RESUMO
Conventional culture of pleural fluid samples frequently provides false-negative results. Universal polymerase chain reaction (PCR) of the 16S ribosomal ribonucleic acid (rRNA) gene (16S PCR) has proven useful in the diagnosis of various bacterial infections. We conducted a prospective study to assess the value of 16S PCR in the etiologic diagnosis of pleural effusion. All pleural fluid samples received for culture were also studied using 16S PCR. Positive samples were sequenced for identification. Clinical records and conventional culture results were analyzed to classify pleural fluid samples as infected or not infected. We studied 723 samples. We excluded 188 samples because they were obtained from a long-term chest tube, there was a diagnosis of mycobacterial infection, or there were insufficient data to classify the episode. Finally, 535 pleural fluid samples were analyzed. According to our criteria, 82 (15.3%) were infected and 453 (84.7%) were not infected. In the infected samples, 16S PCR was positive in 67 samples (81.7%) while conventional culture was positive in 45 (54.9%). There were 4 false positives with 16S PCR (0.9%) and 12 with culture (2.6%). The values for the etiologic diagnosis of bacterial pleural effusion of conventional culture compared with 16S PCR were as follows: sensitivity, 54.9%/81.7%; specificity, 97.4%/99.1%; positive predictive value, 76.3%/94.4%; negative predictive value, 92.6%/96.8%; and accuracy, 90.8%/96.5%.When compared with conventional culture, 16S PCR plus sequencing substantially improves the etiologic diagnosis of infectious pleural effusion. In our opinion, this technique should be added to the routine diagnostic armamentarium of clinical microbiology laboratories.
