The prevalence of possible cross-transmission of selected bacteria (Escherichia coli, Klebsiella pneumoniae, Enterobacter cloacae, Pseudomonas aeruginosa, Staphylococcus aureus, and enterococci) among infected patients was evaluated in five intensive care units (ICUs) over 6 months. A total of 284 isolates from clinical specimens were typed by plasmid profile analysis (E. coli, K. pneumoniae, and E. cloacae), restriction endonuclease analysis of plasmid DNA (S. aureus), and/or pulse-field gel electrophoresis of chromosomal DNA (P. aeruginosa, enterococci, S. aureus, and other bacteria without plasmid DNA). By typing criteria, only 13% of the 177 isolates obtained after > 2 days in an ICU were classified as possibly cross-transmitted. Many patients whose cultures yielded bacteria of an identical type may have been the sources rather than the recipients of these organisms. Episodes of possible cross-transmission were scattered among all ICUs, usually affected only two patients, and were associated with most bacterial species. These data suggest that endemic bacterial cross-transmission in ICUs is relatively infrequent and that cross-transmitted bacteria are not common causes of endemic ICU-related nosocomial infections.

We evaluated test discriminatory power and DNA type alterations among methicillin-resistant Staphylococcus aureus strains by testing 199 sequential isolates from 39 patients collected over 30 to 228 days. Isolates were typed by one or three different methods (restriction endonuclease analysis of plasmid DNA [REAP] with or without pulsed-field gel electrophoresis of genomic DNA [PFGE] and immunoblotting [IB]). REAP was highly discriminatory compared with PFGE and IB. However, the initial isolates from 4 of the 39 patients lacked detectable plasmid DNA and could not be typed by REAP. Typing of individual patient isolates showed that a different REAP type was identified only once every 138 days. Among 25 comparisons, seven sequential isolate pairs demonstrating REAP differences were also different by PFGE and IB. This likely represented the presence of more than one strain. Eighteen other pairs with REAP differences were identical or related to one another by PFGE and IB typing, and 17 of these differences were likely caused by a single genetic alteration within the same strain or clone. The rate of PFGE differences explicable by single genetic alterations among sequential isolates identical by REAP was similar to the overall rate for REAP differences in the whole collection. We conclude that REAP and PFGE typing differences explicable by single genetic alterations are relatively infrequent but not rare. These isolates should be examined by alternative typing systems to further support or refute clonality.

No abstract

No abstract