A simple method to quantify country-specific effects of COVID-19 containment measures

Morten Gram Pedersen; Matteo Meneghini

To parameterize the SIQR model, we fitted analytically derived expressions for Q + R (see Methods) to the data of total COVID-19 cases in seven different European countries, British Columbia (Canada), New South Wales (Australia) and New York City (USA), from the first date when the number of confirmed cases of the single regions passed above 50 through April 2, 2020. (Fig. 1 ). The underlying model was assumed to have a piecewise-constant transmission rate β, taking one value before a given date and another value after that date (date-of-change in the following). This assumption on β translates into a piecewise constant growth rate for the number of infectious individuals I (see Methods). Table 1 shows estimated values for the four free, identifiable parameters of the model: date-of-change (T c ), growth rates before (τ −1 1 ) and after (τ −1 2 ) the date-of-change, and the combined parameter ηI 0 of quarantine rate (η) times the initial value of the number of non-identified positive cases (I 0 ).