Saturday, August 29, 2020

North Dakota, South Dakota: Where the Coronavirus is Catching On: the week ending August 22nd

Last week I speculated as to whether South Dakota would experience an increase due to the Sturgis Motorcycle Rally which took place from August 7 to August 16th. An estimated 140,000 participants took few safety precautions. Although the bikers came from all over, they stayed for a time in South Dakota and may have passed through or originated in nearby states. If the spike in cases in South Dakota and North Dakota are an indication, then the Sturgis Rally was indeed a super-spreader event.

Event: Aug. 7 to 16th.

New cases in South Dakota for the seven day period week ending:

Aug. 8 --- 650 new cases (6944 new tests)

Aug. 15 -- 669 new cases (7877 new tests)

Aug. 22 -- 1002 new cases (7986 new tests)

This just in: Aug. 29 -- 2046 new cases (8891 new tests)

For North Dakota 

Aug. 8 --- 906 new cases (11182 new tests)

Aug. 15 -- 936 new cases (10778 new tests)

Aug. 22 -- 1292 new cases (10893 new tests)

This just in: Aug. 29 -- 1748 new cases (10709 new tests)

The August 29th number for the Dakotas are off the charts. They would rank number one and two for highest infection rate in the country if they were for the week ending August 22nd. 

For the week ending August 22nd, 35 states experienced a drop in their case rates. For the week ending August 22nd, the three states with the biggest nationwide percentage jump in new cases were: #1 Wyoming (bordering South Dakota), #2 South Dakota, and #3 North Dakota. (Wyoming seems to be a blip, and has settled down this past week.)


For the week ending August 22nd, North Dakota moved into second place nationwide in terms of new cases per million population, behind Mississippi. This is unprecedented for such a low population state (North Dakota is the 4th least populous.)


Mississippi

When assembling these numbers week to week, in order to keep my sanity, I start to root for certain states. (I also have a trick of entering the numbers in reverse, from the end of the week back to the beginning: everything gets better.) I root for Mississippi because they have the highest poverty rate for any state in the nation. They are nearly 40% black in population. They have had one of the worst rates of coronavirus infection, that has continued on from week to week.


That said, this week they rank number one in most new cases per million and dead last in testing rate. Unlike states that whiplash up and down in testing, Mississippi has puttered along the last few weeks with a low testing rate, ranking last place last week. This week further decreasing their testing by another 20%. At 5828 tests per million population (over the course of the week), they have the lowest number of any state since the week ending July 4th (South Dakota at that time). This low testing rate and first place in case rate gives Mississippi an eye-popping 32.2% positivity rate.


Mississippi: 11 straight weeks in the top ten for cases. They have reduced their testing by two-thirds since a peak for the week ending July 18th when they ranked 15th in the nation. To Mississippi I give a sad and reluctant DeSantis award.


Testing and New Case Rates

To what degree do testing numbers influence the finding of new cases? I have the feeling that I have enough data to answer that question. I also have the feeling that the answer would take days or weeks of work. If it were a simple matter of extrapolation, as long as the positivity rate remained the same, doubling the tests would double the number of cases.

100,000 new tests with a 10% positivity rate will identify 10,000 cases.

200,000 new tests with a 10% positivity rate will identify 20,000 cases.


That's fine, but the positivity rate does not stay the same. Who is tested is not from a homogeneous pool. Especially when positivity rates are high, the state is selecting those who are strongly suspected to be sick (arriving at the hospital with symptoms) and those known to be exposed. So, how much does it change? That's unknown. It cannot simply be determined by comparing states from week to week and looking at the testing rates and the positivity rates because we don't know what the true rate of infection is for the state and whether it is increasing or decreasing or holding the same. Antibody surveys, which in theory look at both people with ongoing infections and past infections show alarmingly low rates: typically 5% of the population. 

  

Florida and Texas.

Here are the testing rates, the new case rates and the positivity rates for Florida and Texas which have been bobbing up and down in the number of tests week to week over the past four weeks.  

week  ending    Jun. 27    Jul. 4    Jul. 11    Jul. 18    Jul. 25   Aug. 1   Aug. 8   Aug. 15   Aug. 22
Florida case rate 1804.1 2677.5 3001.2 3867.2 3582.4 3381.2 1836.6 2180.8 1264.3
test rate 12489.8 14843.4 15223.8 21263.4 21377.8 15617.6 10211.1 13662.4 8814.9
positivity 14.44 18.04 19.71 18.19 16.76 21.65 17.99 15.96 14.34
Texas case rate 1229.0 1669.9 2023.5 2319.9 2004.3 2080.6 1561.0 1634.7 1527.8
test rate 8588.8 12354.8 12391.3 15161.1 15092.2 16023.3 6596.7 14249.5 7897.2
positivity 14.31 13.52 16.33 15.3 13.28 12.98 23.66 11.47 19.35

Although these two states parallel each other in some ways, Florida has had a decline in its positivity in the past few weeks, while Texas has had its positivity rate jump up and down in line with its testing. The values for Texas suggest that it needs more tests to obtain its true numbers. The meager testing in Florida is still problematic.

What should be the correct rate of testing? 

The first goal of testing is to diagnose disease among those who status is unknown. This provides the information for possibly initiating treatment or monitoring, isolation and safety procedures, for testing those exposed, and for contact tracing. 

My opinion: Under optimal conditions, the testing should be sufficient to identify cases and limit infection. For states with very low positivity rates, let's say 2% or less (9 states), the testing should be at the levels to allow them to stay below 2%. For other states, at the very minimum, if a state has more than the median number of new cases, they should have more than the median number of tests. This week the median testing rate was 12639 per million population. 


Other Major Outbreaks.

I will highlight two other states: Illinois and Hawaii. Illinois is undergoing a dramatic increase in their case rate, up 32.8% week over week. Hawaii, although not doubling in cases, as they had been doing for several weeks, nevertheless continues to creep up. Just in: results for the week ending August 29 shows Hawaii up another 16.5%.


States in order of rate of new infections


I already highlighted Mississippi above.

North Dakota is 20th in testing.

Georgia is 13th in testing. Their positivity rate is 10.6%.

This week Texas ranks 45th in testing. Their positivity rate is 19.4%, second to Mississippi.

In contrast to Texas, Tennessee ranked 6th in testing rate. This allowed their positivity rate to be 5.8%.

Alabama has moderately strong testing program, 15th in the nation. 

Over the past two months, Nevada has cut its testing rate by more than half, currently ranked 39th. They maintain their position as the only "blue" state in the top ten (blue as in voting for Clinton in 2016).

Arkansas' dip down to 17th place for the week ending August 15th was due to reclassifying cases. The jump up therefore is an artifact. They rank 13th in testing.

Missouri ranks 31st in testing. 

Florida, still hanging on to a place in the top ten. They are 40th in testing this week and fifth in positivity.

Iowa continues to hang around the 1000 new cases per million mark. The are 27th in testing.

Illinois had a jump this past week. Its preliminary numbers for this week don't look so good. They are fourth in testing.

Idaho has embraced low testing, 43rd in the nation. Their positivity rate is 4th, just above Florida's.

I'll have to do a study of Republican voting states and case rates. Here is one of the most red-leaning states with a consistently high rate. They are 19th in testing.

Kansas has been consistently low in testing. They've bumped up their testing by 70% over the last four weeks to rank 33rd.

South Carolina reclassified a large number of tests, dropping their figures by 220,000 and thus were not included among the test rankings. 

South Dakota was mentioned above. Their testing rate remained about the same even though they jumped from 46th to 38th in testing, others moving down.

Who would have guessed that Hawaii would rank above Louisiana (1st last week) and California? Hawaii has been increasing their testing, now ranked 26th. 

Louisiana has made a dramatic decline. However, their once great testing rate (ranking 2nd two weeks ago) has since fallen by 36%. They still rank 10th.

The big jump California had last week fell back to about where it was before. The tests declined by 23.6% and yet still they rank 11th. 

North Carolina also ranks 21st in testing rate. Their gradual rise and fall is unusual.


Kentucky ranks 35th in testing. 

I find these states with lingering moderately high case rates to be worrisome, evidence that life has shifted to a new reality. Wisconsin ranks 34th in testing. 

Indiana ranks 32nd in testing.

Nebraska dropped their testing by 17.8% this past week, that drop reflecting in their cases.

Minnesota filled the median spot, number 26th. They are 29th in testing.

Utah has declined in testing to its position in 41st place.

Virginia seems to be getting over its second wave. It is 23rd in testing.

Arizona's cases are down to a fifth of what they were at their peak. Their testing rate, never great, now ranks 36th.

Rhode Island was first in testing this past week.

Montana has greatly decreased its testing, down by 60% in three weeks. They rank 44th. 

Maryland ranks 17th in testing. They've ridden two waves.

Ohio's second wave was greater than their first. They rank 30th in testing.

The jump this past week doesn't seem to continue on in preliminary results of this week. Wyoming ranks 25th in testing, for this week's median.

The District of Columbia is third in testing. 

Washington state is still not providing testing figures through this past week. I could not include them in test rankings or positivity rates.

New Mexico, which avoided the worst of the outbreak that infected the southwest, has greatly lowered its cases without lowering its testing, now ranked 9th.

Delaware ranks 28th in testing.

West Virginia bumped up their testing this past week and now rank 7th.

Michigan performs a decent amount of testing, this week ranked 14th.

Oregon, along with the next two states, is one of the worst in testing, this week ranked 47th.

Pennsylvania is ranked 46th in testing.

. . . and Colorado is ranked 48th.

Now we have a set of North Atlantic and New England states that for the most part have shown significant recovery and continue to have strong testing. Among these Massachusetts has taken a step back and is at 22nd in testing this week.

New York is 5th in testing and third to last in positivity at 0.76%.

New Jersey dropped its testing rate but is still in twelfth place.

Connecticut is 8th in testing and second to last in positivity. 

Note the scale of Maine. They have never gone over 300 new cases per million per week. (the best scale of all of them) They rank 16th in testing.

New Hampshire is stingy on testing, ranked 42nd. 

And Vermont continues to be last in new cases. They are 24th in testing.


0 comments:

Post a Comment