Most episodes of pneumonia resolve spontaneously or with limited medical therapy. Certain groups, those with underlying diseases, those with anatomic or functional abnormalities, or immunocompromised patients, are more prone to complications, but the normal host is not spared. The most concerning are those who require surgical intervention or prolonged therapy. Guideline development has provided effective management options. This chapter is divided into discussions of empyema, lung abscess, and other complications ( Box 23.1 ).
Abscess
Apnea or respiratory failure
Acute respiratory distress syndrome
Atelectasis
Bacteremia and sepsis
Bronchopulmonary fistula
Dissemination (skin, solid organs, joints, central nervous system)
Empyema
Hilar adenopathy
Infarction
Necrotizing pneumonia
Pericardial effusion
Pleural effusion
Pneumothorax
Pneumatocele
Recurrent pneumonia
Pneumothorax
Pericardial effusion
Empyema
Collections of fluid in the pleural space have been described in the literature as transudates, pleural effusions, exudates, purulent pleurisy, parapneumonic effusions, empyema, and complicated empyema. A great deal of inexactitude exists in the use of these terms, and comparing methods of diagnosis and management from one study to another is difficult. The term transudative pleural effusion refers to fluid in the pleural space that is a nonpurulent effusion and typically nonpneumonic in origin. The term purulent effusions refers to effusions that are more cellular (exudative) and typically pneumonic in origin. The term empyema describes purulent effusions with chemical or microbial evidence of a severe process requiring drainage. Complicated empyema describes the processes associated with loculations or a fibropurulent rind requiring aggressive manipulations for cure. Parapneumonic effusion is a general term referring to any pleural exudative process resulting from an inflammatory process in the lung. Although these definitions are arbitrary and the literature is inconsistent in the use of these terms, a standard approach to the classification of these four types of effusions and their management has evolved.
The first description of parapneumonic infection is attributed to Hippocrates, who in the 4th century bce advocated incision and drainage of empyema 2 weeks after the onset of symptoms. Since then, the physiology and microbiology of effusions have been described, and parapneumonic diseases and the management of fluid collections of the pleural space have been defined. In most cases, the directions of Hippocrates still are relevant: “set him upon a stool, which is not wobbly; someone should hold his hands, then shake him by the shoulders and listen to see on which side a noise is heard. And right at this place—preferably on the left—make an incision, and then it produces death more rarely.”
This description of open drainage to normal atmospheric pressures was recognized to be associated with significant mortality rates from hemodynamic instability in 1918, when the Empyema Commission of the United States Army recommended that the practice be abandoned ; thereafter, closed-tube drainage was introduced and mortality rates decreased. Developments in radiology, antimicrobials, and surgery resulted in further improvements in care. Today the major cause of mortality is related to the underlying disease because the management of parapneumonic effusions in children is largely successful and without residual morbidity. The physician must understand the risks for development of pleural effusion and empyema and indications for thoracentesis, the implications of the results of pleural fluid studies, and the optimal medical and surgical management of the effusions and empyema.
Epidemiology
Parapneumonic effusions are known and expected complications in children with respiratory tract infections. The frequency of effusions may be up to 20% of patients with viral or mycoplasmal pneumonia and 75% of patients with proven Staphylococcus aureus pneumonia. Empyema has been reported to occur in 6.3 to 23 per 1000 admissions of children. Reports before the introduction of the 13-valent pneumococcal conjugate vaccine suggested rates were increasing.
Pathophysiology
The pleurae are mesodermally derived tissues approximately 30 to 40 µm thick and permeable to liquid and gas. The parietal pleurae, which adhere to the chest wall, are fed from the intrathoracic and superior phrenic arteries and have sensory innervation. The visceral pleurae are splanchnic in origin, with blood flow from the pulmonic and pericardiophrenic arteries and no sensory innervation.
Parietal lymphatics absorb most of the excess fluids in pathologic situations and play an important role in normal physiology as well, removing 250 to 500 mL/day in adults. They are the only mechanism for absorbing cells and other debris from the pleura.
The raison d’être of the pleural space is unknown. Some mammals do not have a pleural space. In the normal situation, pleural fluid in small amounts is a necessary requirement for optimal lubrication of the pleural space and for mechanical coupling of the lung and chest wall. The accumulation of excess fluid (i.e., effusion) occurs in a limited set of circumstances, through excess production or deficient absorption. Increased production occurs when vessels are leaky (e.g., in septic shock) or active secretion of fluid with mesothelial inflammation (e.g., pleural infection) is present. Decreased absorption occurs with decreased oncotic pressure (e.g., nephrosis), increased pulmonary hydrostatic pressure (e.g., congestive heart failure), or lymphatic obstruction (e.g., malignancy).
The mechanisms behind pleural effusions may vary among different infectious diseases. Effusion can be a “sympathetic” pleural response to a bacterial infection in the lung associated with inflammatory cytokines and altered venous or lymphatic drainage because of local edema. Direct or hematogenous extension of a bacterial process can occur in the pleura. Mycoplasma is particularly pathogenic in patients with sickle-cell disease, presumably because of pulmonary sludging, which increases pulmonic venous drainage pressures and results in accumulation of effusions. In pneumococcal disease, effusions often develop several days after the acute infection, when bacteria no longer can be recovered. These effusions may be related to immune complex disease. In patients with tuberculosis, the most common cause of pleural effusions is thought to be the rupture of an old granuloma into the pleural space, with a hypersensitivity response similar to the skin test response, which partly explains the low yield in cultures.
After an inflammatory process is initiated, it tends to progress through three classic stages. The first, defined as a purulent effusion, is the acute exudative stage, with a thin pleural exudate characterized by normal glucose, lactate dehydrogenase (LDH), and pH. The second transitional fibropurulent stage, categorized as empyema, is characterized by turbid fluid, decreased glucose concentration (<60 mg/dL) and pH (7.2 to 7.35), and elevated LDH (>200 U/L). The third chronic organizing stage is notable for a very low pH (<7.2) and glucose level (<40 mg/dL), LDH concentration greater than 1000 U/L, and development of loculations and peel. This fluid is found in patients with complicated empyema . Intervening early (<4 days) before the organizational stages of disease has been proposed to lead to a better prognosis. However a recent review concludes that, in the modern era, these data correlate poorly with management strategies and may not carry predictive weight.
Analysis of the pleural fluid is most helpful when the underlying disease is unknown or when a primary pulmonic process is suspected. When patients have effusions caused by hydrostatic imbalance, the effusion is a transudate. Its protein and cell count do not exceed the range of normal pleural fluid (5000 cells/mm 3 and <2 g of protein). Patients who have an active inflammatory process may have an exudate, defined by excess protein and cells. In children, the most common cause of exudative pleural processes is pneumonia. In adults, most pleural effusions are related to congestive heart failure or malignancy, but pneumonia is the most common cause of empyema. Table 23.1 summarizes the general differences among pleural effusions.
| Characteristics | Transudative | Purulent Effusion | Empyema | Complicated Empyema |
|---|---|---|---|---|
| Appearance | Serous | Thin exudate | Turbid | Thick pus |
| Mean WBC | 1000 | 5300 | 25,500 | 55,000 |
| PMN (%) | 50 | >90 | >95 | >95 |
| Protein (fluid/serum ratio) | <0.5 | >0.5 | >0.5 | >0.5 |
| LDH (fluid/serum ratio) | <0.6 | >0.6 | >0.6 | >0.6 |
| LDH (IU/L) | >200 | >200 | >1000 | |
| Glucose (mg/dL) | >60 | <60 | <60 | <40 |
| pH a | 7.4–7.5 | 7.35–7.45 | 7.2–7.35 | <7.2 |
| Imaging | Fluid | Fluid | Fluid | Loculations, thick peel, scoliosis |
Box 23.2 lists causes of effusions. Some of these causes, particularly iatrogenic causes, such as invasive procedures and drugs, are important to consider in the differential diagnosis of a difficult case. Others are associated with specific syndromes, such as acute respiratory distress syndrome and yellow nail lymphedema syndrome. Motor vehicle accidents have been identified as a common cause of serosanguineous effusions when disruption of normal mechanical lung function and hematoma occur.
Capillary leak
- •
Sepsis syndrome
- •
Vasculitis associated with immune complex disease
- •
Connective tissue diseases
- •
Inflammatory bowel disease
- •
Malignancy (lymphoreticular, sarcoma, neuroblastoma)
- •
Toxins (e.g., TSST-1)
- •
Drugs (amiodarone, bleomycin, isoniazid, methotrexate, nitrofurantoin, phenytoin)
- •
Myxedema
- •
Trauma
- •
Increased hydrostatic pressure
- •
Congestive heart failure
- •
Sickle-cell disease
- •
Pulmonary venous hypertension
- •
Superior vena cava syndrome
- •
Pregnancy
- •
Decreased oncotic pressure
- •
Nephrosis
- •
Cirrhosis
- •
Protein malnutrition
- •
Obstructed lymphatics
- •
Congenital lymphangiectasia
- •
Yellow nail syndrome
- •
Radiation injury
- •
Neoplasia (metastatic disease)
- •
Pleural inflammation
- •
Pneumonia
- •
Lung abscess with pleural fistula
- •
Pleural infection (e.g., tuberculosis)
- •
Esophageal rupture
- •
Pancreatitis
- •
Iatrogenic
- •
Drugs
- •
Central line misplacement
- •
TSST-1, Toxic shock syndrome toxin–1.
Microbiology
Among children with parapneumonic effusions, studies continue to evolve regarding the frequency with which effusions occur and how many are associated with particular microbes. Although respiratory viruses infrequently cause symptomatic effusions, the sheer number of cases and the presence of asymptomatic cases likely would implicate viral infection as the most common cause. Definite viral disease enhanced by molecular diagnosis has been commonly associated with rhinovirus, RSV, influenza, enterovirus, cytomegalovirus, Epstein-Barr virus, measles, and adenovirus but co-infection of bacterial and viral species is also found. Other pathogens, such as Mycoplasma and Chlamydia, are more difficult to diagnose but may account for a significant number of pneumonic infections in older children and adolescents. Viral, mycoplasmal, and chlamydial organisms uncommonly are identified in patients with effusions requiring intervention.
The bacteriology of empyema is constantly changing. Several past articles have established the role of various bacterial pathogens in childhood effusions. A study of 227 children by Freij and colleagues published in 1984 found S. aureus (29%), Streptococcus pneumoniae (22%), and Haemophilus influenzae (18%) to be the three most frequent causes of parapneumonic effusions. Subsequent studies show that the frequency of these pathogens has been affected by vaccine and antibiotic use. Pathogen frequency in different eras is shown in Table 23.2 . H. influenzae and S. pneumoniae vaccines have led to a decreased incidence of vaccine-specific strains as a cause of empyema, whereas methicillin-resistant S. aureus (MRSA) has emerged as a frequent cause of community-acquired disease, with rates in Texas doubling from 2001 to 2009. Among cases of S. pneumoniae infections, nonvaccine strains are increasing in number and severity. Recent studies using a U.S. national hospital discharge database suggest that S. pneumoniae remains the most common etiology of bacterial empyema. Certain groups of children (e.g., neonates, immunocompromised hosts, patients with preexisting chest tubes that become infected with nosocomial pathogens, patients with a ruptured viscus, and patients with foreign body aspiration) are at higher risk for acquiring gram-negative infections. Some pathogens seem to produce empyema at very high rates such as S. aureus and the Streptococcus milleri group.
| Site/Years (No. Patients) | Staphylococcus aureus | Streptococcus pneumoniae | Haemophilus influenzae | Other Pathogens | Sterile | Reference |
|---|---|---|---|---|---|---|
| Dallas/1964–82 (227) | 29 | 22 | 18 | 8 | 24 | |
| Nashville/1977–89 (61) | 11 | 34 | 3 | 11 | 39 | |
| Washington, DC/1973–85 (33) | 15 | 12 | 21 | 52 | Not reported | |
| Israel/1972–81 (37) | 14 | 41 | Not reported | 35 | 11 | |
| Dallas/1992–98 (135) | 8 | 32 | 1 | 13 | 46 | |
| Houston/2001–02 (47) | 19 | 9 | Not reported | 4 | 68 | |
| Australia/2007–09 a (145) | 9 | 51 | 3 | 2 | 35 |
Administration of antibiotics before the diagnosis of empyema is made influences the recovery of organisms. Reports suggest 43% to 61% fewer positive cultures. Pretreatment with antibiotics may be associated with a decrease in the number of positive blood cultures and in the number of patients from whom S. pneumoniae isolates are recovered.
Anaerobes were sought carefully by Brook and Frazier, who found them infrequently in patients younger than 6 years of age. The anaerobes rarely were found in patients with primary pneumonia, occurring most often in patients with lung abscess and aspiration pneumonia. In older patients (7 to 17 years old), anaerobes were recovered as isolated pathogens in 44% of cases. Virtually every bacterial organism has been associated with pleural effusion at one time or another. Brucella, Francisella tularensis, and Yersinia enterocolitica may be associated with the development of pleural effusions. The diagnosis in such cases often is suggested by a unique history in the patient.
Mycobacterial effusions are uncommon in U.S. children but are well described. In four published reviews, only two patients (from Nigeria) were reported to have Mycobacterium tuberculosis . In a series of 303 children younger than 2 years of age with tuberculosis, 3.3% had an effusion. In adolescents with tuberculosis, the incidence of effusion likely approximates that of adult disease. In one series of adult patients with primary tuberculous disease, pleural effusion occurred in 29% of cases. In another adult series, primarily of reactivation disease, pleural effusion occurred in 1% of the patients. Whether coinfection with human immunodeficiency virus (HIV) is increasing the incidence of effusion is controversial.
Histoplasmosis has been associated with pleural effusion in 0% to 6% of childhood histoplasmosis cases. Blastomycosis has been associated with pleural effusions in 0% to 40% of cases. Effusions resulting from other fungi (e.g., Coccidioides, Aspergillus ) have been described. Parasitic diseases manifesting with effusions are uncommon but are found in patients with Entamoeba histolytica disease, most often from rupture of a hepatic abscess into the pleural space. Echinococcal disease also has been reported.
Pleural effusion associated with adult HIV infection has been reported in 14.6% of hospital admissions in one series in which 67 of 160 cases were infectious. Of those cases, 50 were associated with bacterial pneumonia, 10 with tuberculosis, and 5 with Pneumocystis jiroveci pneumonia. Another report on patients infected with HIV suggested that empyema was seen primarily in patients with intravenous drug abuse. We have not seen empyema in our pediatric HIV-infected patients, perhaps because of the more recent use of more effective antiretroviral therapy.
Drug resistance in community-acquired pneumonia complicates the management of parapneumonic effusions. Intermediate or fully resistant S. pneumoniae was found in 12.8% and 10.1% of isolates in a study from multiple U.S. pediatric centers from 1993 to 2000; 7.5% were cephalosporin resistant. Recent studies of nasal colonization in Israeli infants show that, post introduction of PCV13, the number of new resistant isolates has decreased, as noted in other countries. A 2012–13 U.S. study in middle ear isolates showed 33% nonsusceptible strains, not a change from prior studies. MRSA is a growing concern in empyema. Reports since 2000 have shown that 22% to 78% of isolates of S. aureus were methicillin resistant, with many being community acquired.
Diagnosis
Clinical Presentation
Symptoms most specific for parapneumonic processes are dyspnea and pleuritic pain. Dyspnea occurs when the volume of the effusion mechanically interferes with breathing or when pain prevents adequate gas exchange. Pain occurs with irritation of the parietal pleura and on inspiration (i.e., pleurisy). Fever is generated by the inflammatory response and pathogen-specific components (e.g., lipopolysaccharide, toxins). With an acute bacterial process, the fever can be high and hectic, mimicking the fever that occurs with an abscess. Patients in the chronic organizational phase generally have less fever. Cough and malaise are secondary symptoms. Hemoptysis and purulent sputum also may occur. The onset of symptoms of a purulent effusion may be delayed in time and distinct from the symptoms found at the onset of the pneumonia in older children; infants usually have no symptom-free period ; in the early phases of effusions, the patient may have no symptoms.
The physical examination usually is revealing. The child is tachypneic in more than 70% of cases, but breathing is shallow as a result of the child’s attempt to minimize pain. Fever and cough usually occur in more than 90% of patients with purulent effusions. The patient may appear toxic, with acute infection. Patients often posture toward the affected side. Classically, auscultation reveals a decrease in breath sounds and occasionally detects a pleural rub, but pleural rubs often are absent in a very young child. Rales from an associated pneumonia may be heard. Depending on the stage of the process, percussion may reveal a level of dullness associated with free-flowing effusion. As the process organizes, it may be less evident. Empyemas can erode through the chest wall into the subcutaneous tissue (i.e., empyema necessitatis) or into a bronchus (i.e., bronchopleural fistula).
Imaging
The diagnosis most often is made by radiographic examination of the chest. Consolidation of a lobe of the lung is present, with an effusion obscuring the diaphragm. A standard posteroanterior standing view reveals blunting of the costal diaphragmatic gutter. As fluid tracks along the lateral and posterior chest wall, a meniscus configuration is seen. Distinguishing it from pleural thickening may be difficult, and in such cases, a decubitus or cross-table view of the chest allows free-flowing fluid to layer out on the dependent chest wall. In older children and adults, a decubitus layer of fluid of more than 10 mm is considered a sufficient volume of fluid to attempt to extract by thoracentesis. With large volumes of fluid (>1000 mL), compression of the lung and shift of the trachea away from the effusion ( Fig. 23.1 ) may occur. As an empyema develops and organizes, discrete pockets of fluid (i.e., loculations) may form within the pleural cavity ( Fig. 23.1C ). Occasionally, loculations are confused with lung abscess. Scoliosis also is well defined by the chest radiograph and occasionally is used as an indication for surgery. The observation of an air-fluid level in the pleural space signifies that air has been generated in the pleural space by gas-forming organisms or has entered through a pneumothorax, perforated viscus, or bronchopleural fistula.
Ultrasonography has shown great utility in providing better guidance for thoracentesis of pleural fluid. It is noninvasive, without ionizing radiation, allows empyema to be defined by showing internal echoes and septations ( Fig. 23.2 ), and is the imaging mode of choice for most pleural effusions. Transudates uniformly are anechoic, although approximately one-third of exudates also are anechoic. Ultrasonography is not as precise as computed tomography (CT) in differentiating a lung abscess from an empyema. CT and magnetic resonance imaging (MRI) occasionally are required to distinguish parenchymal from pleural disease, identify a nonopaque foreign body, or locate a fistula. CT particularly is useful in a patient whose chest radiograph shows total opacification of the lung and thus is considered by some physicians the study of choice in this situation. Ultrasound can have false-negative results, and CT can have false-positive findings on examination of pleural effusions. A recent review noted that ultrasound and CT were comparable in managing pediatric effusions.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
