ABSTRACT
Haemophilus influenzae is a part of normal upper respiratory flora of human, which can cause a wide variety of infections. Other members of genus Haemophilus rarely cause human infection but are frequently isolated from clinical specimens, such as sputum. The pathogenicity between H. influenzae and other Haemophilus species is different therefore a reliable method for identification of H. influenzae is essential. The aim of this study was to compare the identification methods for Haemophilus by four phenotypic tests with that by a PCR-based method. A total of 101 Haemophilus isolates were identified by biochemical tests and the XV requirement test by using XV paper strip technique, porphyrin test and Staphylococcus streak technique. The PCR-based method was performed using specific primers for 16SrDNA, p6 genes of H. influenzae and sodA gene of H. parainfluenzae. Using the XV paper strip technique, porphyrin test and biochemical tests, 88 and 13 isolates were identified as H. influenzae and H. parainfluenzae respectively, whereas 54 H. influenzae and 47 H. parainfluenzae were identified by using Staphylococcus streak technique (66.4 % agreement with that of the three tests). The PCR-based method revealed that 83 H. influenzae and 12 H. parainfluenzae were identified, whereas 6 isolates could not be categorized into both species. This study showed that identification of Haemophilus by the XV paper strip technique, porphyrin test and biochemical test gave 93.1 % agreement with that of the PCR method, whereas the Staphylococcus streak technique gave only 71.3 % agreement.
ABSTRACT
Carbapenem resistance mediated by acquired carbapenemase genes has been increasingly reported in Pseudomonas aeruginosa. Class B carbapenemases or metallo-b-lactamases (MBLs) are the most clinical concern. Currently, there is no recommendation available from the Clinical Laboratory Standards Institute (CLSI) for MBL detection. In this study, we attempted to evaluate the performance for phenotypic detection of MBLs in imipenem-nonsusceptible clinical isolates of P. aeruginosa from Srinagarind Hospital. Five MBL-positive control strains, each producing IMP-1, IMP-4, IMP-9, VIM-1, or VIM-2 enzyme, and 74 clinical isolates of P. aeruginosa were screened for the presence of MBLs by combined-disk test (CDT) and double-disk synergy test (DDST) on a single agar plate using imipenem (10 µg IPM) and meropenem (10 µg MEM) disks as substrates and 292 µg of EDTA as an MBL inhibitor. Multiplex PCR for detection of MBL genes was used as a gold standard. Genotypic confirmation revealed that 9 of the 74 clinical isolates (12.2%) carried MBL genes, blaVIM (6 isolates) and blaIMP (3 isolates). Comparison of the MBL phenotypic tests to the multiplex PCR revealed that the CDT using IPM+EDTA/IPM correctly differentiated all MBL-producing isolates (sensitivity of 100%) but gave three false-positive isolates (specificity of 95.4%), whereas that with MEM+EDTA/MEM showed sensitivity and specificity of 92.9% and 92.3% respectively. In addition, the DDST using either IPM or MEM with EDTA gave the same result for all isolates with sensitivity and specificity of 92.9% and 96.9% respectively. These methods are simple and accurate for detection of MBLs in imipenem-nonsusceptible P. aeruginosa isolates from this hospital. Early detection of MBL producers and strict infection control will contribute to prevent further spread of these resistant strains.