ABSTRACT
Several sets of PCR primers have recently been developed for detection of Burkholderia pseudomallei. In this report, the performance of 16S rRNA gene primers (16S), rRNA spacer gene primers (spacer), and 'LPS' primers (LPS) were compared. All primer sets were tested by PCR amplification of the same DNA samples extracted from blood specimens of 46 patients from northeastern Thailand, of which 29 had melioidosis based on blood culture as a gold standard. The sensitivities were 41, 35.7, and 31% while the specificities were 47, 59, and 100% for the 16S, spacer, and LPS primers, respectively. The positive predictive values were 60, 59, and 100%, while negative predictive values were 35, 34, and 46%, for these primers. The low sensitivity of PCR was suspected to be because of small numbers of bacteria in the samples. In addition, one primer set could not detect all B. pseudomallei strains. To make PCR for melioidosis more practical, bacterial concentration steps must be added. Lastly, mixed infection of patients in endemic areas may be the cause of controversial false positive PCR results, and should be further investigated.
