Monterey County Department of Public Health has revised their Covid-19 website, and they now provide two sorting of the cases-by-day information: one by when the case was first reported, and another by the day of onset of first symptoms. The second changes a lot from day to day. Here’s the data by date first reported:
The vertical axis is linear here. I had been using a log scale when it looked like the growth in cases was close to exponential with time, but that hasn’t been true for a while. I will switch back to a log scale if the present growth in cases seems to be becoming exponential. The dotted orange line is a least-squares straight line fit to the data. As long as the orange 7-day moving average is close to it, then the growth is linear. It looks like the last week has been well above linear.
If we look at dates of symptom onset, things don’t look as worrisome, but that data always shows low case incidence in the near past:
The DPH also now provides information about daily tests performed. Using that information, and assuming that the positive results are reported on the same day the test is performed (and the county is not clear about what date they used for when a test is performed; is it the data the sample is taken or the date when the results come in), we can calculate the positive tests to test performed. Because of possible varying reporting timings, the 7-day moving average of this is likely to be more meaningful than the daily data:
You can see that the positivity ratio dropped a lot about three weeks ago. That coincides with an increase in the number of tests performed:
Are the cases in Monterey County increasing because of more testing? Are the additional positives cases that would not have been detected at all a few weeks ago? Or are the new cases indications that the virus has begun to spread more rapidly again? Without additional information about the testing, it’s impossible to answer that question from the daily positivity ratios. It’s appealing to think that a low positivity ratio means that there is a low incidence of Covid-19 in the population, but without knowing who is being tested and why, we can’t jump to that conclusion.
In mid April, the CDPH decided to count in the number of tests performed each and every test performed on each person. That number forms the denominator of the positivity ratio. The numerator is the number of people who tested positive, not the number of times each person tested positive. If the increased number of tests performed are the result of routine tests on medical workers, a low positivity ratio doesn’t say much about the incidence of the virus in the general population. Likewise, when a recovering patient is repeatedly tested for the disease, all those repeated tests count in the denominator, but none of them, even the positive ones, count in the numerator.
So it’s hard for me to draw firm conclusions from the above numbers, but I think the recent uptick bears careful watching.
Here’s some context: according to the LA Times Covid-19 page, Monterey Country’s case doubling rate is 16.8 days, which makes it one of the fastest-doubling counties in the state.
Leave a Reply