The role of the microbiology laboratory is (1) to provide infection control information, so that highly transmissible isolates may be identified and appropriate control measures instigated as rapidly as possible and (2) to provide adequate information to the clinician enabling correct antibiotic choices to be made, particularly in the critically ill. Microbiological data is by definition slow as it is culture dependent: this study focused on the development of genetic, culture-independent methods for detection of resistance in nosocomial pathogens that could be introduced into the routine microbiology department and would fit into the routine workflow with a consequent reduction in time to result.
Initially a duplex real-time polymerase chain reaction was developed for the rapid identification and detection of S. aureus and methicillin-resistance. This was optimised for immediate as-needs testing of positive blood cultures signalling with “Gram positive cocci, possibly staphylococcus” evident on Gram stain, on a random access real-time PCR platform. This technology, allowing early identification of S. aureus and its susceptibility to methicillin, by simple automated methodology, may soon become the standard for all microbiology laboratories servicing the critically ill.
The second part of the study involved the development of a selective broth and multiplex PCR for detection of three important nosocomial isolates at this institution, methicillin-resistant S. aureus (MRSA), carbapenem-resistant Enterobacteriaceae, and multi-resistant Acinetobacter baumannii (MRAB). A multiplex PCR using four primer sets was designed to detect low colonisation levels of these isolates after overnight incubation in selective broth, significantly reducing the time to result and associated costs. This potentially useful epidemiological screening tool is practical, reproducible and sensitive with the potential of moving to an automated test (using real-time PCR, for example) in the future. The availability of early negative results judged by simple visual scanning (or by densitometry), means that the result is less operator-dependent, potentially reducing error rate.
The last part of the study dealt with an important resistance phenotype, aminoglycoside resistance. There had been no recent comprehensive local surveys performed to determine the frequency of aminoglycoside resistance amongst the Enterobacteriaceae, or to identify the genetic determinants and their transmissibility. The isolates collected for the study were all resistant to at least one of gentamicin, tobramycin or amikacin. Identification of integron cassette arrays and use of specific internal primers identified at least one genetic determinant for gentamicin and tobramycin resistance in 22 of 23 isolates. Three isolates had two aminoglycoside resistance genes, and three isolates had three aminoglycoside resistance genes identified (Table 6.1). Transferable gentamicin-resistant plasmids were predominant amongst Klebsiella spp., but less so amongst Enterobacter spp. and E. coli. Gentamicin-resistant Klebsiella spp. were often ESBL positive, the genetic determinants of which were typically co-transferred on a conjugative plasmid. The importance of screening at a local level was demonstrated by the unexpected predominance of aac(6')-IIc amongst Enterobacter spp. and the detection of a new gene (aac(6')-LT).
This part of the study has provided an understanding of the primary aminoglycoside resistance genes present in the local setting and their association with other resistances. This knowledge will allow development of assays for patient screening (clinical isolates and colonising flora), to better understand the epidemiology of aminoglycoside resistance and to allow better choice of antibiotic therapy related to presence or absence of these genes.