| dc.description.abstract | Background: Interleukin-27 (IL-27) has been proposed to be useful for diagnosing tuberculous pleural effusion (TPE). Adenosine deaminase (ADA) has been long used for the same purpose. The aim of this study was to compare the performance of IL-27, ADA, and their product (IL-27.ADA) in the diagnosis of TPE. Methods: Pleural fluid samples from patients with exudative pleural effusions were assessed for IL-27 and ADA levels. Receiver operating characteristic (ROC) curves were constructed to compare the overall diagnostic accuracy of IL-27, ADA, and IL-27.ADA. Curves of false-positive (FPR) and false-negative (FNR) rates as a function of TPE prevalence were also constructed, and mean rates of false results in low (1-10%), intermediate (11-40%), and high (41-70%) prevalences were estimated to evaluate the ability of the three markers in ruling in or ruling out TPE. Results: We studied 121 exudates. IL-27 and ADA were higher in TPEs compared with non-TPEs and they presented similar accuracies for the diagnosis of TPE. The product of IL-27 and ADA (IL-27.ADA) was more accurate than ADA for the same purpose. IL-27 and IL-27.ADA presented the lowest overall FPR and FNR, respectively. The FPR of IL-27, ADA and IL-27.ADA was >9%, even in high prevalence settings. Although their FNR was <2% in low prevalence settings, only IL-27.ADA exhibited sufficiently low FNR (<1%) in intermediate and high prevalences. Conclusions: ADA, IL-27, and IL-27.ADA cannot reliably 'rule in' TPE in any prevalence setting. TPE can be 'ruled out' by each of the biomarkers in low prevalence settings. In intermediate and high prevalence settings, IL-27.ADA is a reliable 'rule out' test in the diagnostic approach to TPEs. | en |
