Deconstructing CA Data

July 12, 2020

Note to readers on social media platforms: the post you’re reading does not have all the tables and charts that pertain to today’s commentary. To see them all, go to the post on my website, zorgi.me .

There is an awful lot of data floating around during the pandemic, and I know I have a tendency to show it all and let you decide what to look at and what to ignore. Today I thought I’d concentrate on one state, California, and deconstruct the 9-metric Vital Indicators chart [ST2].

There’s a lot here, and it may be more comprehensible if we take it apart a bit. Let’s start with daily tests, daily cases, and the positivity rate [ST2a].

This chart is just for a one month period, and it gives us a good idea why there’s so much concern. Daily tests during the past 30 days have increased b y 74%, so we would expect the positivity rate to go down. Instead, it’s gone up by 70%, from 4.7% to 8.0%. Daily cases have increased by almost 200%, from just under 3,000 to over 8,000 a day.

ST2b shows another troubling aspect of the pandemic in CA: a 74% increase in daily patients, and a nearly 50% increase in daily ICU patients. While the HUR is still comparatively low at 10%, the IUR is starting to climb rapidly, and is now almost 30%.

ST2c shows the alarming increase in daily fatalities, from 66 a day to 90 now, a 38% increase. Ordinarily, we’d expect the case fatality rate (CFR) to keep decreasing, because we art getting better at keeping people alive. It has gone down, from 3% to 1.7%, so that’s a welcome sign. But keep in mind that fatalities are a lagging indicator, following cases by 3 to 4 weeks. The CFR in this chart is based on fatalities today and cases 2 weeks prior, accounting for some of that lag.

I’ve added another ratio, which I’m abbreviating as the IFR, representing fatalities today divided by ICU patients one week prior. Note that this ratio fairly closely mirrors the ICU daily patient trend. I call it a “ratio” instead of a “rate” because I don’t want to imply at all that this is the number of ICU patients who die after they enter the ICU. It’s simply an expression of two metrics, and the relationship between them isn’t clear.

Finally, ST2d shows the Rt value over time as well as the Zorgi Score over the past month. Some readers have asked about how this looked, and I modified my program to run it for any date. While the Rt is still above 2.0, indicating a spreading virus, the Zorgi Score fairly closely mirrors the increase in all the metrics that have caused concern in CA over the past month. On June 11, it was a comparatively low 7.3, but now has almost doubled to 13.7. While that’s not the worst among the states I track, it’s not a score that inspires a lot of confidence.

ZS0 shows the scores for all 7 states over the past month. Only 3 of them – CA, AZ, and FL were over 10.0. Today, they all are. And while CA is high at almost 15, it’s nowhere near AZ, TX, FL, and NV – all of them in territory previously occupied only by Brazil. Of note here is OK’s precipitous increase. When Trump took his COVID clown show to OK on June 21, they had 326 cases per day; now they have 591. They were averaging 1 fatality a day; now it’s 3. Daily patients: 186, compared to 444 today. ICU: 86 when Trump came; 191 today. I can’t say for sure that Trump caused all this, but Dr. Bruce Dart, Tulsa City-County Health Director thought it was a likely contributor.

Chart SDT1 shows how all seven states look today.

There are four states that are absolutely in critical condition: AZ, FL, NV, and TX. How would you like to be the parent of a school-age child, or a teacher, or a custodian, or an administrator in a public school in one of those four states? Later this week, I plan to write something on the “plan” to reopen the public schools. So if you’re in one of those groups above, please write me and tell me what the school in your neighborhood is planning to do and how you feel about it. You can email me at zorgi@zorgi.me .

Later this week, I plan to dissect LA County and SD County in the same way as I did the state of CA today. In the meantime, CD1 shows the current Zorgi Scores for the counties.

For my readers in LA, CVS3a is a teaser. And CVS71 is the same for SD readers.

On the city level in SD, CY1 shows the unfolding surprisingly bad situation in Carlsbad. Much of that is due to the incredible 1100%+ increase in daily cases, bringing case doubling days down to a scary 6 days.

By the way, I’ve also added a complete description of the calculation methods for city level Zorgi Scores. I also modified my calculation program so that I can click on the checkbox by a city, and in the CY2 box, it will display every step taken in the calculation of the final score.

I’m now at work on the zip code reorganization. Previously, I covered 13 zip codes. In the future, I’ll be covering 31 zip codes. These are the zip codes in SD county that had the 30 greatest number of cases as of July 11. I also added my hometown zip code – 92024. Unfortunately, it almost made the cut without me having to add it. The Zorgi Score calculation for zip codes will closely mirror that for cities, since all the metrics have to focus on cases.

I’ve been looking for a similar dataset in LA county. So far, the only thing I’ve found is this dashboard, but I don’t see any way to get historical data. If anyone out there can help, I’d greatly appreciate it. I

On my website only, I’m adding the rest of the tables and charts backing up the comments above. If you’re on a social media platform, just click on the link and you can see them all.

Be safe and healthy, everyone, and wear your mask!

8 comments on “Deconstructing CA Data

    1. Thank you so much! I keep adding to the site, so come back often. And please let me know if there are any technical questions about how the site operates, or if you have any questions about the data. I’ll do my best to answer them.

  1. Thanks for putting all this info in one place! Much appreciated.

    Question: does your data set refer to tracking overall (covid only) mortality rates?

    Apologies if Its there and I missed it. 🙂

    1. Thanks, Dano. The fatality numbers come from covidtracking.com on a state level, and CA on a county level (see the Sources page for url’s). These are cumulative numbers, so I construct a daily number from the difference between two days. Then I use a seven day average to get a “smoothed” daily number. That’s the number you see used everywhere, including in formulas like the case fatality rate.

      The fatality numbers are for covid only. From everything I’ve read, they are undercounted.

      Did this answer your question? I may not have understood you correctly.

    1. Thanks so much! It appears that you are writing from Thailand? I hope you are doing OK. You’re sure doing better than we are in the U.S. Our deaths per 1M are 500 times that of Thailand.

  2. Hi! I’m at work browsing your blog from my new iphone
    4! Just wanted to say I love reading through your blog and look forward to all your posts!
    Carry on the fantastic work!

Leave a Reply

Your email address will not be published. Required fields are marked *

 

This site uses Akismet to reduce spam. Learn how your comment data is processed.