The primary objective of this study is to determine if lung ultrasound (LUS) can replace chest x-ray (CXR) when evaluating patients with possible pneumonia. Specifically, we are looking for an overall reduction of CXR when LUS is used first. Our null hypothesis is that LUS cannot replace CXR for diagnosing pneumonia. Our alternate hypothesis is that LUS can replace CXR for diagnosing pneumonia. Our secondary objectives include: (1) a comparison of unscheduled healthcare visits after the index Emergency Department (ED) visit between those subjects who undergo CXR first and those who undergo LUS first, (2) an evaluation of the rate of antibiotic use between the two groups, (3) a comparison of the admission rates, and (4) a comparison of the length of stay in the Emergency Department between the two groups.

Background - Ultrasound is now widely accepted as a diagnostic tool for use in the emergency department, as supported by the American College of Emergency Physicians position statement in 2001 (revised in 2008). Evidence-based guidelines for point-of-care lung ultrasound have recently been published (Volpicelli et al 2012). Lichtenstein et al (2004) performed bedside LUS on 117 critically ill patients to evaluate for alveolar consolidation and compared these findings with CT, the gold standard. Sensitivity of ultrasound was 90% and specificity 98%, indicating that US is a feasible imaging modality for the lungs. Copetti et al (2008) compared the diagnostic accuracy of LUS and CXR in children with suspected pneumonia. 79 children underwent LUS and CXR. Lung ultrasound was positive for the diagnosis of pneumonia in 60 patients, whereas CXR was positive in 53. Copetti concluded that LUS is as reliable as CXR in diagnosing pneumonia plus it has the added benefit of no radiation exposure for patients. Shah et al (2009) found LUS to be superior to CXR in detecting pneumonia. 200 patients with suspected pneumonia were enrolled and underwent LUS and CXR. LUS detected 49 pneumonias whereas CXR detected 36. The 13 cases of radiographically occult pneumonia that were identified by LUS were all less than 1 centimeter in diameter, suggesting that LUS is superior in identifying early and/or small pulmonary consolidations. This particular study found that LUS was able to detect pneumonia with a Sensitivity of 86% and a Specificity of 97%. Additionally, Tsung et all (2009) found that it is feasible to use ultrasound to distinguish viral from bacterial pneumonia, thus indicating another striking advantage to LUS. From these studies, it is clear that lung ultrasound plays a role in the diagnosis of pulmonary pathology and moreover it is possible that LUS may replace CXR as the imaging modality of choice. This study is designed as a comparative effectiveness randomized controlled trial between ultrasound and chest x-ray for diagnosing pneumonia. The study cited above performed by Shah et al 2009 forms the basis of our pilot data in planning this randomized controlled trial. In Dr. Shah's study, there were no missed pneumonias and no over or under treatment of pneumonia when pneumonia was diagnosed on lung ultrasound.

Study Design - Currently CXR is the standard of care for the detection of pneumonia, however, there is published evidence that demonstrates LUS is as reliable as CXR and even surpasses CXR in detecting small and/or early pneumonias as well differentiating viral from bacterial processes as cited above (Lichtenstein et al 2004; Copetti et al 2008; Shah et al 2009; Tsung et al 2012).

The motivation for conducting this study is that we have possibly identified an imaging modality that is better than our current standard of care. It is our primary aim to compare the two imaging modalities to clinical outcomes to see if subjects in the investigational arm have better outcomes than those in the control arm who receive the standard of care.

The attending physician or fellow caring for the patient will determine if the patient is eligible. If the ED provider clearly identifies a pneumonia on the ultrasound then the patient will be diagnosed and treated for pneumonia without being subjected to the unnecessary radiation of a CXR. However, if the provider does not clearly identify a pneumonia on ultrasound or if the LUS fails to detect a pneumonia and the clinical suspicion remains high, then the provider has the option to proceed to the CXR to assist in the diagnosis of pneumonia. Alternatively, all subjects randomized to the control arm will under a CXR first followed by a LUS, because LUS can often provide additional information that CXR does not as noted above (e.g. the ability to differentiate between viral and pneumonia infections).

Inclusion Criteria:

• All patients who present to the ED with respiratory symptoms suspicious for pneumonia
• In whom the treating physician believes would benefit from diagnostic imaging

Exclusion Criteria:

• Patients who arrive at the ED with a previously performed CXR
• Unstable patients with life-threatening injuries who require ongoing resuscitation

Adverse event collected duration of study, up to 2 weeks.