R_{t}, effective reproduction number, is a measure of transmission potential of a disease. It's the average number of people one infected individual ends up infecting.

If R_{t} is above 1 the epidemic becomes more widespread and when it gets below 1 the disease starts to wane in the population.

Data through

Arrange by:

- Only states with > 250 cases and averaging at least 10 cases for 14 days included.
- Dashboard updates every ~2-3 days.
- Uses model kindly shared by rt.live to calculate R
_{t}

**R _{t}** is easy to read by anyone because safe values are < 1 , the more > 1, the faster the spread.

**R _{t}** can be used to evaluate efficacy of outbreak control measures (and consequently respond fast by taking more or less measures). It also has information that can be used to estimate total outbreak size.

The other major metric to characterize the speed of the outbreak is exponential growth, ** r** , or its inversely related "doubling time". These can be inferred directly from the log(cases) graph. But doubling time varies hugely and is not in an easy to analyse range. For example, as the total cases plateau the doubling time explodes to ∞.

**R _{t}** is related to doubling time by the generation interval. Using oversimplified assumptions, one can arrive at approx. thumb rule to relate

Sustained R_{t} | New cases in 2 weeks ^{*} | 2x growth days | 10x growth days |
---|---|---|---|

3 | 164 | 3.5 | 11.5 |

2.5 | 103 | 4.2 | 13.8 |

2 | 58 | 5.5 | 18.2 |

1.8 | 45 | 6.5 | 21.5 |

1.5 | 28 | 9.4 | 31.2 |

1.2 | 16 | 20.9 | 69.5 |

^{*} assuming 10 cases at start

- Model does not incorporate onset delay, serial interval as a distribution.
- Daily cases data from covid19india.org