About

City level estimates of the time-varying reproduction number of Covid-19 produced using EpiNow2 based on data from brazil.io. Regional estimates for Brazil (along with a national estimate) are available here. Detail of the method used is given here, though the estimates shown here were derived using an approximate approach (direct de-convolution of observed cases) rather than the exact method used on epiforecasts.io and therefore should be considered indicative. The code and data supporting these estimates is available here. Estimates are available in a summarised form here.

Data sources

  • Case data is extracted from: brazil.io

  • Estimates were only produced for cities with at least 1000 cases, and 90 days of data, since the start of the data extraction. All time series were truncated by 3 days to account for delayed reporting (the delay from original report to reporting in this dataset not the adjusted for delay from onset to original report). Estimates are only shown from the date when a given region reported at least 10 cases on a single day.

  • Data on the delay from onset to report were extracted for Brazil from here: github.com/beoutbreakprepared/nCoV2019

  • The delay from onset to report was assumed to be log normal and fit using a subsampled bootstrap in order to better represent the uncertainty in the distribution at any time point. See here for details.

  • Estimates for the incubation period and the generation time were based on those used here: epiforecasts.io

Regional Breakdown

Summary of latest reproduction number and confirmed case count estimates by date of infection


Figure 1: Confirmed cases with date of infection on the 2020-11-21 and the time-varying estimate of the effective reproduction number (lightest ribbon = 90% credible interval; darker ribbon = the 50% credible interval, darkest ribbon = 20% credible interval). Regions are ordered by the number of expected daily confirmed cases and shaded based on the expected change in daily confirmedcases. The horizontal dotted line indicates the target value of 1 for the effective reproduction no. required for control and a single case required for elimination. Uncertainty has been curtailed to a maximum of ten times the maximum number of reported cases for plotting purposes.

Reproduction numbers over time in the six regions expected to have the most new confirmed cases


Figure 2: Time-varying estimate of the effective reproduction number (lightest ribbon = 90% credible interval; darker ribbon = the 50% credible interval, darkest ribbon = 20% credible interval) in the regions expected to have the highest number of new confirmed cases. Estimates from existing data are shown up to the 2020-11-21 from when forecasts are shown. These should be considered indicative only. Estimates based on partial data have been adjusted for right truncation of infections. The vertical dashed line indicates the date of report generation.

Confirmed cases and their estimated date of infection in the six regions expected to have the most new confirmed cases


Figure 3: Confirmed cases by date of report (bars) and their estimated date of infection (lightest ribbon = 90% credible interval; darker ribbon = the 50% credible interval, darkest ribbon = 20% credible interval) in the regions expected to have the highest number of new confirmed cases. Estimates from existing data are shown up to the 2020-11-21 from when forecasts are shown. These should be considered indicative only. Estimates based on partial data have been adjusted for right truncation of infections. The vertical dashed line indicates the date of report generation. Uncertainty has been curtailed to a maximum of ten times the maximum number of reported cases for plotting purposes.

Confirmed cases and their estimated date of report in the six regions expected to have the most new confirmed cases


Figure 4: Confirmed cases by date of report (bars) and their estimated date of report (lightest ribbon = 90% credible interval; darker ribbon = the 50% credible interval, darkest ribbon = 20% credible interval) in the regions expected to have the highest number of new confirmed cases. Estimates from existing data are shown up to the 2020-11-21 from when forecasts are shown. These should be considered indicative only. Estimates based on partial data have been adjusted for right truncation of infections. The vertical dashed line indicates the date of report generation. Uncertainty has been curtailed to a maximum of ten times the maximum number of reported cases for plotting purposes.

Latest estimates (as of the 2020-11-21)

Table 1: Latest estimates (as of the 2020-11-21) of the number of confirmed cases by date of infection, the effective reproduction number, the rate of growth, and the doubling time (when negative this corresponds to the halving time) in each region. The median and 90% credible interval is shown.

 

Produced by EpiForecasts and the CMMID