Background & Aim: Pleural effusion is seen in the background of many diseases, two major groups of which are tuberculosis and malignancies. At present pleural biopsy helps us differentiate these two causes from each other, but it is not only an invasive method but also an expensive one. For this reason, investigators are in search of simple and less invasive methods to diagnose the cause of pleural effusion. One of the new methods is measuring adenosine deaminase activity in pleural fluid. This method has been practiced worldwide and has come up with various results. The main objective of this study is to measure adenosine deaminase levels in pleural fluids of patients and compare them with the histopathological results, and if the outcomes are meaningful, this method can be recommended for clinically confusing cases. Patients & Method: This descriptive analytical cross-sectional study was performed on sixty patients who were hospitalized in the infectious and internal wards on suspicion of tuberculosis or malignancy. First, pleural fluid was aspirated to measure the ADA level(cutoff point 35u/lit). Then biopsy was done and the results were compared with each other. Results: The data was analyzed using SPSS 10 software and Chi-square test. Compared to biopsy, the sensitivity and specificity of ADA for the diagnosis of tuberculosis were 68.4% and 92.6% and for malignancies were 91.3% and 37.8% respectively. In comparison to biopsy, the positive predictive value of ADA to diagnose tuberculosis and malignancies was 81.2% and 47.8% respectively, and the negative predictive value of ADA for the diagnosis of TB and malignancies was 86.36% and 87.5% respectively. Conclusion: Based on the obtained data, raised levels of ADA in pleural effusion(over 35u/l) suggest tuberculosis with a more than 90% probability. Although low levels of ADA(below 35u/l) have great sensitivity for the diagnosis of malignancies, as the specificity of ADA level is 37.8%, it does not necessarily propound malignancy.
Rights and permissions | |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |