Accuracy of chest sonography and polymorphonuclear elastase in the assessment of bacterial pleural effusion

Carmen Alemán, José Alegre, Jordi Andreu, Rosa Ma Segura, Lluís Armadans, Delia Sureda, Ana Vázquez, Daniel Iglesias, Tomás Fernández De Sevilla

Research output: Contribution to journalArticleResearchpeer-review

1 Citation (Scopus)


Background: The relationship between chest sonography findings and inflammatory markers for assessing bacterial pleural effusion is not well established. We decided to study the accuracy of chest sonography in determining the nature of bacterial pleural effusion and its relationship with polymorphonuclear elastase (PMN-E) results. Methods: Pleural sonography and PMN-E were evaluated in a prospective study of 144 consecutive patients with pleural effusion of various etiologies: 25 complicated parapneumonic, 18 uncomplicated parapneumonic, 33 tuberculous, 17 malignant, 12 transudates, and 39 of unknown etiology. The sonographer distinguished between anechoic and septated pattern. The relationship between sonographic appearance and inflammatory markers was evaluated. Results: All of the complicated parapneumonic, 11 uncomplicated parapneumonic, and 28 tuberculous effusions were septated. Septated pattern and PMN-E value were independent predictors of infectious pleural disease (p <0.05). The simultaneous presence of a septated pattern and a PMN-E higher than 100 μg/l had a sensitivity of 79.1% and a specificity of 91.1% for the diagnosis of bacterial effusions. Conclusions: PMN-E level and the sonographic pattern of pleural fluid may be useful in the diagnosis of bacterial pleural effusions. © 2004 Elsevier B.V. All rights reserved.
Original languageEnglish
Pages (from-to)89-92
JournalEuropean Journal of Internal Medicine
Issue number2
Publication statusPublished - 1 Apr 2004


  • Etiology
  • Imaging
  • Pleural fluid


Dive into the research topics of 'Accuracy of chest sonography and polymorphonuclear elastase in the assessment of bacterial pleural effusion'. Together they form a unique fingerprint.

Cite this