Background. A non-invasive technique, such as CT scan, to determine whether pleural effusion is transudative or exudative would aid in therapeutic management especially in cases when thoracentesis is contraindicated. Aside from the financial advantage, this method would benefit patients with CHF and concomitant transudative effusion (comprising approximately half of all patients with pleural effusion), since these patients usually do not require thoracentesis, hence preventing the complications inherent to thoracentesis. This study is done to determine the accuracy of Computed Tomography (CT) in characterizing pleural fluid into transudative and exudative based on attenuation values (Hounsfield Units), a non-invasive technique not being used for this purpose at this time.
Methods. This is a criterion-reference based cross-sectional study. Patients with pleural effusion who had thoracentesis and chest CT done within 7 days of each were evaluated retrospectively over a 4 year period. Effusions are classified as transudates or exudates using Light's criteria. The mean Hounsfield unit (HU) of an effusion is determined and a receiver operating characteristic (ROC) curve was constructed to determine threshold values for classification on the basis of mean HU and to examine overall accuracy, using the area under the curve (Az).
Results. Of the 110 exudates and 158 transudates, the mean attenuation of exudates (25.99 HU; [SD] 9.3 HU; range, 12-45 HU) was significantly higher than transudates (11.89 HU; 2.5HU; range, 5-19 HU), (p = 0.00). The overall accuracy of attenuation values for identifying exudates was high, Az = 0.981, standard error = 0.007, with the largest limitation being the overlap of exudates in the 16 to 18 HU range, which constituted 13% (35/268) of the total effusions measured (35/268) of the total effusions measured.
Conclusions. The mean attenuation of exudates is significantly higher than transudates as their accuracy is high. However, pleural effusions with HU's between 16 to 18 represent a gray area where clinical correlation is required prior to thoracentesis.