In many cases, the diagnosis of pneumonia in children is based on clinical criteria, and a good physical examination can usually detect typical symptoms of pneumonia and localize the infection. Some complications such as parapneumonic effusion or abscess may be detectable through changes in percussion note or vocal fremitus, but there is a role for imaging in confirming the diagnosis when physical examination is inconclusive or difficult. The best situations to employ imaging in the setting of pneumonia are when diagnosing the infection in younger children and infants (where the signs and symptoms may be non-specific or subtle), and when assessing possible treatment failure, or complications that may require surgical intervention.
Chest radiography is the primary imaging study used to confirm the diagnosis of pneumonia.
Well-centered, appropriately penetrated, anteroposterior chest radiography is essential (see the image below), although other views may be warranted to clarify anatomic relationships and air-fluid levels.
When considering pneumonia, devote particular attention to the following:
Costophrenic angles
Pleural spaces and surfaces
Diaphragmatic margins
Cardiothymic silhouette
Pulmonary vasculature
Right major fissure
Air bronchograms overlying the cardiac shadow
Lung expansion
Patterns of aeration
Be aware that any image reflects conditions only at the instant at which the study was performed. Because lung diseases, including pneumonia, are dynamic, initially suggestive images may require reassessment based on the subsequent clinical course and findings in later studies.
Go to Pediatric Pneumonia and Congenital Pneumonia for more complete information on these topics.
Chest radiography is not always necessary, or even useful, as an aid in determining the etiology of the infection.
Several studies, in fact, have demonstrated that chest radiography is 42-73% accurate in predicting the etiology of a case of pneumonia. For example, in a study of 168 children with pneumonia, 2 radiologists who independently evaluated all chest radiographs were unable to distinguish whether the agent involved was bacterial, viral, or unidentified.[1]
A retrospective cohort study that included 4708 children who presented to the emergency department with an asthma exacerbation reported that radiographic confirmation of pneumonia was present in only 5.9% (280) of the 4708 children who underwent chest radiography.[2]
Given the frequency of nonspecific findings obtained with imaging, clinical presentation and other laboratory findings must be considered in the diagnosis of pneumonia and in the determination of the etiologic agent.
Chest radiography is indicated in an infant or toddler who presents with fever and any of the following:
Tachypnea
Nasal flaring
Retractions
Grunting
Rales
Decreased breath sounds
Respiratory distress
In older children and adolescents, the diagnosis of pneumonia is often based on clinical presentation and chest x-rays may not be necessary to make the diagnosis.
Chest radiography also helps to confirm the diagnosis of active tuberculosis in a child with positive Mantoux test results. If the chest radiography findings are positive or if the child has other symptoms consistent with the diagnosis of tuberculosis, an attempt should be made to isolate the tuberculous bacilli from early-morning gastric aspirates, cerebrospinal fluid, sputum, urine, pleural fluid, or biopsy specimens.
Chest radiography is indicated in complicated cases in which treatment fails to elicit a response, in patients with respiratory distress, or in those who require hospitalization. Obtain frontal and lateral radiographs, particularly in cases in which the clinical examination findings are equivocal.
In complicated cases of pneumonia, perform chest radiography 6 weeks after treatment to verify resolution of the pneumonia and to screen for any underlying predisposing conditions, such as sequestration.
Numerous radiographic patterns are consistent with pneumonia and a multitude of other pathologic processes.[3] A synthesis of all available information and careful consideration of the differential diagnosis is essential to establishing the diagnosis, although empiric antimicrobial treatment usually cannot be deferred because of an inability to prospectively exclude the diagnosis.
Generalized hyperinflation with patchy infiltrates suggests partial airway obstruction from particulate or inflammatory debris, although the contribution of positive airway pressure from respiratory support must be considered. Pneumatoceles (especially with air-fluid interfaces) and prominent pleural fluid collections also support the presence of infectious processes.
Chest radiographs of infants infected with organisms in utero or via the maternal genital tract may demonstrate a ground-glass appearance and air bronchograms. Diffuse, relatively homogeneous infiltrates that resemble the ground-glass pattern of respiratory distress syndrome are suggestive of a hematogenous process, although aspiration of infected fluid with subsequent seeding of the bloodstream cannot be excluded.
Patchy, irregular densities that obscure normal margins are suggestive of antepartum or intrapartum aspiration, especially if such opacities are distant from the hilus. Patchy, irregular densities in dependent areas that are more prominent on the right side are more consistent with postnatal aspiration.
Except for patients with sickle cell disease, a significant pleural effusion usually indicates a bacterial etiology. Although these patterns are typical, the etiology cannot be reliably identified based solely on chest radiography findings. Other typical findings of bacterial pneumonia include a lobar consolidation with air bronchograms occasionally accompanied by a pleural effusion. Lobar consolidation and pleural effusion are seen in the images below.
Single or multiple prominent air bronchograms 2 or more generations beyond the mainstem bronchi reflect dense pulmonary parenchyma (possibly an infiltrate) highlighting the air-filled conducting airways. A well-defined, dense lobar infiltrate with bulging margins is unusual. Lateral or oblique projections may help to better define structures whose location and significance are unclear.
Although unilateral and/or lobar infiltrates are often seen in bacterial pneumonia (see the image below), several studies have found that the pattern of radiologic features cannot accurately distinguish a bacterial etiology from a viral etiology.[4, 5]
In contrast, a large Finnish series concluded that an alveolar (equivalent to a lobar) infiltrate is an insensitive, but reasonably specific, indication of bacterial infection.[6] Thus, a lobar infiltrate can be seen with viral infections, foreign body aspirations, and mucous plugging that results in atelectasis. Furthermore, pleural effusions, although usually parapneumonic (80%), may be observed in numerous disease processes.
At either extreme (from typical bronchiolitis with scattered infiltrates to dense lobar pneumonia with a large pleural effusion), the level of diagnostic certainty provided by radiologic findings increases.
Pneumatoceles and abscesses are less commonly found but may indicate a Staphylococcus aureus, gram-negative, or complicated pneumococcal pneumonia. Round pneumonia on chest radiographs should raise suspicion that the disease has a bacterial etiology, and particularly, that Streptococcus pneumoniae or S aureus is the causative agent. Round pneumonia is shown in the radiograph below.
The radiographic appearance of Mycoplasma infection varies. Early in the infection, the pattern tends to be reticular and interstitial; as the infection progresses, patchy and segmental areas of consolidation are noted, along with hilar adenopathy and pleural effusions.
For M pneumoniae, 3 radiographic patterns may be observed: peribronchial and perivascular interstitial infiltrates, patchy consolidations, and homogeneous acinar consolidations like ground glass.[7] The lower fields of the lungs are most often affected, and enlargement of the hilar glands is common.
Although no radiographic findings are specific for Chlamydophilapneumoniae (formerly Chlamydia pneumoniae) , a combination of the clinical and radiographic findings strongly suggests the diagnosis before laboratory diagnosis is available.
In a study of 125 cases of Chlamydophila pneumonia, Radkowski et al demonstrated that most chest radiographs showed bilateral hyperexpansion and diffuse infiltrates with various radiographic patterns, including interstitial, reticular nodular, atelectasis, coalescence, and bronchopneumonia. Pleural effusion and lobar consolidation were not seen.[8]
Chest radiography findings in children with tuberculous pneumonia may include hilar or mediastinal lymphadenopathy, atelectasis, or consolidation of a segment or lobe (usually right upper lobe), pleural effusion, cavitary lesions (in adolescents and adults only), or miliary disease.
In viral pneumonias, 4 common radiographic findings are generally detected: parahilar peribronchial infiltrates, hyperexpansion, segmental or lobar atelectasis, and hilar adenopathy.[9]
Children with an asthma exacerbation often have abnormalities visible on chest radiographs; in an undiagnosed asthmatic child, these abnormalities, most commonly appearing in the right middle lobe can occur in any part of the lung, are frequently interpreted as pneumonia.
The overall presence of radiographic pneumonia among children with wheezing is uncommon.[10] Historical and clinical factors, such as fever and hypoxia, may be used to determine the need for chest radiography for wheezing children. Wheezing responsive to bronchodilators, a history of atopy, a family history of asthma, and a history of cough or wheeze with exercise may be helpful in differentiating these patients from those with pneumonia. The authors discourage the routine use of chest radiography for children with wheezing but without fever.
Computed tomography (CT) scanning of the chest is indicated in children with complications, such as pleural effusions, and in those in whom antibiotic treatment fails to elicit a response. This imaging modality is often more sensitive than radiography and demonstrates changes typical for these complications. The information is beneficial when making treatment decisions (eg, whether to perform surgical debridement of organized empyemas or loculated effusions) and in outlining the projected course of the patient's illness.
Ultrasonography is indicated primarily in children with complications, such as pleural effusions, and in those in whom antibiotic treatment fails to elicit a response.
Ultrasonography is used to effectively differentiate between a low-grade (nonfibrinopurulent) effusion and a high-grade (fibrinopurulent and organizing) one.
Ultrasonography may also prove useful for guidance in thoracentesis of a loculated effusion. In addition to a pleural effusion or empyema, other suppurative complications of pneumonia include cavitary necrosis or abscess and purulent pericarditis. A significant number of these complications are not evident on radiographs.
A meta-analysis by Pereda et al summarized evidence on the diagnostic accuracy of lung ultrasound for childhood pneumonia and concluded that current evidence supports lung ultrasound as an imaging alternative for diagnosis.[11, 12]