Estimating the Reproduction Number (Daily Updates)

The effective reproduction number of SARS-CoV-2 is the number of people that are infected by a single infected person.

I estimate the effective reproduction number (Rt) of SARS-CoV-2 the virus that causes COVID-19 for various countries on a daily basis.  Links to these reports are below.  At the bottom I link to the methodology too.

South Africa

Reproduction number estimates for South Africa done using case data by specimen received date made available by the NICD, and by reported date of hospital admissions. An allowance for public holidays and late reporting of cases are made to the data. Estimates are available at a national, provincial and district municipality level in the linked report.

Effective Reproduction Number for South Africa

Australia

I recently added calculation for Australia.  I show R estimated from cases, but include time series on hospitalisation, percentage testing positive and deaths.

Canada

Calculations for Canada are based on cases and deaths by reporting date.  This means that where reporting delays are changing estimates for Canada can be volatile. Estimates are available at a national and provincial level.

Effective Reproduction Number of Canada

United Kingdom

UK estimates are based on cases by date of the test.  There are reporting delays and at present these are allowed for by excluding the last 7 days of data due to these delays.

UK makes lots of data available and estimates are made based on cases, admissions and deaths at various geographies, such as nation, regions and upper tier local authorities.

Worldwide Estimates

Based on cases and deaths as collated by Our World In Data I estimate the effecitve reproduction number for countries worldwide.  These estimates are as good as the data provided and for some countries the data is not of great quality.  Caution is required when interpreting results from this report.

Reproduction number estimated from cases

Methodology

The generation interval is the time on average it takes for one person to infect the next.  There are studies that have established the distribution of this time interval.  Using this information together with number of cases, admissions or deaths one can estimate Rt.  More details on the methodology are available here.  Note the main assumption is that the testing and reporting of cases, admission or deaths is stable.  The method will not be accurate if these are changing.  This is also true of the time from infection and the data capturing date (case, admission or death) is variable.