Abstract
Seven laboratories in six European countries examined 40 isolates belonging to the Acinetobacter calcoaceticus-Acinetobacter baumannii complex to investigate whether standardized protocols and quality-controlled reagents could produce reliable, discriminatory, and reproducible PCR-based fingerprinting results. Four PCR protocols with different primers (primers DAF4, ERIC-2, M13, and REP1 + REP2) were used. The epidemiological conclusions reached by the participating laboratories were substantially correct, with 96.4% of the total isolate grouping allocations agreeing with the consensus view. All laboratories identified the main epidemiological clusters, and each laboratory also identified two non-outbreak-related isolates. There were no significant differences between the isolate grouping results obtained by the different protocols and with the different primers. Visual comparison indicated that the standardized protocols and reagents yielded reproducible fingerprint patterns, but with some variations in particular band intensities. Minor variations in fingerprint profiles were detected, but computer-assisted analysis of PCR fingerprints obtained on agarose gels demonstrated that 88.3 to 91.6% (depending on the source of DNA) of the patterns clustered correctly, while 96.4 to 98.9% of the patterns clustered correctly following automated high-resolution laser fluorescence analysis. Correlation of the patterns for isogenic isolates ranged from 83.3 to 86.6% but was slightly better (mean correlation, 87.1%) for centrally prepared DNA extracts than for DNA extracts prepared by individual laboratories (mean correlation, 84.7%). It was concluded that independently produced PCR fingerprint patterns can be obtained reproducibly for Acinetobacter spp. at the practical level if (i) quality-controlled reagents, (ii) standardized extraction of DNA, and (iii) standardized amplification conditions are used.