Comparison of five different methods for detection of SHV extended-spectrum beta-lactamases.

Five different methods for detection of different types of SHV extended-spectrum beta-lactamases (ESBL) were compared: minimum inhibitory concentration (MIC) determination of beta-lactam with and without clavulanic acid, double-disk synergy test (DDST), inhibitor potentiated disk diffusion test (IPDDT), three-dimensional test (TDT) and PCR/Nhe I test. MIC determination of beta-lactam with and without clavulanic acid was the most sensitive method regardless of the type of beta-lactamase. However the specificity of this method was a little above 90%. IPDDT turned out to be a very sensitive method too but it lacks specificity because 26.9% of ceftazidime sensitive strains (putative ESBL negative), gave a positive result. It is important to put all four disks on the plate because ceftazidime and aztreonam were more sensitive indicators for SHV-5 and SHV-12 beta-lactamase producers while cefotaxime and ceftriaxone were more reliable in detecting SHV-2 beta-lactamase producers. The DDST detected all SHV-5 and SHV-12 beta-lactamase producers and 95.2% of SHV-2, so it was less sensitive than MIC determination but was highly specific, since there were no false negative results observed. The sensitivity of DDST can be improved by using all four disks and placing them at the smaller distance from the central disk (2.5 cm). The TDT was the least sensitive method, particularly for SHV-5 and SHV-12 beta-lactamase producers. The PCR/Nhe I test for detection of ESBL blaSHV genes is a highly sensitive and specific method but it is rather laborious and thus not very practical for use in routine clinical laboratories. Nevertheless it has potential to serve as the gold standard in epidemiological investigations on ESBLs. According to the results of this investigation MIC determination of beta-lactam with and without clavulanic acid, even if only one antibiotic is used and the PCR/Nhe I tests are the most reliable methods for detection of SHV ESBLs.