key: cord-0694184-735htytn authors: El Deeb, O.; Jalloul, M. title: Forecasting the outbreak of COVID-19 in Lebanon date: 2020-09-05 journal: nan DOI: 10.1101/2020.09.03.20187880 sha: 9a6dcb2b7f03285801422178f63b05097c65c9b8 doc_id: 694184 cord_uid: 735htytn in Lebanon using available data until August 25th, 2020 and forecasts the number of infections until the end of September using four different scenarios for mitigation measures reflected in the reproductive number Rt. Mitigation measures in Lebanon date back to early March soon after the first confirmed cases, and have been gradually lifted as of May. Thereafter, the country has witnessed a slow yet steady increase in the number of cases that has been significantly exacerbated after the explosion at Beirut harbor on August 4. Furthermore, we estimate the daily active cases in need of intensive care compared to the available number of beds and we assess accordingly that this capacity will be exhausted within a short span of time, unless severe measures are imposed. The COVID-19 outbreak in Lebanon has started on February 21, 2020 with few conrmed cases during the rst days [1, 2]. Mitigation measures have been made into eect early March, and have been gradually lifted as of May 10, while the airport activity was resumed with one-fth capacity on July 1. A couple of weeks after, the number of infections was on the rise once again. On August 4, a massive explosion hit Beirut's port caused by approximately 3000 tonnes of unsecured Ammonium Nitrate, leading to hundreds of deaths and thousands of injuries, in addition to a wide radius of destruction aecting thousands of buildings, including hospitals, in and outside the city [3] . Hospitals and medical centers We use the STEIR, a novel adaptation of the SEIR models accounting for trave, that was introduced in [5] . We employ data available until August 26, 2020 [1] to construct forecast scenarios of the spread of the disease until the end of September, 2020 (see Figure ( 1)). The model divides the population N into ve categories of people: susceptible S, incoming travellers T , exposed E, infectious I and removed R (through recovery or death). The key parameters that determine the dynamics of spread are: the rate of exit by recovery 2 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted September 5, 2020. or death per-day which is associated to the average illness period (γ), the rate at which exposed individuals become infected associated to the mean incubation period (σ), the reproductive number R t that determines the transmission rate from susceptible to infected and is a proxy for social distancing measures, the daily rate of incoming travellers (τ ), and the average infection rate among the travellers (θ). The numerical values of these parameters are consistent with those used in [5] , apart from the parameterization of the basic reproduction number R t which we update according to the actual cases registered until August 26. We forecast four possible future scenarios for the development of R t corresponding to four levels of social distancing measures (see Figure 2 ), and we parameterize it as follows: 3 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted September 5, 2020. . https://doi.org/10.1101/2020.09.03.20187880 doi: medRxiv preprint (1) where R 5 is the future expected value for the reproduction number. Note that all values of R 0 to R 4 are obtained by tting the reproduction rate according to the available data. We examine four potential values of the reproduction rate: R 5 = 1.5 or 2 corresponding to intensifying the social distancing measures which could include complete lockdown, isolation of certain areas, imposing curfews, etc., 2.4 maintaining the current infection rate, and 2.5 which would lead to a slight increase with respect to the number of cases recorded in the last days. The corresponding scenarios of the spread are plotted in Figure ( 1) until the end of September. Our results presented in Figure (1) and Table ( is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted September 5, 2020. Finally, it is essential to note that the estimated number of beds in the intensive care units used here and by the Ministry of Public Health is not exclusive for treatment of COVID-19 patients. Hence, actual saturation dates of the capacity of the health sector are likely to be earlier than the aforementioned. In this sense, it is of high importance to adopt the reasonable mitigation policies [6] to avoid bearing the consequences of a entirely overloaded health care system. [1] Disaster Risk Management Unit., 2020. Available from: http://drm.pcm.gov.lb. . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted September 5, 2020. . https://doi.org/10.1101/2020.09.03.20187880 doi: medRxiv preprint Beirut explosion: What we know so far, BBC News The dynamics of COVID-19 spread: evidence from Lebanon How to use WHO risk assessment and mitigation checklist for mass gatherings in the context of COVID-19, World Health Organization