Decline in Clostridium difficile-associated disease rates in Singapore public hospitals, 2006 to 2008

BACKGROUND: Clostridium difficile is the major cause of pseudomembranous colitis associated with antibiotic use, and the spread of the hypervirulent epidemic ribotype 027/NAP-1 strain across hospitals worldwide has re-focused attention on this nosocomial pathogen. The overall incidence and trend of C. difficile-associated disease (CDAD) in Singapore is unknown, and a surveillance program to determine these via formal laboratory-based reporting was established. FINDINGS: Laboratory and pharmacy data were collated from one tertiary and two secondary hospitals on a quarterly basis between 2006 and 2008. All hospitals tested for C. difficile using Immunocard Toxins A&B (Meridian Bioscience Inc., Cincinnati, OH) during this period. Duplicate positive C. difficile results within a 14-day period were removed. The CDAD results were compared with trends in hospital-based prescription of major classes of antibiotics. Overall CDAD incidence-density decreased from 5.16 (95%CI: 4.73 - 5.62) cases per 10,000 inpatient-days in 2006 to 2.99 (95%CI: 2.67 to 3.33) cases per 10,000 inpatient-days in 2008 (p < 0.001), while overall rates for C. difficile testing increased significantly (p < 0.001) within the same period. These trends were mirrored at the individual hospital level. Evaluation of antibiotic prescription data at all hospitals showed increasing use of carbapenems and fluoroquinolones, while cephalosporin and clindamycin prescription remained stable. CONCLUSIONS: Our results demonstrate a real decline of CDAD rates in three large local hospitals. The cause is unclear and is not associated with improved infection control measures or reduction in antibiotic prescription. Lack of C. difficile stool cultures as part of routine testing precluded determination of the decline of a major clone as a potential explanation. For more accurate epidemiological trending of CDAD and early detection of epidemic clones, data collection will have to be expanded and resources set in place for reference laboratory culture and typing.