A Surveillance on Nosocomial Acquisition of Clostridium difficile and Comparative Analysis of 3 Molecular Typing Methods including Pulsed-Field Gel Electrophoresis, Arbitrarily Primed Polymerase Chain Reaction, and Ribotyping / 감염

ABSTRACT

BACKGROUND: Clostridium difficile is a major cause of nosocomial infectious diarrhea. Nosocomial clusters of C. difficile disease have been ascribed to the transfer of the organism form patient to patient. The aim of this study was to survey the nosocomial acquisition of C. difficile infection and to evaluate the efficacy and efficiency of epidemiologic typing systems by molecular analysis of the isolates. METHODS: A surveillance study for C. difficile acquisition was performed in patients admitted to neurology ward (NW) and medical intensive care unit (MICU) in an 800-bed tertiary-care hospital from August 1998 to October 1998. Stool specimens were taken weekly for culture of C. difficile. All isolates were examined for toxin B gene by PCR assay. Three molecular typing methods, including pulsed-field gel electrophoresis (PFGE), ribotyping, and arbitrarily primed polymerase chain reaction (AP-PCR) were used to differentiate individual strains of C. difficile isolates. Their performance characteristics were compared according to the consensus guidelines by the European Society for Clinical Microbiology and Infectious Disease. RESULTS: A total of 38 C. difficile strains were isolated from 308 stool cultures. The period prevalence was 7.4/1000 patient-days and 21.2/1000 patient-days in the NW and MICU, respectively (P=0.034). The acquisition incidence of C. difficile infection was 1.85/ 1,000 patient-days and 5.33/1,000 patient-days in NW and MICU, respectively. The toxin B gene was detected in 38% (8/21) of C. difficile isolates; 62.5% from diarrheal patients and 23% from asymptomatic patients. In a comparison of the three typing systems, the typeability was 0.444 by PFGE, 0.972 by AP-PCR and 1 by ribotyping, and the discrimination index was 0.975 by PFGE, 0.810 by AP-PCR and 0.777 by ribotyping. All three typing systems were highly reproducible. AP-PCR was the least costly and most rapid method. CONCLUSION: The relatively high prevalence of C. difficile infection in the hospital might indicate a potential nosocomial spread, even though the acquisition incidence was low. AP-PCR appears to be an efficacious and efficient method for the epidemiologic study of C. difficile infection, and its suboptimal discriminative power may be enhanced by complementary PFGE.