A modified SEIR model to predict the COVID-19 outbreak in Spain: simulating control scenarios and multi-scale epidemics

Leonardo R Lopez; Xavier Rodo

Epidemiological reports for the current situation of COVID-19 out of China indicate Europe is one of the centers of the pandemic, and that nearly 200 countries/regions worldwide have reported cases [1] . Particularly in Italy and Spain a large outbreak of SARS-CoV-2 infection is underway. Given the lack of previous exposure to this virus, future predictions of both the levels of virus spread in the global naive population and the acquired herd immunity cannot be properly anticipated. Data on the last similar pandemic one century ago remains limited for obvious reasons, despite aftermath reports exist [2] . Policymakers in the most affected countries face a difficult situation when trying to balance between draconian public health actions and keeping economy alive, as impact on health of severe economic measures is also well-known [3] . The fact that the 1918 Spanish flu pandemic had a second deadly epidemic wave, presumably caused by mutations in the H1N 1 viral strain [4] , has also stimulated a vivid debate on whether actions to take now should take into account this uncertain future. Under this scenario, optimal action on the COVID-19 pandemic is hard to fathom. And more so, for the extent of pre-symptomatic and asymptomatic infections is not narrowly constrained (e.g. up to 86 percent [5] ). In China, regions as well as local governments, including Hubei, tightened preventive measures to curb the spreading of COVID-19 since Jan. 2020 [6] . Many cities in Hubei province were locked down and many measures implemented, such as tracing close contacts, quarantining infected cases, promoting social consensus on self-protection (e.g. wearing face mask in public area, minimum social distances). However, in other areas, the extent and efficacy of the so-called confinement or self isolation is doubtful, and facebook data on mobile location and movement showed yet massive people displacements under semi-confinement measures. At a time when success of large-scale social distancing interventions is critical, access to accurate information to ascertain mobility is lacking [7] . Similarly, credible serology tests that could show whether someone has had the infection and recovered are not yet massively deployed, thereby assessment of SARS-CoV-2 prevalence in the population is not established [8] .