Although the general method was first described by Carling , our data is the first to document that by using a UV visible marker it is possible to easily assess the compliance of housekeeping staff with the cleaning protocol for patient toilets. This is a significant finding as it will allow for more accurate analysis of the efficacy of environmental interventions since compliance with cleaning can be verified.
Our data using Rodac plates as a means of monitoring the presence of C. difficile on surfaces confirmed that this sample method provides essentially 100% recovery. The count recovered using the Rodac plate was slightly higher than the calculated maximum inoculum but this likely reflects the variability in the method of counting. There have been a number of evaluations of the presence and persistence of C. difficile spores in the environment of patients with CDAD. As early as 1981  there was published data demonstrating that the environment of patients with CDAD had a higher likelihood of having C. difficile spores compared to those patients who did not have CDAD (9.3% of 910 floor and surface sites for CDAD patients environments compared to 2.6% of 497 similar sites for non-CDAD patients). Environmental contamination with C. difficile spores is not surprising as Louie  reported that even during and following CDAD treatment patients may shed up to 104 spores/g feces and Tomiczek  commented on the fecal aerosols created when bedpan sprayers are used for cleaning in patient bathrooms. Furthermore, Kim et al  demonstrated that spores of C. difficile could survive for five months on the floor. They also recognized that utilization of contact isolation precautions for CDAD patients was not effective for curbing nosocomial transmission. Our data collected using Rodac plates demonstrates that even after enrolment and implementation of enhanced housekeeping that C. difficile can still be detected in the toilets of 33% of CDAD patients.
The value of using the UVM to monitor cleaning compliance of patient toilet facilities was immediately apparent in that it helped identify a major flaw in our housekeeping. The commodes were not being cleaned at all on 72% of the days sampled. This finding was reviewed at a meeting with Infection Control and Housekeeping administration and it was determined that the nursing staff thought the housekeeping staff were responsible for the commodes while the housekeeping staff thought the nurses were responsible for doing it since they helped the patients use the commodes. The responsibility issue was resolved and housekeeping staff were assigned this responsibility.
Carling  reported that some degree of UV marker removal was achieved in 80% of toilets that were part of terminal cleaning. However, this is a very crude marker of compliance because it was only done to assess compliance with terminal cleaning (i.e. compliance with daily cleaning was not evaluated). Our prospective daily monitoring indicated that for 32.4% of the days when patients were on CDAD isolation, the toilet cleaning was not done as outlined in the housekeeping policy as the UVM was not removed (UVM score of 3). Patients who were not on isolation precautions appeared to have better toilet cleaning because a score of 3 for the UVM was only present for 14.1% of the samples. The average cleaning score for toilets of patients on isolation precautions was higher (i.e. UV marker not removed as well) than for toilets of patients who were not on isolation precautions. This is surprising as patients on isolation for CDAD were supposed to have had their toilets cleaned twice daily compared to once daily for rooms of patients who were not on isolation precautions. This may be similar to the effect noted by Stelfox et al  where contact between healthcare staff and patients on isolation precautions is greatly reduced compared to the level of contact for patients who are not on isolation precautions. The lack of cleaning may reflect excess workload for housekeepers or a reluctance to enter rooms of patients on isolation precautions due to inconvenience of following precautions necessary for entering such isolation rooms.
Although the number of patients enrolled in this prospective study was low (7 in Arm 1 and 13 in Arm 2), the number of patient days assessed was 102 in Arm 1 and 99 in Arm 2. Despite the housekeeping policy requiring these rooms to have twice daily cleaning, toilets of CDAD patient rooms had toxigenic C. difficile detected on 33% of the days post-implementation of twice-daily cleaning compared to 4% for non-CDAD patient rooms where toilet cleaning was once per day. Our data in Table 1 clearly indicated that the cleaning was suboptimal for the toilets of patients on contact isolation precautions (32.7% with UVM not removed). Even when cleaning was optimal (UVM of 0) there were still high detection rates for toxigenic C. difficile (41.5%). Furthermore, our study indicated that C. difficile spores were detected on the toilets of CDAD patients over prolonged periods as some toilets still had toxemic C. difficile detected on day 28 post-enrolment (data not shown). This suggests that both the physical cleaning action as well as the disinfectant/cleaning agent were ineffective for killing and/or removing C. difficile from toilets.
Wilcox et al  reported that using bleach for environmental disinfection of patient rooms did reduce the incidence of CDAD. As pointed out by Dettenkofer , Wilcox's data on the C. difficile spores in the environment demonstrated that spores persisted at similar levels regardless of which cleaning/disinfecting agent was used. Although Wilcox  documented reduced rates of CDAD this could not be sustained when the wards studied were switched over. These difficulties may well be linked to lack of compliance with the housekeeping protocol. It is impossible to conclusively determine the effect of any housekeeping cleaner/disinfector if the compliance of staff with the physical aspect of cleaning cannot be verified. Although there is some evidence that bleach [16, 17] or Accelerated Hydrogen Peroxide  can help contain nosocomial spread of CDAD these studies did not attempt to correlate the detection of spores in the environment with the reduction in cases of CDAD. Further studies are needed that use UVM (or some other validated means of assessing cleaning compliance) to correlate the presence of spores in the environment with an intervention using a specific cleaner/disinfector that has activity against C. difficile spores.
To determine if the poor compliance with the housekeeping protocol extended to rooms of patients not on isolation precautions regardless of whether they had diarrhea or not, a prospective ward-wide surveillance evaluation was undertaken. The data from this part of the current study (Figure 2) demonstrated that compliance with the routine housekeeping policy was ward dependent. There were dedicated housekeeping staff on each ward therefore; our results likely reflect the compliance of the specific housekeeping staff on each ward. From the initial data it appeared that patients who were on isolation precautions were getting less optimal cleaning compared to the rooms of patients not on isolation. However, the prospective "routine ward" assessment of three other wards for one week indicated that on these wards compliance with cleaning can be even worse than for the isolation rooms. The time of highest risk of transmission of C. difficile from one patient to another is likely when the patient is developing diarrhea prior to being diagnosed with CDAD because they have not yet been treated and have not been placed on isolation precautions. As such compliance with routine housekeeping in rooms of patients who are not on isolation precautions is very important because the frequency of physical cleaning is lower and the agent used would have no activity against this organism. Thus use of UVM to monitor compliance to housekeeping protocols would be valuable in all patient rooms – not just those of patients on isolation precautions.