Abstract
In the context of the COVID-19 pandemic, governments worldwide face the challenge of designing tailored measures of epidemic control to provide reliable health protection while allowing societal and economic activity. In this paper, we propose an extension of the epidemiological SEIR model to enable a detailed analysis of commonly discussed tailored measures of epidemic control—among them group-specific protection and the use of tracing apps. We introduce groups into the SEIR model that may differ both in their underlying parameters as well as in their behavioral response to public health interventions. Moreover, we allow for different infectiousness parameters within and across groups, different asymptomatic, hospitalization, and lethality rates, as well as different take-up rates of tracing apps. We then examine predictions from these models for a variety of scenarios. Our results visualize the sharp trade-offs between different goals of epidemic control, namely a low death toll, avoiding overload of the health system, and a short duration of the epidemic. We show that a combination of tailored mechanisms, e.g., the protection of vulnerable groups together with a “trace & isolate” approach, can be effective in preventing a high death toll. Protection of vulnerable groups without further measures requires unrealistically strict isolation. A key insight is that high compliance is critical for the effectiveness of a “trace & isolate” approach. Our model allows to analyze the interplay of group-specific social distancing and tracing also beyond our case study in scenarios with a large number of groups reflecting, e.g., sectoral, regional, or age differentiation and group-specific behavioural responses.
Highlights
In the context of the COVID-19 pandemic, governments worldwide face the challenge of designing tailored measures of epidemic control to provide reliable health protection while allowing societal and economic activity
Scenario USD2 assumes thorough social distancing leading to an effective reproduction rate R ≈ 1
If we add tracing onto a scenario with uniform or group-specific social distancing (USD1-Trace and GSD3Trace; see Tables 9 and 10 ), we see similar patterns, but in a context where there is substantially less strain on the health system compared to the benchmark scenarios
Summary
This section contains the results and discussion of our simulation outcomes. In order to first draw the bigger picture, we present two selected scenarios that illustrate extreme cases, the scenarios B2 and USD2 (see Table 4). In USD2 and USD3, social distancing is effective to suppress the epidemic with the effect that the number of deaths in the first 500 days is comparatively low In these scenarios the vast majority of the population (up to 99%) is still susceptible after 500 days. (Note that all our uniform social distancing scenarios are based on benchmark B3, which assumes a higher asymptomatic rate among the group with low vulnerability. GSD1 builds on benchmark B2, but drastically reduces all interactions involving highly vulnerable individuals, i.e., contact rates to βHH = βLH = βHL = 0.04 This roughly corresponds to an effective reproduction number of R ≈ 0.6 generated from these interactions only. Compliance Deaths in first 500 days ICU capacity exceeded on Duration Susceptible after 500 days
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
