Light's criteria revisited: consistency and comparison with new proposed alternative criteria for separating pleural transudates from exudates.

OBJECTIVES: The first objective was to assess the diagnostic value of new biochemical criteria proposed to discriminate pleural transudates from exudates and to compare their efficiency with those of Light's criteria. The second objective of the study was to assess the interstudy variability of the parameters repeatedly determinated in two different groups of patients with pleural effusion.

PATIENTS AND METHODS: We recorded clinical characteristics and final diagnoses and measured pleural fluid (PF) and serum levels of protein, LDH, cholesterol and cholinesterase of 243 patients with pleural effusion.

RESULTS: Sixty-one (25%) pleural effusions were transudates and 182 were exudates. The sensitivity (99%) and accuracy (96%) of Light's criteria were higher than those of the other criteria tested, although the differences with those of the PF LDH-cholesterol combination (96 and 93%) did not show statistical significance. Pleural LDH concentration was the criterion with the highest specificity (95%), being significantly higher (p < 0.05) than that of Light's criteria. The sensitivity, specificity and accuracy of most criteria tested did not vary when compared with those obtained in a study performed 5 years previously.

CONCLUSIONS: Light's criteria remain the criteria of choice for segregating exudates from transudates. Based on cost-efficiency reasons, the PF LDH-cholesterol combination appears as an alternative. Because both sets of criteria misdiagnose a substantial percentage of transudates, exceptions based on good clinical judgment and the complementary use of a more specific criterion, as the PF concentration of LDH, must be considered.