key: cord-0852016-1ntaz1ju authors: Mumtaz, G. R.; El-Jardali, F.; Jabbour, M.; Harb, A.; Abu-Raddad, L. J.; Makhoul, M. title: Modeling the impact of COVID-19 vaccination in Lebanon: A call to speed-up vaccine roll out date: 2021-05-30 journal: nan DOI: 10.1101/2021.05.27.21257937 sha: 53d941731e8546f57ae17d378972c9912a18b7f9 doc_id: 852016 cord_uid: 1ntaz1ju Background: Amidst a very difficult economic and political situation, and after a large first SARS-CoV-2 wave near the end of 2020, Lebanon launched its vaccination campaign on 14 February 2021. To date, only 6.7% of the population have received at least one dose of the vaccine, raising serious concerns over the speed of vaccine roll-out and its impact in the event of a future surge. Objective: Using mathematical modeling, we assessed the short-term impact (by end of 2021) of various vaccine roll-out scenarios on SARS-CoV-2 epidemic course in Lebanon. Results: At current immunity levels in the population, estimated by the model at 40% on 15 April 2021, a large epidemic wave is predicted if all social distancing restrictions are gradually eased and variants of concern are introduced. Reaching 80% vaccine coverage by end of 2021 will flatten the epidemic curve and will result in a 37% and 34% decrease in the peak daily numbers of severe/critical disease cases and deaths, respectively; while reaching intermediate coverage of 40% will result in only 10-11% decrease in each. Reaching 80% coverage by end of 2021 will avert 3 times more hospitalizations and deaths over the course of this year compared with 40% coverage. Impact of vaccination was substantially enhanced with rapid scale-up. Reaching 80% vaccine coverage by August would prevent twice as many severe/critical disease cases and deaths than if it were reached by December. Finally, a longer duration over which restrictions are eased resulted in a more favorable impact of vaccination. Conclusion: For vaccination to have an impact on the predicted epidemic course and associated disease burden in Lebanon, vaccination has to be rapid and reach high coverage (at least 70%), while sustaining social distancing measures during roll-out. At current vaccination pace, this is unlikely to be achieved. Concerted efforts need to be put to overcome local challenges and substantially scale up vaccination to avoid a surge that the country, with its multiple crises and limited health-care capacity, is largely unprepared for. Lebanon documented its first severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) case on February 21, 2020. After almost a year of ongoing cycles of imposing and easing restrictions to suppress the epidemic, the country experienced its first large wave in December 2020-January 2021 which stretched its fragmented, under-resourced, and highly privatized health care system beyond capacity. As of May 19, 2021 , over 537,000 confirmed cases and 7,650 deaths have been documented [1] , for a population of 6.8 million inhabitants [2] . While the pandemic has severely impacted economies and societies worldwide, it was compounded in Lebanon by multiple crises including an unpreceded devaluation of the local currency, political uprising against the ruling class, and the world's largest non-nuclear explosion that blasted its capital Beirut in August 2020 causing thousands of casualties and destroying several hospitals with Coronavirus Disease 2019 (COVID- 19) units. With over one million refugees on its territory, Lebanon also hosts the highest per capita number of refugees worldwide [3] , a population highly vulnerable to large SARS-CoV-2 outbreaks [4] . At this stage, mass vaccination with the recently available and highly efficacious COVID-19 vaccines [5] [6] [7] [8] [9] brings hope to control infection spread and mitigate its profound impact, especially with the emerging 'real-world' evidence of their high effectiveness in reducing both symptomatic and asymptomatic disease [10] [11] [12] [13] . On February 14, 2021 the Lebanese Ministry of Public Health launched its vaccination campaign and inoculated its first residents. To date, doses enough to ensure a 73% coverage of the population have been reserved. However, only 7% of these doses have been received so far and, three months into the vaccination campaign, only 6.7% of Lebanese residents have received at least one dose of the vaccine [14], raising serious concerns over supply and the speed of vaccine roll-out. Also, only 21% of the population have registered to receive the vaccine on the ministry's official vaccine platform [14] , raising further concerns over demand, access to and use of the online registration platform, and misinformation among the general public. With the expected relaxation of social distancing restrictions in the upcoming summer months and with the threat of the potential introduction of more transmissible and likely more severe SARS-CoV-2 variants [15] [16] [17] , the objective of this study was to forecast the short term (by end of year) impact of COVID-19 vaccination in Lebanon in order to inform policy and the national vaccination strategy. We simulate what would be the epidemic course if all restrictions are gradually eased and if variants of concern are introduced, and assess the impact of different levels of vaccine coverage, different 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. durations for vaccine scale-up, and different schedules for easing social distancing restrictions on SARS-CoV-2 incidence and associated burden of COVID-19 disease and death. We adapted an existing mathematical model that investigates the generic population-level impact of SARS-CoV-2 vaccination and extended it to be applied to Lebanon [18] [19] [20] . This is a deterministic age-structured Susceptible-Exposed-Infectious-Recovered (SEIR) model that stratifies the population into compartments reflecting their vaccination status, age group, infection status, infection stage, and disease stage. The model consists of a set of coupled nonlinear differential equations that depict population movement between the different compartments. An age-mixing matrix allowed a range of assortativeness for contact patterns between individuals in different age groups. Population demography was assumed to be stable given the short-term timeframe of the predictions (one year). We assumed that the duration of natural immunity (following recovery from a natural infection) and vaccine-acquired immunity is one year [21] [22] [23] [24] [25] . We further assumed that waning of immunity follows a gamma distribution where only a small fraction of individuals lose their immunity either much earlier or much later than one year. The model was coded, fitted, and analyzed using MATLAB R2019a [26] . Details of the model structure and equations can be found in the Supplementary Information Text S1A, S1B, and Table S1 . The model was parametrized with most recent SARS-CoV-2 epidemiological and natural history data. Model parameters, values, and justifications can be found in the Supplementary Information Text S1C and Table S2 . The model was fitted to SARS-CoV-2 case series of infections and deaths in Lebanon [1] . Fitting used a nonlinear least square technique, based on the Nelder-Mead simplex algorithm. Given the large wave that happened near the end of year holidays and informed by unpublished clinical antibody testing reports, we assumed that 20% of the Lebanese population had already been exposed to the infection by January 1, 2021. Since the ultimate objective of vaccination is to allow a return of social and economic life back to normalcy, we gradually eased all social and physical distancing restrictions over a period of four months starting 15 April 2021. We assumed a constant SARS-CoV-2 incidence rate between 1 January 2021 and 15 April 2021, to reflect the observed relative stability of the epidemic during this time where 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 30, 2021. an average of 3,000 confirmed daily cases were being reported [1] . As restrictions were gradually eased starting 15 April 2021, R0 was gradually increased to reach R0=6 at the end of the four months, thus accounting for the increased transmissibility in absence of non-therapeutic interventions of the new SARS-CoV-2 variants of concern [15, [27] [28] [29] [30] [31] . In Lebanon, the BNT162b2 mRNA Covid-19 Vaccine (Pfizer) is the main vaccine rolled out and has the largest fraction of the doses reserved, followed more recently by the ChAdOx1 nCoV-19 vaccine (Astrazeneca). We assumed that the vaccine has 'real-world' effectiveness of 80% at reducing infection (VES=80%), based on recent evidence for the effectiveness of the vaccine in reducing asymptomatic as well as symptomatic infections [10, 12, 32] . After easing all restrictions over four months and introducing variants of concern starting 15 April 2021, we assessed the population-level impact of SARS-CoV-2 vaccination by quantifying and comparing the incident number of infections, severe/critical disease cases, and deaths for the scenarios of 1) no vaccination, 2) vaccine introduced on 14 February 2021 and 80% vaccine coverage is reached by the end of the year, and 3) vaccine introduced on 14 February 2021 and 40% vaccine coverage is reached by the end of the year. We used the World Health Organization (WHO) guidelines for classifying infection severity [33] and determining COVID-related death [34] . We further compared the cumulative number of averted severe/critical disease cases and deaths from 14 February 2021 (date of vaccine introduction) to 31 December 2021 for a range of 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 30, 2021. ; https://doi.org/10.1101/2021.05.27.21257937 doi: medRxiv preprint vaccine coverage attained by end of year, for different durations of vaccine scale-up to reach the target 80% coverage (thus assessing impact of the speed of rollout), and for different durations over which restrictions are eased while reaching the target 80% coverage by end of year 2021. We did not assess the impact of vaccine prioritization by age as it was investigated in earlier studies [18, 35] . Uncertainty analysis was conducted to investigate the impact of uncertainty around our assumed transmissibility of the new variants of concern introduced on 15 April 2021 on the model's predictions of the COVID-19 severe/critical disease incidence for the different scenarios. For a total of 500 simulation runs, we used, in each run, Latin Hypercube sampling [36, 37] to select an R0 within ±30% of the assumed baseline value of R0=6. The resulting distribution for the predicted number of incident and cumulative severe/critical disease cases across the 500 runs was used to calculate the predicted mean and 95% uncertainty intervals around these two estimates. Based on the model assumptions, we estimated that 44% of the Lebanese population have Vaccination did not fully prevent a new epidemic surge because the current contact rate in the community is not sufficiently low and the population immunity level is not sufficiently high, and because of the higher transmissibility of the new variants. The 40% vaccination coverage was not able to noticeably flatten the curve (Figure 1 , orange curve) and hence had a much smaller impact on epidemic course compared with the target 80% coverage rate. This scenario resulted only in an 11.0%, 10.8%, and 10.0% decrease in the peak incidence of infections, severe/critical cases, and deaths, respectively, compared with the no vaccination scenario ( Figure 1 ). Reaching 40% coverage by the end of the year means that most vaccinations would be 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 extent of transmissibility of the variants of concern introduced on April 15 2021 remains uncertain. Analysis factoring in the uncertainty around the assumed R0 of these variants indicates that the value of R0 will affect the scale of the predicted surge ( Figure S1A ) and, hence, will affect the cumulative number of averted infections ( Figure S1B) , with higher R0 leading to more averted infections. However, the beneficial vaccine impact relative to the no vaccination scenario remains largely unchanged. Our analysis predicts that current held immunity levels in Lebanon, estimated by the model at 40% on 15 April 2021, are still not sufficient to go back to normalcy and will result in a large new 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 30, 2021. ; https://doi.org/10.1101/2021.05.27.21257937 doi: medRxiv preprint epidemic surge if variants of concern are introduced and if restrictions are eased. We found that for vaccination to make an impact on this predicted surge, its scale-up has to be rapid; otherwise, a large fraction of people in the community will be infected before they get vaccinated. For substantial impact, high levels of vaccine coverage also need to be achieved, as the gains from vaccination did not increase linearly with coverage and were most pronounced starting at 70% coverage. While beneficial, intermediate coverage will not be sufficient to allow return of social contacts to normalcy in the near term and will not spare substantial covid-19 hospitalizations and deaths. Reaching, for example, 40% coverage by the end of the year -which is the more likely scenario in Lebanon given current vaccination pace -means a roll-out that is too slow against fast moving incidence, resulting in a more limited impact of vaccination. Our analyses also highlight the importance of keeping social distancing restrictions as long as possible during vaccine scale-up to maximize the benefits of vaccination. Therefore, vaccination will flatten the epidemic curve of a predicted summer surge in Lebanon but will not prevent it or delay it, unless scale-up to reach high vaccination coverage (at least 70%) will be very rapid -for example achieved in August rather than December 2021, while maintaining social distancing restriction during roll-out ( Figure S2 ). At the pace vaccination is taking place, where only 3.3% of the population have completed two doses of the vaccine in three months, it is unlikely that vaccination will have a noticeable impact on a potential surge in the near future. The likelihood of this surge is high given the circulation worldwide of highly infectious SARS-CoV-2 variants of concern [13, 17, 38] , the poor borders control implemented in the country, and absence of systematic viral genome sequencing or multiplex PCR variant screening to track the introduction of variants. Social contacts are also expected to increase over the summer months with the arrival of large numbers of Lebanese expatriates and other tourists. This is further motivated by a false sense of security due to the recent marked decline in the number of confirmed cases, as immunity had been building up mainly because of ongoing incidence [1] . There is therefore an urgent need to rapidly scale-up vaccination to avoid a surge that the country, with its multiple crises and limited health-care capacity, is largely unprepared for. Multiple strategies can be followed to overcome supply, administration, and demand challenges to speed up community immunization in Lebanon. While some of the supply bottlenecks result from the high global demand on manufacturers and the prioritization of high-income countries for delivery of reserved doses due to early agreements [39] , the government may adopt simple strategies to maximize the use of available doses. The recent decision by the Lebanese authorities to delay administration of the second dose is one good example [40] . Promoting the WHO-approved extraction of one additional jab per vial of the BNT162b2 vaccine [41, 42] and reinforcing prioritization of available doses for 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 30, 2021. ; https://doi.org/10.1101/2021.05.27.21257937 doi: medRxiv preprint individuals at higher risk of developing severe disease [18, 35] are other suggestions that would capitalize on current supply and maximize its impact. Efforts also need to be put to expand the vaccine distribution campaign to cover all geographical and remote areas, and reach out to the most underprivileged communities. Currently, vaccination is being deployed only through major hospitals. Expanding vaccination sites to include primary health care centers and other existing health care infrastructures, as well as specifically establishing sites such as drive-through vaccination facilities or vaccination centers in sports arenas could be considered. Particular consideration needs to be made to reach the most vulnerable, including refugees, migrant workers, and other underprivileged communities impact. We assumed that 20% had been exposed to the infection on January 1,2021, a sensible estimate based on triangulation of available local data, including the cumulative number of confirmed infections and deaths and routine clinical antibody testing data. We assumed that both natural and vaccineinduced immunity last for one year, as suggested by available data [21] [22] [23] [24] . However, these two important parameters remain unknown. If they prove to be less than the assumed one year, the impact of the vaccine will be reduced. The model did not prioritize vaccination based on age, since younger individuals in the workforce are currently being vaccinated through the private sector. Given that the highest risk elderly population (>70 years) did receive the vaccine first, our estimates for the impact of vaccination are conservative and the number of averted disease cases and deaths may be higher than the ones reported in our study. Finally, analyses were conducted at a national level, whereas vaccine impact will likely differ between geographic locations with different epidemic transmission dynamics, different levels of compliance to preventive measures, and different vaccine uptake levels. 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. In conclusion, amidst all ongoing crises in the country, Lebanon cannot afford the burden of yet another large epidemic wave. Given the dire economic situation, strict lockdowns and extended closures may also be difficult to sustain. At this stage, scaling up vaccination offers the only hope to go back to normalcy with the least deaths and burden on the health care system possible, and with the least damage to the economy in the short and longer term. Our findings reiterate that there is every virtue in speeding up vaccine roll-out and increasing coverage to avoid the potentially devastating consequences of a predicted and likely epidemic surge. With only 6.8 million inhabitants, this goal should be feasible. However, concerted efforts need to be put to overcome local challenges. In the meantime, public messages need to promote adherence to social distancing measures until we approach high levels of vaccine coverage. Text S1A: Model structure Text S1B: Model equations Text S1C: Model parameters Table S1 : Definitions of population variables and symbols used in the model Table S2 . Model assumptions in terms of parameter values Figure S1 : Analysis of the effect of uncertainty around the transmissibility (R0) of the new variants introduced on 15 April 2021 Figure S2 : Impact of SARS-CoV-2 vaccination on number of A) new infections, B) new severe and critical disease cases, and C) new deaths in Lebanon for vaccine coverage of 80% by 31 August 2021 Author Contributions: GRM co-conceptualized the study, co-designed the methodology, and wrote the first draft of the manuscript. MM constructed and parameterized the mathematical model, and conducted the mathematical modeling analyses. FE-J co-conceptualized the study. LJA-R coconceptualized the study and led the design of the methodology. All authors contributed to the discussion and interpretation of the results, and writing of the manuscript. All authors have read and approved the final manuscript. 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. 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. 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. 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. Key Indicators Based on Publicly Available Information United Nations Department of Economic and Social Affairs Population Dynamics, The 2019 Revision of World Population Prospects Regional Refugee Response -4 Vulnerability of Syrian refugees in Lebanon to COVID-19: quantitative insights Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine. The New England journal of medicine Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK. The Lancet Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine COVID-19 Vaccine Frequently Asked Questions Safety and efficacy of an rAd26 and rAd5 vector-based heterologous primeboost COVID-19 vaccine: an interim analysis of a randomised controlled phase 3 trial in Russia. The Lancet Early rate reductions of SARS-CoV-2 infection and COVID-19 in BNT162b2 vaccine recipients Impact of the COVID-19 Vaccine on Asymptomatic Infection Among Patients Undergoing Pre-Procedural COVID-19 Molecular Screening BNT162b2 mRNA Covid-19 Vaccine in a Nationwide Mass Vaccination Setting Effectiveness of the BNT162b2 Covid-19 Vaccine against the B.1.1.7 and B.1.351 Variants Estimated transmissibility and impact of SARS-CoV-2 lineage B.1.1.7 in England Risk of mortality in patients infected with SARS-CoV-2 variant of concern 202012/1: matched cohort study New SARS-CoV-2 Variants -Clinical, Public Health, and Vaccine Implications Epidemiological Impact of SARS-CoV-2 Vaccination: Mathematical Modeling Analyses. Vaccines (Basel) Epidemiological Differences in the Impact of COVID-19 Vaccination in the United States and China. Vaccines (Basel) Epidemiological impact of prioritising SARS-CoV-2 vaccination by antibody status: mathematical modelling analyses Assessment of the risk of SARS-CoV-2 reinfection in an intense reexposure setting Robust neutralizing antibodies to SARS-CoV-2 infection persist for months Humoral response and PCR positivity in patients with COVID-19 in the New York City region, USA: an observational study Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection SARS-CoV-2 antibody-positivity protects against reinfection for at least seven months with 95% efficacy. EClinicalMedicine The Language of Technical Computing Estimation of the basic reproduction number, average incubation time, asymptomatic infection rate, and case fatality rate for COVID-19: Meta-analysis and sensitivity analysis COVID-19 parameter estimates: basic reproduction number Tracking Changes in SARS-CoV-2 Spike: Evidence that D614G Increases Infectivity of the COVID-19 Virus Making Sense of Mutation: What D614G Means for the COVID-19 Pandemic Remains Unclear Genetic Variants of SARS-CoV-2-What Do They Mean? COVID-19 vaccine coverage in health-care workers in England and effectiveness of BNT162b2 mRNA vaccine against infection (SIREN): a prospective, multicentre, cohort study World Health Organization, Clinical management of COVID-19 World Health Organization, International guidelines for certification and classification (coding) of COVID-19 as cause of death Model-informed COVID-19 vaccine prioritization strategies by age and serostatus A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code Uncertainty and sensitivity analysis of the basic reproductive rate -Tuberculosis as an example India's COVID-19 emergency Challenges in ensuring global access to COVID-19 vaccines: production, affordability, allocation, and deployment Accelerate COVID-19 Vaccine Rollout by Delaying the Second Dose of mRNA Vaccines Overview of the implementation of COVID-19 vaccination strategies and vaccine deployment plans in the EU/EEA -EN. . 2021, European Centre for Disease Prevention and Control Why are there extra doses of vaccine in the vaccine vial? The authors are grateful for support provided by the Biomedical Research Program