Chapter 36: Pneumonia

• The primary predictor of the etiologic agent for infectious pneumonia is the patient's age, and empiric antibiotic therapy should be based on the most likely etiologic organisms.

• Presenting signs and symptoms of pneumonia in infants and children may be nonspecific.

Pneumonia is an inflammatory process affecting the lung parenchyma and usually is due to an infectious etiology. Physical or chemical agents are noninfectious causes (Table 36-1). Various signs (such as rales) and symptoms (especially cough and fever) may lead to a presumptive “clinical” diagnosis of pneumonia, although pneumonia often is determined by an abnormal chest radiograph. The clinical spectrum of pneumonia is highly variable whether in infants, children, or adults. It ranges from a mild illness to a life-threatening disease with significant morbidity and mortality. Considering the large numbers of microorganisms and other agents that can cause pneumonia, and the limitations of diagnostic testing, the exact cause is often difficult to determine. However, a constellation of clinical, radiologic, and ancillary/laboratory findings may suggest a likely pathogen, and therefore, appropriate therapy (Table 36-2).

Acute upper respiratory tract infections are the number one diagnosis for emergency department (ED) visits in pediatric patients in the United States, accounting for 11.6% of all pediatric ED visits.^1,2 Pneumonia accounts for about 15% of all acute respiratory tract infections.² About 1.8% of all pediatric ED visits are for pneumonia with approximately 21.4% admitted to the hospital.^1,3

In the United States, pneumonia and influenza are the leading causes of deaths from infectious disease.⁴ In children (excluding infants age <1 year and adolescents), respiratory diseases are the most common cause of hospitalization.⁵ Respiratory diseases accounted for almost 38.4% of all hospital discharges among children aged 1 to 4 years and 26.8% of all hospital discharges (for children aged 5 to 9 years).⁵ Worldwide, acute respiratory diseases accounts for over 5 million deaths annually with approximately 2 million deaths in children from pneumonia each year.^6,7

Pneumonia causes significant morbidity and has a substantial impact on patients and their family's lives, including financial burden. On an annual basis, approximately 1% to 4% of children (4% in children <5 years old, 2% in children 5–9 years, 1% in children >9 years old) are diagnosed with pneumonia in the United States.⁷ The attack rates are approximately 35 to 40 per 1000 infants (age <1 year), 30 to 35 per 1000 in the preschool-age children (2–5 years), and 15 per 1000 school-age children (age 5–9 years) versus 6 to 12 per 1000 in children older than 9 years.⁸ A recent study in the United States of office-based physician practices, hospital-based outpatient clinics, and EDs found that “the incidence of pediatric ambulatory community acquired pneumonia (CAP) () visits has not changed significantly between 1994 and 2007, despite the introduction of heptavalent pneumococcal conjugate vaccine in 2000.”⁹

The number of US pediatric deaths from pneumonia has decreased by 97% from 1939 to 1996 due to several factors: the introduction of antibiotics, the availability of vaccines, and improved access to medical care for poor children.¹⁰

Unfortunately, pediatric pneumonia remains a major cause of morbidity and mortality worldwide, especially in children younger than 5 years.¹¹ Worldwide, pneumonia is the number one killer of children and kills more children than AIDS, malaria, and measles combined.^11,12 Annually, the incidence of pneumonia in the developing countries is nearly 19 times that in the developed world.¹¹ The estimated annual new episodes of pneumonia are 150 million in developing countries alone and 2.1 million in developed countries according to the World Health Organization (WHO).^12,13

The clinical presentation of pneumonia in children is quite variable and is dependent on many factors including age, comorbidity, risk factors (Table 36-3), disease severity, and causative microorganism. The classic triad of fever, cough, and rales that is often present in adults or adolescents with pneumonia is rarely present in the infant or young child.¹⁴ In the neonate, nonspecific signs and symptoms such as lethargy, irritability, apnea, vomiting, poor feeding, isolated fever or hypothermia, and poor muscle tone are more common and frequently occurs in the context of a sepsis syndrome.

Similarly, infants often lack the typical findings of pneumonia. They may present with nonspecific signs and symptoms. They may present as “a fever without source,” sepsis, or vital sign abnormalities (fever, hypothermia, bradycardia, tachycardia, and tachypnea); gastrointestinal symptoms (poor feeding, vomiting, diarrhea, and abdominal pain); or grunting, lethargy, and shock. Infants with a bacterial pneumonia may be febrile with respiratory distress manifesting as tachypnea, retractions, and hypoxia, whereas infants with Chlamydia pneumonia may be afebrile and have a cough as their only symptom with a normal physical examination.

The toddler with pneumonia frequently has a fever and cough; although gastrointestinal complaints, such as vomiting or abdominal pain, are common and may be the presenting symptom or chief complaint.¹⁴ The clinical presentation in older children and adolescents is similar to that in adults with cough, sometimes chest pain (which is usually pleuritic), and often generalized symptoms from abdominal pain to headache.

Although there is much overlap, two clinical presentations have been delineated: “typical” and “atypical” pneumonia. Typical pneumonia, presumed to be due to a bacterial microorganism, is characterized by sudden symptom onset with fever, chills, pleuritic chest pain, productive cough, a toxic appearance, and rales on lung examination. Atypical pneumonia is usually attributed to a viral etiology, mycoplasma, or Chlamydia with a gradual onset of low-grade fever, nonproductive cough, malaise, headache, and physical examination findings that may include wheezing, a viral enanthem, an upper respiratory infection (URI) with rhinitis, pharyngitis, and conjunctivitis. However, determination of the etiologic agent based on clinical presentation alone is difficult.

The clinical presentation usually indicates the severity of the pneumonia and the need for hospitalization or outpatient therapy. The lethargic infant who is not feeding or has respiratory distress will need hospitalization more so than the playful nontoxic, well-appearing infant. Any infant or child with respiratory distress (e.g., hypoxia, cyanosis, grunting, flaring, retractions), or an altered mental status (whether lethargic or irritable, inconsolable, or unresponsive), with pneumonia will require hospitalization. Patients with any risk factors such as immunosuppression or chronic diseases (including chronic lung disease, congenital heart disease, and sickle cell disease) tend to have a more serious life-threatening pneumonia (Table 36-2).

Although rales, decreased breath sounds, and wheezing may be heard in children with pneumonia, such findings may be absent with “normal” auscultation of the lungs. Rales may not be heard in infants or young children because of poor inspiration, poor ventilation of affected areas, transmission of sounds throughout the chest, which precludes localization, and noisy upper-airway sounds. Respiratory distress and signs of increased work of breathing, retractions, grunting, flaring, head bobbing, or paradoxical (seesaw) breathing may occur in children with pneumonia. Abdominal pain and/or distention from swallowed air, an ileus, or diaphragmatic irritation from lower lobe pneumonia may occur. Meningismus, without meningeal infections, can occur with upper lobe pneumonia.

Other physical examination findings may be useful in detecting the source of the pneumonia as with contiguous or hematogenous spread from other sites, and in diagnosing comorbidity (e.g., immunosuppression, chronic diseases). Extrapulmonary findings suggesting specific etiologic agents may occur: such as conjunctivitis with Chlamydia, pharyngitis with streptococcal or mycoplasma, a preceding or coexistent URI, and skin exanthems with bacterial or viral pathogens. The physical examination may reveal complications of pneumonia, such as dehydration, pleural effusion, respiratory failure, or even sepsis.

Tachypnea may be an otherwise isolated finding in pediatric patients with pneumonia, especially in infants and young children. The WHO defines tachypnea as: respiratory rate (RR) >60 breaths/min in infants <2 months old, RR >50 in infants 2 to 12 months old, and RR >40 breaths/min in children >1 to 4 years, and RR >30 breaths/min in children >5 years old.¹⁵ Using this definition, tachypnea was found in 50% to 80% of pediatric patients with CAP confirmed by chest roentgenogram (74% sensitivity, 76% specificity).¹⁶ Some studies have noted that the presence versus the absence of tachypnea increases the likelihood of pneumonia.^17–21

However, one group from the same institution has published multiple studies suggesting that the presence of tachypnea is not a predictor of pneumonia in infants and children.^22,23 In another study of bacteremic pediatric patients with pneumococcal pneumonia, tachypnea was found in only 19% of patients.²⁴ Despite these conflicting reports, the absence of tachypnea has been purported as the best finding for ruling out pneumonia²⁵ and “tachypnea, remains the most consistent clinical manifestation of pediatric pneumonia above all in developing countries.”¹⁴

A Canadian task force suggested evidence-based guidelines for the diagnosis of pediatric pneumonia. The absence of respiratory distress, tachypnea, crackles, and decreased breath sounds excluded pneumonia.²⁶ An attempt was made to validate these guidelines in pediatric patients in an urban ED. The results were dismal: only 45% sensitivity, 66% specificity, 25% positive predictive value, and 82% negative predictive value for the diagnosis of pneumonia.²⁷

Another study also tried to determine predictive factors for pneumonia in a pediatric ED and had very poor specificity (8%) but excellent sensitivity (98%). The statistically significant variables associated with pneumonia were a history of fever, fever, tachypnea, decreased breath sounds, crackles, grunting, and or retractions. Fever and tachypnea had 93% sensitivity in predicting pneumonia.²⁸

Tachypnea is a nonspecific sign and has many causes: fever, pain, anxiety, respiratory disease other than pneumonia, heart disease, and even metabolic disease. Thus, the results are not surprising. Tachycardia also has many causes but significant tachycardia that is above the normal range for the child's age and temperature may suggest pneumonia. Mean pulse oximetry was not a predictor of pneumonia in another study.²⁹

A study comparing the etiology of CAP in hospitalized children looked at numerous variables. They found that fever (>38.4°C) and the presence of a pleural effusion had a 79% sensitivity and a 59% specificity for bacterial pneumonia and that wheezing was more frequent in viral or atypical pneumonia (41%) versus bacterial pneumonia (14%).³⁰ Rales, RR >60 breaths/min, and an absolute band count >1500/mm³ had an 85% sensitivity and a 59% specificity and predicted 85% of lobar pneumonia in febrile infants <60 days of age.³¹ Leukocytosis, defined as a WBC >20,000 mm³ in febrile (>39.0°C) children <5 years old, detected 26% of patients with lobar pneumonia who had no physical signs of pneumonia.³²

There are numerous factors that increase an individual's susceptibility to pneumonia. Any process that negatively impacts host defense mechanisms predisposes to pneumonia. Such factors include congenital (such as a tracheoesophageal fistula) or acquired (e.g., human immunodeficiency virus [HIV]), a primary pulmonary disorder (e.g., bronchopulmonary dysplasia, chronic obstructive pulmonary lung disease, cystic fibrosis) versus a “nonpulmonary” disease (e.g. muscular dystrophy, cerebral palsy, or congestive heart failure) or a “systemic” disease (e.g., malignancy) and even iatrogenic (anesthesia/sedation/medications or invasive procedures to the upper airway or lungs and chest) (Table 36-3). Environmental factors, such as smokers in the household, are also associated with both an increased incidence and severity of respiratory infections.

Perhaps the most common condition that predisposes to pneumonia in children is a preceding viral URI. Such viral URIs suppress the host's normal respiratory anatomic and physiologic mechanisms by destroying the normal respiratory epithelium that acts as a cellular barrier, altering the native bacterial flora in the upper respiratory tract, impairing the mucociliary system, and even inhibiting phagocytosis.

Numerous host defense mechanisms are employed by the body to keep the lower respiratory tract (below the vocal cords) sterile. Microbes most commonly reach the lung parenchyma after infective particles are aspirated from the upper respiratory tract. Less frequently, the organisms can spread hematogenously from a bacteremia or from a distant focal infection. Contiguous spread and iatrogenic spread postoperatively following thoracic surgery or after invasive chest or upper airway procedures can also occur, but are less common.

The respiratory epithelium serves as a mucosal barrier. The mucociliary apparatus works to clear foreign materials including microorganisms from the airway. Coughing is a physiologic reflex, which also acts to clear foreign material from the respiratory tract. If microorganisms reach the lung parenchyma, the lymphatic channels that drain to regional lymph nodes, and macrophages work to clear microorganisms.

With bacterial infections, neutrophils migrate into the lung, while viruses and intracellular pathogens (such as Mycobacterium) are attacked by the host's cell-mediated immunity.

Secretory immunoglobulins, particularly IgA, augment bacterial lysis as well as neutralizing various toxins and certain viruses. The accumulation of inflammatory cells and fluids produces the clinical manifestations and radiologic findings of pneumonia.

When a pulmonary infection occurs, neutrophils travel from pulmonary capillaries into the lung's airspaces and kill microorganisms by two mechanisms: phagocytosis or by neutrophil extracellular trap (NET).³³ After ingestion of the microbes, the neutrophil uses reactive oxygen species, antimicrobial proteins, and/or degradative enzymes to kill the ingested microorganism. Neutrophils also extrude NET, which is a meshwork of chromatin that contains antimicrobial proteins. The NET traps and then kills the extracellular microbe. Pathogenic organisms have evolved ways to escape the body's defense mechanism. For example, pneumococci are adept at avoiding recognition by the lung epithelial cells, perhaps by misleading or avoiding the pattern recognition receptors on lung cells. Streptococcus pneumoniae also contains the virulence factor, pneumolysin, which is a pore-forming protein that allows the microbe to kill the host cell. All of these properties of the invading microorganisms may explain why neutrophils can fail to contain and kill pneumococci. These various processes, in turn, activate the complement system, which leads to an inflammatory response.

Unfortunately, although inflammation is essential for host defense mechanisms, it may also result in injury to the lungs. The substances produced by neutrophils also kill lung cells and destroy host tissues. The host's inflammatory response causes cellular migration and the secretion of fluids that can interfere with the exchange of gases across the alveolar capillary membrane and cause noncardiogenic pulmonary edema, respiratory distress/failure, and precipitate the acute respiratory distress syndrome (ARDS). When respiratory epithelium is infected, host cells may be killed, protective ciliary activity is destroyed, and the dead cells may slough into the airway. This narrows the airway causing hyperinflation and air trapping as well as increased dead space ventilation. The inflammatory response along with the destroyed epithelial cells may lead to atelectasis with increased intrapulmonary shunting, hyaline membrane formation, and noncardiogenic pulmonary edema.

Unfortunately, the causative microorganism is usually unknown to the clinician, at least initially. The specific etiologic agent of pneumonia is determined, at best, only about one-third of the time, although this percentage may improve in the future with better diagnostic testing.³⁴ Recent studies, using extensive state-of-the-art diagnostic tests, identified a respiratory pathogen in 79% to 85% of pediatric patients with a lower respiratory tract infection. However, this was in hospitalized patients after an “expanded diagnostic armamentarium,” which included the following: bacterial cultures, viral cultures, direct fluorescent antibody (DFA) test, PCR test, and serology; done on nasopharyngeal swabs and blood, as well as on tracheal aspirate and/or pleural fluid.^35,36 This extensive testing is not cost-effective and is infrequently done in clinical practice.

Two important caveats need to be remembered: detection of a pathogen does not necessarily mean that organism is the cause of the current infection. An elevated acute serology sample may denote a prior infection, thus the need for paired serology samples. Nasal carriage of some organisms occurs and may not be the etiologic agent of the pneumonia. The yield for even “good” sputum and blood cultures may be low, especially for fastidious, difficult-to-grow organisms. Some organisms need special culture material and handling, and may take days or weeks to grow. Second, coexistence of pathogens is quite common with concomitant viral and bacterial infections occurring in infants and children about 41% of the time.³⁷ The isolated virus or bacteria may be part of a coexistent infection or a “bystander” and not a cause of the pneumonia.³⁷

The term atypical pneumonia arose at the onset of the antibiotic era, in the 1940s, when the sulfonamides and penicillins were introduced. It referred to those cases of pneumonia that did not respond to these two antibiotics and no organism could be identified using the available testing (bacterial gram stain and culture). Over the years, with the new field of virology and other improved diagnostic techniques, many organisms that cause atypical pneumonia have been identified. These include viruses, Mycoplasma pneumoniae, Chlamydophilia (previously known as Chlamydia pneumoniae), Legionella, and Pneumocystis to name a few. Currently, atypical pneumonias are characterized by: failure to identify an etiologic agent on routine gram stain or sputum culture, and chest radiograph appearance showing a nonlobar, patchy, or interstitial pattern.

The prevailing pathogens responsible for pneumonia in pediatric patients are dependent on the patient's age, comorbidity, immunizations, day care attendance, and epidemiologic factors including seasonal variations/recent local outbreaks. There is a seasonal variation with the incidence of pneumonia peaking in the winter months in the Northern Hemisphere. If the causative organism cannot be determined, age is the best predictor of the pathogen (Table 36-4).

In pediatric patients of all ages, S. pneumoniae is the most common bacterial cause of pneumonia, whereas respiratory syncytial virus (RSV) is the most frequent viral cause, and M. pneumoniae is the most common cause of “atypical” pneumonia, with Chlamydophilia (Chlamydia) pneumoniae the second.^36–38 According to several studies, the most common organisms identified in ambulatory children with CAP regardless of age were S. pneumoniae (17.5%–30%), viruses (20%–39%), M. pneumoniae (7%–29.5%), and Chlamydophilia (Chlamydia) pneumoniae (6%–15%), with RSV being the most common virus (21%–28%).^37,38 The findings were similar in hospitalized children with CAP. In all age groups excluding neonates, the most common pathogens causing severe pneumonia necessitating an intensive care unit admission are S. pneumoniae, Staphylococcus aureus, group A streptococcus, Haemophilus influenzae type B, adenovirus, and M. pneumoniae. Coinfections of S. pneumoniae with other organisms were common: bacterial–viral (often with RSV) or bacterial–bacterial (often with M. pneumoniae or C. pneumoniae).

Human metapneumovirus has recently been identified as a leading cause of respiratory tract infection in the first few years of life. It has a clinical spectrum similar to RSV: with bronchiolitis occurring in 59%, croup in 18%, asthma exacerbation in 14%, and pneumonia in 8%. It occurs most frequently during the winter months. The mean age of infected children is 11.6 months.³⁹

It is likely that the pneumococcal vaccine altered the relative frequency of the various pathogens, as occurred after the introduction of widespread immunization with the Hib vaccine.⁴⁰ Prior to the Hib vaccine, H. influenzae type B was a frequent cause of pneumonia. After widespread immunization with Hib, other H. influenzae nontype B pathogens emerged with increasing frequency and H. influenzae type B infections had a precipitous decline.⁴¹

Neonate

Neonatal (birth to 1 month) pneumonia is usually due to aspiration of maternal genital organisms acquired during labor and delivery. The predominant organism is group B Streptococcus followed by enteric gram-negative bacilli, generally Escherichia coli and Klebsiella pneumoniae. Other less common etiologic agents include Listeria monocytogenes, other streptococci (group A and B-hemolytic species), S. aureus, Bordetella pertussis, anaerobic bacteria, and viruses (most often, herpes simplex, cytomegalovirus, and then rubella).

Infants: 1 to 3 Months

Infants aged 4 to 8 weeks may still have pneumonia caused by organisms predominantly found in neonates; specifically, group B streptococci, gram-negative enteric bacteria, and L. monocytogenes. In addition, viruses (especially RSV and parainfluenza), Chlamydia (afebrile pneumonia of infancy), and staphylococci may be a cause of pneumonia in infants 1 to 3 months of age. The predominant organisms causing pneumonia differ between early infancy (aged 4–12 weeks) versus older infants. In infants aged 4 to 12 weeks, the prevailing pathogens are (in order of prevalence) Chlamydia trachomatis, viruses (RSV and parainfluenza), S. pneumoniae, and B. pertussis.^37,38,42

“Afebrile pneumonia of infancy” or atypical pneumonia typically occurs in infants from 3–4 weeks to 16 weeks of age. The classic presentation is that of an afebrile infant 1 to 4 months of age with a prominent dry cough who looks well. There is gradual symptom onset beginning with nasal congestion followed by a dry cough. Conjunctivitis either preceding or concurrent with the respiratory symptoms is found in approximately half of the infants. The infant appears nontoxic with a normal physical examination, except for conjunctivitis and a cough that may be paroxysmal. One may hear occasional rales. The chest radiograph usually demonstrates an interstitial pneumonia while the total WBC count is normal but eosinophilia may be present.⁴³

Generally, “Chlamydia pneumonia” is a mild disease but the cough may interfere with usual activities (e.g., feeding, sleeping) and significant respiratory distress can occur with tachypnea, retractions, and hypoxia noted. Apnea is also a major concern with Chlamydia pneumonia in young infants. It can occur as well with other causes of respiratory infections (such as RSV or pertussis) which explains why admission, whether as an inpatient or to an observation unit, is often considered in infants younger than 6 months. A macrolide erythromycin, azithromycin, or clarithromycin can be given for Chlamydia pneumonia. Azithromycin may be better-tolerated (e.g., less gastrointestinal side effects) than erythromycin or clarithromycin,⁴⁴ although erythromycin is still recommended for Chlamydia pneumonia by the Red Book.⁴³

Infants: 3 to 24 Months

In older infants (3–12 months) and very young children (12–24 months), viruses are the most common etiologic agent causing pneumonia. The viruses implicated in this age group are RSV, human metapneumovirus, parainfluenza; less frequently, adenovirus, rhinovirus; and numerous other viral agents.⁴⁵ The most common pathogens responsible for bacterial pneumonia in this age group are S. pneumoniae, H. influenzae (nontype B), Moraxella catarrhalis, M. pneumoniae, group A streptococci, and S. aureus.^37,38,42 There has been a precipitous decline in H. influenzae type B invasive disease, including pneumonia, since the widespread use of Hib.⁴¹

Toddler/Preschooler

Viruses remain the most common cause of pneumonia in this age group. The most frequent viral pathogens are RSV, parainfluenza, adenovirus, and influenza; and less frequently, adenovirus and rhinovirus. The bacterial pathogens include S. pneumoniae, H. influenzae, and nontype B, followed by the “atypicals” M. pneumoniae and Chlamydia pneumoniae; and less frequently S. aureus, group A streptococcus, and M. catarrhalis.^37,40,45 The Hib vaccine does not confer immunity against nontype B H. influenzae, which has increased in frequency as an etiologic agent of bacterial pneumonia and other infections.⁴¹

School Age and Adolescents

The “atypical” bacterial organisms, M. pneumoniae, and less commonly Chlamydophilia (Chlamydia) pneumoniae are the predominant pathogens causing pneumonia in this age group followed by S. pneumoniae and viruses (RSV, parainfluenza, influenza, adenovirus, and rhinovirus). Mycoplasma accounts for approximately one-fifth of all cases of pneumonia in the general population. The incidence of M. pneumoniae increases from school-age children (9%–16%), to adolescents (16%–21%), and to college students and military recruits (up to 30%–50%).⁴⁶

Severe, Life-Threatening Pneumonia

In pediatric patients excluding newborns, the most common organisms that can lead to severe, life-threatening pneumonia needing an intensive care unit admission are S. pneumoniae, S. aureus, group A streptococcus, H. influenzae b, adenovirus, and M. pneumoniae.^{37,40,41,45,47}

Streptococcus Pneumoniae

S. pneumonia is the most common cause of bacterial pneumonia in all ages of pediatric patients excluding neonates. The classic presentation is the sudden onset of fever, chills, productive cough with blood-tinged sputum (or hemoptysis in adolescents and adults), pleuritic chest pain, and dyspnea. Patients with functional (e.g., sickle cell disease) or anatomic asplenia are at increased risk for pneumococcal disease. Routine immunization with pneumococcal conjugate vaccine PCV7 (Prevnar) and also PPV23 (Pneumovax) in high-risk children is recommended. The incidence of all invasive pneumococcal infections has decreased by 80% for children <2 years of age and approximately 90% for vaccine/vaccine-related serotypes. It has also decreased in older children and adults since the recommendation in 2000 for the routine use of PCV7 in infants.⁴⁰

Staphylococcus Aureus

S. aureus is an uncommon but rapidly progressive fulminant illness. S. aureus pneumonia may be primary with no extrapulmonary site of infection or secondary with ≥1 nonpulmonary site of infection. Blood cultures are positive in 20% to 30% of patients with primary S. aureus pneumonia. Pleural effusions and pneumatoceles are noted in 45% to 60% of patients.⁴⁷

Haemophilus Influenzae

H. influenzae infections, including pneumonia, are most prevalent in young children. Extrapulmonary infections (such as otitis media, meningitis, pericarditis, and epiglottitis) are common in patients with H. influenzae type B pneumonia. There has been a marked decline in infections due to H. influenzae type B since the use of Hib vaccine. However, H. influenzae type B remains a significant life-threatening pathogen and H. influenzae nontype B pneumonia is a growing threat.⁴¹

Mycoplasma Pneumoniae

Generally, M. pneumoniae is characterized by a low-grade fever and nonproductive cough. Other symptoms that may occur include headache, myalgias, abdominal pain, and vomiting. The pulmonary examination may reveal rales or decreased breath sounds, or may be normal. Typical laboratory studies include normal WBC, elevated transaminases, and positive cold agglutinins. Neurologic complications including meningoencephalitis, transverse myelitis, Guillain–Barre syndrome, and cerebellar ataxia occur in approximately 7% of patients with mycoplasma pneumonia.⁴⁶

Respiratory Syncytial Virus

RSV is the most common pediatric viral pneumonia, particularly in infancy. Although the most common presentation of RSV infection is bronchiolitis, RSV pneumonia can occur and can coexist (up to one-third of infants with RSV bronchiolitis will also have pneumonia). The typical findings with an RSV infection include a low-grade fever; signs of URI, cough, wheezing, and/or rales; WBC is normal or mildly increased with a lymphocytosis; and variable chest radiographic findings with air bronchograms, atelectasis, and/or hyperinflation.^37,45

Newly Identified Pathogens

Human metapneumovirus (hMPV) was identified in 2001 from young children with respiratory tract disease. This virus is very similar to RSV as humans are its only host; it affects similar age group and is prevalent during same season. Children <5 years old are most susceptible to hMPV and primary infection in those <2 years of age can be severe. Essentially all children have been infected by hMPV by 5 to 10 years of age as indicated by serologic testing. According to PCR identification, hMPV is responsible for 4% to 7% of acute lower respiratory illnesses (ALRI) and 5% of CAP.⁴⁸

Human Boca virus (HBoV) was recognized in 2005 and its clinical significance is yet to be determined. However, PCR testing indicates that by school age nearly all children have been infected with HBoV. Antibodies from the mother are transmitted to newborns. The incidence of HBoV is 3% to 19% for ALRI and for mixed infections at 19% to 83%, whereas the incidence of HBoV in pediatric pneumonia was found to be 5% to 14% and for mixed infections 69% to 83%.⁴⁸

Other Specific Pathogens and Unusual Causes of Pneumonia

Gram-negative bacilli, including pseudomonas, should be considered in neonates, in patients who have been recently hospitalized, and in children with cystic fibrosis. Group A streptococcus can cause invasive disease ranging from pneumonia and empyema to necrotizing fasciitis. Pneumonia secondary to anaerobes can occur with any condition that predispose to aspiration (such as neurologic disorders or congenital anomalies) (Table 36-4). Mycobacterial infection should always be in the differential diagnosis of pneumonia. Tuberculosis is a particular concern in patients with HIV/AIDS and in immigrants.

Other unusual etiologic agents causing pneumonia are associated with a specific geographic locations, animal exposure, or hobbies. Recent travel to or residence in an endemic region may suggest pneumonia secondary to coccidioidomycosis (southwestern United States), histoplasmosis (Mississippi/Ohio river valley), or hantavirus (southwestern United States). Exposure to/contact with an infected individual(s) and travel or residence in an endemic area may be a clue to severe acute respiratory syndrome (SARS) or tuberculosis (TB). Exposure to pets or other animals as well as occupational or avocational exposure can be associated with certain pneumonias (Table 36-5).

Laboratory studies may not be necessary in every patient with suspected pneumonia, such as the well-appearing afebrile patient presenting with a cough that has a good oral intake and is in no distress. However, in some patients laboratory studies may be useful in identifying the underlying disease and potential complications. The WBC is usually normal or mildly elevated with a viral pneumonia.⁴⁹ A normal WBC is characteristic of M. pneumoniae or C. pneumoniae. Lymphocytosis may be present with a viral, Chlamydia, or pertussis pneumonia. A markedly elevated total WBC and lymphocytosis is characteristic of pertussis (see Chapter 37). Eosinophilia often occurs with a parasitic infection.

Leukopenia may portend a poor prognosis. Leukocytosis (>15,000 cells/mm³) with bandemia is generally present with a bacterial pneumonia.^24,50,51 A markedly increased WBC (>25,000 mm³) is highly suggestive of a bacterial etiologic agent, frequently with a bacteremia and a greater complication rate.⁵⁰ There is a conflicting report indicating that an elevated WBC >15,000 did not differentiate between viral or bacterial pneumonia.³⁶ In a study of febrile children presenting to an ED, 26% of febrile pediatric ED patients with a WBC >20,000/mm³ had an occult pneumonia on the chest radiograph.³² There may be other factors that affect the WBC count, such as coinfections, time of presentation, or treatment given prior to the ED visit.

Acute phase reactants, such as the erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), or serum procalcitonin, “cannot be used as the sole determinant to distinguish between viral and bacterial causes of CAP.”⁵¹ Positive blood cultures for hospitalized infants and children with pneumonia have been reported in the range of 10% to 15% and 6.5% in ED patients, which is comparable to that in adult ED patients with 7% positive.^34,50,52 The higher rate for positive blood cultures in hospitalized versus ambulatory patients may reflect a higher incidence of bacterial pneumonia (such as S. pneumoniae) in hospitalized patients compared with a greater incidence of viral and atypical pneumonia in outpatients.³⁴ Although the yield from blood cultures is low and their use has been questioned by some,⁵² most recommendations include obtaining blood cultures in febrile toxic or ill-appearing hospitalized patients with the understanding that mandating blood cultures in all hospitalized patients (versus targeting more critically ill patients) is probably neither necessary nor cost-effective.^51,53

Sputum cultures have limited utility, as children younger than 10 years of age rarely produce sputum and obtaining high-quality sputum uncontaminated by mouth/pharyngeal flora is difficult.³⁴ However, sputum samples may be useful in selected cases and to confirm tuberculosis. Nasopharyngeal and throat cultures for bacteria are not useful and may be misleading because of contamination with oral flora.

Conversely, nasopharyngeal and/or throat cultures for viruses, pertussis, and the atypicals (e.g., mycoplasma and Chlamydia) may help identify the pathogen causing pneumonia.

Rapid viral antigen testing (such as for RSV, influenza, and parainfluenza) and fluorescent antibody testing for B. pertussis and C. trachomatis are available and have the advantage of yielding results more quickly than a culture. Bacterial antigen testing for pneumonia, however, has limited utility because of poor sensitivity and specificity. Serologic testing, which usually involves acute and convalescent sera may be helpful in selected patients in whom the diagnosis is uncertain or the pneumonia persistent and unresponsive to therapy.^51,53

Positive cold agglutinins occur in 70% to 90% of patients with M. pneumoniae but may be negative in young children and may be positive with viral infections. The bedside cold agglutinin test is confirmed by putting several drops of blood in a blue coagulation tube then placing the tube in ice water for 15 to 30 seconds. Course floccular agglutination that disappears with rewarming is a positive test.⁵² Currently, with Legionella pneumonia, the diagnostic test of choice is Legionella urinary antigen which remains positive weeks after the infection but it only detects Legionella pneumophila and not other Legionella species. Tuberculosis skin testing may be indicated with apical and/or cavitary pneumonia, but this is usually not done in the ED.⁵³

Invasive diagnostic testing including transtracheal aspiration, bronchoalveolar lavage, percutaneous, open-lung biopsy, or video-assisted thoracic surgery (VATS) is generally reserved for patients with severe disease unresponsive to therapy, especially if they are immunocompromised and in intubated patients.⁵¹

The chest radiograph is noninvasive, painless, easily obtained, readily available, fairly inexpensive, has a low risk of radiation (although it is not zero), reliable, can yield valuable information, and is indicated in most patients with a clinical suspicion of pneumonia.^7,34 It may help make the diagnosis by eliminating other possible infectious and noninfectious causes for the patient's symptoms (Table 36-6). The chest roentgenogram is essential in confirming the diagnosis of pneumonia, may indicate the likely pathogen, and is valuable in detecting complications such as a pleural effusion, abscess, or pneumatoceles. An abnormal chest radiograph, characteristic of a specific type of pneumonia, may identify those patients who will (or will not) benefit from specific antimicrobial therapy. The well-appearing low-risk child/infant with clinically mild pneumonia who has close follow-up and reliable caregiver may not need a chest radiograph, but it is indicated in the patient with respiratory distress or hypoxemia and the hospitalized patient.^51,53

There are three major categories of pneumonia: lobar (airspace), bronchopneumonia (lobular), and interstitial (Fig. 36-1).^7,54 Bacterial infections, specifically S. pneumoniae and Klebsiella, tend to have a lobar configuration or consolidation, ^36,54 whereas bronchopneumonia (lobular) pneumonia has small fluffy infiltrates and peribronchial markings (a patchy or stringy consolidation) and is commonly due to mycoplasma and other bacteria, particularly S. aureus and many gram-negative bacteria.⁵⁴ Initially, lobular pneumonia may be segmental since it originates in the airways rather than the airspaces but commonly spreads to other areas and may be bilateral. Acute interstitial pneumonia is characterized by hyperinflation, peribronchial cuffing (thickening), and increased bronchovascular markings, and may have a nodular appearance. Viral pneumonia usually has an interstitial pattern or bilateral bronchoalveolar or peribronchial infiltrates; however, lobar or segmental consolidation can also occur.^34,54

A “round” pneumonia is a large solitary consolidated spherical-shaped (round) lesion and is typically due to S. pneumoniae.⁷ The presence of pneumatoceles, pleural effusion, and/or empyema suggests S. aureus. Pneumonia due to C. trachomatis generally appears as diffuse alveolar or perihilar interstitial infiltrates with hyperexpansion. Classically, M. pneumoniae appear as lower lobe streaky or patchy infiltrates, although 10% to 25% of the time different patterns occur including lobar infiltrates.

Studies have documented that interobserver agreement among radiologists' readings of chest radiographs regarding pneumonia is poor.^55,56 Moreover, radiologists' readings of chest radiographs in febrile 3- to 24-month-old children are biased by information from the treating physician.⁵⁷ Overall, using a chest radiograph to differentiate viral from bacterial pneumonia has widely varying results with sensitivities of 42% to 80% and specificities of 42% to 100%.²⁵

Occasionally, a patient with clinical signs and symptoms of pneumonia may have a negative chest radiograph and later after treatment, a repeat radiograph may be positive for pneumonia. Similarly, a patient may have recovered from pneumonia and have clear lungs on auscultation yet the radiograph may still demonstrate pneumonia, which may take 6 to 8 weeks for the radiographic findings to resolve.

The author of a study with only 16 patients with documented radiographic findings of pneumonia questioned whether infants with just a fever but no clinical findings of pulmonary disease (e.g., rales, rhonchi, wheezing, tachypnea, grunting, flaring, and cough coryza) warrant a chest radiograph.⁵⁸ This study has been criticized for major methodologic flaws.⁵⁹ Unfortunately, although this approach of using a combination of clinical variables may have negative predictive value, it may overlook many cases of pneumonia. In one study, 28% of patients with bacteremic pneumococcal pneumonia documented by positive blood cultures had no respiratory symptoms whatsoever.²⁴ In a study of bacteremic pneumococcal pneumonia patients, 81% of patients did not have tachypnea. The “best” physical examination finding was positive only 19% of the time, and would miss 81% of the bacteremic pneumonia patients.²⁴ It has also been documented that young infants with a high fever and leukocytosis may have an occult pneumonia (in 26%), so a chest radiograph may be warranted in these patients.³²

In summary, although the radiographic patterns may be suggestive of a specific etiologic agent, overlap occurs and it is not possible to make the diagnosis of bacterial versus viral pneumonia solely on a chest radiograph. However, the chest radiograph can add valuable information and, along with other data, may suggest the likely pathogen.^6,7,54 In selected patients with equivocal chest radiographs, other plain films may be helpful. Computed tomography (CT) scan is rarely needed but may be useful in complicated patients or those in whom the diagnosis is uncertain.

The differential diagnosis of pneumonia is extensive and includes both infectious and noninfectious etiologies (Table 36-7). The history and physical examination and the chest radiograph along with ancillary studies in some patients (Table 36-2) are the key to establishing the etiology and initiating appropriate therapy (Tables 36-8 and 36-9).

Management of pneumonia centers on supportive care, appropriate use of antimicrobials, and hospitalization for some patients. Determine the severity of illness and the need for hospital admission (Table 36-10). Any patient with altered mental status, hypoxia, respiratory distress, or respiratory failure or one who has severe pneumonia warrants admission. Indicators of respiratory distress include cyanosis, hypoxemia, retractions, grunting, flaring, abdominal (seesaw) respirations, head bobbing, altered mental status, restlessness (secondary to hypoxia), somnolence (possibly due to hypercarbia), and vital sign abnormalities such as tachycardia, tachypnea, bradycardia, bradypnea, and apnea. Obtain pulse oximetry in all patients suspected of having pneumonia and provide oxygen to those with low oxygen saturation.⁵¹ There is almost universal agreement that a previously healthy child with pneumonia with a pulse oximetry <90% should be admitted, and a majority of physicians would also admit children with a pulse oximetry level <93%.⁶⁰

Supportive care includes respiratory care, control of fever, maintaining hydration (may need intravenous fluids), and treatment of complications. Respiratory care may involve supplemental oxygen, suctioning, cardiorespiratory monitoring, bronchodilators for wheezing, humidification, and, in a few cases, intubation and mechanical ventilation. Apnea and/or respiratory failure may occur precipitously in infants younger than 6 months, so monitoring and inpatient observation may be indicated.

Complications of pneumonia range from dehydration, pleural effusions, empyema, pneumatoceles, shock, pneumothorax, bacteremia, sepsis, shock, apnea to bronchiolitis obliterans.⁶¹ Additional foci of infection may be present such as meningitis, epiglottitis, septic arthritis, pericarditis, soft-tissue/skin infections, and otitis media. Thoracentesis of a pleural effusion or empyema may be diagnostic and therapeutic.⁶¹ Selected pleural effusions (e.g., a large effusion that is interfering with respiration) and any empyema require drainage with a chest tube.

Most patients with bacterial pneumonia improve rapidly (with defervescence and symptom improvement within 1–2 days) once appropriate antibiotics are started, unless complications occur, although resolution of chest radiographic abnormalities may take ≥6 to 8 weeks. Failure to improve or worsening symptoms suggest possible complications or a misdiagnosis such as another infection (including TB, fungal, or parasitic infection), foreign body, other pulmonary disease, and underlying systemic conditions (e.g., malignancy, HIV, autoimmune disorder, immunologic disorder) as the cause. Many patients with pneumonia, including most atypical pneumonias (e.g., viral, mycoplasma, C. pneumoniae), and some bacterial pneumonias in nonimmunosuppressed older-age children may be treated as an outpatient. All pediatric patients discharged from the ED with pneumonia should have close follow-up in 1 to 2 days with their primary care provider. Another option to inpatient hospitalization is admission to an ED observation unit, especially if it is early in the clinical course or difficult to assess the severity of the pneumonia or there are any other concerns.⁶²

Since the specific etiologic agent is likely unknown, empiric therapy is the usual approach and is determined by the most likely pathogen(s) based on the patient's age, need for hospitalization/illness severity, and epidemiologic factors (Table 36-8). In patients who need hospital admission, the first dose of antibiotic(s) should be administered in the ED. Newborns should be hospitalized and double antibiotic coverage begun with ampicillin for Listeria and enterococci and either a third-generation cephalosporin (e.g., cefotaxime) or an aminoglycoside (e.g., gentamycin). Similarly, hospitalization is generally recommended for young infants (aged 1–3 months) because of immature immune system; pneumonia is often part of a sepsis syndrome, signs/symptoms of sepsis are subtle and nonspecific, and complications such as apnea and/or respiratory failure can occur. For the 1- to 3-month-old, a third-generation cephalosporin (e.g., cefuroxime) is one option and, if an afebrile pneumonitis or pertussis is suspected, a macrolide or a sulfonamide should be administered.

For the older infant and preschool child (3 months up to 5 years), a third-generation cephalosporin and a macrolide for inpatients is one antibiotic regimen which would provide empiric coverage for the atypicals (macrolide for mycoplasma and Chlamydia) and ceftriaxone for other bacterial organisms, whereas amoxicillin–clavulanate is an option for outpatient treatment. For school-age children and adolescents (age >5 years up to age 18 years), inpatient therapy could be ceftriaxone and a macrolide, whereas outpatient therapy could be with a macrolide alone (Table 36-8).

In a patient of any age, if there is any concern for a staphylococcal pneumonia, especially if the patient is seriously ill, then add an antistaphylococcal agent such as vancomycin, clindamycin, or even nafcillin/oxacillin/methicillin (depending on resistance patterns). Vancomycin is a reasonable choice if methicillin-resistant S. aureus is a possibility, and clindamycin or one of the “antistaphylococcal cillins” (e.g., oxacillin, methicillin, or nafcillin) is an appropriate antibiotic for vancomycin-resistant enterococci until sensitivities are known.⁴⁷

Fulminant viral pneumonia can occur, particularly in immunocompromised patients, and empiric viral therapy may be needed: such as acyclovir (for varicella pneumonia), ribavirin for high-risk patients with RSV pneumonia, ganciclovir for cytomegalovirus pneumonia, amantadine for influenza, or prednisone and zidovudine for HIV patients with lymphocytic interstitial pneumonia.

Pneumonia is one of the most common illnesses encountered in infants and children. The clinical spectrum ranges from mild to severe life-threatening disease with management ranging from outpatient therapy to ED observation, to inpatient hospitalization, and intensive care admission. As there are a vast number of microorganisms that can cause pneumonia and diagnostic testing has limitations, the specific etiologic agent is generally not known in the ED. However, the most likely pathogen and the recommended therapy may be suggested by clinical and ancillary data. Management includes supportive therapy and often empiric antibiotic therapy, which is based on many variables: patient age, comorbidity, clinical picture, ancillary (radiology and laboratory) studies, and epidemiology.