Emergence and control of fluoroquinolone-resistant, toxin A-negative, toxin B-positive Clostridium difficile.

BACKGROUND Clostridium difficile is a major cause of infectious diarrhea in hospitalized patients. Between August 2003 and January 2004, we experienced an increase in the incidence of C. difficile-associated disease. We describe the investigation into and management of the outbreak in this article. METHODS A total of 73 consecutive patients with nosocomial C. difficile-associated diarrhea were identified. C. difficile isolates were characterized using toxin-specific enzyme immunoassays, a tissue-culture fibroblast cytotoxicity assay, polymerase chain reaction (PCR), and antimicrobial susceptibility tests. Rates of recurrence and of C. difficile colitis were recorded. Changes in antibiotic use and infection control policies were documented. RESULTS The incidence of C. difficile-associated diarrhea peaked at 21 cases per 1,000 patient admissions. Of the C. difficile isolates recovered, 85 (95%) were identical toxin A-negative and toxin B-positive strains, corresponding to toxinotype VIII and PCR ribotype 017. All clonal isolates were resistant to multiple antibiotics, including ofloxacin, ciprofloxacin, levofloxacin, moxifloxacin, and gatifloxacin (minimum inhibitory concentrations [MICs] of greater than 32 micro g/mL) and erythromycin, clarithromycin, and clindamycin (MICs of greater than 256 micro g/mL). Recurrent C. difficile-associated disease occurred in 26 (36%) of the patients. At least 10 (14%) of the patients developed C. difficile colitis. Additional infection control measures introduced included the use of ward memos, a hand-hygiene awareness campaign, increased environmental cleaning, attention to prescribing practices for antibiotics, increased awareness of diarrheal illness, and early isolation of affected patients. Total use of fluoroquinolones did not change throughout the study period. Despite persistence of this toxin-variant strain, the incidence of C. difficile-associated disease in our institution decreased to fewer than 5 cases per 1,000 admissions. CONCLUSIONS We report on the emergence of a fluoroquinolone- and clindamycin-resistant, toxin A-negative, and toxin B-positive strain of C. difficile associated with an outbreak of C. difficile-associated disease in our institution during a 6-month period. We found that careful attention to improvement of infection control interventions was the most important means of controlling this nosocomial pathogen.