Ozone disinfection for viruses with applications in healthcare environments: a scoping review

: The aim of this scoping review was to provide sufficient information about the effectiveness of ozone gas in virus inactivation of surfaces and objects under different environmental conditions. The review was performed according to the list of PRISMA SrC recommendations and the JBI Manual for Evidence Synthesis for Scoping Reviews. The review was registered in Open Science Framework (OSF). EMBASE (Ovid), Lilacs, LIVIVO, MEDLINE (PubMed), SciELO, Scopus and Web of Science were primary sources, and “gray literature” was searched in OpenGray and OpenThesis. A study was included if it reported primary data on the effect of ozone gas application for vehicle-borne and airborne virus inactivation. No language or publication date restriction was applied. The search was conduct on July 1, 2020. A total of 16,120 studies were screened, and after exclusion of noneligible studies, fifteen studies fulfilled all selection criteria. Application of ozone gas varied in terms of concentration, ozone exposure period and the devices used to generate ozone gas. Twelve studies showed positive results for inactivation of different virus types, including bacteriophages, SARS-CoV-2 surrogates and other vehicle-borne viruses. Most of the studies were classified as unclear regarding sponsorship status. Although most of the population has not yet been vaccinated against COVID-19, disinfection of environments, surfaces, and objects is an essential prevention strategy to control the spread of this disease. The results of this Scoping Review demonstrate that ozone gas is promising for viral disinfection of surfaces.

Introduction SARS-CoV-2 spread from China worldwide in less than 4 months, directly impacting the economy, health and lifestyle of affected countries. Transmission can occur via human-to-human contact, contact with infected surfaces (fomites or skin-to-skin), or airborne transmission, which occurs through the mouth, nose, and eyes or through inhalation of small respiratory droplets suspended in the air. 1,2 There is currently no approved and effective antiviral treatment against SARS-CoV-2, 1,3 and ongoing COVID-19 vaccination programs have not reached most of the population; thus, prevention remains the main strategy for controlling the spread of the disease. However, considering the difficulty in directly conducting research with SARS-CoV-2, which requires biosafety level 3 (BSL3) or higher, many studies using surrogate viruses have been conducted to develop methods to prevent SARS-CoV-2 infection. 4 Hence, efforts have focused on disinfection of environments and objects as well as individual protection and preventative measures to avoid disease spread and potential outbreaks. 1 Aerosol, droplet, and fomite transmission are important routes for the spread of many viral diseases. 5 Among the numerous disinfection and sanitization methods proposed, 6 devices that generate ozone are commonly applied due to the ability of gaseous ozone to easily penetrate into all areas of a room, furniture and other objects. 7 Ozone is a gas that forms chemically via an unstable triatomic molecule of oxygen (O 3 ) that it quickly degrades to its stable state (diatomic oxygen), 8 leading to formation of secondary oxidants (hydroxyl radicals) with high reactivity and a short reaction time. 9,10 In brief, the disinfection mechanism is based on the reaction of ozone with organic compounds containing double bonds. 11 Virus inactivation occurs as a consequence of envelope protein denaturation, impairing virus adhesion to cells, oxidation of unsaturated fatty acids present in the lipid envelope, and destruction of single-stranded RNA. 7 However, the effectiveness of ozone as a virucide is related to various factors, including the ozone concentration, 8,12 exposure time, 8 and temperature 8,12,13 and relative humidity of the environment. 12,13 In the current pandemic of SARS -CoV-2, identifying a disinfection process that interrupts virus transmission routes, including contaminated surfaces, is of utmost importance. Therefore, the present scoping review aims to provide sufficient information about the effectiveness of ozone gas for inactivation of vehicle-borne and airborne viruses under different environmental conditions.

Methodology Protocol registration
The scoping review protocol was performed according to Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) 14 and was registered in Open Science Framework (https:// osf.io/2a3nu/ -DOI 10.17605/OSF.IO/2A3NU). This review is reported following Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines -extension for scoping review 15 and was conducted according to Joanna Briggs Institute Critical Appraisal tools in JBI Manual for Evidence Synthesis for Scoping Reviews 16 . The Arksey and O'Malley methodological framework 17 was employed to conduct this scoping review: a) identifying the research question, b) identifying relevant studies, c) study selection, d) data charting process and e) summarizing and reporting results.

Research question and eligibility criteria
Eligibility criteria were based on the following research question: "Is there evidence of the effectiveness of the application of ozone gas in inactivating vehicle borne and airborne viruses with applications to healthcare environments"? The "PCC" (Population, Concept, and Context) mnemonic was used to guide this scoping review, where P denotes vehicle borne and airborne viruses, C ozone disinfection, and C reducing surface transmission.

Inclusion criteria
A study was eligible for inclusion if it reported primary data on the effect of ozone gas application on inactivation of vehicle-borne (blood, fluids and fomites) and airborne viruses. No language or publication date restriction was applied. In vitro studies, clinical trials, and experimental and observational studies (prospective and retrospective) were included.

Exclusion criteria
Exclusion criteria were as follows: a) studies involving virus inactivation in water for sewage treatment; b) studies involving food and agriculture; c) studies using ozone application as a coadjuvant method for virus inactivation or studies in which ozone was a byproduct from a different decontamination system (not gas); d) case reports and case series; e) systematic reviews; f) conference abstracts, letters, and editorials; and g) personal opinions and books and/or book chapters.

Literature search and study selection
The following databases were searched on July 1, 2020: MEDLINE (PubMed), Scopus, EMBASE, Lilacs, LIVIVO, SciELO, and Web of Science. Part of the "gray literature" was searched in OpenGray and OpenThesis. Descriptors were selected using Medical Subject Headings (MeSH), Descriptors in Health Science (DeCS), and Embase Subject Headings (Emtree). Boolean operators (AND and OR) were used to combine descriptors and improve the search strategy by means of different combinations ( Table 1). The search strategy for MEDLINE was adapted for the other databases, respecting their rules of syntax.

Data charting process and data items
After study selection, the reviewers performed a calibration exercise that consisted of selecting and extracting data from three randomly selected articles. The studies were analyzed independently by two reviewers (MSI and GLS). All of the data were checked by another author. The following information was extracted: meta data (author, date, year of publication), virus type, treatment groups, virus carrier/contact surface, method of assessing virus inactivation, ozonation method, device, ozone concentration (ppm), pH, temperature, relative humidity (RH), exposure time to ozone and key findings.

Sponsorship status evaluation
Information regarding the source of funding of the selected studies was also assessed. These data were extracted because industry sponsorship might be associated with risks of publication, reporting and selection biases. 18 The sponsorship status was classified as follows: 19 a. Unclear -when it was not possible to affirm the sponsorship status even after an attempt to contact the authors by email; b. Nonsponsored -when the authors declared that the study did not receive any type of financial support from companies related to ozone-generating devices; c. Sponsored -when the authors reported any financial contribution (financial support, provision of equipment or supplies, discounts, etc.) from companies related to ozonegenerating devices.
The main text and acknowledgments were checked to collect this information. In cases of missing information or unclear data, the authors were contacted twice by e-mail at an interval of one week.

Synthesis of results
The results of eligible studies were summarized in a descriptive/narrative manner, including study characteristics, and virus inactivation efficiency of interest relative to the healthcare environment was synthesized. Studies reporting ozone concentrations expressed in mg/L were transformed to part per million (ppm) for analysis.

Selection of evidence sources
During the first phase of study selection, 16,120 results among nine electronic databases, including the gray literature, were obtained. After removing repeated/duplicate results, 9937 articles remained for analysis of titles and abstracts. After detailed analysis, only 59 studies were eligible for full-text analysis. One study was included as an expert suggestion. The references of the potentially eligible studies were evaluated carefully, and one additional article was selected, resulting in 61 studies for full-text assessment. After reading the full text, studies that did not fulfill the inclusion criteria were eliminated. Ultimately, 15 studies were included in the scoping review. Figure 1 illustrates the search process, identification, inclusion and exclusion of eligible studies.

Characteristics of evidence sources Overview
The studies included were published between 1990 and 2020, and most were performed in North America (46.6%) ( Figure 2). These studies analyzed the effectiveness of ozone application using different study designs. Most tested the effectiveness of ozone   (Table 2).

Virus type
The effect of ozone application was investigated on several viruses, including hepatitis B, 26 influenza H1N1, 28 H3N2, 7 human respiratory syncytial RSV, 28 lentiviral vector, 33  All of the studies included in this scoping review published in 2020 and 2021 27,28,30,31,33 focused on viral strains similar to SARS-CoV2. The authors described the following viruses as biosafe substitutes for SARS-CoV-2: lentiviral vector, 33 HCoV-229E, 27 influenza A virus (IAV; strain A/WSN/33) and human respiratory syncytial virus (RSV; strain A2). 28 Three of these studies 28,30,31 reported good results of virus inactivation after ozone exposure for at least 40 minutes.

Ozone application
Standardization of the unit of measurement used to assess the ozone concentration in ppm was obtained using the formula 1 ppmv = 0.002 mcg/ml for conversion, as suggested by Bocci (2011). 34 Different types of equipment were used for ozonation, all of which use oxygen from the air as the source; an exception was Tseng and Li, 21,32 in which pure oxygen was the source used to feed the ozone generator OZIPCS-V/SW (Ozotech Inc., Yreka, USA). The devices produce ozone through electrical discharges or through plasma (Table  3). Five studies evaluated the effect of low ozone concentrations (< 1.3 ppm), 21,22,23,31,32 and two of them 31,32 used aerosol generators to simulate virus spread. Considering the results of these studies, 40 minutes of ozone application with high RH (85%) was most effective. Two studies showed that increased exposure time (44 minutes 21 to 180 min 22 ) was required when using low ozone concentrations for virus inactivation. One study 23 demonstrated that a low ozone concentration (< 1.3 ppm) was not effective in reducing virus survival.
Most of the experiments 20,21,24,27,28,30,31,32 utilized high ozone concentrations and were carried out in chambers and cabinets to avoid ozone leakage. One study demonstrated that ozone cannons were not efficient for disinfecting the surfaces of ambulances. 33 The efficacy of ozone gas in an office, a laboratory, a hotel room and a cruise liner cabin was also tested. 7,25 Overall, the variation in ozonation protocols among the studies did not allow direct comparison of results.

Relative humidity (RH)
Ten articles reported RH during the tests. 7,20-24,28,31,32,33 Among them, eight evaluated the influence of RH on virus degradation and/or virus infectivity. 7,21,24,25,28,29,31,32 The results of these experiments controlling RH demonstrated that RH plays a key role in ozone reactivity, and in most studies, the optimum efficacy of ozone treatment was achieved under high RH (> 50%). All of these studies reported a positive effect of elevated RH on the overall efficacy of ozone treatment.

Industry sponsorship status
The sponsorship status is shown in Table 4. None of the studies reported any type of financial support from companies related to ozone-generating devices. In 7 studies, the authors declared no conflicts of interest. Two studies 29,33 provided information by e-mail. Unclear information was observed in 8 of the selected studies, and in these cases, it was not possible to affirm sponsorship status even after attempts to contact the authors by email.  Ineffective: No statistically significant differences in cell viability were noted among the control group and 0.5 or 1 ppm ozone exposed infected cells (Fig. 3B). The 1.5 or 2 ppm-exposed cells had 45-40% viability.

Discussion
Our search identified articles with sufficient evidence for a survey of the effectiveness of ozone gas for inactivating some vehicle-borne viruses. This scoping review provides useful insight for the design, implementation, and effectiveness of ozone gas applications for virus inactivation on surfaces and objects under different conditions, including health services.
The outbreak of SARS-CoV-2 has reinforced the need to develop methods for disinfection, especially those applicable to health care environments. Several methods have been proposed, such as aerosolized hydrogen peroxide, H 2 O 2 vapor, ultraviolet C light, pulsed xenon, and gaseous ozone. 35 Among these, the efficient penetrability of a gas allows the decontamination of inaccessible locations and disinfection of much more than just surfaces, such as crevices, fixtures, and the undersides of furniture.
In addition, air treatment should be considered to reduce the infectivity of airborne diseases. 7 Thus, several devices that generate ozone for this purpose have been developed in recent months.
To narrow the search to microorganisms (or their surrogates) commonly transmitted in health care environments, vehicle-borne and airborne viruses were considered in this scoping review. Thus, studies involving sewage treatment, decontamination in the food and agriculture fields were excluded. Disinfection with ozone gas was effective against many types of viruses with or without an envelope, and decreased virus infectivity was reported in most of the studies (80%). Gaseous ozone was efficient in reducing the infectivity of several common viruses acquired in hospital settings, such as norovirus, 7,20,31 influenza, 7, 28 and RSV. 28 Such effectiveness has also been recently demonstrated for SARS-COV2 biosafe surrogate viruses. As handling of SARS-CoV-2 requires biosafety level (BSL) 3 facilities, 36 surrounding viruses are commonly used to facilitate research. Twist synthetic SARS-CoV-2 RNA, 30 HCoV-229E, 27 influenza A virus (IAV; strain A/WSN/33) and human respiratory syncytial virus (RSV; strain A2) 28 are considered biosafe substitutes for SARS-COV-2 due to their similarity in form, structure and function. Although the analyzed studies employed surrogates for SARS-CoV-2, some ,21,24,32 published before the pandemic showed the efficiency of ozone gas application against viruses that are actually considered surrogates for SARS-CoV-2, such as MHV (mouse hepatitis virus) Plaque assays Effective: When face masks experimentally contaminated with a human coronavirus (HCoV-229E) as a surrogate were exposed to ozone gas (approximately 120 ppm) produced by the plasma generator for either 1 or 5 min, the virus lost infectivity.  and Phi6. 4 Dubuis et al. 31 demonstrated the effect of ozone as a disinfectant against multiple phages with different features to represent a broad range of eukaryotic viruses and their resistance when airborne and when exposed to disinfecting agents. Nevertheless, three studies 23,26,33 showed evidence that gaseous ozone was ineffective against specific types of viruses. For example, Cía et al. 33 found that gaseous ozone treatments applied for emergency vehicles do not significantly affect virus viability; in this study, the effect of an ozone cannon device (10 ppm) against HIV-1-derived lentivectors inside an emergency vehicle was investigated. Guo et al. 26 demonstrated that the application of ozone (140 ppm) to disinfect HBV-contaminated hospital linen was also ineffective, even when the disinfection time was prolonged to 80 min. Lin et al. 23 observed approximately 85% viability for cells exposed to 1.5 and 2 ppm ozone, and increasing the ozone concentration to 2.5 and 3 ppm achieved 40 and 50% cell mortality with 1 and 2 h exposure, respectively. Interestingly, the study involving the emergency vehicle 33 used low RH (37%-48%), but the other studies 23,26 did not mention the RH of the experiment. The literature is clear regarding the role RH plays in the disinfection process using ozone. 37 Blanchard et al. 28 suggested that humidity may act through two mechanisms: (a) water may promote the generation of highly reactive hydroxyl radicals from ozone, and (b) greater humidity may facilitate ozone interacting with surface-bound species, such as by diffusion, to reach viruses. Therefore, the evidence of this scoping review found optimum efficacy of ozone treatment under high RH (> 50%).

RT-qPCR
It is not appropriate to determine the effectiveness of a specific disinfection method by relying only on the resistance of the virus being analyzed. Indeed, each virus has different tolerances for various protocols. 31 The ozone generation mechanism may also affect results. Ozone is obtained by transforming oxygen via electrical discharges, and there are three different systems exist for generating ozone gas: ultraviolet, plasma or corona systems. The ozone gas produced by ultraviolet light is obtained at low concentrations and acts as a secondary product complementary to the UV effect. The cold plasma system is very commonly used to purify water and air in public environments, and the corona discharge system, which is mainly used in health care settings, produces ozone more effectively and at a greater concentration. The mechanism of artificial ozone production occurs through electrochemical discharge known as the corona effect. Another advantage of this method is that it can be applied in water. 38 The type of material or virus carrier applied varied among the studies, which did not allow for direct comparisons. The experimental setup and methodology used most commonly were virus stock spread on glass dishes, 7,20,22,23,24,33 stainless steel disks, 29 or gelatin medium 21 and allowed to dry before ozone exposure. In general, disinfection of hospital equipment was the main focus after the SARS-COV2 outbreak. Disinfection of face masks, 27,28 Tyvek (spun high-density polyethylene) fabric used in disposable gowns and PAPR (powered air-purifying respirator) hoods 28 , N95 respirators, 28 and samples of fabrics and carpet 25 has been demonstrated. No substantial loss of filtration performance was found for ozone-treated masks 27,28 . Thus, in times of shortages, ozone treatment has proven to be an effective method for disinfecting personal protective equipment (PPE) for safe reuse. Viral quantification can be performed by the standard plaque assay method 39 or viral titration for quantitative assay of the infectivity of virus recovered in monolayer cells of selected bacterial strains as hosts 23,40,41,42 . RT-PCR is another quantification method that measures a defined sequence of the viral genome; as one would expect this method be more resistant to the damaging effects of ozone gas than infectivity, it probably underestimates the effectiveness of antiviral agents. Thus, it would be worthwhile to use two methods to analyze infectivity and viral degradation. In fact, the importance of these two methods of quantification was clear in the study of Lee et al. 27 In this experiment, no viable HCoV-22E was recovered from face masks, and the virus was shown to lose infectivity in a human cell line (MRC-5) when exposed for a short period of time (1 min) to ozone gas produced by a DBD plasma generator. However, the short exposure may not fully degrade the viral RNA because no significant difference in the amount of amplifiable RNAs between treated and untreated masks was observed. These results suggest that loss of infectivity could be due to damage to the viral envelope or envelope proteins, resulting in failure of the virus to attach to host cells.
The concentration of ozone required for efficacy also depends on the exposure time. The ozone concentration in the majority of the studies included in this review was several magnitudes higher than the exposure limits defined for Occupational Safety and Health regulations. For this reason, most of them were performed in a proper chamber or cabinet. The major disadvantage of the use of ozone is its potential toxicity at high concentrations; thus, precautions must be taken to avoid such exposure. In general, limits for workers are < 0.1 ppm or 0.2 mg/m 3 on average over 8 h, < 0.3 ppm if the exposure time is only 15 min; PPE is needed if the concentration is > 0.3 ppm. Challenges can be overcome by ensuring that the area is emptied of staff, closed off, and sealed. 37 Ozone gas decays quickly to oxygen, with a half-life of approximately 20 min. Another plausible option is the use of a catalytic converter (scrubber) near closed doors and inside them to accelerate ozone degradation and avoid leakage. Ozone concentrations below 0.1 ppm may be feasible for treating air inside unoccupied hospital rooms. 31 More studies using lower concentrations but longer exposure periods to treat air in an unsealed unoccupied room should be performed to avoid worker and patient toxicity.
It should also be noted that most studies used ozone generators with industrial and domestic applications that obtain ozone from ambient air. Such equipment produces ozone but can also release other compounds derived from gases present in the atmosphere. Similar to ozone, the concentration of these contaminants in air varies according to temperature and air humidity. 43,44 It is important to note that there is a possibility that such devices produce ozone concentrations higher or lower than those indicated by the manufacturer, which can also affect results. Thus, such rates and concentrations should be measured. 45 Devices that use sterile oxygen for ozone generation, on the other hand, will not only provide byproducts free of such gases and compounds but will also have effective flow regulation. 43 Future studies should be carried out to detect compounds generated by domestic and industrial generators, as well as the exact concentrations of the generated ozone in both systems. Comparative studies between generators that use ambient air and medicinal oxygen should also be performed.
Another limitation was possible sponsorship bias present among the selected studies. Most of the studies 7,20-25,32 were classified as unclear with regard to sponsorship status. Authors did not state whether they received any type of funding or provision of devices from companies related to ozone generators, and they did not mention no conflict of interest. Sponsorship from companies, mainly in studies evaluating specific devices or products, might lead to bias. Risks of publication, reporting and selection biases as well as publication of selected results have been associated with industry sponsorship. 18,46 Therefore, the results of this study should be interpreted with caution, and future studies should be performed to assess the presence of sponsorship bias among studies evaluating the effectiveness of ozone generators as disinfecting agents.

Conclusion
Our findings suggest that ozone should be considered as an effective method to decrease the infectivity of several viruses commonly acquired inside hospitals and other healthcare environments. This scoping review provides direction for a future systematic review to investigate not only the effectiveness but also a better protocol for the use of ozone as a disinfectant or sterilization method for environments and surfaces, thus contributing to disease prevention in both emergency situations and pandemics, as well as in daily routines.