uncertainty.^88-90

Radiographs also reveal the presence of complications. Opacification of the pleural space and blunting of the costophrenic angle indicative of parapneumonic effusion can be seen on upright chest radiographs; larger effusions may opacify an entire lung or demonstrate mediastinal shift to the unaffected lung (Figure 102-3). A lateral decubitus film may also be helpful for discerning effusion size and organization, although ultrasound is superior in this regard. Cavitation with air-fluid levels usually indicate lung abscess and/or pneumatocele formation with necrotizing pneumonia, and may also be seen with any of the above described infiltrate patterns.

Although not absolute, radiographic patterns may offer clues to microbiologic etiology. Alveolar infiltrates are strongly suggestive of typical bacterial pathogens.^91,92 Interstitial infiltrates are more often associated with viral and atypical bacterial pathogens; however, typical bacterial pathogens may also produce a similar pattern. Specific radiographic findings are also occasionally helpful. The presence of large parapneumonic effusions and empyema (Figure 102-4) are strongly indicative of infection caused by S. pneumoniae, S. aureus, or S. pyogenes, whereas small effusions may indicate a typical or atypical bacterial or viral cause.⁹³ Pneumatoceles are more commonly reported following pneumonia caused by S. aureus, gram-negative bacteria, and occasionally S. pneumoniae.^49,94 S. pneumoniae is suggested by the presence of round infiltrates (“round pneumonia”). Prominent hilar adenopathy as a distinguishing feature is commonly associad with H. capsulatum and other endemic fungi, M. tuberculosis, and occasionally M. pneumoniae, but may also be indicative of a noninfectious process such as malignancy.

Follow-up Chest Radiographs

Once the diagnosis of pneumonia is confirmed by chest radiograph, follow-up studies are not routinely indicated. Radiographic findings may progress for several days despite clinical improvement and may remain abnormal for months.⁹⁵ Repeat chest radiographs are suggested for children not responding or worsening, despite 48 to72 hours of appropriate antimicrobial therapy, to evaluate for the development of complications as well in children with known complications who require ongoing intervention. Radiographic resolution of severe or complicated pneumonia generally takes weeks to months, and pleural thickening months after the illness is not uncommon. Cohen et al evaluated 82 children at 1, 6, and 12 months following hospital discharge for pneumonia complicated by empyema, and although 25% of children had clinical, radiographic, and/or pulmonary function abnormalities at 1 month, few children had persistent morbidity at 12 months.⁹⁵

Chest Ultrasound and Computed Tomography

Chest ultrasound provides an accurate estimate of effusion size and location and can detect the presence of stranding and degree of loculation—important factors for assessing the need for surgical intervention. More recently, chest ultrasound has been investigated as a means for the primary diagnosis of pneumonia, performing as well as or better than chest radiograph for detecting pulmonary infiltrates, and was superior for detecting effusion.^96-99 In a 2013 study among 200 children with suspected pneumonia (18% of whom had a positive chest radiograph), ultrasonography demonstrated sensitivity of 86% and specificity of 89% for detection of lung consolidation and air bronchograms indicating pneumonia. This study is of particular interest because the ultrasounds were performed at the point of care by clinicians with only 1 hour of focused ultrasonography training. Chest ultrasound also has the benefit of non-exposure to ionizing radiation, making it especially attractive for use in children.

In contrast, chest computed tomography (CT) exposes children to high levels of ionizing radiation and should not be used routinely to diagnose pneumonia. The benefit of chest CT is that it offers a highly detailed view of the lung parenchyma, bronchial tree, and pleura. Thus CT may be useful in instances of severe or complicated disease, especially if surgical intervention is planned, and in children with recurrent or non-responding pneumonia to evaluate for underlying pathology.

Laboratory Diagnostics for Microbiologic Etiology

Culture-Based Methods

Blood. Blood cultures are positive in ~5% children with pneumonia who have blood cultures obtained.^38,100-107 Blood cultures are positive in 10% to 13% of children with parapneumonic effusion.^102,108 However, blood cultures remain our best possibility for establishing a bacterial cause, and thus are recommended for children hospitalized with moderate or severe pneumonia in the absence of more sensitive bacterial diagnostics.⁸⁵

Culture of sputum often aids in the diagnosis of bacterial pneumonia in adults; however, obtaining high quality sputum in children is hindered by the young age of most children hospitalized with pneumonia. Sputum and tracheal aspirates. Sputum cultures could contribute meaningfully to the diagnosis of bacterial pneumonia in school-age children provided a high quality specimen is obtained (less than 10–25 epithelial cells and/or greater than 25 polymorphonuclear cells per low power field and culture isolation of a predominant organism at a concentration of 10⁶/mL).¹⁰⁹ Hypertonic saline may be helpful to induce sputum in younger children, although the yield of this approach is uncertain.^14,110 Culture of tracheal aspirates among intubated children is similarly hampered by concerns regarding upper airway contamination as well as rapid colonization of tracheal tubing, making interpretation problematic unless samples are obtained at the time of intubation (or under direct laryngoscopy) and qualitative assessments similar to sputa are employed.

Pleural fluid and other lower respiratory tract samples. Among children with pneumonia complicated by parapneumonic effusion, culture of pleural fluid is positive in up to 25% of cases.^{38,94,101,102,111,112} With the exception of pleural drainage, sampling directly from the lower respiratory tract (e.g. bronchoalveolar lavage) is rarely indicated except in instances of non-responding or very severe pneumonia.

Antigen Detection

Antigen testing by immunofluorescence from nasal or nasopharyngeal samples can be performed rapidly and is helpful for the detection of viruses. A positive test is usually highly specific; however, sensitivity is variable, particularly for influenza (sensitivity 19%–71%) and adenovirus, and is likely influenced by virus type and strain, viral load, and the test itself.^113-115 Urinary antigen testing is available for both Streptococcus pneumoniae and Legionella pneumophilia, and both are useful tests in adults; however, in immunocompetent children, L. pneumophilia is rare except in localized outbreaks, and although test sensitivity for S. pneumoniae is generally excellent, frequent false positives due to nasopharyngeal colonization limit the clinical utility of urinary antigen detection.^116,117 However, the pneumococcal urinary antigen tests could be useful for the detection of pneumococcus from normally sterile sites (e.g. blood, pleural fluid) with much less concern regarding test specificity. Studies have demonstrated sensitivities of >95% among children with documented pneumococcal bacteremia (from urine) and/or empyema (from pleural fluid),^118,119 although test sensitivity is ultimately dependent on organism density within the fluid tested.

Serology

Serologic testing is available for a number of different viral and bacterial pathogens and is used widely in epidemiologic studies. However, this method is hindered by the need for convalescent titers, making serologic testing impractical in most clinical scenarios. Testing for cold agglutinins to detect M. pneumoniae is no longer recommended due to poor test specificity since many conditions, including viral infections, may elevate cold agglutinin titers.⁸⁵

Molecular Diagnostics

Nasopharynx. Highly sensitive molecular diagnostics (e.g. polymerase chain reaction [PCR]) are available for a wide range of viral and bacterial respiratory pathogens. Viruses are readily detected from nasopharyngeal samples and testing is highly specific; a positive test may also persuade one to withhold antimicrobial therapy for a child without strong indications of bacterial coinfection, or to start antivirals for influenza.^120,121 However, these tests are often expensive and not rapidly available at the point of care, barriers that must be overcome to increase their clinical utility. Frequent bacterial colonization of the upper respiratory tract makes detection of bacterial pathogens from nasopharyngeal samples problematic and is not advised, although quantitative PCR to assess bacterial load may prove useful for differentiating carriage from causal pathogen as well as predicting disease severity.^122-124 In contrast, M. pneumoniae is not considered a colonizer of the upper respiratory tract, and specificity of nasopharyngeal PCR is high (>95%), although test sensitivity is variable (9%–93%).^125,126

Blood. PCR testing of blood or pleural fluid for bacterial pathogens is less likely to produce false positive results and does not require viable organisms for detection, presumably reducing the influence of antimicrobial therapy on test performance compared to culture-based methods. However, lack of a well-defined reference standard for bacterial pneumonia makes interpreting clinical test characteristics problematic. Using culture-confirmed pneumococcal bacteremia as the reference standard, clinical sensitivity of whole-blood pneumococcal PCR ranges from 40% to 100%; however, sensitivity decreases to 26% to 32% in adult studies, which consider a broader reference standard (positive blood culture, sputum culture, or urinary antigen test).^127-132

Pleural fluid. PCR of pleural fluid samples is more likely to reveal a causative pathogen compared to whole-blood PCR for the same reasons pleural fluid culture is superior to blood culture. Several studies demonstrate the substantial improvement in diagnostic yield using both broad-range (16S ribosomal RNA) and species-specific PCR (69%–100%) compared with pleural fluid culture (19%–58%).^119,133,134 Reliance on culture alone may also bias our understanding of the importance of individual pathogens, particularly resistant phenotypes. Studying 63 children with pneumonia complicated by empyema, the etiologic yield increased from 35% with culture alone to 84% using pleural fluid PCR for specific bacteria.¹³³ The typically drug-resistant pneumococcal serotype 19A and staphylococcal pneumonias accounted for 23% and 18% of culture positive cases, respectively, but only 2% each of culture negative cases. In contrast, S. pyogenes, recovered in 5% by culture, was detected in 11% of children by PCR. Overall, detection of S. pneumoniae doubled from 35% by culture alone to 71% using PCR, and was largely due to increased yield of nonresistant serotypes.

Other Laboratory Testing

Peripheral white blood cell count, C-reactive protein, erythrocyte sedimentation rate, and other acute-phase reactants are often abnormal in children with pneumonia, although the magnitude of abnormality is variable and likely predicted by a number of factors including host characteristics, timing of illness, and infecting organism. Bacteria can induce a robust immune response with substantially elevated acute phase reactants, leading many to associate elevated inflammatory markers with bacterial infection. However, studies have failed to demonstrate adequate test sensitivity or specificity to reliably distinguish bacterial from viral etiologies limiting their clinical utility.^92,135-141 Nevertheless, in children with severe pneumonia, serial measurements may be helpful to assess response to therapy.

Biomarker discovery and the burgeoning “omics” fields are areas of intense research focus with implications for a wide range of disease states, including childhood pneumonia. A precursor to calcitonin, serum procalcitonin levels often rise substantially in response to serious bacterial infections and do not increase appreciably in the setting of viral infections or noninfectious inflammatory states.^142-145 Specific to pneumonia, procalcitonin levels are typically higher among children with pneumonia caused by typical bacterial pathogens compared to viral or atypical bacterial pathogens.^{137,138,146-150} A randomized trial among children hospitalized with pneumonia demonstrated reduced antimicrobial prescribing (85% vs. 100%, p <.05) and duration of antibiotic exposure (5 vs. 11 days, p <.05) when procalcitonin (cutoff 0.25 ng/mL) was used to inform treatment decisions compared to usual care.¹⁵¹ Similarly, a review of 14 randomized trials among adults with respiratory tract infections and sepsis reported consistent reductions in antimicrobial prescribing and duration of therapy when procalcitonin-based algorithms were used to guide treatment decisions compared to usual care.¹⁵² Pathogen and disease-specific biosignatures are also under development using whole-blood gene expression patterns and urinary metabolite profiles;^153-156 whether or not this area of research translates to improved care and outcomes at the bedside remains to be determined.

MANAGEMENT

Antimicrobial Therapy

Bacterial Pneumonia

Neonates and Young Infants (Table 102-2). Treatment for neonatal pneumonia should include coverage for both gram-positive and gram-negative bacteria commonly encountered in neonates (e.g. group B Streptococcus, E. coli, Listeria monocytogenes). This is usually accomplished with combination therapy consisting of parenteral aminopenicillin plus gentamicin or cefotaxime. Antistaphylococcal therapy should be considered in select cases. A third-generation cephalosporin is usually sufficient for infants 1 to 3 months of age, unless the clinical history raises suspicion for infection with methicillin-resistant S. aureus, Bordetella pertussis, or Chlamydia trachomatis.