Abstract

We examined the pleural liquid and serum samples of 79 patients with pleural effusion of well defined etiology to determine the efficacy of the Light’s criteria and several biochemical parameters for differentrating pleural exudate and transudate. Based on the etiologic diagnosis, 62 patients had pleural effusions classified as transudate and 17 patients had pleural effusions classified as exudate. Sensitivity and specificity for Light’s criteria were 100% and 56%, respectively; for pleural fluid cholesterol, 89% and 100%; for the pleural fluid/serum cholesterol ratio, 95% and 93%; for the albumin gradient between serum and pleural effusion, 84% and 94%. In seven patients taking diuretics, transudative pleural effusions were falsely classified as exudates accordig to Light’s criteria. The best results with combinations of several biochemical parameters were achieved when the pleural fluid cholesterol was combined with pleural fluid/serum LDH ratio or pleural fluid/serum protein ratio (sensitivity; 100%, specificity; 100%). In conclusion, Light’s criteria may be insufficient for differentiating transudates from exudates, especially in patients taking a diuretic. In these situations, albumin gradient between serum and pleural fluid, pleural fluid cholesterol measurement and the combined use of several biochemical parameters may help clinicians.