Impact of different empirical antibiotic treatment regimens for community-acquired pneumonia on the emergence of Clostridium difficile

Abstract

Treatment of community-acquired pneumonia (CAP) with newer fluoroquinolones may contribute to selection for Clostridium difficile. We studied the prevalence of C. difficile carriage and C. difficile infection (CDI) on admission, and nosocomial acquisition rates in patients hospitalized for CAP and compared different empirical treatment strategies. In a prospective study among patients admitted for antibiotic treatment of CAP, consecutive stool and skin samples were collected and cultured for C. difficile. Cultured isolates were typed by PCR ribotyping and characterized for toxinogenicity. In total, 20 of 107 (18.7%) patients included carried C. difficile. Various ribotypes were found and 14 (70%) isolates were toxinogenic. On admission, prevalence of C. difficile carriage was 9.4% (n=9), of which 22% also carried C. difficile on the skin and one patient had mild CDI with persistent positive cultures. The overall nosocomial acquisition rate of C. difficile carriage was 11.2%. No nosocomially acquired CDI occurred. Acquisition rates of C. difficile were 11.9% (5/45) in moxifloxacin-, 11.1% (5/47) in β-lactam- and 9.0% (1/14) in β-lactam plus macrolide- or fluoroquinolone-treated patients (P=0.84). Risk factors for C. difficile carriage were intravenous antibiotic treatment >7 days [odds ratio (OR) 3.89; 95% confidence interval (CI) 1.30 to 11.79] and hospitalization during the past 3 months (OR 4.08; 95% CI 1.40 to 11.90). In a non-outbreak setting with a low endemic rate, the prevalence of C. difficile carriage in patients admitted because of CAP is high and nosocomial acquisition rates for C. difficile colonization are 11%. Fluoroquinolones were not associated with increased acquisition rates for C. difficile as compared with other empirical regimens for CAP.