PURPOSE: Discriminating cardiogenic pulmonary edema (CPE) from acute lung injury (ALI) or acute respiratory distress syndrome (ARDS) is a serious problem, but the ability of chest radiographs to distinguish between these entities is limited. The aim of this study was to compare computed tomography (CT) findings in the acute phase of CPE with those of ALI/ARDS.
METHODS: Outpatients with acute respiratory failure presenting to emergency departments with bilateral pulmonary infiltrates on radiographs were enrolled. Using the clinical diagnostic definition to rule out the CPE triggered by airway-infections, two intensivists, who were blinded to the chest images, independently reviewed the clinical records in order to determine a diagnosis of CPE, ALI/ARDS, or mixed edema. CT findings of patients, excluded those the patients diagnosed with mixed edema, were independently evaluated by two radiologists, who were unaware of any clinical information.
RESULTS: Forty-one patients with CPE and 20 patients with ALI/ARDS (9 intrapulmonary, 11 extra-pulmonary causes) were assessed. Upper lobe dominant ground-glass attenuation (p = 0.025), thickening of the bronchovascular bundle (p = 0.007), and right dominant pleural effusion (p = 0.047) were significant predictors for distinguishing CPE from ALI/ARDS in multivariate models. When upper lobe dominant ground-glass attenuation and thickening of the bronchovascular bundle were concurrently present, all of the patients met the definition of CPE.
CONCLUSIONS: Chest CT in the emergency department may be a useful tool in differentiating causes of acute respiratory failure.
CLINICAL IMPLICATIONS: In this study, CT findings of upper lobe dominant ground-glass attenuation, thickening of the bronchovascular bundle, and right dominant pleural effusion were significant predictors of CPE. These findings may enable radiographic discrimination between CPE and ALI/ARDS.
DISCLOSURE: The following authors have nothing to disclose: Hiroshi Ishii, Kosaku Komiya, Fumito Okada, Junji Murakami, Hidehiko Yamamoto, Jun-ichi Kadota
No Product/Research Disclosure Information