It is not clear why the albumin gradient would be more reliable than Lights criteria. Possibly the total fluid protein level is relatively increased after diuresis, due to a greater percentage of nonalbumin protein that originates in the pleural space. Alternatively, the mathematics of a gradient may be more representative of the protein diffusion than a ratio. The fluid-to-serum albumin ratio was much less effective in separating exudates from transudates.
The major weakness in our study was that the diagnoses of congestive heart failure were made clinically. We cannot rule out some subclinical exudative cause for their pleural effusions; however, all of these patients had resolution of their pleural effusion with diuretic therapy. A more exact way to compare the albumin criteria to Lights criteria would be to repeat the experiment of Chakko et al on patients with congestive heart failure and do serial thoracentesis after diuretic therapy. The present study was necessary first to show that the albumin gradient was effective in discriminating exudates from transudates.