There are a number of studies which have demonstrated the effect of ultraviolet light disinfection in reducing environmental microbiological contamination [3,4,5] and healthcare-acquired infections by MRSA [6, 7], VRE [7, 8] and Clostridioides difficile [7,8,9,10,11,12]. However, all but one study [5] was performed in the United States, where most of the patient rooms are private. Furthermore, that study [5] did not evaluate clinical outcomes.
Our study is the first hospital-acquired infection outcome study evaluating the clinical effectiveness of the PX-UV device in a non-US healthcare setting with an open-style ICU. In this setting, patient beds were separated by privacy curtains, not by walls. PX-UV emits intense visible light and creates a sound while disinfecting. The light and sound can be seen and heard outside the privacy curtains. This was initially not well tolerated by healthcare staff, some of whom raised concerns about sensitivity to the light and sound. We also experienced faults from the pulse oximeter when PX-UV was used adjacent to patient beds. To overcome these challenges, we provided goggles and earplugs to healthcare staff and ordered blackout curtains from the PX-UV vendor that hung inside the privacy curtains during PX-UV operation. The blackout curtains worked well and eliminated the problems stated above.
Adding PX-UV disinfection to routine terminal cleaning after patient discharge increased the turn-around time of ICU beds by approximately 20 min. Manual cleaning by sodium hypochlorite solution took about 50 min, so the increase in time by adding PX-UV was not significant and was well accepted by ICU staff and physicians as a routine workflow.
The effectiveness of PX-UV is expected to be maximized when environmental contamination is a major factor in transmission of the specific pathogen. In this context, pathogens such as Clostridioides difficile and multidrug-resistant Acinetobacter are more likely to be controlled. The effectiveness of PX-UV in controlling transmission of Clostridioides difficile is well studied and demonstrated [7,8,9,10,11,12], but that of Acinetobacter has not been well investigated.
Sporadic transmission of 2DRA has been observed in our ICU for the last five years. In Japan, antimicrobial resistance of Acinetobacter is not as serious. According to the Japanese national microbiological surveillance report, 97, 97 and 87% of Acinetobacter isolates were reported to be susceptible to meropenem, amikacin and levofloxacin, respectively [13]. Therefore, most of the newly identified 2DRA in our hospital seemed to be acquired by horizontal transmission. Until mid-2018, this situation has not been well controlled, despite our efforts for elevated compliance to hand hygiene, strict contact precaution of patients with this pathogen, terminal cleaning using bleach, and in some occasions, restriction of new admission to the ICU. However, by introducing the PX-UV, transmission of 2DRA in the ICU halted. As of June 2019, no new isolation of 2DRA from patients in the ICU has been observed for 11 months (August 2018 to June 2019). Generally, terminal cleaning by bleach is effective in preventing transmission of pathogens via environmental route, however, the result from the environmental sampling in this study, approximately 80% reduction in CFU by manual cleaning, indicated that cleaning by bleach is partially effective in reducing the bioburden in the patient care area. The halt of transmission of 2DRA in our ICU after introducing PX-UV suggests its adjunctive effect.
The effectiveness of ultraviolet light disinfection is maximized when performed after every patient discharge from the targeted ward. In the only multicenter, randomized controlled study conducted by Anderson et al. [8], ultraviolet disinfection was performed in only isolation rooms occupied by known MDRO or C. difficile colonization or infection. They did not observe any statistically significant effect on the incidence of MRSA and multidrug-resistant Acinetobacter. The effect of the addition of PX-UV to terminal cleaning by bleach was not observed in the study by Anderson et al. [8]. The differences between our study result and the results from Andersen et al. could be due to differences in the healthcare setting (proportion of private rooms), in the way the device is implemented (disinfecting isolation rooms in multiple units vs every discharge/transfer on single unit), or in the intervention fidelity (proportion of eligible rooms that are disinfected). We operated PX-UV after every patient discharge or transfer from the ICU, regardless of their colonization status, and obtained a statistically significant reduction in the incidence of healthcare-associated transmission of MRSA and 2DRA. This difference may be because undetected carriers of MRSA or Acinetobacter could serve as a source of environmental contamination even under the operation of ultraviolet light disinfection in the targeted rooms.
No new isolation of 2DRA from patients in the non-ICU ward has been observed for 10 months (September 2018 to June 2019). We believe that PX-UV successfully decreased or eradicated the environmental Acinetobacter bioburden of not only the targeted ward (ICU), but also of the other wards, perhaps indirectly through the decrease in bioburden throughout the hospital and also of colonized/infected patients with 2DRA. Anderson et al. experienced a similar decrease in the transmission of MRSA and VRE throughout the hospital by using ultraviolet light disinfection in only isolation rooms after patients with targeted pathogen were discharged [8].
Our study has several limitations. First, the effect of PX-UV was evaluated using historical controls when we were not using this technology. We have only one ICU in our hospital, so we were not able to have a non-intervention arm in this study. Second, not all MRSA or 2DRA identified were necessarily acquired in the ICU by horizontal transmission. However, we screened all patients on the day of their admission into the ICU and if they were positive for MRSA and/or 2DRA we regarded it as prior acquisition and excluded them from the acquisition in the ICU. Therefore, we believe that most of the identified MRSA and 2DRA in this study were acquired by horizontal transmission. Third, the microbiological effect of PX-UV was evaluated by comparing the number of colonies from sampling the same frequently touched surface, but with different sites adjacent to each other. If there were significant differences in contamination between sites on the same surface, the result may not accurately reflect the effect of cleaning and PX-UV. We, however, believe that by sampling over 100 surfaces, we can minimize the effect by the heterogeneity of environmental contamination. Fourth, we did not measure the environmental contamination of MRSA or 2DRA by using selective media to sample the environment. For this reason, we are unable to demonstrate the direct relationship between environmental disinfection by PX-UV and the decrease/eradication of new isolation of MRSA or 2DRA. However, it is well known that contamination of the patient care area by the same pathogen that is isolated from the patient is common [14]. We have previously observed contamination by 2DRA on the keyboard in a patient care area where 2DRA was isolated from the patient (data not shown). Furthermore, MRSA, 2DRA and other common environmental flora are susceptible to bleach and PX-UV disinfection. Therefore, it is plausible that environmental contamination by MRSA and 2DRA was controlled by adding PX-UV as an adjunct to manual cleaning, which led to the decrease/eradication of the transmission of these pathogens.