Good morning Mission. And welcome to Virus Village, your daily data dump.
It gets kind of foggy in Virus Village. This is San Francisco after all; even the numbers come shrouded in a thick mist.
We changed the headline. Why? “New cases” does not quite cut it. The cases referred to like today’s 90 aren’t “new” in the sense they came from newly performed tests. They’re actually old test results, newly added to the overall total. How old? We don’t know.
The time lag between when a test is performed and when it’s analyzed and reported is a key figure in understanding whether the virus is spreading, stabilizing or fading. The interval is also key in gaining control over the virus (by enabling contact tracing). Yet DPH does not report the length of delays or reason for them.
For more on the fog of the City’s testing numbers, be sure to read Joe and Lydia’s piece today on the difference between San Francisco’s reported testing “capacity” and what actually exists.
We get our local figures from the City’s Department of Public Health. DPH has not updated hospital figures since July 21 because “on July 22, there were significant changes to the state’s hospital data due to new federal reporting requirements” — to satisfy Dr. Donald Strangelove .
Now onto today’s numbers.
HiGeorge, a data visualization startup, developed some new visualizations for Mission Local, which we will be using and fine-tuning in the days to come.
The Mission District added 14 cases today raising our total to 872.
Instead of looking at the number of cases added to the totals each day, DPH suggests a better idea is to look at the seven-day rolling average. The “rolling average is the average of new cases for a particular day and the prior six days, which shows the trend of new cases (smoothing out daily fluctuations).” Sounds good.
But don’t forget the time delay between the test and when a case reported. As a reader pointed out, the average numbers for the prior six days keep growing as more old tests are added. What this means is the trend line compares last week’s adjusted numbers with today’s unadjusted numbers. The number is falling, but not as steeply.
The average number of new cases per day per 100,00 residents is 8.4 which has fallen from last week, but still in the red zone. High Alert.
I hope to be soon talking with DPH about this and other data issues that have surfaced.
Don’t freak out on the positivity rate. Look at the extremely low number of tests on which the percentage is based. This number will also change in time.
There have been no new deaths in the City since July 18.
Last week, after performing some second-grade math, I reported that the infection fatality rate for COVID-19 in San Francisco would be “around the same” as the national case fatality rate for the flu. My bad. Dr. Carl Bergstrom from the University of Washington graciously, and patiently, explained I was comparing apples to oranges, i.e. the infection fatality rate for COVID to the case fatality rate for the flu, both around .1 percent.
But infections and cases are not the same. Infection fatality rates are based on the many people who get infected, but never get tested or go to a hospital (because they are asymptomatic or mildly symptomatic). Cases are established, positive cases.
The error I made has consequences – social and political because it portrays the fatality rate for COVID-19 as comparable to that of the flu, which is not true. That wasn’t my intention. I apologize. Thanks Dr. Bergstrom.
The R number has been stable for the past week or so, indicating that the virus continues to spread exponentially.
Approximately 75 percent of the cases reported in San Francisco come from people below the age of 50. Almost 93 percent of the deaths were people over 50.