Sorry I missed an update yesterday; it’s the first day I didn’t post since March 28, and I figured a day off for Father’s Day was a necessary thing.
The reorganization of the way I present state data was a success judging from all the terrific comments I got on several subs. So I’m trying to apply the same methodology to counties, and work my way down.
One of the striking things about this pandemic is that there is no “national” outbreak. Each locality has its own characteristics. That makes it not as useful to compare one state with another, and it starts to make more sense to look at each state in light of its own metrics.
The same is true, it turns out, with counties in California. Previously, I covered counties all over the state, with an eye to comparing them. I’ve decided to change my approach, and concentrate on counties just in Southern California, the thinking being that while the “anchor” subs I post in are San Diego and Los Angeles, people travel to and from all the counties in the region.
The counties I’ll be covering from now on are: Imperial, Kern, Los Angeles, Orange, Riverside, San Bernardino, San Diego, San Luis Obispo, and Ventura.
This has required a complete reorganization of my spreadsheets, as well as the creation of new charts. I’m about halfway done, so I hope you’ll bear with me over the next week or so while I get everything back in good order.
One thing that is extremely aggravating is the lack of data available. I can download data for cases, fatalities, patients, and ICU with no problem. Beyond that, it’s a mess. Only three of these counties publish testing data: Los Angeles, Orange, and San Diego. Yet one of the big metrics public health officials keep talking about is the Positivity Rate – the ratio of cases to tests. How is anyone supposed to calculate that without having access to testing data? They say that transparency is essential during a pandemic. As good as California has done in many respects, we still have a long way to go.
Hospital bed availability and utilization rates are another metric that virtually everyone agrees is important. Try getting these stats. You’ll see that I’ve included a hospital bed utilization rate (HUR) and an ICU utilization rate (IUR) for each county. There is no easy way to get up to date data. I was able to find a source for ICU utilization rates before the pandemic, but for HUR’s, I had to make an estimate of 65% for every county but San Diego.
When you see the HUR and the IUR on a chart, that includes the HUR and IUR for COVID only, PLUS the pre-pandemic HUR and IUR. I did this to give a better visual of how close a county is to maxing out on its healthcare system. I realize this is nothing more than an estimate, but I thought it might be more useful that just showing the COVID HUR and IUR alone.
Some counties, like Imperial, have a very low death count in absolute numbers (even though its rising quickly), so I provided another metric that shows up the chart – a 7 day fatality count. This is a rolling 7 day SUM of the fatalities for the previous week; it is not a daily count.
Since so many of the counties lack data on testing, making it impossible to show a positivity rate, for those I added case doubling days and fatality doubling days to help complete the picture of what’s going on.
I’ll provide a “Z-score” for each county, on a scale from 1 to 10, with 1 signifying the least risk and 10 signifying extreme risk. I’m sure you realize that this is a totally subjective assessment. Also, it’s not a grade on how the county is doing.
The charts with today’s update are:
- CVS0 – Z-scores for counties
- CVS1 – Imperial County Vital Statistics
- CVS2 – Kern County Vital Statistics
- CVS3 – LA County Vital Statistics
- CVS4 – Orange County Vital Statistics
- CVS7 – San Diego County Vital Statistics
- CGS1 – Daily Cases per 1M People
- CGS2 – Daily Fatalities, 7 Day Avg.
- CGS3 – Daily Fatalities per !M People,
- CGS4 – Prevalence Index
- Analysis – screenshot of the numbers used for data commentary
- County Info on hospital stats + sources
- VS3 – Southern California Snapshot
- VS1 – Vital Statistics, San Diego Are (SD Only)
The following commentary is based on a starting date of May 20 and an ending date of June 20. Whenever I refer to “the beginning” or “at the start” that means May 20. Please refer to the “Analysis” chart if you would like to check the numbers.
Imperial County – see CVS1
This is a county where a lot of people are struggling to get by. The average median value of owner occupied housing for the 10 SoCal counties is $438,840; the median value in Imperial is $177,100. 62.4% of the population is non-white. There are lots of farm workers who work in the county, but can’t afford to live there, so they go back and forth between Mexico and the US every day. Mexico is not faring well in the pandemic, so this naturally affects the situation in Imperial.
Cases have jumped by 122% over the last month. Since testing data isn’t available, there is no way to assess it’s impact on case expansion. Case doubling days have increased by 90%, from 10 days to 19 days. The fatality doubling rate decreased, which is a very bad thing, from 15 days to 14 days and fatalities per 1M have increased by 425% in the past month. Since absolute number are still quite low, less than 1 a day, that isn’t as worrisome as it would be in LA. The real worry hear is that ICU beds are in short supply, and my estimate is that the IUR a month ago was 115% and is now 107%. I don’t think this is far off the mark, because we’ve had continual reports of overflow patients from Imperial going to surrounding counties.
Z-score for Imperial County: 7.5 out of 10
Kern County – see CVS2
Kern has seen even more of an increase in cases than Imperial – 140%, from 36 cases a day to 85 cases a day. Again, we have no testing data, so we have to rely on other metrics. Case doubling days have only risen by one day, from 25 to 26 days, so case growth isn’t slowing down much at all. Fortunately, fatality doubling days have gone up, from 15 to 27 days, and daily fatalities have stayed exactly even.
Hospitalizations have increased by 107%, from 56 to 116, bringing the HUR at the end of the period to 73% – up there, but nothing like Imperial. The IUR went from 63% to 75%, but there is still capacity remaining.
Z-score for Kern County: 5 out of 10
Los Angeles County – see CVS3
LA is one of the three counties that provide testing data, so we can use the positivity rate as a metric. Daily cases increased from 908 to 1,381, or 52%. That was the smallest rise of all ten counties. The positivity rate jumped, but that may be just a two day aberration, since it had been holding steady at around 7% for quite a while. Testing only went up 9% from 11,494 to 12,510 a day, a 9% increase. That means that a maximum of 71.6% of the increase in cases could theoretically be tied to testing increases. [Please note the word theoretically. I am not saying testing created these cases or was responsible for them. This only means that if you were to use the argument that the increase in cases was due to testing, that would only explain 71.6%, even if every single case came from a new test] Looking at it another way, if cases stayed the same, with the testing changes, we would have had a maximum positivity rate of 7.4%. Since it was 11.0%, that means new sick people are showing up.
Fortunately, the HUR decreased, from 75% to 74%, reflecting a 5% drop in hospitalizations, from 2,067 to 1,969. Daily fatalities also decreased from 44.4 to 31.4 – a drop of 29%. That’s a very good sign.
Z-score for Los Angeles County: 3.5 out of 10
Orange County – see CVS4
Another county providing testing data. Here we have double the increase in cases – 102% – that LA had, from 120 a day to 243 per day. Testing did increase from 2,533 per day to 3,758 per day, a 48% increase. That means the maximum theoretical increase from testing would be 73.3% of the new cases. If cases had remained the same, the positivity rate at the end of the period would have been 3.2%. Instead, it was 6.5%. So there are some clear indications that cases are starting to multiply on their own. This is reflected in the daily fatalities which increased by 178%, from 2.6 per to 7.1.
Hospitalizations also went up, from 388 to 446, bringing the final HUR from 71% to 72%. There’s still room for growth there, but less so in ICU’s. There was a 28% increase in ICU’s, from 130 to 167, bringing the final IUR to 81%, getting a little close for comfort.
Z-score for Orange County: 6.0 out of 10
San Diego County – see CVS4
The third and last county providing testing data. Here we’ve had one of the lowest increases in cases, only 36%, resulting from cases going from 123 a day to 167. Case doubling days went up by a healthy 10 days, from 29 days to 39. The increase in fatality doubling days was even better – from 26 days to 45 days. Testing increased by 65%, from 3,246 to 5,349. And yes, you could theoretically say that all the new cases came from more testing.
Again, I’m pretty sure they didn’t, but this is the only one of the five counties covered today where one could even make a half credible argument for that.
There was also a nice drop in hospitalizations, from 404 to 330 per day, a change of negative 21%, bringing the HUR to the lowest of all ten counties, 55%. ICU dropped slightly as well, from 150 to 148, resulting in a final IUR of 85%.
Z-score for San Diego County: 3.0 out of 10
The remaining five counties will be covered in the next couple of days. None of them provide testing data, so the analysis will all be like that for Imperial and Kern.
Some last thoughts on the county comparison charts – CGS1, 2, 3, & 4. CGS1 shows the outsized proportion Imperial’s cases are having on the situation. Numbers are not huge yet, but on a normalizd basis, Imperial is taking up much more space than any other county, including LA.
If you look at daily fatalities in terms of absolute numbers, as in CGS2, LA county clearly is huge. But as soon as you normalize the numbers, just as in CGS1, Imperial County in CGS3 dominates the landscape. Even more worrying is the steepness of the increase. It’s clearly an area where significant problems could develop very quickly.
The Prevalence index of 42 for Imperial is the lowest number I’ve seen for any locality I’ve tracked. Even in LA, where there have been thousands and thousands of cases, the prevalence index is five times that of Imperial. How worried would you be if one out of every 42 people in your county had direct experience with CV19? I’m pretty sure you wouldn’t be out there protesting that wearing a mask was a “deep state plot.”
I’ve included the other “Vital Statistics” tables just to keep you updated, but commentary on that will have to wait.
By the way, today I again got a Trumper “calling me out” on my political bias and telling me I had “Trump Derangement Syndrome.” If you’re a TrumpTroll, save your energy and either skip my posts altogether or just read the data section.
Stay safe and healthy, everyone!
By the way, many of you have offered to give me money, coffee, burritos, and other material rewards. I really appreciate that. If you’re still so inclined, I’d urge you to donate to the Equal Justice Initiative.