Abstract
Introduction: To describe the incidence of discordant exudate effusions, their underlying etiologies and their impact on the accuracy of the Light's criteria to diagnose exudate effusions.
Methods: A retrospective review of pleural fluid analysis from a cohort of patients who underwent thoracentesis. Pleural effusions were defined as exudative based on the Light's criteria. The effusions were further classified in concordant or discordant exudates.
Results: From 847 pleural effusion analysis, 588 (69.4%) were finally diagnosed as exudate and 259 (30.6%) as transudate. In the 13% of cases there was discordancy between the final diagnosis and the pleural fluid analysis by the Light's criteria. 281 (44%) of the effusions classified as exudate by Light's criteria were discordant exudates (62 transudates and 219 exudates). 182 (75%) were protein discordant (44 transudates and 138 exudates), and 99 (35%) were LDH discordant (18 transudates and 81 exudates). The positive predicted value and positive likelihood ratio of the Light's criteria for the diagnosis of exudate effusion decreased from 95.7% and 12 respectively when the exudates were concordant to 77.9% and 3.4 respectively if they were discordant.
Conclusions: In a significant percentage of patients, there is discordancy between the results of the pleural fluid analysis and the final clinical diagnosis. The decrease in the accuracy of Light's criteria to identify exudate effusions when they are discordant exudates suggest that, in these cases, physicians should look aggressively to identify the underlying etiology of the effusion, taking in account the clinical history and presentation.
- Copyright ©ERS 2015