key: cord-0847654-2ccxiqea
authors: Silva, António; Azevedo, Maria; Sampaio-Maia, Benedita; Sousa-Pinto, Bernardo
title: The effect of mouthwashes on SARS-CoV-2 viral load: a systematic review
date: 2021-12-30
journal: J Am Dent Assoc
DOI: 10.1016/j.adaj.2021.12.007
sha: 979e5cef3d0519ebb02d583d413a8bb26ba8a020
doc_id: 847654
cord_uid: 2ccxiqea

Background Considering the oral cavity a major entryway and reservoir for SARS-CoV-2, the aim of this study was to perform a systematic review of in vivo and in vitro studies to assess the effectiveness of mouthwashes on SARS-CoV-2 viral load. Types of study We searched PubMed, Web of Science, Scopus, MedRxiv, and bioRxiv databases, including in vitro and in vivo studies assessing the virucidal effect of mouthwashes on SARS-CoV-2 or surrogates. From a total of 1622 articles retrieved, 39 were included in this systematic review. Results Povidone-iodine (PVP-I) was the most studied mouthwash (14 in vitro and 9 in vivo studies), frequently showing significant reductions in viral load in vitro assays. Similarly, cetylpyridinium chloride (CPC) also showed good results, although evaluated in fewer studies. Chlorhexidine gluconate (CHX) and hydrogen peroxide (H2O2) showed conflicting results on SARS-CoV-2 load reduction in both in vitro and in vivo studies. Practical implications PVP-I-based mouthwashes appear to be the best option as an oral pre-rinse in dental context for SARS-CoV-2 viral load reduction. Although the results of primary studies are relevant, there is a need for more in vivo studies on mouthwashes, in particular randomized controlled clinical trials, to better understand their effect on SARS-CoV-2 viral load and infection prevention.

INTRODUCTION 1 intervention and control group, virus strain, type of mouthwash, concentration, number 23 of mouthwashes per day, rinsing duration, treatment duration, and decrease in viral 24 load. For in vitro studies, the cell lineage used, and existence of interfering substances 25 were also assessed. 26

Assessment of the risk of bias (RoB) of included randomized controlled trials (RCT) 28 was carried out independently by two reviewers according to Cochrane Collaboration 29 tool for assessing RoB 17 . Disagreements between reviewers were resolved after 30 J o u r n a l P r e -p r o o f discussion and analysis. No RoB assessment was performed on in vitro studies or 1 observational before-after studies due to a lack of consensually accepted tools for 2 assessing RoB in those specific studies. 3

We considered all outcome measures directly evaluating SARS-CoV-2 viral load. Main 5 outcome measures presented in this systematic review are viral load expressed in 6 logarithmic (log) reduction value, copies per milliliter (copies/mL), and Relative Light 7 Units (RLU). When primary studies used a mouthwash with known concentration and 8

presented the viral load decrease in logarithmic scale, such results were interpreted 9 following the European Norm EN-14476, which recognizes antiseptics virucidal 10 capacity when achieving a reduction on viral load equal or greater than 4 log10 18 . 11 Therefore, results of the primary in vitro studies when expressed in log scale were 12 classified according to three levels considering virucidal activity (viral load reduction): 13 high efficacy (≥4 log10; +); moderate efficacy (≥3 log10 and <4 log10; ±); and low efficacy 14 (<3 log10; -). To simplify the comparison between studies results expressed in Molar 15 were converted to a percentage (%, in g/100mL). Results presented as a percentage 16 of inactivation or fold reduction were converted to a logarithmic scale. 17

Due to methodological diversity of included primary studies, it was not possible to carry 19 out a meta-analysis. 20

A total of 1560 articles were retrieved from bibliographic databases (MEDLINE, 24 Scopus, and Web of Science), and 62 from preprint databases. The study selection 25 process is described in Figure 1 . From a total of thirty-nine included studies, thirty-three had been published as peer-1 reviewed articles and six were preprints (Appendix Table 1 ). Twenty-four of the 2 published articles were performed in vitro and ten were in vivo, five of which were RCT 3 while the remaining were uncontrolled before-and-after studies. Five of the included 4 preprints were performed in vitro and one was in vivo. 5

In vivo studies included COVID-19 positive hospitalized patients [19] [20] [21] [22] [23] [24] [25] [26] [27] , and home-6 isolated patients 22, 28 . All in vivo studies quantified SARS-CoV-2 viral load via 7

Polymerase Chain Reaction (PCR), targeting genes E [19] [20] [21] [22] 24 , RNA-dependent RNA 8 polymerase (RdRP) 20,22,24 , nucleo-capsid (N) 22-24,26,27 , S and R 23 . Three in vivo 9 studies used water as a control 21, 24, 27 , one used RNA from trizol-inactivated virus 26 . 10

One used a similar solution regarding aspect and content but without virucidal 11 components 28 . In vivo studies evaluated the reduction of SARS-CoV-2 in viral titers: 12 four presented the results with cycle threshold (Ct) fold changes 21, 23, 24, 27 , three in the 13 form of a logarithmic reduction value 20, 22, 25 , one in the form of a logarithmic reduction 14 percentage scale 28 , one in a percentage scale 26 , and one in copies per milliliter 19 . 15

Regarding SARS-CoV-2 strains used across in vitro studies, several used well-16 characterized strains, being the most used USA-WA1/2020 [29] [30] [31] [32] [33] [34] [35] [36] [37] . Four studies used a 17 SARS-CoV-2 strain directly obtained from an infected patient 38-41 , while one study did 18 not report the strain employed 42 . In vitro studies were performed under dirty 43-47 , 19 clean 29,31-35,37-39,41,42,48-54 , or both conditions 36,40,55,56 , being these terms referring to the 20 existence of interfering substances. Two in vitro studies did not provide information 21 about the existence of interfering substances 30,57 . 22 In vivo and in vitro studies applied the intervention solution for a pre-determined period 23 mouthwash contact time, most commonly ranging from 15 to 120 seconds. Seven in 24 vitro studies included periods of application of 5 minutes or more 30, 34, 41, 42, 51, 53, 57 . 25

Two RCT were marked as a high RoB study 21, 27 , while the other three were marked 27 as low RoB studies 24,26,28 (Appendix Table 2 ). The other five in vivo studies were 28 "uncontrolled before-after" studies including a low number of participants and for which 29

Five in vivo studies showed the virucidal efficacy of PVP-I solutions on SARS-CoV-2 1 (Appendix Table 3 ). Seneviratne, et al. 21 conducted a RCT and reported a 30-second 2 rinse with 0.5% PVP-I conducted on a group of four hospitalized patients resulted in a 3 significant reduction of viral load 6 hours post-rinse when compared to water. 4

However, no significant differences were found 5 minutes and 3 hours after rinsing. 5 After using the same concentration of PVP-I, but by performing two consecutive 30-6 second rinses, Chaudhary, et al. 26 verified a 61% reduction on viral load after 15 7 minutes and a 97% reduction after 30 minutes. The RCT conducted by Elzein, et al. 24 8 found a significant mean Ct difference increase between the paired samples before 9 and after a 30-second 1% PVP-I rinse. In an uncontrolled before-after clinical study, 10

Lamas, et al. 22 reported a 60-second 1% PVP-I rinse led to a significant drop (≈5 log10) 11

in viral load in one of the four patients evaluated, sustained for at least three hours. 12 Jayaraman, et al. 25 found 1% PVP-I could reduce viral load in saliva up to 1.8±1.1 13 log10. Significant reductions were observed after 20 and 60 minutes. 14 In vitro studies demonstrated PVP-I-containing mouthwashes have a virucidal effect 15 on SARS-CoV-2 (Appendix Table 4 ). Table 2 summarizes the results found in 16 different studies with application times up to 60 seconds and interpreted following the 17 EN-14476. Concentrations up to 0.75% showed moderate-to-high efficacy in reducing 18 SARS-CoV-2 viral load 29, [31] [32] [33] 44, 49, 52, 53, 55 . The 60-second application of PVP-I with 19

concentrations between 0.5% and 0.58% presented high efficacy results in the 4 20 studies evaluating this condition 31, 49, 53, 55 . Concentrations of PVP-I between 1.25% and 21 2.5% consistently showed moderate-to-high efficacy results 29,31-33 . Applying 22 concentrations of PVP-I greater than 2.5% showed low 46 (PVP-I 7.5%), moderate 43,53 23 (PVP-I 5% and 7.5%), and high efficacy 44,53 (PVP-I at 7.5% and 10%) within 15 to 24 30 seconds. The 60-second application also reached moderate-to-high efficacy 25 results (PVP-I concentrations ranging from 5% to 10% ) 43, 53 . 26 Regarding H2O2, Gottsauner, et al. 19 conducted an in vivo study assessing virucidal 27 efficacy of a 30-second H2O2 (1%) rinse with. No significant difference was found 28 between baseline and the viral load 30 minutes after rinsing. Chaudhary, et al. 26 found 29 that two consecutive 30-second H2O2 (1%), led to a 90% reduction after 15 and 30 30 minutes. Jayaraman, et al. 25 reported a 30-second H2O2 (1.5%) rinse could decrease 31 the viral load up to 1.6±1.5 log10 after 60 minutes. A 60-second H2O2 (1.5%) rinse led 32 to a significant reduction on viral load immediately after and 30 minutes after rinsing, 33 but not after 60 minutes 27 . In vitro studies on the virucidal effect of H2O2 showed very 1 limited success (Table 3 and Appendix Table 4 ). 2

Chlorhexidine gluconate mouthwashes virucidal efficacy was evaluated with in vivo 3 and in vitro studies (Appendix Tables 3 and 4 ). In an RCT, Seneviratne, et al. 21 Table 3 ). CPC 0.075% 26 mouthwash significantly reduced viral load within 5 minutes of use. Compared to the 27 control group, the viral load reduction with CPC was maintained for 3 and 6 hours. In 28 vitro studies demonstrated CPC-containing mouthwashes have a virucidal effect on 29 SARS-CoV-2 (Appendix Table 4 ). Considering application times between 30 and 30 60 seconds (Table 3) , concentrations of up to 0.3% showed low-to-high 31 efficacy 43, 46, 48, 50, 54, 56 . The 20 second application of CPC had moderate-to-high 32 efficacy 54 . Meyers, et al. 43 reported a 120-second application of 0.07% CPC showed 33 J o u r n a l P r e -p r o o f moderate-to-high efficacy. et al. 38 reported moderate results with a 1 120-second application of CPC at a concentration of up to 10mM (0.3%). 2

Other mouthwashes, either more complex or with less frequently used active 3 compounds, were studied in vivo and in vitro by several authors (Appendix Tables 3  4   and 4) . Carrouel, et al. 28 studied the effect of a 60-second CDCM rinse, a Citrox, and 5 ß-cyclodextrin containing mouthwash. This study reported a significant decrease in 6 viral load of approximately 13% when using the mouthwash, compared to a 7% 7 decrease observed in the placebo group. Eduardo, et al. 27 conducted a RCT studying 8 the effect of performing a 60-second H2O2 (1.5%) (Peroxyl ® ), combined with a 30-9 second CHX (0.12%) (PerioGard ® ) rinse. This combined rinse only achieved minor in 10 Ct values when compared to the placebo group. However, when rinsing with a 11 mouthwash containing CPC (0.075%) and Zinc Lactate (0.28%) a significant decrease 12 in salivary viral load was achieved for up to 60 minutes. On an uncontrolled before-13 after study, Schürmann, et al. 23 studied the effect of a 60-second Linola ® sept rinse 14 and reported a mean value increase of Ct values of 3.1 (basal versus after-rinsing). 15

In vitro studies included a diversity of complex mouthwashes. Listerine ® mouthwashes 16 were studied by several authors, although each formulation was only assessed in one 17 study, apart from Listerine ® Cool Mint ® that was assessed by two studies. Listerine ® 18 mouthwashes showed variable efficacy 43, 45, 46, 49 (Table 4) . 19 20

Summary of evidence 23

In this systematic review, we included primary studies assessing the virucidal effect of 24 mouthwashes regarding SARS-CoV-2, that presented a diverse set of methodologies 25 and assess a wide range of mouthwashes. PVP-I was most frequently studied To the best of our knowledge, this is the first systematic review analyzing information 1 both from in vivo and in vitro studies. A previous systematic review had assessed in 2 vitro studies, with results consistent to those displayed in this study 15 . 3 Considering mouthwashes as antiseptics, they should follow regulating norms. The 4

International Organization for Standardization (ISO) defines on ISO-16408:2015 the 5 chemical and physical properties of oral rinses, as well as of their test methods, but 6 guidelines for microbiological analysis are specific to mold, bacteria, and yeast, lacking 7 virus instructions 58 . There seems to be a lack of standardization on the evaluation of 8 mouthwashes regarding virucidal properties. According to the European Standard EN-9 14476, an antiseptic is effective when it reduces viral load ≥4 log10 18 . Although EN-10 14476 is not specific towards oral rinses, due to the lack of more appropriate 11 regulation, we decided to compare our results in light of this European Norm for 12 assessing mouthwash virucidal properties. 13

Included primary studies displayed substantial diversity in their methodologies and 14 results presentation, limiting our capacity of comparing different mouthwashes. PVP-15 I-based mouthwashes appear to have potential for reducing SARS-CoV-2 in the oral 16 cavity. Nonetheless, these results must be cautiously interpreted. The use of CPC mouthwashes for reducing the viral load also showed encouraging 26 results. Of note, CPC is also capable of inactivating influenza viruses both in vitro and 27 in vivo, but only after 10 minutes of contact time 59 . 28

In the included primary studies, H2O2 and CHX-based mouthwashes produced a 29 varied effect on SARS-CoV-2 viral load. As their effect was inconclusive, 30 recommending their use may not be adequate. CHX and H2O2 are already currently 31 used in some oral care products, with CHX displaying broad-spectrum antimicrobial 32 activity 60 , including against anaerobic oral bacteria 61 . Worldwide government agencies 1 and professional associations currently advise the use of pre-procedural rinse with 2 H2O2 mouthwashes to reduce oral SARS-CoV-2 viral load mouthwashes 10-14 , so there 3 may be a need to reconsider these directives. 4

Some complex mouthwashes like Listerine ® Total Care, Listerine ® Advanced, and 5

Listerine ® Antiseptic showed promising results in reducing SARS-CoV-2 viral load in 6 the oral cavity, although they were evaluated by only one or two studies each. Using 7 these mouthwashes as a coadjutant in oral health is well established, contributing to 8 the reduction of dental biofilm and gingivitis 62 . 9

The included primary studies have the limitation of only evaluating the presence of 10 viral particles and not their viability or infectious capacity, therefore using other 11 techniques as viability-PCR could be employed to study the infectious potential of the were also able to quantify the generated viral particles detected from a distance higher 24 or equal to two meters. These results highlight the importance of preventive measures 25 such as pre-rinse antiseptic mouthwash but also a rubber dam isolation given that both 26 strategies can significantly reduce aerosol pathogen load 65,66 . 27 In addition to the wide diversity of study methodologies, and of results presentations, 28 a major limitation of this systematic review is the scarcity of RCTs, with only five 29 meeting eligibility criteria 21, 24, [26] [27] [28] . The validity of the conclusions is affected by the bias 30 of the included primary studies, in this case, regarding the high RoB of two of the RCT. 31 Besides, the other five in vivo studies have important limitations in their designs, 32 including the absence of randomization or even a control group, and a relatively low 1 number of included patients; this prompts a low level of evidence and hampers the 2 precision of their estimates, respectively. Although in vitro studies are part of the tests 3 proposed by EN-14476 18 , their results cannot be directly transposed to in vivo 4 application of these mouthwashes. In vivo studies should be RCT conducted with a 5 better study design, including a higher number of patients, include a control solution, 6 and express their results as virus log reduction allowing a better interpretation of 7 results with a greater level of evidence. 8 A recurrent inadequacy found in selected studies was the existence of studies that 9 include times of application not feasible in clinical practice. Some in vitro studies had 10 application times of 30 minutes 30 , and one preprint article also considered an 11 application with a duration of 72 hours 51 . We find these application times unrealistic 12 and not adequate for clinical practice since patients are normally only able to gargle 13 for a short period 67 , usually up to 60 seconds. 14

There is a need for more in vivo and in vitro studies on different mouthwashes that 16 consider adequate and realistic application times, of up to 60 seconds. Well-designed 17 RCT with a larger number of patients should be considered a priority when it comes 18 to design of in vivo studies. Based on results from already published primary studies, 19 future studies should mainly focus on PVP-I and CPC-based mouthwashes. 20

Furthermore, the studies should present their results in form of a logarithmic reduction 21 that can be compared according to EN-14476. Studying mouthwash-induced 22 cytotoxicity should be a concern when assessing virucidal properties of different 23 mouthwashes with different concentrations. Studying viral viability post-rinse and viral 24 presence in aerosols should be considered to better assess the real impact of virus 25 dissemination in the dental setting. Overall, guidelines for the standardized evaluation 26 of the effect of mouthwashes on viruses are needed. 27

In conclusion, considering the current knowledge, using PVP-I-based solutions as a 29 preprocedural rinse in dental setting appears to be potentially effective in reducing 30 SARS-CoV-2 oral load. There are no powerful arguments to consider using of H2O2 31 and CHX effective regarding SARS-CoV-2 virus and their use as a pre-procedural 32 mouthwash aiming to reduce SARS-CoV-2 oral load should be revised. accordingly to EN-14476, considering a reduction on viral load greater or equal than 11 4 log10 as a high efficacy ( ), a reduction greater than 3 log10 and lower than 4 log10 12 as a moderate efficacy ( ), and a reduction lower than 3 log10 as a low efficacy ( ). 13 14 Table 2 . H2O2, CHX, and CPC mouthwashes in vitro effect on SARS-CoV-2 oral viral 15 load. Results interpretation accordingly to EN-14476, considering a reduction on viral 16 load greater or equal than 4 log10 as a high efficacy ( ), a reduction greater than 3 17 log10 and lower than 4 log10 as a moderate efficacy ( ), and a reduction lower than 18 3 log10 as a low efficacy ( ). 19 20 Table 3 . Other mouthwashes in vitro effect on SARS-CoV-2 oral viral load. Results 21 interpretation accordingly to EN-14476, considering a reduction on viral load greater 22 or equal than 4 log10 as a high efficacy ( ), a reduction greater than 3 log10 and lower 23 than 4 log10 as a moderate efficacy ( ), and a reduction lower than 3 log10 as a low 24 efficacy ( ). Table 1 . Studies characterization.

Uncontrolled before-andafter studies

Anderson, et al. 44 x Bidra, et al. 29 x Bidra, et al. 33 x Carrouel, et al. 28 x Frank, et al. 32 x Gottsauner, et al. 19 x Hassandarvish, et al. 55 x Jain, et al. 39 x Koch-Heier, et al. 50 x Lamas, et al. 22 x Meister, et al. 45 x Meyers, et al. 43 x Pelletier, et al. 31 x Schürmann, et al. 23 x Seneviratne, et al. 21 x Xu, et al. 30 x Yoon, et al. 20 x Almanza-Reyes, et al. 51 x Davies, et al. 49 x Elzein, et al. 24 x Muñoz-Basagoiti, et al. 56 x Steinhauer, et al. 42 x Zoltán 35

x Santos, et al. 41 x Shewale, et al. 37 x Shet, et al. 53 x Kariwa, et al. 52 x Tiong, et al. 40 x Meister, et al. 47 x Komine, et al. 54 x Santos, et al. 57 x Chaudhary, et al. 26 x Eduardo, et al. 27 x Pre-print Green, et al. 48 x Jayaraman, et al. 25 x Mantlo, et al. 34 x Muñoz-Basagoiti, et al. 38 x Statkute, et al. 46 x Anderson, et al. 36 x Water Ct values detected in all 16 patients were within the range of 15.6-34.5, with a mean value of 27.7±4.8; Results are presented in form of fold change calculated as a ratio between Ct value at different timepoints and Ct value at baseline. PVP-I: significant increase in fold change was obtained only at 6h (ratio=1) post-rinsing with PVP-I in comparison with water (p<0.01). In comparison to the water group, the PVP-I group patients had higher fold increases in Ct value after 5min (ratio=1.1) and 3h (ratio=1.2) of postrinsing, but no significance was achieved. CHX: patients demonstrated a varied effect among saliva Ct values after 5min rinsing and hence further studies with a larger sample size are required to determine its significance. CPC: significant increase in fold change of Ct value at 5min (ratio=1) and 6h (ratio=0.9) was observed post-rinsing with CPC mouth-rinse compared to the water group patients (p<0.05). Although the fold changes in Ct values were higher at 3h (ratio=0.9) in the CPC group, no significance was achieved (p=0.20). The viral load in the saliva decreased transiently for 2h after using the CHX mouthwash, but it increased again at 2-4h post-mouthwash. On day 3, viral load was not detected at 1h and 2h post rinse, on both patients. One of the patients showed a baseline viral load of 6.9 log10 and the other of 4.9 log10. On day 6, one hour after using the mouthwash, there was no reduction in viral load in one patient.

Uncontrolled before-after study Mean values showed an increase of the Ct-values of 3.1±3.6, which translated into a significant reduction of the viral load in the pharynx of about 90%. Most patients exhibited a ten-fold reduction of viral load, independently of the initial viral load. The viral load required approximately six hours to recover to the initial viral load. Moreover, highly infectious patients were able to restore their initial viral load during this time, while less infectious patients were not able to restore their initial infectivity 6h after gargling.

Uncontrolled before-after study Hospitalized patients diagnosed with COVID-19. Single rinse performed in a single day.

Saliva (Passive drool) and Exhaled respiratory droplets, via RT-PCR PVP-I (1%); H2O2 (1.5%); CHX (0.2%). Duration of the rinse not available -The reduction was significantly higher in respiratory droplets (92%) than in whole saliva samples (50%; p=0.008). PVP-I: -Saliva 20min: 1.8±1.1 log10 reduction 60min: 1.3±0.9 log10 reduction -Respiratory droplets 20min: 2.5±0.4 log10 reduction 60min: 1.6±1.9 log10 reduction H2O2: -Saliva 20min: 1.2±0.3 log10 reduction 60min: 1.6±1.6 log10 reduction 90min: 1.5±1.5 log10 reduction 180min: 0.9±0.8 log10 reduction -Respiratory droplets 20min: 3.5±3.7 log10 reduction 60min: 2.5±2.8 log10 reduction 90min: 1.9±1.6 log10 reduction 180min: 3.0±0.03 log10 reduction CHX -Saliva 90min: 1.6±1.2 log10 reduction 180min: 0.4±1.5 log10 reduction -Respiratory droplets 90min: 1. Sensitive fluoride toothpaste achieved a 2.26 log10 reduction with application times of 30s 60s, and 120s.

SARS-CoV-2 hCoV-19/Germany/BY-Bochum-1/2020; Vero E6 Oral sprays: A) Carragelose® (1.2 mg/mL), Kappa-Carrageenan (0.4 mg/mL)

B) Sodium chlorite (0.9%), Panthenol

C) Xylometazolin hydrochloride (1 mg/mL), Dexpanthenol (50 mg/mL)

Lithiummagnesium-sodium-silicate

F)Hydroxypropyl methyl cellulose, Succinic acid, Disodium succinate

I) Anise oil, Eucalyptus oil, Levomenthol, Myrrh extract, Clove oil, Peppermint oil Ratanhia root extract

In general, oral sprays led to a >1 log10 reduction: A) 0.53 log10 reduction; B) 0.13 log10 reduction; C) 0.09 log10 reduction; E) 0.20 log10 reduction; F) 0.18 log10 reduction. Oral spray G) led to no reduction, while oral spray D) led to a 2.21 log10 reduction.Nasal spray H) led to no reduction on viral load. Nasal spray I) led to a ≥3.03 log10 or ≥ 4.69 log10 (large volume plating: to reduce cell toxicity) Colgate Plax ® Fruity Fresh: 5 log10 reduction for all test times and conditions; Thymol ® mouthwash by Xepa: 0.75 log10 reduction after 60s (clean conditions), 0.5 log10 reduction after 30s (clean conditions), and after 30s and 60s (Dirty conditions); Bactidol ® : 5 log10 reduction for all test times and conditions; Salt water: no effect on SARS-CoV-2 viral load. 

(mouthwash* OR "mouth rinse" OR "oral rinse" OR rinse OR gargl* OR "gargle lavage" OR "oral irrigation" OR "oral lavage") AND (COVID-19 OR COVID19 OR sars-cov-2 OR 2019-nCoV OR COVID OR coronavirus)

( mouthwash* OR "mouth rinse" OR "oral rinse" OR rinse OR gargl* OR "gargle lavage" OR "oral irrigation" OR "oral lavage" ) AND ( covid-19 OR covid19 OR sars-cov-2 OR 2019-ncov OR covid OR coronavirus )

TS=((mouthwash* OR "mouth rinse" OR "oral rinse" OR rinse OR gargl* OR "gargle lavage" OR "oral irrigation" OR "oral lavage") AND (COVID-19 OR COVID19 OR sars-cov-2 OR 2019-nCoV OR COVID OR coronavirus))

COVID-19 AND mouthwash J o u r n a l P r e -p r o o f