Speed and strength of an epidemic intervention

Jonathan Dushoff; Sang Woo Park

While both speed-and strength-based frameworks can give the same conclusion about the outcome of an intervention, sometimes one provides a clearer understanding of a given measure. For example, we expect the speed-based framework to be clearer for characterizing newly invading pathogens: when an epidemic is growing exponentially, the reproductive number cannot be estimated with confidence (Weitz and Dushoff, 2015) , especially when there is large uncertainty in the shape of the generation-interval distribution (Park et al., 2020) . Conversely, we expect the strength-based framework to be clearer for evaluating established pathogens (based on the effective proportion of the population susceptible). For interventions, we expect the speed-based framework to be clearer for evaluating intervention strategies that work at the individual level, like test-and-treat for HIV (Granich et al., 2009) , or contact-tracing and quarantine for COVID-19 (Hellewell et al., 2020) ; we expect the strength-based framework to be clearer for intervention strategies that seek to reduce the overall transmission at the population level, like condom use. In other cases, such as real-time rollout of vaccines during an outbreak, both speed and strength approaches might be similarly uncertain because the result depends both on the speed of the rollout and the (strength-like) final coverage (Shah et al., 2018) .