Automated ultraviolet-C (UV-C) room decontamination devices are increasingly used as an adjunct to standard cleaning and disinfection in healthcare facilities.Reference Rutala and Weber 1 Under experimental conditions, these devices have demonstrated efficacy in killing a variety of bacterial pathogens, including Clostridium difficile spores.Reference Owens, Deal and Shoemaker 2 – Reference Rutala, Gergen, Tande and Weber 10 In hospital rooms, use of the devices has been shown to reduce the burden of pathogens on surfaces.Reference Owens, Deal and Shoemaker 2 – Reference Rutala, Gergen and Weber 4 , Reference Boyce, Havill and Moore 6 – Reference Nerandzic, Thota and Sankar 8 , Reference Jinadatha, Quezada, Huber, Williams, Zeber and Copeland 11 – Reference Anderson, Gergen and Smathers 13 There have also been some reports of use of UV-C devices associated with reductions in healthcare-associated infections.Reference Levin, Riley, Parrish, English and Ahn 14 – Reference Simmons, Morgan, Hopkins, Helsabeck, Stachowiak and Stibich 16
In addition to factors such as cost and ease of use, healthcare facilities considering the purchase of a UV-C room decontamination device must weigh claims regarding the effectiveness of different devices. However, few data are available on the comparative effectiveness of different UV-C devices.Reference Nerandzic, Fisher and Donskey 5 , Reference Nerandzic, Thota and Sankar 8 Moreover, in contrast to the situation for surface disinfectants, 17 there are currently no standardized methods recommended for testing of UV-C devices. Thus, a variety of methods have been used in experimental testing. For example, inoculated surfaces have included glass, steel, and Formica carriers or bench countertops, spreading of the inoculum has varied from areas of 4 mm2 to 25 cm2, and inoculated carriers have been oriented in parallel or perpendicular to the bulbs.Reference Owens, Deal and Shoemaker 2 – Reference Rutala, Gergen, Tande and Weber 10 Although it is known that UV-C efficacy is affected by distance and shading, it is not known if variation in other test conditions affect pathogen killing. Here, we compared the effectiveness of 2 UV-C room decontamination devices and evaluated the effect of variation in test methods on performance of 1 of the devices.
METHODS
C. difficile and Methicillin-Resistant Staphylococcus aureus (MRSA) Strains
One strain of each pathogen was studied. The C. difficile strain was American Type Culture Collection (ATCC) strain 43598. Spores were prepared and stored as previously described.Reference Sorg and Sonenshein 18 The MRSA strain was a clinical isolate of pulsed-field gel electrophoresis type USA800. MRSA and C. difficile were cultured on selective media as previously described.Reference Nerandzic, Thota and Sankar 8
The UV-C Disinfection Devices
The Tru-D device (Lumalier) has been described previously.Reference Nerandzic, Cadnum, Pultz and Donskey 3 – Reference Nerandzic, Fisher and Donskey 5 The Clorox Healthcare Optimum-UV System (Clorox) has a 36 inch–diameter circular base and a height of 76 inches. The device is a wheeled unit containing four 64-inch low-pressure mercury lamps emitting 254-nm UV-C light. The device’s cycle time is determined by the manufacturer on the basis of room size and configuration. The system contains a motion sensor connected to a safety-rated relay, aborting the cycle if someone enters the room during use.
Comparison of the Optimum-UV and Tru-D Devices for Killing of the Pathogens
We compared the efficacy of the 2 devices against the pathogens on steel disk carriers in a 10×10 foot room as described previously.Reference Nerandzic, Fisher and Donskey 5 For each pathogen, 10-µL aliquots containing 1×106 colony-forming units in phosphate-buffered saline were spread to cover 1-cm2 area of stainless steel carriers and allowed to air dry. The carriers were placed perpendicular to the vertical lamps (ie, horizontal) 4 feet from the devices at a height of 4 feet and exposed to cycles of 5, 10, 20, or 40 minutes. For comparison, the Tru-D reflected doses of 12,000 (vegetative cycle) and 22,000 (spore cycle) µWs/cm2 in single patient rooms require approximately 20 and 45 minutes, respectively.Reference Nerandzic, Cadnum, Pultz and Donskey 3 To quantify viable organisms, treated and untreated control carriers were submersed in 1 mL of phosphate-buffered saline and vortexed vigorously, and dilutions were plated onto selective media and processed as described previously.Reference Nerandzic, Cadnum, Pultz and Donskey 3 , Reference Nerandzic, Fisher and Donskey 5 , Reference Nerandzic, Thota and Sankar 8 Because initial experiments demonstrated similar efficacy of the devices, subsequent experiments were conducted only with the Clorox Healthcare Optimum-UV System.
Effect of Inoculum Dispersal on Killing of the Pathogens by the Optimum-UV System
To assess the impact of dispersal of the inoculum, 10-µL aliquots of the pathogens were inoculated onto 10-mm2 steel disk carriers (Cole Palmer) either without manual spreading (~4-mm2 area) or with spreading (10-mm2 area) or onto 22-mm2 steel disk carriers with spreading. The carriers were placed perpendicular to the lamps 4 feet from the device at a height of 4 feet in the direct field of radiation and exposed to cycles of 5 or 10 minutes. Reductions were compared with untreated control carriers.
Effect of Orientation of the Carriers Relative to the UV-C Lamps
We tested the effect of the angle of UV-C incidence by altering the orientation of 22-mm2 steel disk carriers relative to the Optimum-UV lamps. The carriers were positioned in parallel with the vertical lamp (ie, vertical orientation directly facing the lamp), perpendicular to the lamp, or at a 45° angle from the lamp. The carriers were placed 4 feet from the device at a height of 4 feet and exposed to 10 minutes of UV-C.
Effect of Different Types of Organic Load
We examined the effect of 3 commonly used simulated organic materials (5% and 10% fetal calf serum and ASTM E2197 standard organic load consisting of 5% tryptone, 0.4% mucin, and 5% bovine serum albumin) on killing of MRSA.Reference Sorg and Sonenshein 18 The 10-µL inoculum of MRSA prepared in each of the test organic materials was spread to cover 22-mm2 steel disk carriers or allowed to dry after placement as a central droplet (~4-mm2 area). Carriers were exposed to 3 minutes of UV-C.
Effect of Carrier Type
We examined reduction in the pathogens inoculated onto steel disks, plastic (Nunc), glass slides (Fisher Scientific), and Formica (Diller). For each carrier, a 10-µL inoculum was spread to cover 22 mm2. The carriers were placed 4 feet from the device at a height of 4 feet and exposed to 5 minutes of UV-C. Log reductions were enumerated against untreated control carriers of the same material.
Effect of Variation in Carrier Height
We tested killing of MRSA at different heights at a distance of 6 feet from the Optimum-UV device. Steel disk carriers were placed on the floor as well as 1, 3, 4, and 6 feet above the floor and exposed to UV-C for 5 minutes.
Effect of Interrupted vs Uninterrupted Dosing
Because some manufacturers recommend running 1 cycle on each side of the bed, we examined the impact of interrupted UV-C cycles on germicidal activity against MRSA. Steel disk carriers were placed 4 feet from the device at a height of 4 feet. Reductions in MRSA were compared for 4 different exposure cycles: (1) uninterrupted 16-minute exposure; (2) 3 exposures of 5.33 minutes each with each exposure interrupted by a 1-minute pause; (3) 3 exposures of 5.33 minutes each with each exposure interrupted by a 20-minute pause; and (4) 8 exposures of 2 minutes each with each exposure interrupted by a 1-minute pause between cycles.
Statistical Analysis
A 1-way analysis of variance was performed to compare the mean log reductions. A post hoc Tukey honest significant difference test was used to test pairwise differences between group means. For each assessment of variation in test methods, a comparison group was chosen as the standard for comparison: spreading of the inoculum, reference equals 4 mm2; orientation of steel disk carriers, reference equals horizontal orientation perpendicular to the lamps; organic load, reference equals no organic load; carrier material, reference equals steel; height, reference equals 4 feet; and cycle interruption, reference equals uninterrupted 16-minute exposure. For statistical analysis, we used R software. 19
RESULTS
As shown in Figure 1, there were no significant differences between the 2 devices in reduction of either pathogen after exposures of 5, 10, 20, and 40 minutes. For MRSA, a greater than 3 log reduction was achieved within 5 minutes. For C. difficile spores, log reductions increased with increasing exposure time, with 20 minutes of exposure required to achieve a reduction of approximately 2 log.
As shown in Figure 2, spreading of the inoculum over a greater surface area significantly enhanced killing of both pathogens. For C. difficile spores, spreading of the inoculum to cover 22 mm2 resulted in a reduction of approximately 2 log within 10 minutes of exposure at 3 feet. For MRSA, spreading of the inoculum to cover 22 mm2 resulted in a reduction of approximately 5 log within 10 minutes. Positioning of the carriers vertically in parallel with the lamp also enhanced killing of both pathogens in comparison with a perpendicular orientation (Figure 3).
Figure 4 shows the effect of different types of organic load on UV-C efficacy against MRSA. The ASTM standard organic load and 10% fetal calf serum significantly reduced killing of MRSA when the inoculum covered a 4-mm2 surface area, whereas 5% fetal calf serum did not. However, none of the organic load suspensions significantly affected killing of MRSA when the inoculum was spread to cover 22 mm2.
As shown in Figure 5, the log reductions of the pathogens achieved on plastic, Formica, and glass slides were not significantly different from the reductions on steel disks. Of note, when the 10-µL inoculum was not manually spread, the reduction of pathogens was significantly greater on glass slides than on steel disks (data not shown); it was observed that natural spreading of the inoculum occurred on glass but not on the steel disk. At 6 feet from the device, there were no significant differences in log reductions of MRSA achieved at different heights ranging from floor level to 6 feet above floor level (P=.88). There were also no significant differences in log reductions of MRSA with 16 minutes of UV-C exposure that was uninterrupted versus administered in interrupted doses.
DISCUSSION
In experimental testing, we found that the Tru-D and Clorox Healthcare Optimum-UV devices were equally effective for killing of MRSA and C. difficile spores. Using the Optimum-UV device, we demonstrated that variation in experimental testing conditions can have a significant impact on the measured efficacy of UV-C. Killing of the pathogens was significantly enhanced by spreading of the inoculum over an increased surface area and orientation of carriers in parallel rather than perpendicular with the UV-C lamps. The choice of organic load also impacted the measured efficacy of UV-C. Our findings highlight the need for development of standardized methods for testing the efficacy of UV-C devices. Such standardization is needed to facilitate comparative effectiveness evaluations, particularly since there may be significant differences in performance of different UV-C devices.Reference Nerandzic, Thota and Sankar 8
The most important practical implication of our results is that relatively short UV-C cycles could potentially be effective in reducing pathogens, including C. difficile spores, on surfaces. We found that 20 minutes was required to achieve a 2 log reduction in C. difficile spores spread to cover 10 mm2, whereas only 10 minutes was required to achieve a similar reduction in spores spread to cover 22 mm2. Similarly, Rutala et alReference Rutala, Gergen, Tande and Weber 10 found that a 10-minute UV-C exposure was sufficient to reduce C. difficile spores by approximately 3 log when the spore inoculum was spread to cover a Rodac template (~25 cm2). Such cycle times would be much shorter than the approximately 45-minute cycles currently recommended for the Tru-D when used in rooms of patients with C. difficile infection.Reference Nerandzic, Cadnum, Pultz and Donskey 3 – Reference Rutala, Gergen and Weber 4 Given that pathogens are typically present in relatively low concentrations on hospital surfaces,Reference Nerandzic, Cadnum, Pultz and Donskey 3 , Reference Nerandzic, Thota and Sankar 8 further studies are indicated to validate the effectiveness of relatively short UV-C exposure times in reducing the burden of C. difficile spores in isolation rooms.
Our findings have some additional practical implications. Since organic load may reduce efficacy of UV-C, our findings support the recommendation that surfaces be cleaned prior to UV-C exposure.Reference Rutala, Gergen and Weber 4 , Reference Zhang, Nerandzic, Kundrapu and Donskey 9 The finding that interrupted and uninterrupted UV-C exposures were similarly effective suggests that the practice of running 2 cycles in different locations in the room will not adversely affect performance.Reference Jinadatha, Quezada, Huber, Williams, Zeber and Copeland 11 – Reference Stibich, Stachowiak and Tanner 12 Finally, positioning of objects such as call buttons or tables in close proximity to UV-C devices has been recommended to take advantage of the fact that efficacy is reduced as distance from the device increases.Reference Nerandzic, Cadnum, Pultz and Donskey 3 – Reference Nerandzic, Thota and Sankar 8 Our results suggest that orientation of objects such that the largest surface area is in parallel with the UV-C lamps may enhance efficacy.
Our study has some limitations. We evaluated only 2 pathogens and tested only 1 strain of each. Moore et alReference Moore, Ali and Cloutman-Green 20 reported that spores of an epidemic ribotype 027 strain were less susceptible to UV-C than a ribotype 001 strain. Similarly, our group reported some variability in killing of different strains of C. difficile and vancomycin-resistant Enterococcus by UV-C.Reference Nerandzic, Cadnum, Pultz and Donskey 3 To standardize testing of chemical disinfectants against C. difficile spores, the Environmental Protection Agency has recommended that 1 clinically relevant strain (ATCC 43598) be used and that a standard method be used for spore preparation. 21 The C. difficile strain used in our study was ATCC 43598. Additional studies are needed to assess the impact of different methods of preparation and handling of C. difficile spores because factors such as triggering of germination or heat activation could have a significant impact on UV-C efficacy.Reference Nerandzic and Donskey 22 – Reference Nerandzic and Donskey 23
ACKNOWLEDGMENTS
Financial support. Department of Veterans Affairs (Merit Review grant to C.J.D.); and Clorox. Clorox provided a Clorox Healthcare Optimum-UV System to be used for the study. Clorox did not provide any input on data analysis or writing and editing of the manuscript.
Potential conflicts of interest. C.J.D. reports that he has served on advisory boards for Clorox and 3M and has received research grants from 3M, EcoLab, GOJO, and Clorox. All other authors report no conflicts of interest relevant to this article.