Extended Data Fig. 7: Sensitivity of estimated averted/delayed infections to the choice of γ and σ in an SIR/SEIR framework.
From: The effect of large-scale anti-contagion policies on the COVID-19 pandemic
The sensitivity of total averted/delayed cases presented in Fig. 4 to alternative modelling assumptions. We compute total cases across the respective final days in our samples for the six countries presented in our analysis. The figure displays how these totals vary with eight values of γ (0.05–0.4) and four values of σ (0.2, 0.33, 0.5, ∞), where the final value of σ (∞) corresponds to the SIR model. a, The simulated total number of infections under no policy. b, Same as in a, but using actual policies. c, The difference between a and b, which is the total number of averted/delayed infections. d, Same as c, but on a logarithmic scale similar to Fig. 4 (a–c are on a linear scale, trimmed to show details). Figure 4 uses γ = 0.079, which we calculate using empirical recovery/death rates in countries for which we observed them (China and South Korea; see Methods). If we assume a 14-day delay between infected individuals becoming non-infectious and being reported as ‘recovered’ in the data, we would calculate γ = 0.18. Figure 4 assumes σ = ∞.
