The differentiation between exudates and transudates is the initial step in the analysis of pleural effusions as it often gives an indication of the underlying pathophysiologic process, the differential diagnosis, and the need for further investigations. Four classifications have been suggested in the literature: Light's criteria, serum-effusion albumin gradient, effusion cholesterol concentration, and pleural/serum bilirubin concentration.
Aim of study: To compare the various biochemical parameters used to identify exudates.
Patients and methods: A study was carried out from February 1993 to March 1994 at Tygerberg Hospital, South Africa. Five hundred pleural effusions and serum specimens were analyzed. After discharge, the hospital records of all patients were reviewed for a diagnosis.
Results: A reliable diagnosis could be made in 393 cases (270 exudates and 123 transudates). Using the criteria of Light and associates 93% of the effusions were correctly classified, yielding a sensitivity and specificity of 98% and 83%, respectively, to detect exudates. The serum-effusion albumin gradient at a cutoff level of 12 g/L yielded the following results: accuracy, 89%; sensitivity, 87%; and specificity, 92%. Using a cutoff level of 1.55 mmol/L, the effusion cholesterol concentration yielded results of 70%, 54%, and 92%, respectively. The results improved at lower cutoff levels for effusion cholesterol level. Use of the pleural/serum bilirubin ratio as a means for identifying exudates produced results of 75%, 81%, and 61%, respectively.
Conclusion: The criteria of Light et al remain the best method for distinguishing exudates from transudates. The serum-effusion albumin gradient is useful when patients are receiving concurrent diuretic therapy.