A Large Outbreak of<i>Clostridium difficile</i>–Associated Disease with an Unexpected Proportion of Deaths and Colectomies at a Teaching Hospital Following Increased Fluoroquinolone Use

<jats:title>Abstract</jats:title><jats:sec id="S0899823X0019782X_abs1"><jats:title>Background and Objective:</jats:title><jats:p>Fluoroquinolones have not been frequently implicated as a cause of<jats:italic>Clostridium difficile</jats:italic>outbreaks. Nosocomial<jats:italic>C. difficile</jats:italic>infections increased from 2.7 to 6.8 cases per 1,000 discharges (<jats:italic>P</jats:italic>< .001). During the first 2 years of the outbreak, there were 253 nosocomial<jats:italic>C. difficile</jats:italic>infections; of these, 26 resulted in colectomy and 18 resulted in death. We conducted an investigation of a large<jats:italic>C. difficile</jats:italic>outbreak in our hospital to identify risk factors and characterize the outbreak.</jats:p></jats:sec><jats:sec id="S0899823X0019782X_abs2"><jats:title>Methods:</jats:title><jats:p>A retrospective case-control study of case-patients with<jats:italic>C. difficile</jats:italic>infection from January 2000 through April 2001 and control-patients matched by date of hospital admission, type of medical service, and length of stay; an analysis of inpatient antibiotic use; and antibiotic susceptibility testing and molecular subtyping of isolates were performed.</jats:p></jats:sec><jats:sec id="S0899823X0019782X_abs3"><jats:title>Results:</jats:title><jats:p>On logistic regression analysis, clindamycin (odds ratio [OR], 4.8; 95% confidence interval [CI<jats:sub>95</jats:sub>], 1.9-12.0), ceftriaxone (OR, 5.4; CI<jats:sub>95</jats:sub>, 1.8-15.8), and levofloxacin (OR, 2.0; CI<jats:sub>95</jats:sub>, 1.2-3.3) were independently associated with infection. The etiologic fractions for these three agents were 10.0%, 6.7%, and 30.8%, respectively. Fluoroquinolone use increased before the onset of the outbreak (<jats:italic>P</jats:italic>< .001); 59% of case-patients and 41% of control-patients had received this antibiotic class. The outbreak was polyclonal, although 52% of isolates belonged to two highly related molecular subtypes.</jats:p></jats:sec><jats:sec id="S0899823X0019782X_abs4"><jats:title>Conclusions:</jats:title><jats:p>Exposure to levofloxacin was an independent risk factor for<jats:italic>C. difficile-associated</jats:italic>diarrhea and appeared to contribute substantially to the outbreak. Restricted use of levofloxacin and the other implicated antibiotics may be required to control the outbreak.</jats:p></jats:sec>