This is the first study to explore the epidemiology of both MRSA and C. difficile in the general ward environment of community hospitals. By investigating MRSA and C. difficile contamination prospectively and under endemic conditions, the information collected can be used by various hospital personnel for surveillance and infection and environmental control measures to reduce the transmission and dissemination of these HA pathogens within the healthcare setting.
Overall, 11.8% and 2.4% of surfaces in the general environment of the medical and surgical wards were contaminated with MRSA and C. difficile, respectively. In the literature, the prevalence of MRSA in the hospital environment has ranged from 0.6% to 54% [11, 17, 35–37] and for C. difficile the prevalence has ranged from 2.8% to 50% [14, 16, 21, 38, 39]. These wide variations in the reported prevalence of MRSA and C. difficile contamination may be attributed to different study designs, including sampling times (endemic versus outbreak settings), the presence of colonized and/or infected patients during sampling, sampling in different hospital wards, sampling pre- and post-disinfection, sampling surfaces once versus multiple times, and the use of different sampling techniques and culture methodologies. Furthermore, the majority of studies investigating MRSA and C. difficile have been conducted almost exclusively in patient rooms [12, 15, 17–22] as opposed to this study which examined the general hospital environment. Therefore, care must be taken when comparing prevalence data between studies.
With respect to the general ward environment, limited information on MRSA and C. difficile contamination exists in the literature. Dancer and colleagues  screened the computer keyboard, desk, and patient notes located in nursing work stations in two surgical wards during their year-long investigation. In nurses’ work stations, 2.2% (95% CI 0.9-4.6%) of samples were positive for MRSA. Dumford and colleagues  conducted a point-prevalence culture survey for toxigenic C. difficile strains in physician and nurse work areas, specifically targeting telephones, tabletops, computer keyboards, and door knobs on eight different wards. Results from their investigation revealed that 31% (95% CI 15.3-50.8) and 10% (95% CI 0.3-44.5%) of surfaces in physician and nursing work areas, respectively, were contaminated with C. difficile. In the present study, the overall prevalence of MRSA and C. difficile on surfaces located in areas accessible only by hospital personnel ranged from 5.8-16.7% and 0-6.4%, respectively. Surfaces sampled included patient charts, computer keyboards, chair backs, and telephones. These surfaces are repeatedly touched by staff throughout the day or immediately after patient contact or leaving an isolation room. Therefore, contamination may be attributed to a number of factors including lack of hand hygiene, failure to use personnel protective equipment, the inability to properly disinfect a surface, inconsistent cleaning schedules, or the quality of cleaning and disinfection protocols.
The dispersal of MRSA in air has been documented in rooms with MRSA positive patients [17, 40]. However, information with respect to MRSA outside of patients’ rooms is limited and the role of airborne or aerosolized MRSA in infection or colonization of patients or healthcare workers is not known. In the present study, MRSA was identified in air samples from a nurses’ work station as well as the area located outside the elevator doors. As hospital personnel, patients, and visitors were present in these areas during sampling, it is possible that dust or shed squames containing MRSA from nearby surfaces, the ventilation system, the disposal of linen, or an individual colonized with MRSA may have introduced MRSA in these areas.
In the three participating hospitals, MRSA or C. difficile were recovered from unused (clean) linen and/or isolation gowns. Possible reasons for these surfaces being identified with HA pathogens may include contamination from staff, patients, or visitors, cross-contamination from a contaminated storage cart, or contamination during the laundry process. Therefore, it may be necessary to clean and disinfect storage carts to prevent cross-contamination in addition to conducting an audit of the laundry process to ensure that linen and isolation gowns are not contaminated prior to use.
Chair backs, hand rails, isolation carts, and sofas, had an increased odds of being contaminated with MRSA compared to computer keyboards. Model-based contrasts also demonstrated that these four surfaces had increased odds of MRSA contamination compared to other surfaces that were commonly found in the general environment of hospital wards. In addition, Figure 1 illustrates that sofas are especially prone to MRSA contamination. Many of the sampled surfaces were common hand-touch sites not only by hospital personnel but also patients and visitors, thereby increasing the rates of contact along with the number of different people having contact, and subsequently the potential for contamination and transmission. Furthermore, certain surfaces may also be at increased odds for contamination due to the inherent difficulty in cleaning them. Surface material was statistically significant in the univariable analysis, but not in the final model for MRSA contamination. While no significant association was identified, fabric, laminate, and plastic surfaces were found to be contaminated with MRSA and C. difficile on multiple occasions and experiments conducted using swatches of fabric and plastic have demonstrated that staphylococci can survive days to months after drying on these types of surfaces . Furthermore, as these surfaces can differ in their texture, they may be more difficult to adequately clean and disinfect. Accordingly, further study of the influence of surface type on the prevalence and persistence of contamination is indicated.
In Canada, epidemic MRSA clones have been designated using pulsed-field gel electrophoresis (PFGE) [26, 42]. While PFGE was not performed in this study, a PFGE clone can be inferred from spa typing . In Hospital A, spa types consistent with CMRSA-2 predominated in both the environment and patients, which is unsurprising since CMRSA-2 is the leading cause of HA-MRSA in Canada . In Hospital B, CMRSA-2 predominated in the environment while both CMRSA-2 and CMRSA-5 were found in patients. CMRSA-5 is an uncommon human epidemic clone typically associated with HA-MRSA infection , but is common in people that have contact with horses . Interestingly, in Hospital C, CMRSA-5 was most common in the environment while CMRSA-2 and CMRSA-10 were predominant patient clones. CMRSA-5 was not identified in any patient in Hospital C. The apparent disconnect between patient and environmental MRSA clone distribution in Hospital C is interesting and may suggest unidentified reservoirs or sources, such as hospital staff, visitors, or unscreened patients.
To our knowledge, this is the first report of spa type 539/t034 contaminating the hospital environment in Canada. This spa type is associated with the livestock-associated (LA) sequence type (ST) 398, although recent evidence indicates it is actually most likely a human S. aureus clone that moved into livestock and subsequently acquired methicillin-resistance . This spa type is an important cause of community-associated MRSA infection in some regions, particularly northern Europe, most often in individuals with contact with pigs [46–48]. Although this MRSA strain is endemic in the swine population in Canada [49, 50] and has also been found in a horse  and dogs , its role in human infections is unclear. A limited number of human infections have been reported , and it remains to be determined whether this is a rare endemic infection, an under diagnosed problem, or an emerging disease in Canada. The relatively high prevalence of environmental contamination with LA-MRSA was surprising given that reports of human infection and colonization in Canada are uncommon and the absence of this strain in patients in this study. The three participating hospitals serve rural communities where pig farming is present, which may increase the likelihood of LA-MRSA exposure. Despite environmental contamination, HA infection with ST398 MRSA was not identified, which is perhaps because this particular clone is known to be relatively inefficiently transmitted in hospitals , and is probably less infectious than typical human epidemic clones. Therefore, it is not unreasonable to suspect that spa type 539/t034 could be present in colonized patients, healthcare workers, or visitors in the absence of recognized disease in a facility.
In the present study, eight different C. difficile ribotypes were identified in the general environment, with six being toxigenic and therefore clinically relevant. However, care must be taken when interpreting the C. difficile typing data due to the small number of isolates. Overall, toxinotype 0 (MOH-T and MOH-0) was the most prevalent, but was identified in Hospital C only. The hypervirulent ribotypes 027 and 078 were also identified in the environment of Hospital A and/or B, along with ribotypes MOH-AD and MOH-C. The predominance of toxinotypes 0 and III (ribotype 027) in the environment is not surprising since these were the most common toxinotypes in an earlier study of hospitalized patients in Ontario . However, in that particular study, the highest prevalence of toxinotype 0 strains was associated with ribotype 001, which was not identified in the present investigation.
Currently, there is no universally accepted standard for the surveillance of pathogens from surfaces in the hospital environment [56–58]. A variety of sampling techniques have been employed to recover MRSA and C. difficile from the hospital environment including dry or moistened swabs [12, 56], sterile gauze pads , sterile electrostatic cloths , and contact plates [17, 56]. Experiments have demonstrated that contact plates had a higher efficiency compared to culture swabs for the recovery of C. difficile. However, for MRSA, reports of contact plates  and specific types of culture swabs  have been identified as efficient and sensitive sampling techniques, respectively. Although contact plates can be processed more quickly than culture swabs , contact plates are typically limited to sampling flat surfaces [56, 58] whereas culture swabs can be used on irregular surfaces . Like culture swabs, electrostatic cloths can be used to sample irregular surfaces and are also inexpensive, simple to use, and easy to sterilize . In a recent study evaluating C. difficile contamination in households, Weese and colleagues  isolated C. difficile from 5.3% (95% CI 3.8-7.0%) of surfaces using similar sampling and culturing methods as the present investigation.
The present study has several limitations. First, caution should be exercised when interpreting some of the results. For example, although sofas were significantly associated with MRSA contamination, they only constituted seven surfaces in the entire study. Second, not all environmental surfaces were sampled each week. Reasons for surfaces not being sampled four times included equipment/surfaces that had been cleaned by housekeeping the morning of sampling or equipment that was being used by hospital personnel or patients at the time of sampling. This lack of sampling follow-through may bias the overall prevalence of MRSA and C. difficile contamination in the hospital environment. Lastly, the discordance of MRSA strains and C. difficile ribotypes between the general environment and patient specimens may be attributed to isolates that were not collected and/or patients with MRSA or C. difficile that were not identified during the study period.