The MIDAS Webinar Series presents the research of MIDAS members and is open to the public.
Date/Time: Friday, May 1, 12:00–1:00 PM EDT
Topic: Modeling Possible Long-Term Intervention Strategies for COVID-19
Speaker: Dr. Erin Mordecai, Assistant Professor of Biology, Stanford University
My research focuses on the ecology of infectious diseases. I aim to understand how climate, species interactions, and climate change affect the dynamics of infectious diseases in human and natural ecosystems. This research combines mathematical modeling and empirical work. I completed my PhD in 2012 at the University of California, Santa Barbara, in Ecology, Evolution, and Marine Biology. Afterwards, I received a 2-year NSF postdoctoral fellowship in the Intersection of Engineering, Physics, Mathematics, and Biology at the University of North Carolina, Chapel Hill, and North Carolina State University. I have been at Stanford since January 2015.
Abstract: With isolation and social distancing measures in place across the U.S., and hospitalizations and deaths beginning to stabilize in some regions, the most urgent question now is: what set of exit strategies will allow a return to some form of normal collective life without triggering the resurgence of a large epidemic?
We created a SEIR model for COVID-19 epidemic dynamics that includes compartments for susceptible; exposed; infectious but pre-symptomatic; asymptomatic; mildly symptomatic; severely symptomatic; hospitalized; deceased; and recovered individuals. The model captures key intervals between exposure, symptom onset, hospitalization, and death, as well as transmission pathways, including asymptomatic infectious individuals, and it is parameterized using MIDAS panel values.
First, we investigated possible scenarios for non-pharmaceutical interventions, including first and second waves of social distancing, which varied in start date, duration, and intensity, as well as testing, quarantine, and contact tracing scenarios. By allowing the transmission parameter (beta) to vary over time, we can implement a wide variety of social distancing strategies.
We demonstrate that even with strong social distancing (a 60% reduction in social contacts) for five months or more, resurgence is highly likely once the intervention is lifted unless accompanied by other measures, such as adaptive social distancing (the “light switch” method), or increased testing and quarantining of symptomatic individuals.
Next, we fitted the model to daily COVID-19 death data reported by Santa Clara and other counties in California. We show that Santa Clara County has an estimated R0 = 3.39 (95% CI: 2.83–4.22), that a 71% reduction in social contact, on average, would reduce R0 to 1, and that restrictions on non-essential services as the sole intervention would need to remain in place for approximately five months to end the epidemic (without accounting for new outbreaks triggered by imported cases).
We are now working with public health departments in the San Francisco Bay Area to fit the model to municipal-level COVID-19 mortality data to help inform intervention decisions.