Wrestling with Reff in Northern California

In my last blog post I noted that Northern California was performing worse than Southern California in terms of COVID-19 reproduction numbers. Two days later the situation remains the same. One of the features of the data in some Northern California counties like Alameda and Placer is the data seems to be published to Johns Hopkins every other day or so. So on August 13, Alameda recorded 10 new active cases. On August 12, 481 cases and on August 11, no new cases.........

Nevertheless our method still picks up an effective reproduction number (Reff) in Alameda on August 13 of 2.7, albeit with wide confidence intervals (the grey shading around the single dot plot points).

We generally see the Reff value, as a leading indicator, reflected in the daily new cases 10 to 14 days later. So, as a rough rule of thumb, if a place has 100 cases per day and an Reff value of 2.0 then you can expect a doubling of new case numbers in 10-14 days. That's why Alameda's Reff of 2.7 is a warning signal which demands changes in behaviors and changes in policies. An Reff value that high is going to lead to exponential growth in daily new cases, particularly since Alameda has not had its reproduction number consistently under 1.0 since 22 April.

Other Northern Californian counties where Reff is 1.5 or higher on 14 August include Placer, Contra Costa, Santa Mateo, Santa Cruz, Merced, Madera and Kings. Together their performance has pushed the State of California's Reff above 1.0 (remember above 1.0 and the disease grows like a pandemic and below it the disease falls away). The map below taken from our https://covid19.closeassociate.com/region/us-ca on 14 August highlights the problem. It's time to act Northern Californians!

Finally, note that our method for calculating Reff is published and is entirely in the public domain. Click here for access to the key technical report: https://www.covid19-blog.closeassociate.com/post/improvements-to-the-sir-b-mathematical-methods

Next blog we will look at the hotspots in the US Northeast and then turn our attention to Europe which has had high Reff values in our modeller with increasing daily case numbers for more than two weeks.