Diagnostic Accuracy of Pleural Fluid Adenosine Deaminase for Diagnosing Tuberculosis. Meta-analysis of Spanish Studies

Abstract

Objective: To evaluate the usefulness of pleural fluid adenosine deaminase (ADA) for diagnosing tuberculous pleural effusions in the Spanish population, according to laboratory technique and cut-off point, and to compare the results with other populations.

Methods: Meta-analysis of diagnostic studies on pleural fluid ADA in the Spanish population, extracted from the PubMed and Embase databases from inception until July 2017, with no language restrictions. The overall diagnostic accuracy of ADA and that of each of the measurement techniques (Giusti, manual and automated kinetic methods) and selected cut-offs were analyzed. The QUADAS-2 tool was used to evaluate the quality of studies. A bivariate random effects model was used. Results were compared with those obtained from previous meta-analyses in non-Spanish populations.

Results: Sixteen studies in a total of 4,147 patients, 1,172 of whom had tuberculous pleural effusions, were included. ADA had 93% sensitivity, 92% specificity, positive likelihood ratio of 12, negative likelihood ratio of 0.08, and an area-under-the-curve of 0.968 for identifying tuberculosis. There were no differences in diagnostic accuracy between the techniques used for ADA measurement or the selected cut-offs. In 73 studies from non-Spanish populations a trend toward lower ADA sensitivity (88%, 95% CI:86%-90%) and specificity (88%, 95% CI: 86%-90%) was noted, but differences did not reach statistical significance.

Conclusions: Pleural fluid ADA in the Spanish population shows good diagnostic accuracy (regardless of the measurement technique or cut-off), similar to that reported in non-Spanish populations.

Keywords: Adenosina desaminasa; Adenosine deaminase; Derrame pleural tuberculoso; Líquido pleural; Meta-analysis; Metaanálisis; Pleural fluid; Tuberculous pleural effusion.