Objective: To investigate an outbreak of Burkholderia cepacia.
Design: Observational study and chart review.
Patients: Adult non-cystic fibrosis (CF) patients.
Setting: Intensive care units (ICUs) at a university-affiliated teaching hospital.
Methods: As part of the epidemiological investigation, we conducted a chart review and collected environmental samples. A review of work schedules of healthcare workers also was performed. We used B. cepacia selective agar for preliminary screening for all isolates, which subsequently were confirmed as members of the B. cepacia complex by polyphasic analysis employing conventional biochemical reactions and genus- and species-specific polymerase chain reaction assays. Pulsed-field gel electrophoresis, randomly amplified polymorphic DNA typing, and automated ribotyping were used to genotype the isolates. As part of the intervention, contact isolation precautions were initiated for all patients identified as having had a culture positive for B. cepacia.
Results: Between September 1997 and September 1999, B. cepacia was isolated from 31 adult patients without CF in ICUs at a university-affiliated teaching hospital. Based on geographic clustering and genotypic analysis, three distinct clusters were observed involving 20 patients. Isolates from 17 of these patients were available for testing and were found to be of the same strain (outbreak strain). Further taxonomic analysis indicated that the outbreak strain was B. cepacia complex genomovar III. Twelve (71%) of the 17 patients were judged to be infected, and 5 (29%) were colonized with this strain. Six of 200 environmental cultures from multiple sources in the hospital's ICUs yielded B. cepacia. Two of these isolates, both recovered from rooms of colonized patients, were the same genotype as the outbreak strain recovered from patients.
Conclusion: Despite an extensive investigation, the source of the B. cepacia clone involved in this outbreak remains unknown. The spatial and temporal pattern of cases suggests that cross-transmission of a genetically related strain contributed to clustering among patients. The initiation of contact isolation may have limited the extent of this transmission. Additional studies are needed to elucidate better the epidemiology of nosocomial B. cepacia infection among non-CF adult patients.