Pediatric lower respiratory infection may be caused by a variety of bacterial and viral pathogens. In addition, the incidence of the atypical pathogens, which include Mycoplasma pneumoniae, Chlamydia pneumoniae, and Legionella pneumophila, increase as children get older. With such a variety of potential pathogens, clinicians need to develop a method for diagnosing pediatric lower respiratory disease and to differentiate self-limiting viral illness from those caused by bacteria that require specific antibiotic therapy.
Certain generalizations regarding the etiology of pediatric pneumonia can be made. Viruses cause most lower respiratory diseases in younger children and include respiratory syncytial virus, influenza A and B, parainfluenza, and adenovirus. Respiratory syncytial virus and influenza viruses have their peak incidence in the fall and winter months, whereas parainfluenza dominates in the spring and summer. The presence of wheezing is more common in patients with viral pneumonia as compared with bacterial disease. Bacterial pathogens commonly associated with pneumonia include Streptococcus pneumoniae, nontypeable Haemophilus influenzae, and Moraxella catarrhalis. Many clinicians consider bacterial pneumonia, particularly S. pneumoniae, to be the likely cause of lower respiratory infection if the clinical history is characterized by acute onset of symptoms such as cough and high fever. In regard to the atypical pathogens, there is an age-related decline in the incidence of viral pneumonia accompanied by an increased incidence of these infections as children approach adolescence.
Recent reviews have suggested that the presence of an increased respiratory rate may be the best method to distinguish lower respiratory tract infection from the more common upper respiratory tract infections. The World Health Organization has issued guidelines for the clinical diagnosis of pneumonia in developing countries; the guidelines state that tachypnea and intercostal retractions are the best indications of lower respiratory tract disease.
Basic Diagnostic Approach
The proportion of children with pneumonia who are diagnosed with a specific etiology is low. Unlike adults, children usually do not produce adequate sputum specimens for Gram stain and culture. Blood cultures have a yield of less than 10% in patients with bacterial pneumonia. “Lung puncture” studies that are conducted in developing countries are obviously not met with enthusiasm in general pediatric practices. Prospective studies that have employed sensitive antibody tests and polymerase chain reaction techniques have suggested that in up to 20% of pediatric community-acquired pneumonias, the infection is “mixed” (i.e., both S. pneumoniae and M. pneumoniae or C. pneumoniae); in these cases, the primary pathogen is not clear. Authors of these studies have also suggested that mixed infection with bacteria and respiratory viruses is likely to be common as well.
Many studies have looked at causes of pediatric pneumonia as it relates to certain readily available laboratory measurements. Many clinicians consider S. pneumoniae to be the likely cause of the lower respiratory infection if the picture is characterized by acute onset of high fever, lobar pneumonia on chest radiograph, leukocytosis, and a rapid response to β-lactam antibiotics. Numerous studies have found that chest radiographs do not readily distinguish between bacterial, atypical bacterial, and viral pneumonia. A variety of laboratory tests have been used in the attempt to distinguish bacterial from viral pneumonia, including the C-reactive protein and absolute neutrophil counts. One problem in using “screening” tests is that specific cutoff levels have often not been established. A recent study done in Europe found that although white blood cell count and C-reactive proteins were statistically higher in patients with pneumococcal infections, other clinical and laboratory and radiographic studies were of little value.
Given the clinical, epidemiologic, and laboratory difficulties in pinpointing the cause of pediatric pneumonia, an additional approach is to divide patients by age.
The primary bacterial pathogen in neonatal pneumonia is group B streptococci, although Escherichia coli and Listeria monocytogenes have also been reported. The mechanism is similar to that in neonatal sepsis, where colonization from the mother results in neonatal colonization and breakthrough infection.
Chlamydia trachomatis is the most common sexually transmitted infection in the United States. The organism may reside in the genital tract of pregnant women and be transmitted in about 60% of cases to infants at the time of delivery. About one half of infants who acquire the organism develop conjunctivitis, and 20% eventually develop lower respiratory disease.
Pneumonia caused by bacteria such as group B streptococcus typically occurs in the first weeks of life, presenting with fever, increased work of breathing, and hypoxia. C. trachomatis infection usually occurs between 2 and 19 weeks after birth. The infants are afebrile, have increased respiratory rate, and cough. Children with chlamydial pneumonia often have hyperinflation, and bilateral infiltrates on chest x-ray, eosinophilia, and elevated serum immunoglobulin levels.
Cultures of the blood, urine, and even cerebrospinal fluid are often obtained and intravenous antibiotic started. C. trachomatis can be diagnosed by culture or direct fluorescent antibody staining of nasopharyngeal secretions.
The management of the febrile tachypneic neonate suspected of having pneumonia is similar to that of neonatal fever. Empiric intravenous antibiotics are started until culture results are final. Empiric treatment usually consists of ampicillin combined with gentamicin or a third-generation cephalosporin. Treatment of C. trachomatis is with oral erythromycin, 50 mg/kg per day in four divided doses for 2 weeks. In the past, erythromycin was given to neonates exposed to C. trachomatis at the time of delivery. Recently, there has been an association reported between oral erythromycin and the subsequent development of hypertrophic pyloric stenosis in infants younger than 6 weeks of age. The current recommendation is to treat with oral erythromycin, 50 mg/kg per day in four divided doses for 14 days all infants with chlamydial conjunctivitis and pneumonia. Patients who are exposed at the time of delivery are not presumptively treated, but rather monitored closely for the development of disease. Routine screening of all
pregnant women for sexually transmitted disease is helpful in reducing disease by C. trachomatis.
Pneumonia in the First 3 Months of Life
Respiratory Syncytial Virus
The peak incidence of this viral pathogen is in the first 6 months of life. Respiratory syncytial virus (RSV) typically occurs annually during the winter months. The spectrum of disease includes significant bronchiolitis and pneumonia in infants and younger children to a mild upper respiratory infection in older children. Patients with underlying conditions such as bronchopulmonary dysplasia, congenital heart disease, or underlying immunodeficiency are at risk for a more severe course.
RSV is diagnosed rapidly using a direct fluorescent antibody on nasal secretions. An aerosolized antibiotic agent, ribavirin, has been used in the treatment of RSV disease in infants. The use of ribavirin remains the subject of continuing debate. Citing new evidence, the American Academy of Pediatrics changed its recommendation in the 1990s regarding the use of ribavirin and now has a less stringent “may be considered” recommendation for its use in RSV infections in children with underlying conditions such as immunodeficiency, congenital heart disease, or chronic lung disease. Children with less serious disease need only supportive treatment.
Parainfluenza viruses are very similar to the disease caused by RSV infection, but usually seen in the summer months. These viruses frequently cause croup but may also cause lower respiratory disease.
Streptococcus pneumoniae Infection
Pertussis is seen in this age group owing to absent or incomplete immunizations. Patients can present first with coryza (catarrhal stage) and then progress to cough (paroxysmal stage). Infants may also present with apnea or seizures. Lymphocytosis is frequently seen. The organism is difficult to culture and often not present
during the paroxysmal phase of the illness. The diagnosis is made by direct immunofluorescence or polymerase chain reaction of nasal secretions. Treatment is with erythromycin. Newer macrolide antibiotics can be used, although there is less experience with these drugs.
Pneumonia in Children 4 Months to 5 Years of Age
Viral pathogens again predominate in this age group, with RSV, parainfluenza, influenza, and adenovirus being common pathogens. The primary bacteria causing pneumonia in infants and children remains S. pneumoniae. Some studies also report M. catarrhalis, and nontypeable H. influenzae as pathogens.
Pneumonia in Children 5 Years of Age and Older
In this age group, the atypical pneumonias begin to be important agents. S. pneumoniae also remains a major cause of lower respiratory infection in this age group.
The atypical pathogens are treated differently from the other bacterial pneumonias, relying on the use of either tetracycline or macrolide antibiotics. Efforts have been made to devise a clinical scoring system that will identify the atypical pathogen in the moderate to severely ill patient with community-based pneumonia. Some studies have found high temperature and previous unsuccessful therapy with β-lactam antibiotics as being predictive of atypical pneumonia. Other studies have not been able to differentiate reliably between the two groups of lower respiratory infection. If there is a concern regarding etiology in a moderately to
severely ill patient, specific testing by serology, urinary antigen, or bronchoscopy is advised.
Treatment of Community-acquired Pneumonia
The evaluation and empiric treatment of pediatric community-acquired pneumonia has been the subject of numerous reviews. Investigators agree that clinical evaluation forms the foundation for practice. If a child is nontoxic and has obvious signs of a viral infection (such as rhinorrhea and nonexudative pharyngitis), no antibiotics are required, and close follow-up is advocated.
In a young child who is suspected of having a bacterial illness (i.e., persistent fever), but who has good hydration status and adequate oxygenation, empiric antibiotics are appropriate. Treatment with oral antibiotics is usually directed against S. pneumoniae because this is the most common cause of bacterial pneumonia in children. During the past decade, alterations in penicillin-binding proteins have led to increasing resistance of the pneumococci to both penicillin and the cephalosporins. Increasing minimal inhibitory concentration (MIC) to β-lactam antibiotics has resulted in new recommendations for the treatment of S. pneumoniae meningitis. Lower respiratory infection differs from meningitis in that there is no blood–brain barrier and therefore no reduction of antibiotic concentration in the infected space. The MIC of S. pneumoniae can thus be interpreted differently in pneumonia than in meningitis. In regard to penicillin, S. pneumoniae is reported as susceptible (MIC ≤ .06 µg/mL), intermediate (0.12 to 1.0 µg/mL), and resistant (>2.0 µg/mL). Unlike the therapy of meningitis, lower respiratory infections caused by intermediate strains will usually respond to penicillin and other β-lactam antibiotics. Pneumonia caused by resistant strains may not be effectively treated by penicillin or even third-generation cephalosporins. In these cases, alternate agents such as vancomycin or fluoroquinolones may be required. An S. pneumoniae isolate with an MIC to cefotaxime of 2.0 µg/mL or less can be treated with a third-generation cephalosporin.
In children older than 5 years of age, consideration of M. pneumoniae is required. Empiric treatment with a macrolide has been proposed for this age group. Pneumococcal resistance to macrolide antibiotics is also increasing, with two separate mechanisms of resistance identified.S. pneumoniae resistance to macrolide may be secondary to alterations in drug-binding sites or the development of active drug efflux. Although the overall in vitro resistance rate of S. pneumoniae to macrolides approaches 30%, it is not clear whether this correlates with clinical failure. It has been suggested that the clinical effect of in vitro macrolide resistance may depend on the precise mechanism of resistance present in the infecting organism or the presence of concurrent bacteremic disease. Macrolide therapy for presumed S. pneumoniae lower respiratory disease is reasonable in a stable outpatient population; children who are toxic, bacteremic, or fail to improve after macrolide monotherapy may warrant combination therapy with β-lactam antibiotics.
TABLE 13.1. Empiric Treatment for Community Acquired Pneumonia
In children who are toxic appearing, hypoxic, or require intravenous hydration, treatment is directed toward both atypical pathogens and bacteria causing severe pyogenic pneumonia, including Streptococcus pyogenes, S. pneumoniae, and Staphylococcus aureus. In treating pneumonia in a toxic-appearing child, the clinician must remember the increasing incidence of resistance in both community S. aureus(MRSA) and S. pneumoniae (MIC to penicillin ≥ 2.0 µg/mL). Vancomycin and a third-generation cephalosporin are reasonable initial therapy in the case of a potentially life-threatening lower respiratory infection. For children who are not in critical condition, a third-generation cephalosporin, clindamycin, or ampicillin-sulbactam (Unasyn) is an acceptable choice. A macrolide antibiotic will also be needed for treatment of the atypical pneumonia pathogens (Table 13.1).
Parapneumonic Effusions and Empyema
Bacterial pneumonia can have a variety of complications. A parapneumonic effusion refers to pleural fluid that accumulates in association with bacterial pneumonia. A certain percentage of these effusions will undergo a secondary process in which the fluid becomes purulent and, if untreated, will actually form a pleural peel that adheres to the surface of the lung. At this stage, the parapneumonic effusion is typically referred to as an empyema.
The three major bacteria responsible for parapneumonic effusion and empyema are S. aureus, S. pneumoniae, and S. pyogenes (group A streptococci). The rate of empyema varies with each particular bacterium; group A streptococcus pneumonia progresses to empyema in up to 40% of patients, whereas less than 5% of patients with pneumococcal pneumonia develop an empyema.
Clinical and Radiographic Features
Patients with parapneumonic effusions or empyema often continue to have spiking temperatures despite appropriate antibiotics. Chest x-ray is usually the initial step in evaluating this condition. If there is extensive effusion, the chest radiograph may appear as a “whiteout,” the entire side of the lung becomes opaque. In these cases, computed tomography of the chest is excellent in distinguishing pleural from parenchymal disease. Computed tomography can also detect the presence of large loculations (Figs. 13.1 and 13.2).
FIG. 13.1. Plain radiograph showing complete opacification of left lung consistent with empyema.
FIG. 13.2. Computed tomography scan revealing large left pleural empyema.
Once a pleural effusion has been identified, analysis of the pleural effusion is necessary. A common mistake is to delay evaluation and drainage of pleural effusion; this can ultimately lead to further formation of loculations and greater difficulty in ultimately clearing the infection.
Because treatment of empyema requires not only antibiotic therapy but also surgical drainage, there is great interest in the pleural fluid parameters that define the diagnosis of empyema. Aspiration of frankly purulent material, a positive Gram stain, or positive pleural culture is enough to make a definitive diagnosis. In the absence of frankly purulent material, there are changes that, of themselves, warrant consideration for drainage. A pleural fluid pH of less than 7.2, a glucose level of less than 40 mg/dL, and a lactate dehydrogenase (LDH) level of more than 1,000 IU identify a complicated parapneumonic effusion that requires drainage.
Pleural fluid should be obtained under sterile conditions and sent for a variety of specific tests. It should be plated on both aerobic and anaerobic media. Determination of the pH of the pleural fluid is vital and should be collected anaerobically and transported on ice to the laboratory. Cell count and chemical analysis are also important.
TABLE 13.2. Diagnosis and Management of Empyema
Treatment of Empyema
In patients in whom an empyema is diagnosed, either by pH, LDH, glucose, Gram stain, or the documentation of gross pus within the pleural space, at the very least a chest tube is required for continued drainage. If the initial suspicion for empyema is high, initial diagnostic thoracentesis may be replaced by immediate placement of a chest tube. In most cases, chest tubes can be removed when the amount of pleural fluid draining from the tube has decreased and the effusion has resolved on plain x-ray. There will be a percentage of patients who do not clear the empyema with chest tube drainage alone. These patients are candidates for further surgical intervention. Patients who require surgical intervention typically have persistent fever, toxicity, and minimal chest tube drainage. These patients often have developed loculations that are not amenable to drainage by chest tube.
In the past, thoracotomy and decortication was done. In this procedure, the chest is opened, pleura removed, and purulent material evacuated from the pleural space. Video-assisted thoracic surgery (VATS) is being increasingly used; this procedure has the advantage of a smaller surgical incision and fewer complications. A greater number of pediatric surgeons are advocating earlier use of VATS, even before placement of a chest tube, for the initial treatment of pediatric empyema. As experience with VATS increases, it may be prudent to involve an experienced pediatric surgeon as soon as a pleural empyema is diagnosed (Table 13.2).
Campbell JD, Nataro JP. Pleural empyema. Pediatr Infect Dis J 1999;18(8):725–726.
Gaston B. Pneumonia. Pediatr Rev 2002;23(4):132–140.
McCracken GH. Diagnosis and management of pneumonia in children. Pediatr Infect Dis J 2000;(9):924–928.
Rodriguez JA, Hill CB, Loe WA, et al. Video-assisted thoracoscopic surgery for children with stage II empyema. Am Surg 2000;66(6):569–572.