Spread of viruses, which measures are the most apt to control COVID-19?

Abstract

The persistent debate about the modes of transmission of SARS-CoV2 and preventive measures has illustrated the limits of our knowledge regarding the measures to be implemented in the face of viral risk. Past and present (pandemic-related) scientific data underline the complexity of the phenomenon and its variability over time. Several factors contribute to the risk of transmission, starting with incidence in the general population (i.e., colonization pressure) and herd immunity. Other major factors include intensity of symptoms, interactions with the reservoir (proximity and duration of contact), the specific characteristics of the virus(es) involved, and a number of unpredictable elements (humidity, temperature, ventilation…).

In this review, we will emphasize the difficulty of “standardizing” the situations that might explain the discrepancies found in the literature. We will show that the airborne route remains the main mode of transmission. Regarding preventive measures of prevention, while vaccination remains the cornerstone of the fight against viral outbreaks, we will remind the reader that wearing a mask is the main barrier measure and that the choice of type of mask depends on the risk situations.

Finally, we believe that the recent pandemic should induce us in the future to modify our recommendations by adapting our measures in hospitals, not to the pathogen concerned, which is currently the case, but rather to the type of at-risk situation.

1. Introduction:

The COVID-19 pandemic raised many questions about preventive measures in community and in-hospital settings. These questions persist more than 2 years after the onset of the pandemic. The contradictory data in different published studies have added to the confusion and caused concern. Awareness the modes of transmission of a pathogen is an important, if not essential step in controlling the risk of its spread. Like all respiratory viruses, SARS-CoV-2 is transmitted by the respiratory and airborne routes. However, it is important to specify not only the role of each transmission route, but also the so-called droplet or airborne character of the respiratory route. In this work, we will take a critical look at the data in the literature, and try to provide “food for thought” and aid to decision-making.

2. What we previously knew about respiratory viruses

Prior to the pandemic [1], [2], existing data could have allowed us to approach the initial phases more serenely in terms of preventive measures designed to decrease risk of transmission. The fear aroused by the lack of knowledge of SARS-CoV2, justified on account of clinical consequences, appears less appropriate as regards modes of transmission. Previous data concerning the modes of transmission of respiratory viruses made it possible to anticipate risks and to propose preventive measures [3], [4]. One of the major difficulties in understanding transmission phenomena lies in the need to take into account a multitude of confounding factors. At the individual level, the probability of being infected involves both the probability of being exposed to the pathogen and frequency of exposure. For example, a route with a low probability of acquisition by exposure but occurring frequently may contribute more to the total number of acquisitions than the ones with rare exposure but with a high probability of transmission by exposure.

At the collective level, several factors contribute to the risk of transmission: the collective reservoir or colonization pressure [i.e., disease prevalence] [5], the individual reservoir e [i.e., symptom intensity, proximity and duration of contact], and the immunity of exposed hosts [6]. In addition, the frequency of infected and asymptomatic individuals is variable for different viruses and makes it more difficult to understand the modes of transmission. In fact, the greater the number of asymptomatic individuals, the greater the need for systematic screening of contacts in order to define secondary acquisitions and transmission routes. These are some reasons why it seems even more difficult to define the modes of transmission during pandemic periods.

3. Modes of transmission of respiratory viruses:

As for other pathogens, different modes of transmission of respiratory viruses exist [7], [8]. For a very long time, we artificially divided transmission into 3 different modes, forgetting that for the same pathogen, there were multiple modes of transmission that varied over time. While one mode may dominate, it is now clear that different modes overlap.

Transmission can be indirect, through the contaminated hands of an individual, whether in a hospital environment [health care workers or visitor] or in a community environment [colleague or relatives]. This contamination can be the result of contact with a human reservoir [infected person] or with a contaminated environment [surfaces]. In the latter case, it will depend on the frequency of environmental contamination linked to a given reservoir, on the virus' ability to survive in this environment, and on the hand disinfection products used.

The second mode of transmission is direct, airborne, from the reservoir [9]. It requires direct contact with the reservoir and duration of exposure. This mode of transmission is divided into two distinct modes: droplet transmission and airborne transmission. This definition is meant to differentiate between short-range [droplet transmission] and long-range [airborne transmission] particles. It is generally based on the size of the exhaled droplets, the former being particles of more than 5 µm that deposit very quickly on surfaces and the latter being particles of less than 5 µm, therefore lighter and able to “float” longer in the air. Droplet transmission refers to direct inhalation of the virus exhaled by an infected person when a contact person is close to him/her. Airborne transmission refers to the inhalation of small droplets of aerosol exhaled by an infected person at a distance exceeding two meters.

Numerous works on respiratory viruses preceding the crisis had questioned this definition and underlined the complexity of the phenomenon [2], [10]. Several factors could explain the mixed character [droplets and air] of the transmission of respiratory viruses. They can be summarized as factors related to the type of reservoir [intensity and symptomatology of the emitting source], external factors [physio-chemical phenomena that explain the possibility of desiccation of the emitted particles], and conditions of care. For example, aerosol generating procedures [AGP] such as intubation or nebulizer therapy [3] generate a large quantity of particles with variable flow and speed. These factors may explain why the risk cannot be categorized as droplets or air. Several studies of patients infected by the influenza virus have highlighted the presence of particles of different sizes at a variable distance from the emitting source [2], depending on the intensity of the symptoms [2], [11] and the conditions under which the samples were taken. These findings suggest that particles > 5 µm [supposed to be responsible for droplet transmission] and particles < 5 µm [supposed to be responsible for airborne transmission] are emitted by the same patient at the same moment [12]. If there is a continuum in the size of the emitted particles, this cannot explain why some viruses are supposed to be highly infective in an airborne route, whereas others are associated with a droplet route. This being said, the risk should be considered in relation to viral load and the distance from the source. In fact, it has been shown that the small droplets produced by an infected individual have a low viral load [13]. They do not pose a serious threat of infection, whereas larger droplets have a higher viral load with higher risk for healthy individuals, but they settle quickly. This phenomenon could be more complex as these larger droplets partially evaporate during their ballistic trajectory before settling on the ground. In this situation, the droplets shrink due to friction forces and evaporation in atmospheric air. This could transform a large particle > 5 µm into various particles, of which some become suspended in the air but still retain their pre-evaporated viral load. These infectious droplets can travel longer distances in the air and have a high enough viral load to infect healthy individuals [13]. It is important to highlight that many factors can impact the behavior of respiratory droplets: temperature and humidity, [12], [13], [14], and air flows due to ventilation systems or human activity. Indeed, the co-presence of a ventilation system and of an infectious source is at the origin of airflow disturbances that could modify the trajectory and the range of the particles [15].

In the community, the risk of transmission/contamination seems to be much higher than in a hospital environment. There are many reasons for this, including duration of contact with the potential reservoir, proximity of the reservoir, the absence of protective measures and the viral load in the patient’s airways [high at disease onset]. The risk is certainly different at home [where contact times are longer] compared to other living environments [including public transportation]. However, this difference is modifiable by the number of reservoirs present [i.e., colonization pressure], promiscuity, and level of air renewal. Similarly, the risk is related not only to direct but also to indirect transmission.

In the hospital, the risks differ, depending on the type of structure. In long-term care facilities, they seem greater than in acute care. There are many reasons for this, including colonization pressure, promiscuity and shared activities. In acute care, risks are variable and depend on colonization pressure [the higher it is, the greater the risks], the earliness of patient hospitalization [in relation to the onset of symptoms], and the extent of care load and the duration of contact with infected patients.

4. Benefits of wearing a mask

For more than a decade there has been a scientific debate about the usefulness of wearing a mask and its effectiveness during periods of viral circulation. While medical masks are designed to block droplets, N95 respirators are tailored to capture 95 % of particles up to 0.3 µm in diameter. Numerous studies have suggested the importance of wearing a mask to prevent the risk of infection. A prospective cluster-randomized study conducted during the 2006 winter season found that compliance with mask use significantly reduced the risk of influenza-like illness [ILI] in households [16]. Furthermore, a blinded cluster randomized study [17] was aimed at determining whether hand hygiene and/or face mask use prevented influenza transmission in households. The authors found that hand hygiene, with or without wearing a face mask, appeared to reduce influenza transmission, but differences from the control group were not significant. That much said, infection appeared to be reduced when interventions were implemented within 36 hours of symptom onset, a finding suggesting that hand hygiene and masks may prevent transmission of influenza virus at home when implemented earlyt after symptom onset in the index patient.

In a meta-analysis including 6 randomized and 23 observational studies, Offedu and colleagues [18] suggested a protective effect of masks and N95 respirators against clinical respiratory illness (CRI) and Influenza-like illness (ILI) in healthcare settings. Moreover, the authors suggested that N95 respirators conferred superior protection against CRI and laboratory-confirmed bacterial infections, but surprisingly not against viral infections or ILI.

However, a previous study evaluating continuous or targeted N95 respirators during high-risk procedures involving 1669 hospital-based healthcare workers in Beijing (China) during the winter of 2009–2010 found a higher clinical respiratory illness rate in the medical mask group, followed by the group wearing targeted N95 respirators and the group continuously wearing them. Furthermore, after adjustment for confounders, compared with targeted N95 respirator use, only continuous N95 respirator use remained significant against CRI [19].

Most guidelines recommend masks for droplet-transmitted infections. As “airborne precautions,” N95 masks offer superior protection against droplet-transmitted infections. To ensure the health and safety of healthcare workers, the superiority of N95 respirators in preventing respiratory infections should be highlighted in infection control guidelines.

5. Why should prevention measures for SARS-CoV2 be different?

5.1. The virus, from the prevention standpoint

Concerning viruses, several characteristics specific to the virus suggest a variable risk of secondary transmission. Among these characteristics, the ability to survive in the environment and resistance to detergents and disinfectants appear primordial. Indeed, adding an environmental reservoir to the human reservoir would increase the risk of transmission. At the beginning of the pandemic, Van Doremalen et al. evaluated the stability and desiccation rate of the SARS-CoV2 virus compared to SARS-CoV1 on different surfaces, suggesting 3 h survival of the virus in aerosols and viability on surfaces lasting up to 3 days [20]. Survival in the environment was found to vary, depending on initial concentration, types of surfaces and temperature and humidity conditions [20], [21]. Many studies with contradictory results have analyzed survival under different conditions. For example, iphysico-chemical conditions act as an amplifier of risk. Experimental data show that SARS-CoV-2 activity decreases with increasing temperature. SARS-CoV-2 can survive for 14 days at 4 °C, 1 day at 37 °C and only 30 minutes at 56 °C. Similarly, a person is more likely to be infected when the relative humidity is low, between 10 and 20 %. In addition, experiments conducted over the past 60 years have indicated that the activity of viral pathogen-carrying droplets is negatively related to ambient humidity and that dryness promotes virus spread. SARS-CoV-2 is sensitive to disinfection [22] and exhibits stability over a wide range of pH values (pH 3–10) at room temperature [21].

In fact, SARS-CoV-2 is sensitive to a wide variety of disinfectants. Lipid solvents, including ethanol (>75 %), formaldehyde (>0.7 %), isopropanol (>70 %), povidone-iodine (>0.23 %), sodium hypochlorite (>0.21 %), or hydrogen peroxide (H2O2; >0.5 %), can be used to inactivate SARS-CoV-2 [23].

5.2. What do we know about SARS-CoV2 transmission?

Recent data on SARS-CoV2 suggest that the risk of transmission begins during the period of asymptomatic infection, 48–72 hours before the onset of symptoms, and progressively decreases until in most situations it disappears by the 8th day after the first symptoms. Respiratory samples taken from infected symptomatic and asymptomatic patients, as well as epidemiological and clinical studies, have demonstrated viral shedding beginning 48 to 72 hours before the first symptoms and persisting up to 8 days after. Transmissionwould begin 2 to 3 days before the onset of symptoms, reach its peak approximately 1 day before the onset of symptoms [24], and rapidly decrease within the first 7 days [24], [25], [26].

SARS-CoV-2 viral loads in the respiratory tract decline rapidly after symptom onset. The highest loads shift from the upper to the lower respiratory tract [27], [28] While patients with severe disease have higher respiratory viral loads than those with mild disease, all loads decline over time [29]. Researchers in China have estimated the duration of RNA shedding from various sites based on detailed analysis of samples from 49 patients with COVID-19 and reported a median duration of shedding from the nasopharynx of 22 days for mild cases and 33 days for severe cases. The median duration of excretion from the nasopharynx was likewise 22 days for mild cases and 33 days for severe cases, with some individuals excreting for more than 2 months [30]. It should be noted that the period of infectiousness is much shorter than the duration of detectable RNA. For mild to moderate cases, infectious virus can be isolated from specimens only until approximately the 8th day of symptoms. Multiple studies have found virtually no viable virus in patients with mild to moderate disease after 10 days of symptoms, despite continuous and frequent RNA shedding [27], [31], [32]. In a study that included patients from 0 to 21 days after symptom onset, viable virus was isolated in 26 out of 90 samples, but no viral growth was found when the cycle threshold was greater than 24 or the patient had more than 8 days of symptoms [33]. A group from the Netherlands evaluated 129 hospitalized patients, 89 of whom required intensive care, and collected upper and lower tract samples [34]. Isolation of infectious virus occurred on average 8 days after the onset of symptoms. The probability of isolating infectious virus was less than 5 % at 15.2 days, and decreased with increasing time since symptom onset, lower viral loads, and higher neutralizing antibody titers. The “tardiest” isolation of infectious virus occurred 20 days after symptom onset. Despite the late isolation of infectious virus, no late transmission was documented, including in health care settings. Some field observations have confirmed that after 6 days of symptoms, the risk of transmission from an index case to HCW is low [35].

To summarize, the period of greatest risk seems to be between Day 2 and Day 3 of the first symptoms, which would explain a greater risk of contamination in the community than in the hospital, since patients are rarely hospitalized as soon as the first symptoms appear, in a situation where the risk of transmission is decreasing. As with many other viruses, pre-symptomatic patients as well as symptomatic patients, are at the origin of transmission. However, the risk of transmission seems to be lower in the former than in the latter, and the period during which pre-symptomatic patients are infectious remains unknown to this day, as do the dynamics of excretion in asymptomatic patients. Among those who develop symptoms, the secondary attack rate may increase with the intensity of the source case's symptoms. However, many authors have suggested the presence of super spreaders [36].

5.3. Transmission and animal Model:

Numerous animal models [37] have been used to understand the modes of transmission and pathogenicity of SARS-Cov2. Transmission by direct contact remains the most frequently identified mode of transmission regardless of the animal model used [38]. However, in ferrets, authors have emphasized the importance of airborne transmission over short distances [animals separated by a maximum of 10 cm] more frequently than over long distances [animals separated by more than one meter][39]. Similarly, numerous studies carried out in the hamster model and testing the different variants of SARS-CoV2 [40] have shown a risk of transmission by contact [40] and by air, with risks varying according to the different variants. In one study, the authors demonstrated [41]that infected hamsters produced aerosolized SARS-CoV-2 infectious particles both before and during the onset of a mild disease. The average emission rate in this study was 25 infectious virions/hour on days 1 and 2 after inoculation, with average levels of viral RNA 200-fold higher than infectious virus in the aerosol particles. In this work, the majority of the virus was contained in particles.

5.4. Human data

In a recent study, the authors demonstrated a correlation between the viral load of index cases and the occurrence of secondary cases [42]. In this study including 212 index cases and 365 contacts, 19 % tested positive after their exposure, and after adjustment for cough, time between test and exposure risk of transmission to a close contact was significantly associated with the viral load. In a home setting, an original study based on the detection of exhaled mRNA showed a correlation between the threshold of mRNA positivity and secondary transmission [43]. Frequency of home transmission was associated with viral load.

Among community activities with high transmission risk, it seems that private gatherings (family meals) or densely populated public places lead to higher viral diffusion [44]. Indeed, in a study including 32 [45] studies involving 68,206 participants, the authors highlighted lower attack rates in the context of extra-familial gatherings. For example, in this work, transportation (RR 10.55, 95 % confidence interval (CI) 1.43–77.85), medical care (RR 11.68, 95 % CI 1.58–86.61), and work or study locations (RR 10.15, 95 % CI 1.40–73.38) had lower attack rates. All in all, attack rates were highest at home (95.3 %), meals or gatherings (81.4 %), public places (58.9 %), daily conversation (50.1 %), transportation (30.8 %), medical care (18.2 %), and work or study sites (15.3 %). In another study of home attack rates [46], including 276 index cases and 644 contact cases, the authors found 200 cases of secondary SARS-CoV-2 infection, representing an attack rate of 45.7 % (95 % CI: 39.7–51.7 %) per household. Asymptomatic or mildly symptomatic index cases had a lower risk of transmission. This work also showed that the majority of transmissions occurred early after the introduction of SARS-CoV-2 into a household.

5.5. Is transmission related to surfaces or to the air?

During the pandemic, numerous studies sought to identify the risk factors associated with environmental contamination and to distinguish between direct transmission through the air and indirect transmission through contaminated surfaces [47].

Environmental contamination depends not only on virus characteristics and patient-related factors but also on several other factors such as temperature, exposure to UV, moisture or surface characteristics [22]. The results of the various studies carried out during the pandemic are contradictory, in terms of both air sampling and surface sampling. In addition, many limitations to the different studies must be underlined before considering the frequency of contamination of the different types of samples. Firstly, most published studies have initially searched by RT-PCR for viral RNA and not for the presence of viable virus, neglecting the possibility that the results obtained do not reflect actual infectivity. Secondly, when interpreting their results few studies have taken into account certain confounding factors, such as the intensity and precocity of symptoms and the duration of exposure of the environments to the reservoir. Thirdly, the heterogeneity of the populations studied make it difficult to interpret the results. Finally, during the pandemic, several sampling devices and methods were used for surfaces or for air, and they were not standardized (Table 1 ).

Table 1.

Confounding factors accounting for the discrepancies in the literature.

Patient-related	Intensity of symptoms (coughing, sneezing) Sampling time (compared to the onset of symptoms) Colonization pressure
Physico-chemical data	Humidity Temperature
Ward-related	Surface and volume of the rooms Ventilation sytems (Air change rate, Air filtration)
Sampling method	Sampling techniques, Positioning of sampling systems Duration and quantity of air intake Microbiological methods (RT-PCR vs Viral isolation)

5.6. Frequency and risk factors of surface contamination

The data in the literature are variable and contradictory suggesting contamination on between 5 % and 50 % of contaminated surfaces. In an initial study carried out in China [48], the authors highlighted surface contamination in 57 % of situations. In this study, they noted high surface contamination in ten (66.7 %) of the 15 patients during the first week of illness, and in only three (20 %) after the first week of illness, suggesting like others [48] a decrease in the frequency of environmental contamination. Similarly [49], in another prospective study the authors correlated the cycle threshold (Ct) for reverse transcription polymerase chain reaction (RT-qPCR) of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) on a nasopharyngeal swab (test taken closest to the day of collection) with the percentage of RT-qPCR-positive surfaces [Spearman rank correlation coefficient of −0.48 (P < 0.001)]. In a more recent study [50], the authors assessed environmental contamination of inpatient rooms in a dedicated COVID-19 unit. Among the 347 samples, they found a contamination rate of 5.5 % based on RT-PCR results and only 0.3 % based on cell culture. They confirmed that despite the frequent identification of viral RNA by RT-PCR, the viable nature of the virus in surface samples remains rare [51]. Finally, a meta-analysis including about fifty in-hospital studies suggested that contamination was more frequent in the ICU than outside and, specifically, in patient rooms compared to non-patient areas [52]. Numerous factors explain these discrepancies, and analysis of the data in the literature underscores several points concerning the contamination of surfaces, which depends on (1) the earliness of sampling compared to the onset of symptoms, (2) the intensity of symptoms noted at the time of sampling and (3) the type of surface sampled [53]. Also, the longer the time from the onset of symptoms, the more negative the results of environmental sampling will be. All in all, environmental contamination is less frequent in asymptomatic patients compared to symptomatic ones. However, recent study suggested a low risk of viral transmission, in most instances by fomites [54].

5.7. Air contamination

Aerosol transmission was better understood during the pandemic. Several studies have tried to detect SARS-CoV-2 in the air. As mentioned above, one of the difficulties of interpretation is related to the different sampling methods applied. In a systematic review including 25 studies [55], while the authors identified 15 studies with at least one positive sample, only one of them suggested the presence of viable viruses in the air [56]. In a more recent prospective study conducted in an acute care setting between March and May 2020, the authors suggested that air contamination was rarely identified (2 % among 146 samples) [51]. In a meta-analysis of in-hospital data, the authors suggested the presence of SARS-CoV2 in 16 % of air sampling [52]. In this meta-analysis, SARS-CoV2 RNA was more frequently detected in patient areas, and the highest prevalence was found in ICU patients. However, the viral concentrations were lower in ICU patients’ rooms when given in copies and did not differ from non-ICU patients in CT-values. Finally, the most recent systematic review[57], including 73 published papers, suggested lower air contamination outdoors compared to hospitals and care facilities, suggesting a lower acquisition risk.

An issue widely debated in the literature is that of the risk associated with ventilation support. To date, several studies have suggested that there is no increased risk with CPAP or HFO [58], [59], whereas, as suggested in the aforementioned meta-analysis suggested a higher risk of detecting surface or airborne SARS contamination when AGPs are performed.

Despite some discrepancies, surface contamination has been more frequently identified than air contamination. While many authors have emphasized the frequency of surface contamination, few have proven the viability of the virus. Indeed, most of the studies are limited to the identification of viral RNA, and do not deal with the viability of the virus by cell culture [60].

Few studies have attempted to identify the proportion of surface transmission compared to airborne transmission. Despite debated data in the literature, there are some indirect arguments to suggest a low risk of transmission through contaminated surfaces. Some authors have even suggested, with the help of video recordings, that there is no transmission from surfaces in the community [61]. Similarly, although the frequency of positivity of surface samples in RT-PCR is high, the thresholds obtained suggest a low viral concentration and the cytopathic effect has rarely been demonstrated. However, it has become increasingly clear that the fomite route is less important than previously thought [62].

Evidence has been accumulated, especially since the outbreak of COVID-19, of the primacy of the airborne transmission route. The most important evidence of airborne transmission concerns long-range airborne transmission, especially in poorly ventilated indoor environments [15], [63]. Relevant studies include analyses of COVID-19 outbreaks on a bus, a cruise ship, a gym, and a restaurant. It has become clear that poor ventilation increases long-range airborne transmission. On the other hand, in the majority of the articles included, this scoping review shows that the risk of SARS-CoV-2 infection via contaminated surfaces is low [61].

5.8. Air or droplets?

This debate seems outdated. In fact, the pandemic has reminded us that there exist several modes of transmission; the question is to know what is the most appropriate mode of prevention according to the different situations. In a very specific moment, a human being infected by SARS-CoV-2 would exhale a very high number of droplets not all of which would be of the same size. During a single act of exhaling, the size of droplets emitted by a human being follows a normal distribution around a mean. Depending on vesico-elastic properties of the respiratory mucus and saliva, both of which are modified by the natural history of the respiratory infection, the mean of the normal distribution changes. A single patient can exhale a continuum of particles from aerosols to droplets, and that the combination can change along time. Numerous models and reviews published during the pandemic pointed out that the transmission of respiratory viruses in general and SARS-Cov2 in particular is mixed, including both droplet and airborne transmission [64]. Similarly, many authors have cited factors and variables associated with the different modes of transmission. In a recent systematic review including 18 community-based studies conducted during the pandemic [65], the authors concluded that long-distance airborne transmission was possible in indoor environments such as restaurants, workplaces, and choral settings. They identified factors such as inadequate air exchange as likely contributors to transmission. These findings reinforce the need for mitigation measures in indoor environments, particularly adequate ventilation. Numerous studies have sought to identify the presence of SARS-Cov2 in air. Despite conflicting data, the presence of SARS-CoV2 has rarely been demonstrated in RT-PCR or cell culture samples. However, these contradictory data are essentially related to the different sampling techniques and patient profiles [66].

5.9. Which mask for which risk?

The debate about the effectiveness of surgical masks or N95 respirators was one of the highlights of the pandemic. While all laboratory studies have demonstrated the expected superiority of filtration by N95 respirators, their superiority in clinical practice remains debated. One disadvantage of the N95 respirators is the increased risk of leakage if the mask is not fitted, and lower adherence has been reported due to higher rates of adverse events. Despite numerous methodological biases, many studies have tried to answer the question of the effectiveness of different masks in terms of prevention for health care workers. A meta-analysis including 4 randomized studies [67] suggested, with a low level of certainty, no difference for either clinical or microbiologically proven infections. Similarly [68], a second meta-analysis including 6 studies suggested that there are insufficient data to definitively determine whether N95 respirators are superior to medical masks in protecting against acute respiratory infections. Further randomized trials are needed to compare the above respiratory protection methods in the context of COVID-19 incidence. The meta-analysis by Maccintyre et al. included the 2 randomized controlled studies performed by the same team in China and including 3591 subjects while comparing 4 interventions: (i) continuous use of an N95 respirator, (ii) targeted use of an N95 respirator, (iii) use of a medical mask, and (iv) control arm. In the adjusted analysis, this study highlighted laboratory-confirmed bacterial colonization rates (RR 0.33, 95 % CI 0.21–0.51), laboratory-confirmed viral infections (RR 0.46, 95 % CI 0.23–0.91), and droplet-transmitted infections (RR 0.26, 95 % CI 0.16–0.42) were significantly lower in the continuous N95 group. Rates of laboratory-confirmed influenza were also lowest in the continuous N95 group (RR 0.34, 95 % CI 0.10–1.11), but the difference was not statistically significant. Rates of laboratory-confirmed bacterial colonization (RR 0.54, 95 % CI 0.33–0.87) and droplet-transmitted infections (RR 0.43, 95 % CI 0.25–0.72) were also lower in the targeted N95 group, but not in the medical mask group.

6. Discussion-Conclusion:

Despite initial uncertainty, the aerosol route is now recognized as the principal path of transmission of COVID-19. Several arguments favor airborne transmission: the effect of ventilation, the difference between indoor and outdoor transmission, the possibility of transmission despite the use of masks and goggles, animal data, and airflow simulations[69]. Several works have suggested more frequent contamination of surfaces compared to airborne viruses. However, it is difficult if not impossible to standardize different clinical situations and many factors, which vary over time, contribute to the risk of transmission. Concerning the clinical elements, it is important to note that most of the studies were carried out on a given day, even though the infection and the risks of contamination continue to evolve. In several studies, the time lapse between sample taking and onset of symptoms remains a first limiting factor. In fact, the longer the time elapsed, the lower the probability of finding the virus in the environment. That said, there exists a window during which the possibility of finding the virus seems higher (D-2 to D + 8). Secondly, the fact that patient profiles are not taken into account makes it difficult to interpret the studies and explains some of the discrepancies. Furthermore, whatever the time of sampling, the intensity of respiratory symptoms is associated with environmental contamination. The third study limitation is related to the unit in which sampling is carried out. For example, intubated and ventilated patients in intensive care units are less at risk of aerosolization than those hospitalized in medical units. Fourthly, the non-standardization of sampling methods further complicates the interpretation of literature dataTable 2 .

Table 2.

Preventive measures according to different situations.

Setting		Mode of transmission		Preventive measures			Comments
		Surface	Air	Vaccination	Hand Hygien	Mask (type)
Community	Home	++	++	++	++	-	At home, airborne transmission occurs early (even before symptoms). Wearing a surgical mask is interesting for symptomatic patients, less for contact patients.
	Gathering	+	++	++	+	+ (surgical)	The risk depends on the colonization pressure and the volume and ventilation level of the room
	Public transport	+	++	++	+	+(cloth)	The risk is low compared to the situations described above. In ventilated spaces, virus survival on surfaces is low.
Hospital	Emergency department	++	++	++	++	++ (surgical)	At risk because of the number of patients managed, the proximity of care and the earliness of management in relation to the onset of symptoms. N95 respirators if AGP
	Medical Ward	++	++	++	++	++ (surgical)	Less risk than ED, due to patient “tracking” and isolation. However, the delay in management from the onset of symptoms is the major risk factor. N95 respirators if AGP
	ICU ward	+	+	++	++	++	Management of ICU patients is usually delayed relative to the onset of symptoms. Patients are generally less shedding. N95 respirators if AGP

Like clinical studies, fundamental studies encounter numerous limitations: the lack of standardization of the diffusion methods, the viral concentrations emitted, and the difficulty of taking into account the physicochemical data associated with the risks of diffusion in the air. Despite the contradictory data during the pandemic [70], there are several clinical and theoretical arguments in favor of N95 respirators in contact with infected patients. Studies during the pandemic suggested the superiority in terms of filtration of N95 respirators compared to surgical masks, and in some cases [71], a lower risk of contamination of health care workers. It is now clear that the presence of fine particles exhaled by infected patients and the risk situations in hospitals [3] are two factors that argue in favor of N95 respirators. However, N95 respirators require a seal [72], [73], and previous work suggested that in terms of the risk of penetration, peripheral leaks had a greater impact than level of filtration.

In conclusion, while airborne transmission of respiratory viruses is a certainty, air vs droplet is in our opinion an outdated debate. Conditions associated with patient status, ventilation data or therapeutic procedures, can transform droplet risk into air risk. If transmission is mainly airborne, it is difficult to measure the environmental part linked to surface contamination. However, studies suggest that the further one moves away from the onset of symptoms, the more this risk decreases. Future work should discuss not so much preventive measures (mask and hand hygiene) as the situations in which the systematic wearing of a mask is useful and the type of mask which, in our opinion, should be adapted not to types of virus, but rather to the care situations that may expose a HCW to greater or lesser risk.