OBJECTIVE: To review phage typing of 12 clusters of nosocomial Klebsiella infections which occurred between 1974 and 1997, and to compare phage typing and K serotyping. Materials and methods A total of 489 clinical and laboratory Klebsiella isolates were phage typed using 110 different phage preparations and K typed by counter current immunoelectrophoresis against 77 K antisera. RESULTS: A total of 152 phage types (PT) and 82 K types were found. Thirty-six phage types and 14 K types were represented only by the reference type strains. Of the remaining 68 K types, 60 could be subdivided into from two to 10 phage types. Ten out of 12 clusters of nosocomial Klebsiella infections could be verified as outbreaks by phage typing, whereas two clusters were found to be accumulations of sporadic cases. K typing performed retrospectively confirmed these results. In addition, for a subset of 104 epidemiologically unrelated isolates, O typing and pulsed field gel electrophoresis typing data were available. Based on these results the discriminative power of phage typing was found to be comparable with that of K typing, but phage types were less stable and reproducible. CONDITIONS: In an outbreak situation, phage typing was found to be very useful, although it seems less suitable for long-term surveillance purposes.
OBJECTIVE: To review phage typing of 12 clusters of nosocomial Klebsiella infections which occurred between 1974 and 1997, and to compare phage typing and K serotyping. Materials and methods A total of 489 clinical and laboratory Klebsiella isolates were phage typed using 110 different phage preparations and K typed by counter current immunoelectrophoresis against 77 K antisera. RESULTS: A total of 152 phage types (PT) and 82 K types were found. Thirty-six phage types and 14 K types were represented only by the reference type strains. Of the remaining 68 K types, 60 could be subdivided into from two to 10 phage types. Ten out of 12 clusters of nosocomial Klebsiella infections could be verified as outbreaks by phage typing, whereas two clusters were found to be accumulations of sporadic cases. K typing performed retrospectively confirmed these results. In addition, for a subset of 104 epidemiologically unrelated isolates, O typing and pulsed field gel electrophoresis typing data were available. Based on these results the discriminative power of phage typing was found to be comparable with that of K typing, but phage types were less stable and reproducible. CONDITIONS: In an outbreak situation, phage typing was found to be very useful, although it seems less suitable for long-term surveillance purposes.