COVID-19 Interest Group
March 29, 2020
Report to the club:
CIG Bulletin #2
This is bulletin #2 from the CIG (Covid-19 Interest Group of Berkeley Rotary). We are an open group, meeting on Zoom at 4:00 PM on Mondays. If you would like to tune in and participate contact Tina Etcheverry who is hosting the meeting on Zoom. This bulletin includes: a contribution from Arlin Peters who is tracking and helping to interpret the data on the growth of the pandemic.
Item 1. Tracking and Analysis of COVID-19. By Arlin Peters
The data, information and estimation of future cases presented below is based on public data and my own personal mathematical analysis of it. It has not been peer reviewed and has not been officially endorsed by the Berkeley Rotary Club. It is intended to help you understand the Covid-19 situation in the Greater Bay Area and its comparison to a few other locations. I have had a life-long interest in mathematics and spent years professionally working with data and models. If you have any questions, please call me at 510-501-0298. If you are interested in a copy of my Excel spreadsheet for the analysis of GBA data, please e-mail me at arcapeters@comcast.
A summary of cumulative confirmed Covid case data as of the evening of March 26th and an analysis is given below. These are Covid-19 cases confirmed by a positive test result, and are the total number as of the date. It is thought that due to the lack of testing especially for mild and moderate cases, the actual number of cases on a particular day is 7 to 10 times the number of confirmed cases.
For various locations, the number of cases per million population is shown. The Greater Bay Area (GBA 7 counties w 7.1 million people) has 1,418 cases or 200 cases per million (cpm), on the low side of many locations. New York City is among the hardest hit areas at 2719 cpm, more than all of Italy, but smaller parts of Italy like Lombardy probably have a larger cpm than NYC.
Doubling time (DT) is crucial to understanding the exponential growth of the Covid epidemic. It is based on past data and is the number of days it has taken for the number of cases to double. DT is a measure of how well the public is using good hygiene and social behavior to slow the transmission of Covid-19 cases.
Smaller DT’s are bad because the number of cases doubles in fewer days compared to larger DT’s. DT calculated from data over the past 10-14 days can be compared to the DT over the last 4 days to see if public transmission rates are decreasing. For example, Italy’s 8.6 day DT over the past 4 days is a big improvement on its 2 week DT of 5.1. The GBA DT has improved from 4.1 days to 4.5 days over the past 4 days. With a 4.5 day DT, I estimate the GBA’s number of cumulative confirmed cases will rise to about 12,000 in two weeks. Improving DT’s over the next two weeks will reduce the number of cases.
I estimate from the growth of the number of confirmed cases and several assumptions, that the number of contagious cases on Mar 26 was 0.3% of the GBA’s 7.1 million population. The risk of transmission is low, but it will increase as the 4.5 day DT means more infected and contagious people in the future. I estimate that in a week from Mar 26, 0.7% of the GBA population will be contagious.
COVID Cases Summary Data 3/26/20 COVID Confirmed Long term ~2 wks Population Confirmed Cases per Doubling DT DT over millions
last 4 days
World 7,770 551,300 71 DT is # of days for # of cases to double China 1,410 81,667 58 3/25 data USA 331 86,012 260 2.3 increasing 3.4 Germany 84 37,273 444 3/25 data UK 68 9,640 142 3/25 data France 65 25,600 394 3/25 data Italy 60 80,589 1343 5.1 increasing 8.6 Spain 47 57,786 1229 Hungary 10 226 23 3/25 data Switzerland 8.7 10,897 1260 3/25 data NY State 20 39,140 1957 NYC 8.5 23,112 2719 1.5 increasing 3.3 NJ 9.4 6,876 731 Washington 7.8 3,207 411 Calif 40 4,033 101 GBA 7.1 1,418 200 4.1 increasing 4.5 GBA Estimated Cumulative cases in 2 weeks = 12,000 GBA Estimated currently contagious = 0.3% of population Alameda & 2.8 309 110 3.8 increasing 4.4 Contra Costa Co Limited poor quality data