Dr Frankie Chi-fung CHOY, Department of Medicine, Tseung Kwan O Hospital, Hong Kong
Background: Pleural effusion is a common clinical condition. Light’s criteria are frequently employed for distinction of pleural exudates from transudates. This study was carried out to determine the accuracy of Light’s criteria. Further analysis of its components was carried out in search of the best single criterion for identification of pleural exudates.
Method: A 3-year retrospective study was carried out in a regional public hospital. All adult patients with thoracocentesis performed in the study period were recruited consecutively. After applying exclusion criteria, detailed review was performed on the clinical records of the resulting 336 patients. Accuracy, sensitivity, specificity, likelihood ratios, positive and negative predictive values of Light’s criteria and other possible parameters were calculated and compared. Receiver Operator Characteristic curves were constructed for individual parameters for the best single criterion for differentiation.
Results: The accuracy of Light’s criteria was 90.2%, with sensitivity and specificity of 97.6% and 67.9% respectively. With more components of Light’s criteria were fulfilled, the chance of getting correct prediction for exudates becomes higher. Tuberculous and parapneumonic effusion rarely presented with transudative effusion by Light’s criteria. As a single criterion for categorization, pleural fluid lactate dehydrogenase level is the best parameter with high sensitivity and accuracy. Comparing with fluid lactate dehydrogenase, Light’s criteria were less specific but more sensitive.
Conclusion: Light’s criteria are the most sensitive method in differentiating pleural effusion, and are superior to currently available diagnostic tests. If single criterion with high sensitivity is used, fluid lactate dehydrogenase level is suggested.