When is CXR necessary for the diagnosis of CAP in the ambulatory setting?
-We strongly recommended based on moderate quality evidence not obtaining a CXR in the ambulatory setting for a patient suspected of bacterial pneumonia on clinical grounds that has mild disease and does not require hospitalization.
The diagnosis of community acquired pneumonia can be made on clinical grounds, specifically in a patient with fever, lower respiratory tract symptoms and focal auscultatory findings on exam. Chest radiograph is not necessary to establish a diagnosis as there is currently no gold standard, including CXR, for the diagnosis of bacterial pneumonia (Lynch, Bialy et al. 2010). Individual studies looking at clinical outcomes for patients after CXR was obtained did not find significant changes in treatment or need for hospitalization (Swingler, Hussey et al. 1998; Virkki, Juven et al. 2002; Swingler and Zwarenstein 2005). Additionally, the likelihood of a CXR interpretation being consistent with bacterial pneumonia is increased when provider’s clinical impression of pneumonia is high (Neuman, Monuteaux et al. 2011).
The British Thoracic Society Guideline recommends that CXR not be considered routine, and that obtaining CXR’s have not been shown to improve clinical outcomes in patients in the ambulatory setting (Harris, Clark et al. 2011). PIDS/IDSA guideline also do not recommend CXR in patients well enough to be treated in the outpatient setting, arguing that CXR results do not reliably distinguish viral from bacterial CAP and cannot distinguish between different bacterial pathogens (Bradley, Byington et al. 2011).
When is CXR necessary for the diagnosis of CAP for patients with respiratory distress, to determine disposition or for a concern of complicated pneumonia?
-We strongly recommended based on low quality evidence obtaining a CXR in a patient with significant respiratory distress, need for hospitalization or concern for complicated pneumonia.
Chest radiograph for patients with significant respiratory distress or need for hospitalization have the potential to identify complicated pneumonia and significantly alter management. The PIDS/IDSA guideline identifies that CXR should be performed in patients with significant respiratory distress as defined as: tachypnea for age, dyspnea, retractions, grunting, nasal flaring, apnea, altered mental status and/or hypoxia (sat <90% on RA). CXR should also be performed in patients who fail initial appropriate antibiotic therapy to evaluate for pneumonia complications including effusions, necrotizing pneumonia and pneumothorax (Bradley, Byington et al. 2011).
What type of CXR is necessary for the diagnosis of CAP?
-We recommended based on low quality evidence obtaining a PA and lateral CXR when obtaining CXR.
The British Thoracic Society and PIDS/IDSA guideline disagree on obtaining a lateral radiograph in a patient with illness significant enough to require hospitalization. The British Thoracic Society recommends NOT obtaining a lateral view, arguing they are not necessary to change clinical management and expose children to additional radiation. The guideline cites a study by Rigsby et al that found a PA CXR had 100% sensitivity for identification of a lobar infiltrate. However, in the study 15% of non-lobar infiltrates would be missed without the lateral view (Rigsby, Strife et al. 2004). The BTS guideline questions the variability of what is considered radiographic pneumonia, especially the 15% non-lobar infiltrates that were missed. The PIDS/IDSA guideline has a strong recommendation based on moderate quality evidence that frontal and lateral radiographs be obtained in all patients requiring hospitalization to evaluate for the size and character of parenchymal infiltrates as well as evaluate for pneumonia complications. Despite making the recommendation on moderate quality evidence, the guideline does not further discuss primary evidence for obtaining a lateral view.
One primary study not included in either guideline is from Lynch and colleagues, 2004, where pediatric emergency physician interpretations of CXR in patients concerning for pneumonia were studied. Participating ED physicians were given either an anterior or anterior and lateral CXR to interpret. The authors found that sensitivity and specificity was not improved with the addition of a lateral XR (Lynch, Gouin et al. 2004).
CMH’s CPG recommends both anterior and lateral XR for hospitalization patients. This recommendation is based on identification of patients with pneumonia complications that will require a change in therapy based on results.