Distinguishing an exudate from a transudate is the cornerstone of the evaluation of pleural effusions. The criteria proposed by Light et al in 1972 have become the standard method of making this distinction. These criteria were retrospectively designed to be close to 100 percent sensitive in identifying exudates, since these effusions imply pleural pathology that requires further evaluation. Out of 47 patients with transudates, Light et al described one patient with congestive heart failure who met two of three criteria for an exudate. This observation of exudative range protein levels in patients with congestive heart failure was first noted by Pillay in 1965 and recently was confirmed by Chakko et al, who showed that diuresis will increase the effusion protein.
The problem of high-protein transudates is more common in the evaluation of ascites and has led to the development of the serum-ascites albumin gradient (serum albumin level minus the ascites albumin level). A gradient of less than 1.1 g/dl has been shown to be the best predictor of exudative ascites and has become an accepted method of distinguishing exudates from transudates. The albumin gradient has not previously been evaluated in pleural effusions. Therefore, we prospectively measured serum and pleural effusion albumin levels to calculate a serum-effusion albumin gradient (serum albumin level minus the effusion albumin level) in 59 consecutive patients undergoing diagnostic or therapeutic thoracentesis and compared the serum-effusion albumin gradient to the criteria of Light et al.