key: cord-1027943-fl5ar971 authors: Arav, Y.; Klausner, Z.; Fattal, E. title: Understanding the indoor pre-symptomatic transmission mechanism of COVID-19 date: 2020-05-17 journal: nan DOI: 10.1101/2020.05.12.20099085 sha: fae601e0dcbd00aa3b4a655b8a62e77b0f7e8f2b doc_id: 1027943 cord_uid: fl5ar971 Discovering the mechanism that enables pre-symptomatic individuals to transmit the SARS-CoV-2 virus has a significant impact on the possibility of controlling COVID-19 pandemic. To this end, we have developed an evidence based quantitative mechanistic mathematical model. The model explicitly tracks the dynamics of contact and airborne transmission between individuals indoors, and was validated against the observed fundamental attributes of the epidemic, the secondary attack rate (SAR) and serial interval distribution. Using the model we identified the dominant driver of pre-symptomatic transmission, which was found to be contact route, while the contribution of the airborne route is negligible. We provide evidence that a combination of rather easy to implement measures of frequent hand washing, cleaning fomites and avoiding physical contact decreases the risk of infection by an order of magnitude, similarly to wearing masks and gloves. During the months following the emergence of the COVID-19 pandemic in December 2019, it became evident that sharing an indoor space is the major SARS-CoV-2 infection risk (1) (2) (3) . These studies also found that the members of the same households has the highest risk of infection among people in different modes of close contact. This conclusion is based on the secondary attack rate (SAR), the percentage of household contacts who were later confirmed to be infected with SARS-CoV-2. Estimates of the SAR made in China, South Korea, Taiwan and the United States ranges between 10.2 − 16.2% (2) (3) (4) (5) (6) . Due to the fact that most of these estimates were made in countries that lead a public health policy of immediate isolation of cases upon symptoms' onset, these estimates represent the effect of pre-symptomatic carrier transmission. In fact, pre-symptomatic transmission was recently referred to as the Achilles' heel of COVID-19 pandemic control, as symptom-based detection of infection is less effective in comparison to the control of the SARS epidemic in 2003 (7) . However, the question of understanding the mechanism that enables seemingly healthy individuals to transmit the virus, was left unsolved. This is the motivation of this study. Generally, respiratory viruses, such as SARS-CoV-2, spread via three transmission routes: contact, droplet and aerosol transmission. In contact transmission an infected person gets virus on his hands and transfers this virus either directly, e.g., via a handshake, or indirectly via an intermediate object, to the hands of an infectee, who then places his hand into his facial membranes, thus exposing himself to the contamination on his hands. Transmission of virus through the air can occur via droplets or aerosol. Droplets generated in a cough or a sneeze travel less than 1.5m before they settle on close contacts or environmental surfaces (8) . Aerosols remain suspended in the air and may infect a susceptible individual once they deposit in his upper or lower respiratory tract. The commonly accepted cutoff is 5µm (9) . However, droplets that are smaller than approximately 100µm evaporate to their droplet nuclei size before they hit the ground (8) . Following (8) , we assume that the respiratory fluid is a physiological saline solution with anion and cation concentration of about 0.9%w/v. Therefore, the droplet shrinks to about 3 √ 0.009 ≈ 0.2 of its original size. Thus, a 100µm droplet would reduce to a droplet nuclei of 20µm before reaching the ground. Here we have used a cutoff size of 100µm between droplets and aerosol. This cutoff size results in an conservative estimation of the contribution of the aerosol transmission route as larger volume of aerosols is considered. The relative importance, if any, of these routes differ for each infectious disease, depending on its specific parameters. In the current study, we model the mechanism of indoor transmission with an individualbased stochastic mechanistic model ( Figure 1 ). The model describes the basic interaction of two individuals, a pre-symptomatic primary (infetor) and a secondary (infectee) individuals. 3 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 17, 2020. face. In addition, the primary sneezes and coughs in a rate characteristic to healthy individuals (11, 12) . We have performed each until the primary developed symptoms in order to address the question of pre-symptomatic transmission, and in accordance with the public health policy that isolates the primary when his symptoms appear. That is, the duration of each realization is the primary's incubation time, that distributes log-normally with a mean of 5 days and standard deviation (SD) of 0.45 days (13) . We assumed an exponential growth law of the viral load with time (10) which reaches its maximal level when the symptoms appear (14) . The probability that the secondary will be infected is inferred from the dose-response curve that was reported for SARS-CoV-1 (15) . We relied on the recent available literature to-date to determine empirically plausible values for the model parameters. A complete list of the model parameters and their values is presented in Table S1 . The reference simulation uses parameters that describe a normal, pre-epidemic, behaviour (see Table 1 ). Details and sensitivity analysis on key parameters such as the dose response, viral loading and shedding, room dimensions, transfer coefficients 4 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. A necessary validation criteria for a model such as the one described in this study is to correctly simulate the SAR and the distribution of the serial interval. The serial interval is the time period between the symptoms' onset of primary and the secondary. Its distribution is closely associated with the estimation of the reproductive number and key transmission variables in epidemic models as well as important in the optimization of quarantine and contact tracing (16, 17) . The serial interval distribution of COVID-19 was estimated in many countries and was usually found to be gamma distributed with mean between 4.03 to 6.3 days and standard deviation between 3 and 4.2 days ( Figure 1A , shaded area) (2, (18) (19) (20) . (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 17, 2020. We have also analyzed the contagious period of pre-symptomatic patients by examining the cumulative SAR over time ( Figure 2C ). As seen, the contagious period begins approximately 30 hours before symptoms' onset, with increasing probability as the onset of the symptoms approaches. This result is consistent with the estimation of He et al. (14) that inferred from data of 77 transmission pairs (i.e., primary and secondary) a contagious period of approximately 2 days before symptoms' onset. The fact that contact transmission is the main route of pre-symptomatic transmission, suggests that the hygienic and behavioral measures (HBMs) advised to the public should focus on HBMs to diminish the contamination on the hands or somehow interrupt the virus transfer from the hand to the facial membranes. We decided to examine five HBMs: Washing hands, cleaning fomites, maintaining social distancing (i.e avoiding physical contact), wearing a mask and 6 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 17, 2020. . https://doi.org/10.1101/2020.05.12.20099085 doi: medRxiv preprint gloves. Naturally, conservative precautions measures would be to implement all these at once. However, strict adherence to all these HBMs would be hard to endure and persist on doing over a long period of time. Therefore, we have tried to sort out few combinations of HBMs that will enable practical implementation by the public, while significantly lowering the risk of infection. As the SAR is a proportion, it is appropriate to compare the HBMs in terms of odd ratio (OR), i.e., the odds that the secondary Will be infected when a given combination of HBMs is taken, compared to the reference scenario in which no HBM is applied. Generally, any HBM that results in OR less than 1 decreases the risk of infection (i.e., provide smaller SAR than the reference) (23) . However, in practice the lower the OR, the better HBM combination is at lowering the risk. The values brought here are in terms of OR alongside with 95% confidence interval (95% CI) Washing hands is known to remove the viruses from the hands of both individuals and it is the simplest measure to implement. Our simulations show that washing hands every hour rather than 3 times a day, as in the reference simulation (Table 1) , results in OR of 0.71 (95% CI 0.62-0.8) ( Figure 3A, column H) . This result is consistent with intervention studies that have shown that increased hand washing decreased respiratory illness by 20%, albeit different viruses were studied (24) . This phenomenon seems counter intuitive, as we found that more than 99% of the viruses are transmitted through the hands and it was expected that washing it would remove the contamination. In order understand the reason for the relatively limited effect of hand hygiene, we have examined the dynamics of the virus concentration on the hands of the secondary individual ( Figure 3B ). This concentration exhibits a periodic behaviour, that is governed by touching events in fomites and the face. Spectral analysis reveals that the hand concentration cycle is characterized mainly by frequencies that are with time scale of 50 minutes (see Supplementary text). Therefore, hand washing is expected to dramatically reduce the risk for infection if it occurs at at higher frequency than 50 min. Unfortunately, such frequent 7 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 17, 2020. . washing is unrealistic. Cleaning the fomites more frequently reduces the virus repositories that are available for intake. Cleaning of the fomites 10 times a day rather then 2 times a day, as in the reference simulation, results in OR of 0.82 (95% CI 0.72-0.93), rather similar to washing hands more 8 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 17, 2020. (Figure 3A , column M+G). This result is surprising, as it was expected that protecting the hands and mouth will provide as the best HBM combination. The fact that the combination of all other HBMs provided better OR means that following these HBMs meticulously may save people the discomfort and limitation that is associated with having to wear constantly a mask and gloves in indoor scenarios. Our analysis, as with all modeling exercises, has several limitations and requires certain assumptions. At this point, the model does not account for contact patterns that prevail in households with young children and does not take into account the diurnal cycle of activity. The model parameters, such as the dose response curve, the viral shedding coefficients and 9 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 17, 2020. . transfer coefficients were chosen on the basis of prior knowledge of the SARS, other strains of coronavirus or other bacteria (15, 30) . Although the model is stable to variations in these parameters, more information on the key characteristic of the disease would considerably reduce uncertainties. To conclude, we have analyzed the possible routes of pre-symptomatic transmission in indoor scenarios. Using a validated model, we were able to identify the main transmission mechanism as contact associated, mostly directly but also mediated by fomites. Frequent hand washing and fomite cleaning coupled with avoiding physical contact result in a similar risk for infection as wearing gloves and a mask. Our findings can provide an important tool for decision makers while advising the public of the HBMs that are necessary to impede the epidemic. As it seems that the initial wave of pandemic may be closing to its end, many countries are gradually lifting the restrictions on society, such as the re-opening of schools and workplaces. However, recurrent outbreaks (the so called second wave) may occur in the coming year (31) . Under such reality, the model presented in this study can be used to quantify the contribution of different measures in mitigating the risk of infection in workplaces or schools scenarios. Transmission routes of respiratory viruses among humans Air, Surface Environmental, and Personal Protective Equipment Contamination by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) from a Symptomatic Patient Include acknowledgments of funding, any patents pending, where raw data for the paper are deposited, etc.