Abstract

Aims: As part of the UK antimicrobial resistance strategy and action plan, the Public Health Laboratory Service (PHLS) is required to collect antibiotic susceptibility data so that resistance trends and patterns can be monitored. Most laboratories report urine Gram negative isolates, as “coliforms” according to morphological appearance, but without an acceptable identification system the antimicrobial surveillance data will be meaningless. Commercially available identification systems tend to be expensive and time consuming. Chromogenic agars, which claim to improve the detection of mixed cultures and identification of organisms from urine, have now become available and may provide a cost effective alternative. The primary aim of this study was to compare the performance of cystine lactose electrolyte deficient (CLED) agar with a chromogenic agar (Oxoid urinary tract infection medium; CUTI) in terms of isolation rates and ability to detect mixed cultures. Secondary aims were to evaluate the correlation of “presumptive” identification of isolates from chromogenic media with that of two commercial identification systems and to appraise the sensitivity of the semiquantitative loop and filter paper strip culture techniques.

Method: One thousand, four hundred and sixty six urine samples were examined in four laboratories using the semiquantitative culture methods of 1 μl loop and filter paper strip. The degree of accuracy of organism identification was measured by comparing the presumptive identification using colony colour supplemented with simple bench tests, with identification obtained from two more complex commercial systems.

Results: There was no significant difference between the performance of the loop and filter paper strip methods on the CLED agar, but the CUTI agar performed significantly better than the CLED agar for the detection of significant isolates and mixed cultures. This difference was greater using the loop method. Identification of the organisms using the commercial systems gave > 99% agreement and was therefore considered suitable as a standard against which to compare the presumptive CUTI identification. Using the manufacturer's colony colour criteria in combination with a bench indole test, the CUTI medium was 99% specific for Escherichia coli, although this was reduced to 97% if the indole test was omitted. Citrobacter spp were the most commonly misidentified organisms, giving false presumptive identification as E coli. By testing oxidase activity to differentiate Pseudomonas spp and the absence of indole production to support the identification of Proteus mirabilis, the CUTI medium provided a suitable identification for 86.8% of Gram negative isolates. The remaining 13.2% would require further identification.

Conclusion: CUTI medium improves the detection of mixed cultures, thereby improving the reliability of reporting of significant isolates when compared with CLED agar. When supplemented with simple bench tests it provides an identification system capable of speciating 86.8% of Gram negative isolates and providing a valuable cost effective mechanism for antimicrobial resistance surveillance.