Résumé
BACKGROUND: Over the decade, Klebsiella pneumoniae resistant to broad-spectrum oephalosporins have been involved in hospital outbreaks, particulaly in intensive care units. Betwem March 20 and June 12. 1900. an outbreak of sepsis caused by multiresistant K. pneumoniae in the neonatal intensive care unit (NICU) of Asan Medical Center. This paper describes bacteriologic, molecular and epidemiologic features of the outbreak. METHODS: For surveillance purpose, stool specimens were obtained from all patients, nurses and house staff in NICU and cultured onto MacConkey agar medium containg cefotaxim, (10 microgram/ml). All K. pneumoniae isolated blood culture isolates form patients with sepsis were tested for antobiogram by microbroth dilution method and for detection of extended-spectrum beta-Iactamase (ESBL) by double disk synergy test and ESBL Etest. Restriction profiles of total genomic DNAs were compared by pulsed filed gel electrophoresis(PFGE) after cleavage by Xbal. beta-Lactamase was tested using nitroefin disks and characterized by transconjugation to Escherichia coli and isoelectric focusing. For infection control, all infected or colonized patients and nurses were cohorted into a separate room and strict barrier precautions were enforced. RESULTS: The outbreak involved 7 patients with sepsis form whom multiresistant. K. pneumoniae were isolated. Surveillance culture revealed that 9 of 37 patients and 2 of 48 nurses and house staff were colonized. The 18 isolates showed 8 different antimicrobial resistance patterns with cefotaxime resistance in all. Test for ESBL was positive in all 18 isolates but only 15 isolates by ESBL Etest. PFGE analysis showed that 6 of the 7 blood isolate from infected patient and 9 of the 11 fecal isolates from surveillance cultures were of the identical or very similar pattern. beta-Lactamase activities were transferable by conjugation in all but one isolate. No additional case of multiresistant. K. pneumoniae infection had been reproted for 6 months since the introduction of strict barrier precautious and other infection control measures. CONCLUSION: The outbreak was caused by ESBL-producing K. pneumoniae which appeared to be introduced into the NICU from multiple sources as was indicated by PFGE patterns. An optimal laboratory method for screening for ESBL remain to be developed as the double disk synergy test and ESBL Etest did not show complete agreement. As for infoction control our results emphasize the necessity of early recognition of outbreaks, cohorting of not only infected but also colonized patients and reinforcement of the barrier precuations for the prevention of further spread of cross-infections.