CLINICAL AND MOLECULAR EPIDEMIOLOGY OF SPORADIC AND CLUSTERED CASES OF NOSOCOMIAL CLOSTRIDIUM-DIFFICILE DIARRHEA

PURPOSE: A prospective clinical and molecular epidemiologic study was conducted to define the frequency of nosocomial Clostridium difficile patient-to-patient transmission in an urban tertiary referral hospital . PATIENTS AND METHODS: Over a B-month period, environmental cultures for C difficile were obtained from patients with new positive stool cy totoxin assay (index cases); stool samples were obtained from selected patient contacts (the roommate, occupants of adjacent rooms, and the patient occupying the index room after discharge of the index case); a nd hand cultures were obtained from personnel contacts. C difficile is olates were analyzed by pulse-field gel electrophoresis (PFGE) or, for isolates that were nontypeable by PFGE, by restriction enzyme analysi s. RESULTS: During the study period, we identified 98 index cases of C difficile toxin-associated diarrhea, including focal outbreaks on two wards totaling 26 cases within a 2-month interval. Environmental cont amination was detected at greater than or equal to 1 sites in 58% of r ooms and often involved wide dispersed areas. Among 99 prospectively i dentified patient contacts, C difficile was cultured from the stool of 31 (31%), including 12 with diarrhea and 19 who were asymptomatic. C difficile was cultured from the hands of 10 (14%) of 73 personnel. Mol ecular analysis resolved 31 typing profiles among the index isolates; the most common profile (designated strain D1) was represented by 30 i solates. Among the isolates from patient contacts, 5 of 12 from sympto matic contacts matched the corresponding index isolate, and only 1 of 19 from asymptomatically colonized contacts matched. Transmission to p ersonnel or patient contacts of the strain cultured from the correspon ding index case was correlated strongly with the intensity of environm ental contamination. Strain D1 was frequently represented among isolat es associated with heavy environmental contamination, with personnel c arriage, and with development of symptomatic illness among prospective ly identified contacts. CONCLUSIONS: Intense environmental contaminati on and transmission to close personnel and patient contacts represente d coordinated properties of an individual epidemic strain. For most ep idemiologically linked contacts, positive cultures for C difficile did not result from transmission from the presumed index case.