Bacterial Pneumonia Workup

Diagnostic testing in patients with suspected pneumonia is driven mostly by the possibility that the results would significantly alter empiric therapy and management decisions and whether the test is likely to have a high yield.^{[16, 46]} Diagnostic testing is also useful in classifying the severity of illness and site-of-care decisions (outpatient vs inpatient vs intensive care unit [ICU]). The most obvious indication for extensive diagnostic testing is in the critically ill patient.^{[16, 47]}

Various tools to assess the severity of disease and risk of death exist and are in wide use, including the PSI/PORT (ie, pneumonia severity index/Patient Outcomes Research Team score), the CURB-65 system (ie, confusion, urea, respiratory rate, blood pressure, and age >65 y), and the APACHE (ie, acute physiology and chronic health evaluation), among others discussed under Risk Stratification in the Clinical Presentation section. A number of laboratory values are commonly used in the calculation of these risk indices.

Hyponatremia (sodium level < 130 mEq/L) and microhematuria may be associated with Legionella pneumonia. Sputum examination may be supplemented by using a Legionella -specific fluorescent antibody. However, this technique has a high false-negative rate.

Urinary antigen testing for Legionella serogroup 1 organisms is accurate. However, as many as 30% of infections are not caused by serogroup 1 organisms. Pneumococcal antigen tests for serum, urine, and saliva samples have been developed. A Legionella serum antibody titer of 1:128 or more is suggestive of the diagnosis. Antigen-antibody testing has little clinical effect in an emergency department setting, although it may help in recalcitrant or unclear cases.

Imaging studies are generally helpful in detecting suspected pneumonia and identifying the presence of complications; however, only occasionally do radiologic studies suggest specific pathogens.^[48]

The following laboratory tests may not be useful for diagnostic purposes but are useful for classifying illness severity and site-of-care/admission decisions^{[32, 38, 49, 50]} (see Risk Stratification under Clinical Presentation):

• Serum chemistry panel (sodium, potassium, bicarbonate, blood urea nitrogen [BUN], creatinine, glucose)
• Arterial blood gas (ABG) determination (serum pH, arterial oxygen saturation, arterial oxygen pressure) – Hypoxia and respiratory acidosis may be present.
• Venous blood gas determination (central venous oxygen saturation)
• Complete blood cell (CBC) count with differential
• Serum free cortisol value
• Serum lactate level

A pulse oximetry finding of < 95% indicates significant hypoxia, and an elevated C-reactive protein (CRP) level may be predictive of more serious disease.^[51] However, CRP has not been clearly shown to differentiate bacterial versus viral illness.

CBC count with differential

Leukocytosis with a left shift may be observed in any bacterial infection; however, its absence, particularly in patients who are elderly, should not cause the clinician to discount the possibility of a bacterial infection.

Leukopenia (usually defined as a WBC count < 5000 cells/µL) may be an ominous clinical sign of impending sepsis.

Coagulation studies

An elevated international normalized ratio (INR) has been associated with more severe illness. This finding may herald the development of disseminated intravascular coagulation.

Blood cultures

Blood cultures should be obtained before administering antibiotic therapy. These cultures require 24 hours (minimum) to incubate. When the findings are positive, they correlate well with the causative agent.

Unfortunately, blood cultures show poor sensitivity in pneumonia; findings are positive in approximately 40% of cases. Even in pneumococcal pneumonia, the results are often negative. Their yield may be better in patients with more severe cases.

The findings probably have minimal clinical effect in treating bacterial pneumonia. Indeed, the use of blood cultures only rarely dictates a change in antibiotic use.

Sputum Gram stain and culture should be performed before initiating antibiotic therapy (if a good-quality, contaminant-sparse specimen containing < 10 squamous epithelial cells per low-power field can be obtained). The white blood cell (WBC) count should be more than 25 per low-power field.

A single predominant microbe should be noted at Gram staining, although mixed flora may be observed with anaerobic infections.

However, often, patients cannot produce an adequate specimen. Many specimens produced are so contaminated by oral materials that they are unusable.

Cultures of the sputum have similar limitations. To be accurate, only specimens that have been examined microscopically and that have satisfied the criteria above should be submitted for culturing.

In intubated patients admitted to the ICU, some researchers suggest that upper airway samples and cultures obtained initially on admission may aid in directing antibiotic therapy should ventilator-associated pneumonia (VAP) ensue during the first several days of admission.^[52]

Fiberoptic bronchoscopy has largely replaced transtracheal aspiration for obtaining lower respiratory secretions.

Chest radiography is considered the criterion standard for diagnosing the presence of pneumonia: The presence of an infiltrate is required for the diagnosis. However, it must be noted that the accuracy of plain chest radiography for detecting pneumonia decreases depending on the setting of infection (see Background).

In H influenzae pneumonia, pleural effusion is present in approximately half of infected individuals.

Go to Imaging Typical Bacterial Pneumonia and Imaging Atypical Bacterial Pneumonia for complete information on these topics.

Lobar pneumonia

Radiographically, lobar pneumonia, or focal or nonsegmental pneumonia, is manifested as nonsegmental, homogeneous consolidation involving one, or less commonly, multiple lobes. Larger bronchi often remain patent with air, creating the characteristic air bronchogram. Lobar consolidation is pathologically the result of the rapid production of edema fluid with minimal cellular reaction, occurring initially in the lung periphery and then spreading between acini through the pores of Kohn and canals of Lambert.

S pneumoniae infection is characterized by homogenous parenchymal lobar opacities with air bronchograms. This condition can occasionally manifest as a round opacity stimulating a pulmonary mass, called round pneumonia. Frank consolidation and air bronchograms have been associated with a higher incidence of bacteremia. The classic findings are in the right lower lobe, but aspiration pneumonia also has a characteristic distribution based on patient positioning.

Aspiration pneumonia radiographic findings may be seen in the gravity-dependent portions of the lungs (affected by patient positioning); the right lung is affected twice as often as the left lung. In recumbent patients, the findings are in the posterior segments of the upper lobes, and, in upright patients, the basal segments of the lower lobes are often affected.

K pneumoniae infection may show radiographic evidence of lobar expansion with bulging of interlobular fissures due to voluminous inflammatory exudate, as well as cavitations. Klebsiella has a tendency to occur in the upper lobes.

Legionella has a predilection for the lower lung fields. Radiologic resolution tends to lag far behind clinical improvement (8 wk to clear).

The following radiographs depict examples of lobar pneumonia.

Bronchopneumonia

Bronchopneumonia, also known as multifocal or lobular pneumonia, is radiographically identified by its patchy appearance, with peribronchial thickening and poorly defined air-space opacities. As illness becomes more severe, consolidation involving the terminal and respiratory bronchioles and alveoli results in the development of centrilobular nodular opacities or air-space nodules; this consolidation can develop further and coalesce to give a lobular or lobar pattern of involvement.

Typically, air bronchograms are absent. The pathogens known to cause this pattern of pneumonia are particularly destructive; thus, abscesses, pneumatoceles, and pulmonary gangrene are often seen. Pathologically, bronchopneumonia stems from inflammation of large airways (bronchitis) with patchy (lobular) involvement.

In S aureus pneumonia, lobar enlargement with bulging of interlobular fissures can be seen in severe cases. Abscesses, cavitations (with air-fluid levels), and pneumatoceles are commonly seen, and 30-50% of patients develop pleural effusions, half of which are empyemas. Note that cavitation and associated pleural effusions are also observed in cases of anaerobic infections, gram-negative infections, and tuberculosis.

In P aeruginosa infection, the radiographic findings tend to be nonspecific and difficult to differentiate from underlying lung disease. Usually all the lobes are involved, with a predilection for the lower lobes, and necrosis and cavitation occur frequently. In addition, pulmonary vasculitis can produce areas of pulmonary infarction that radiographically resembles invasive aspergillosis.

Interstitial pneumonia

Interstitial pneumonia is classified as focal or diffuse. Pathologically, the radiographic pattern results from edema and inflammatory cellular infiltrate into the interstitial tissue of the lung. The pathologic development of interstitial pneumonia generally takes 1 of 2 forms: (1) an insidious infectious course that results in lymphatic infiltration of alveolar septa without parenchymal abnormality or (2) acute or rapidly progressive disease that results in diffuse alveolar damage affecting the interstitial and air spaces. Radiographically, the disease manifests with a reticular or reticulonodular pattern.^{[53, 54]}

The role of computed tomography (CT) scanning in the diagnosis of pneumonia is not yet well defined. For inpatients, CT scanning may identify pulmonary infections earlier than plain radiography.^[47] In most cases, it can be helpful in the analysis of more complex lung findings and the evaluation of other intrathoracic structures. In unclear cases, high-resolution CT scanning of the lungs may aid in the diagnosis.

CT patterns of disease may be broken down into abnormalities that cause either increased or decreased lung opacity.^[55] Abnormalities that cause increased lung opacity include the following:

• Nodular pattern, based on the anatomy of the secondary pulmonary lobule – A centrilobular pattern is further characterized by the presence or absence of tree-in-bud morphology, the presence of which is almost always seen in infection; its absence is likely to expand the differential beyond infectious processes. Other nodules include perilymphatic nodules and random nodules.
• Linear patter - Interlobular septal thickening (smooth, nodular, irregular), parenchymal bands, subpleural lines, and irregular linear opacities may be seen.
• Reticular pattern
• Ground-glass opacity
• Consolidation

Abnormalities that cause decreased lung opacity include the following:

• Bronchiectasis
• Emphysematous change (centrilobular, panlobular, paraseptal, irregular)
• Honeycomb lung and cystic disease
• Mosaic perfusion and inhomogeneous lung opacity

Go to Imaging Typical Bacterial Pneumonia and Imaging Atypical Bacterial Pneumonia for complete information on these topics.

Ultrasonography is useful in evaluating suspected parapneumonic effusions, especially if septations are present within the fluid collection that may not be visible on CT scans. Ultrasonography also has great utility for direct operation by physicians at the patient bedside.^[48]

Lung tissue can be visually evaluated and bronchial washing specimens can be obtained with the aid of a fiberoptic bronchoscope. Protected brushings and bronchoalveolar lavage (BAL) can be performed for fluid analysis and cultures.

BAL can also be performed without the use of a bronchoscope.

This is an essential procedure in patients with a parapneumonic pleural effusion. Obtaining fluid from the pleural space for laboratory analysis allows for the differentiation between simple and complicated effusions. This determination may help guide further therapeutic intervention.

Pleural effusions and frank empyema fluid should also be sent for Gram stain and culture.

• Urine assays also available for the rapid detection of Legionella and pneumococcal antigens. These fast card-type assays have been developed in recent years, may be performed at the bedside, and may be useful in unclear cases or when the choices for antimicrobial therapy are limited.
• Sputum and/or urinary antigen tests are available for Legionella pneumophila.
• Sputum, serum, and/or urinary antigen tests are available for Streptococcus pneumoniae.
• Immune serologic tests have been developed for Mycoplasma pneumoniae, Chlamydophila pneumoniae, L pneumophila, and Coxiella burnetii. However, the results are usually not available until several weeks after the infection, which makes these tests less useful.
• Nucleic acid detection (eg, polymerase chain reaction [PCR]) is still in development, but this test poses a major weakness in its extreme sensitivity and potential for false-positive results, also rendering it less useful than other tests.

Histologic inflammatory lung changes are best described according to the pattern of infection.^[56]

Lobar pneumonia

Four stages of inflammatory response are classically described, as follows:

1. Congestion: This stage is characterized by vascular engorgement, intraalveolar fluid, and numerous bacteria. The lung is heavy, boggy, and red.
2. Red hepatization: In this stage, massive confluent exudation develops, with red blood cells, leukocytes, and fibrin filling the alveolar spaces. The affected area appears red, firm, and airless, with a liverlike consistency.
3. Gray hepatization: This stage is characterized by progressive disintegration of red blood cells and the persistence of a fibrin exudate.
4. Resolution: The consolidated exudate within the alveolar spaces undergoes progressive enzymatic digestion to produce debris that is later resorbed, ingested by macrophages, coughed up, or becomes organized by fibroblasts growing into it.

Bronchopneumonia

Bronchopneumonia typically consists of foci of consolidation resulting from a suppurative, leukocyte-rich exudate that fills the bronchi, bronchioles, and adjacent alveolar spaces. In terms of gross appearance, well-developed lesions may be 3-4 cm in diameter, dry, granular, and grayish-red to yellow, with poorly demarcated margins.

Interstitial pneumonia

The typical lung inflammatory response to the atypical bacteria results in an interstitial picture. Alveolar septa become widened and edematous and usually have a mononuclear inflammatory infiltrate of lymphocytes, histiocytes, and plasma cells. Neutrophils may also be present in acute cases. Pleuritis may result if the underlying inflammation extends to the pleural surface of the lung.