Use of Ultraviolet Irradiation in Addition to Commonly Used Hospital Disinfectants or Cleaners Further Reduces the Bioburden on High-Touch Surfaces

Abstract

Background

The reduction in microbial bioburden achieved by terminal disinfection in a hospital may vary considerably by type of disinfectant or cleaner and by environmental service (EVS) personnel. This study estimated whether supplemental ultraviolet (UV) irradiation after disinfection or cleaning reduced bioburden and whether the supplemental effect persisted after adjusting for disinfectant or cleaner type and EVS personnel.

Methods

Environmental samples for aerobic bacterial colonies (ABC) and methicillin-resistant Staphylococcus aureus (MRSA) were obtained from 5 high-touch surfaces in patient rooms at 3 time points: before manual cleaning, after manual cleaning using 1 of 3 disinfectants or 1 cleaner, and after UV irradiation.

Results

For ABC, the model-estimated mean (95% uncertainty interval) counts were 56% (48%–63%) lower for postmanual + UV compared with manual clean alone, and for MRSA they were 93% (62%–99%) lower for postmanual + UV vs manual clean alone.

Conclusions

Although UV supplementation provides incremental benefit in bioburden reduction across all cleaners or disinfectants tested, it provides the greatest benefit when supplementing cleaners or disinfectants with the lowest disinfection properties. UV irradiation provides additional bioburden reduction to manual cleaning or disinfection, even when accounting for variability introduced by different disinfectants and EVS personnel.

Reducing microbial bioburden through terminal disinfection of patient rooms is a crucial component in preventing health care–associated infections (HAIs) [1]. However, the level of disinfection may vary with the type of disinfectant used and the individual environmental services (EVS) personnel using it [2, 3]. Cleaners but not disinfectants are used in some countries other than the United States by EVS personnel [4]. Several no-touch disinfection technologies (NTD) have recently been implemented to improve disinfection beyond what is achieved by manual cleaning alone. NTD reduces the impact of human variation and improves overall disinfection results [5, 6]. One such technology is the use of ultraviolet light (UV) after manual disinfection [6, 7]. The purpose of this study was to estimate the effect of supplemental UV compared with standard manual disinfection or cleaning using 3 commonly used hospital disinfectants and 1 cleaner and whether the result persisted after adjusting for disinfectant or cleaner type and EVS personnel.

METHODS

The study was conducted in a 120-bed acute care Veterans Affairs hospital located in Temple, Texas. The study was reviewed and approved by the research committees. The room selection criteria included single-occupant rooms that had been occupied for a minimum of 48 hours with a discharge time before 3:00 pm on a day when research staff and EVS staff were available.

Environmental Sampling

Research staff collected environmental samples for aerobic bacterial counts (ABC) and methicillin-resistant Staphylococcus aureus (MRSA) using nonselective RODAC contact plates (Hardy Diagnostics, Santa Monica, CA, USA) from 5 high-touch surfaces (bedrail, call button, toilet seat, bathroom handrail, and tray table) before the room was cleaned [6]. Trained EVS staff, dedicated to research only, then performed a terminal cleaning of the room using 1 of 4 preselected disinfectant or cleaning agents: sodium hypochlorite 10% solution (SH10; Dispatch, Clorox Healthcare Services, Pleasanton, CA, USA), hydrogen peroxide with peracetic acid (HPA; Oxycide, Ecolab, St. Paul, MN, USA), quaternary ammonium compound (QAC; Virex II 256, Diversey Inc, Sturtevant, WI, USA), or detergent (DT; Dawn Dish Soap, Proctor & Gamble, Cincinnati, OH, USA) using standard manufacturer protocols. After the terminal cleaning, research staff collected postcleaning environmental samples from the same high-touch surfaces but in adjacent locations [6]. After collection of the postcleaning samples, a UV device (Xenex Disinfection Systems, San Antonio, TX, USA) was used 3 times—a 5-minute cycle on each side of the patient bed and a 5-minute cycle in the restroom [8]. Post-UV samples were taken after completion of the UV irradiation. Plated colonies were counted and recorded by a laboratory technician in a blinded manner [6].

Statistical Analyses

Bacterial counts were compared between manual clean and manual clean + UV in a Bayesian negative binomial multilevel regression model of ABC counts with the following predictors: treatment (manual clean or manual + UV), cleaner or disinfectant chemical, sample surface location, precleaning bacterial count (z-transformed), a varying intercept for sample ID, and a varying intercept for EVS staff; data were analyzed using a similarly constructed zero-inflated negative binomial multilevel regression model for MRSA. The models were run in R, version 3.2.3, using the ‘brms’ package [9].

RESULTS

Six hundred precleaning, 600 postcleaning, and 600 postcleaning + UV samples were taken. The same number of samples (150 pre, 150 post, 150 post + UV) were taken for each disinfectant or cleaner and for each surface location (60 pre, 60 post, 60 post + UV). Our study included 30 rooms per disinfectant/cleaner arm, for a total of 120 rooms. Seventy-four out of 120 rooms (61.67%) were MRSA isolation rooms. The mean, median, and sums for ABC and MRSA counts for each cleaning chemical for each EVS staff person at precleaning, post–manual cleaning, and post–manual cleaning + UV are shown in Figures 1 and 2. Fourteen different EVS staff participated in the study, but not all staff cleaned the same number of rooms or used each of the disinfectants or cleaners.

Figure 1.

Mean bacterial counts and summary statistics for aerobic bacterial colonies for each EVS staff and cleaner type for precleaning, post–manual cleaning, and postmanual + ultraviolet. Please note that not all EVS staff cleaned with each of the cleaner types. Abbreviations: ABC, aerobic bacterial colonies; DT, detergent; EVS, environmental service; HPA, hydrogen peroxide with peracetic acid; QAC, quaternary ammonium compound; SH10, sodium hypochlorite 10% solution; UV, ultraviolet light.

Figure 2.

Mean bacterial counts and summary statistics for MRSA for each environmental service staff and cleaner type for precleaning, post–manual cleaning, and postmanual + ultraviolet. The y-axis for MRSA is different from aerobic bacterial colonies. Abbreviations: DT, detergent; EVS, environmental service; HPA, hydrogen peroxide with peracetic acid; MRSA, methicillin-resistant Staphylococcus aureus; QAC, quaternary ammonium compound; SH10, sodium hypochlorite 10% solution; UV, ultraviolet light.

The model-estimated mean (95% uncertainty interval) ABC counts were 56% (48%–63%) lower for postmanual + UV compared with postmanual, adjusting for baseline postmanual counts for each chemical arm and holding all other predictors and random effects constant. The standard deviation of the varying intercept for EVS staff was 0.91 (0.40–1.64) on the log-odds scale.

The model-estimated mean MRSA counts were 93% (62%–99%) lower for postmanual + UV compared with postmanual, adjusting for baseline postmanual counts for each chemical arm. The standard deviation of the varying intercept for EVS staff was 0.96 (0.03–3.21) on the log-odds scale.

Real-world estimated effects of UV on ABC and MRSA counts were obtained via model prediction, conditioning on the cleaning chemical used, the mean precleaning count, and a sample surface location of the bathroom handrail. The model-estimated mean ABC counts and 95% uncertainty intervals for the mean and for each disinfectant or cleaner, with and without the use of UV, can be seen in Figure 3.

Figure 3.

Model estimated mean aerobic bacterial colony counts and 95% uncertainty intervals for manual and postmanual + ultraviolet for each cleaner/disinfectant. Abbreviations: ABC, aerobic bacterial colonies; DT, detergent; HPA, hydrogen peroxide with peracetic acid; QAC, quaternary ammonium compound; SH10, sodium hypochlorite 10% solution; UV, ultraviolet light.

Model-estimated mean MRSA counts were <0.1 postmanual and <0.01 postmanual + UV for all cleaning chemical arms, and lower and upper uncertainty bounds for MRSA counts were 0 when rounded to the nearest count for all cleaning chemical arms. Actual summed MRSA colony counts in our study decreased from 145 to 0 from postmanual to postmanual + UV in the QAC arm, from 102 to 11 in the DT arm, from 5 to 2 in the HPA arm, and from 2 to 0 in the SH10 arm.

DISCUSSION

The results of our study indicate that supplementing manual disinfection or cleaning with UV provides additional bioburden reduction, findings consistent with our previous studies [6–8]. Greater residual ABC and MRSA after UV supplementation for the QAC and DT arms as compared with HPA and SH10 (Figures 1 and 2) is due to the inherent lower or no disinfection potency for this disinfectant (QAC) and cleaner (DT) [2]. For example, DT lacks disinfectant properties, so any bioburden reduction is purely due to mechanical removal of organic material. As seen in Figure 3, postmanual + UV ABC counts are still higher for the QAC and DT arms compared with SH10 and HPA alone or after supplementation with UV, indicating that UV supplementation by itself does not completely make up for a lack of disinfectant use in the manual clean but does provide incremental benefit. Our study results also suggest that supplementing manual disinfection or cleaning with UV provides further reduction of MRSA on a surface regardless of the cleaner or disinfectant used, although the very low counts after manual disinfection using SH10 and HPA make this incremental benefit less practically important. Our model-predicted mean counts were >0.01, and the actual summed counts from our data after use of UV were 13 colonies out of 600 samples (Figure 1). This large effect for UV for MRSA compared with the effect for ABC is likely due to the very low level of MRSA present at preclean compared with substantial bioburden levels at preclean for ABC.

Our study is unique in that it is the first head-to-head assessment of the efficacy of UV using 3 popular disinfectants and a cleaner, while accounting for the performance variability in EVS personnel. Variability in EVS staff performance was included in the analysis by allowing the baseline mean bacterial colony counts to vary with each individual staff member. This was accomplished by including a varying intercept in the statistical model. Accounting for this variability gives us a more realistic estimate of the efficacy of a disinfectant as well as any additional effect of UV. As can be seen in Figures 1 and 2, not all EVS staff clean equally well given the same disinfectant or cleaner. The group-level variability in performance among EVS personnel given the same product suggests that standardized training for EVS personnel could improve disinfection, as has been suggested in other studies [10]. The variability in product performance on bioburden including HAI-causing pathogens such as MRSA suggests that UV may be more beneficial for facilities using certain disinfectant or cleaning products. If QAC is used, supplementing with UV provides a large incremental benefit. Although DT alone is probably not used in any facilities in the United States for cleaning, supplementing it with UV does provide additional benefit, important information for countries that use detergent-based cleaning, especially as manual cleaning alone seems inadequate. These results further inform the long-time debate regarding the value of detergent-only-based cleaning seen in the United Kingdom and northern Europe, as opposed to disinfectant-based cleaning seen in the United States and Australia [4].

Our study had limitations. All samples were collected at a single hospital, so the results may not be generalizable to other facilities in different geographic locations. In addition, EVS staff were not randomized to the cleaning chemical arm. The recovery of MRSA counts in high-touch surfaces was low, thus limiting the analysis of the MRSA results. Larger studies are needed to validate our single-center study results. Another limitation to the study is the use of EVS personnel dedicated to research only. This could have blunted the potential additional impact of the disinfectants and the UV on ABC and MRSA counts. But, on the other hand, this study represents the best case scenario.

CONCLUSIONS

Supplementing manual cleaning or disinfection with UV provides additional bioburden reduction for aerobic bacteria, possibly including MRSA. Although UV supplementation by itself does not completely make up for a lack of disinfectant in manual cleaning, it does provide incremental benefit. The practical importance of UV supplementation is largely related to the amount of bioburden left after manual cleaning, so if an effective disinfectant such as HPA or SH10 is used, the bioburden may be so low that the additional effect of UV is minimal.

Prior presentation. Part of this manuscript was submitted as a poster to ID Week; October 2017; San Diego, CA, USA.