Kansas COVID-19 Update, Week 7

 

This week's summary will be short and sweet, in part because two of my three key metrics are largely the same.

The Good: The test positivity rate in Kansas last week was 11.4%, according to Johns Hopkins. Now, normally having 11.4% of your state's COVID-19 tests come back positive would not qualify as "good" by any stretch of the imagination (it should be 5% or less to be reasonably confident you're testing enough to identify most cases). But this is all about context. Kansas' test positivity rate has been increasing steadily since I started this blog almost two months ago and topped out at 12.8% last week. Any change in that trajectory is very welcome. One caveat about last week's numbers, though: the University of Kansas recorded nearly 21,800 tests as it prepared to welcome students back to campus. It's very possible that was enough to bring down the test positivity for the entire state, given how little the state has been testing. We'll have to wait and see whether this is a true downward trend or just a one-week blip.

The Bad: The reproduction rate (Rt) was 1.06 in Kansas as of Aug. 29. That's the same as it was the week before, but it still qualifies as "bad" because the longer that number stays above 1.0, the more cases we will end up with. If it's below 1.0, and it stays the same, that's good. If it's above 1.0 and staying the same, that's bad — especially as kids start going back to school, because gatherings tend to increase the Rt.

The Neutral: Nothing "ugly" this week, thankfully. Once again, hospital ICU capacity held steady at about 37% availability (with 120 hospitals reporting). It's remarkable how stable that number has been over the last month or so. 

Kansas COVID-19 Update, Week 6

 

As we add another week of COVID-19 updates, you may see a pattern developing in the good, bad and ugly trends. I see it too, but as much as I would like to mix things up, these are the directions the data are pointing. This week's update includes an intriguing proposal to ramp up testing, and a check-in on my home county, Johnson.

The Good: Hospital ICU capacity continues to hold steady, clocking in at 37% availability on Aug. 20, with 118 hospitals reporting. That's exactly what it was a week earlier and only one percentage point lower than it was the week before that. In fact, that number has held remarkably steady for about a month now. As I've said before, our COVID-19 cases in ICU seem to have settled in at a relatively high (compared to May/June), but manageable level of about 100 patients statewide. Manageable for now, that is. It might be a different story if hospitals are still treating that many critically ill COVID-19 patients when flu season hits.

The Bad: The infection reproduction rate, or Rt, ticked up again, to 1.06. It's the second straight week that has gone up, following several weeks of it trending down the way we'd like. That's discouraging, because I had hoped it was just a one-week blip. Remember: Rt below 1.0 means we're slowly choking off the virus. Rt above 1.0 means we're spreading it, and the farther above 1.0 we go, the faster it spreads. Social distancing, avoiding large gatherings and wearing masks are the best ways to lower it.

The Ugly: According to Johns Hopkins, the test positivity rate in Kansas took another jump last week, rising to 12.8%. It was 12.2% the week before and 11.8% the week before that. Not a good trend. To get that number lower we need more testing, and it appears there may be a plan developing to get it. Kansas health secretary Lee Norman acknowledged during a hearing this week that the state lab does not have enough capacity to do the necessary amount of testing (this is no surprise; very few states have the kind of public infrastructure required to handle current COVID levels, and Kansas budgets have been particularly tight for almost a decade). To increase capacity, Norman proposed that the state partner with private labs like Quest and LabCorp, which are currently running tests ordered by individual doctors (not public health departments). It's a good idea, and then Sen. Jim Denning added an element that made it much better: only paying the private labs if they get the test results done within 48 hours. That requirement should be written into any contract, in order to prevent the long delays that have plagued the private labs (Norman said the state lab returns results in about 1-3 days, while the private labs often take 7-14). I would also consider offering them a bonus if they get the results back in under 12 hours, or even 24.

Bonus: In Johnson County there has been an awful lot of talk about schools lately. Will they open up online-only, fully in-person, some combination of the two? Will there be sports? The county health department has developed a data-driven set of criteria for the school districts, but the actual decisions  are up to each school board.

Local journalist Kyle Palmer of the Shawnee Mission Post provided the health department's "gating criteria" recently on Twitter. 

BREAKING: @jcdhe says county is currently in "red" zone for school reopening, as part of its final guidance to JoCo schools before the start of classes this fall.

That would put @theSMSD in "remote only" learning for all students to start the year. But this is NOT official. pic.twitter.com/Ol138Jo1JD

— Kyle Palmer (@smpostkyle) August 18, 2020

As you can see, the criteria is based on a combination of test positivity and whether overall cases are increasing or not. According to the health department the test positivity rate in the county is currently about 11.5% and new cases are increasing (though slowly). That puts us in the "red zone." But we're not very far from the "yellow zone." After stay-home orders expired, new cases and test positivity rose quickly in Johnson County throughout June and early July. But the county adopted the state mask mandate that went into effect July 3 and after about one incubation period, case growth and positivity slowed considerably. 

That's similar to what we've seen in Douglas County and Sedgwick County, which also required masks. Douglas County has actually gotten its test positivity rate all the way down to 4% now, from a high of about 10%. But remember, Douglas County also closed bars. I don't know if there's any political will to do that in JoCo, or if it would reduce COVID-19 rates to the same degree it seems to have in Douglas County (I would guess there are more bars per capita in Douglas than there are in Johnson). Closing bars would obviously be controversial, and maybe it's not necessary. Maybe all JoCo needs to get that test positivity rate down is better adherence to mask wearing and social distancing and more testing. But with case rates and positivity rates where they are currently, there's a good chance of sports/school-related outbreaks. 

I will end on a hopeful note: last week I stopped into the Blue Moose in Prairie Village to pick up some takeout. The restaurant's outdoor patio was pretty full, but indoors there were hardly any patrons, and the few that were eating inside were very spread out, in a pretty well-ventilated space (the garage door leading to the patio was open). People seem to be getting the message: stay outdoors as much as possible. Doing so should only get easier in the months to come, as temperatures hopefully start to decrease. Perhaps in lieu of closing bars, the county could give them no-interest loans to rent tents and folding tables/chairs to set up outside, a la German beer gardens. That would help not only the bars, but also the party rental industry, which I assume is also hurting because of the pandemic.

Kansas COVID-19 Update, Week 5

 

This week Kansas' infection reproduction rate is almost unchanged, COVID-19 cases in hospital ICUs  are holding steady at a relatively high but manageable level, and test positivity keeps getting worse. This week's bonus is another chart that (inadvertently) makes a strong case for masks.

The Good: The state's ICUs are stable, with 110 hospitals reporting 37% capacity available on Aug. 13. That's almost identical to last week (38%) and the week before (36%). Cases numbers keep rising, but over the last few months doctors and hospitals nationwide have gotten better at keeping people from becoming ill enough to require an ICU bed. It's one of the few success stories of the pandemic in the U.S. so far.

The Bad: After a couple weeks of improvement, the infection reproduction rate (Rt) in Kansas ticked up a notch from 1.04 to 1.05. That's not good, but I'm hopeful it's just a brief plateau and we can continue the downward trend next week and get below that magic number of 1.0, which is when you start decreasing overall infections. Social distancing and wearing masks helps. It's also vitally important that people stay home if they're sick. Let's make sure each infection doesn't cause a bunch of other infections.

The Ugly: The test positivity rate is bad and getting worse. According to Johns Hopkins, an average of 12.2% of COVID-19 tests in Kansas came back positive last week. That figure has risen every week for the past month, leaving us farther and farther from the 5% threshold that the World Health Organization recommends before lifting restrictions on public gatherings. Deborah Birx, the top White House official in charge of COVID-19 response, said this during an Aug. 15 visit to KCK:

“Kansas has rising test positivity. This is the moment to get it under control. Wear a mask. Close bars. Decrease indoor dining. Increase outdoor dining. Every single person needs to commit to not having parties and family gatherings that are going to spread this virus.”

Bonus: The extra content this week takes us back to Michael Austin, the former Brownback Administration economist who is adamantly opposed to mask mandates. Last week Austin pulled apart a chart from the state health department to show at a glance that Kansas counties without a mask mandate had lower overall COVID-19 rates than counties with a mandate. But in the process he also showed quite clearly that COVID-19 rates dropped after counties enacted mask mandates and the rates have held steady in the counties that didn't.

This week Austin was back with a new chart that showed that COVID-19 cases initially rose in the counties with mask mandates and didn't start dropping until about a week after the mandates went into effect.

Thanks to KORA, we have evidence @GovLauraKelly allowed @SecNorman to rig not only THE LEVEL of #COVID19 cases in their favor but also THE TREND.

Norman HID the fact mask counties had a 112% COVID case JUMP so he could say mask counties had a 38% DECLINE. #ksleg pic.twitter.com/L9YhyCdzei

— Michael Austin (@KSTaxEconomist) August 14, 2020

What Austin left out was that the novel coronavirus has an incubation rate of up to 14 days. So the effects of any intervention — including mask mandates — will never be seen immediately, because exposures from before the intervention will cause infections for weeks afterwards.

Indeed, in the graph Austin made (above, at right), the entire "112% JUMP" that he trumpets happens within the incubation period. 

Look closer and with more context and you see that he's actually made quite a strong case for mask mandates with this latest chart. The Kansas mandate went into effect July 3 for the counties that adopted it. According to the latest available data, the median incubation period for COVID-19 is 7.7 days, meaning a majority of infections won't have shown up until eight days after exposure. Eight days after July 3 would be July 11. Look at what day the infection rate starts to drop precipitously on Austin's chart: July 12.

Let me say that again: The median incubation period for a COVID-19 infection is about eight days. And Austin's chart shows that the COVID-19 infection rate begins to drop nine days after the mask mandate. 

In his rush to accuse the Kansas governor and health secretary of misleading people about the effectiveness of wearing masks, it was actually Austin who was misleading people (either willfully, to score cheap political points, or simply through his ignorance of COVID-19). In his zeal to show mask mandates don't effectively reduce COVID-19 transmission, he once again inadvertently provided pretty compelling evidence that they do. 

Unfortunately, a right-wing website called "The Sentinel" picked up Austin's fatally flawed analysis and ran with it. Like Austin, the site failed to mention that a COVID-19 incubation period even exists. That raises serious questions about the site's editorial standards and level of intellectual honesty. But if you want to address those questions to the people who bankroll the site, good luck: they're anonymous.

So if you see anyone passing around Austin's chart on social media as evidence that the Kansas mask mandate didn't work, explain to them the incubation period, and how it actually makes the chart a strong argument in favor of mask mandates. 

Kansas COVID-19 Update, Week 4

 

Last week's update was a good reminder for me to watch out for outliers in the data and be skeptical of large swings. The positivity rate in Kansas has been trending up, but if it had jumped to 18.6% that would have been truly remarkable — in a bad way. Thankfully it had not. But it's still too high. We'll check in on that number again this week, and also break down a mask mandate graphing controversy.

The Good: There was a significant improvement in the reproduction rate (Rt) in Kansas this week, as it dropped from 1.07 to 1.04. This is great news! If we can get that figure below 1.0, the number of active COVID-19 cases will begin to decrease. Getting Rt below 1.0 and keeping it there is the key to stabilizing the pandemic until we have a vaccine or effective treatment. The latest drop in Rt means we're doing a better job of keeping our germs to ourselves by doing things like wearing masks and maintaining a safe distance from each other, especially indoors. As noted in previous weeks, Rt is very hard to pin down. While one group of data analysts has now pegged it at 1.04 in Kansas, another group still has us at 1.1 (and both groups say it could potential range anywhere from 0.77 to 1.20). More testing would help us get a more accurate number.

The Neutral: Hospital capacity is holding steady for the second week in a row, with 113 Kansas hospitals reporting about 38% of their ICU beds available. That's actually a slight improvement from last week, and we're still well within a range where we should be able to give each patient the best care possible. 

The Bad: No "ugly" news this week, but the test positivity rate still doesn't qualify as "good" or even "neutral." According to Johns Hopkins, 11.8% of COVID-19 tests in Kansas came back positive last week. Two weeks ago that figure was 11.0% and the week before that it was 10.7%. The trajectory is not good. Again, more testing would improve this measure.

Bonus: During a news conference this week, KDHE Sec. Lee Norman said the rate of new COVID-19 infections is decreasing in the 15 Kansas counties with mask mandates and is not decreasing in the 90 counties without mandates. Then he showed this graph:

It's a bad graph. You shouldn't combine two different Y-axes, especially with different scales, on the same chart. At a glance, it looks like the mask mandate counties now have lower COVID-19 rates than the non-mask counties. Only if you really look closely can you tell that's not the case. 

 

Michael Austin, an economist formerly with the Brownback Administration who opposes mask mandates, re-created the graph with a single, standard Y-axis.

If someone manipulated and presented data in the private sector to the upper management, they’d be fired for misleading their employer. #ksleg #ksecon #COVID19 @SecNorman works for Kansans, and he was trying to mislead us, plain and simple. #MaskMandate pic.twitter.com/oGZk3tgAaM

— Michael Austin (@KSTaxEconomist) August 7, 2020

That second graph shows quite clearly that non-mask counties have a lower COVID-19 rate than mask counties. But it still shows that counties that enacted mask mandates have since seen their daily cases go down (by about 40%), while counties that didn't enact mandates have had their case rates remain relatively flat. In other words, while Austin's graph is an effective argument against KDHE's chart, it's not really an effective argument against masks. In fact, it's evidence that mask mandates work, unless there's some other variable that explains that difference.

It's also important to remember that Kansas is just one small state in a global pandemic, and there is a lot more data out there. This study published in Health Affairs (a peer-reviewed health policy journal), found a similar phenomenon throughout the U.S.: states that enacted mask mandates "had a greater decline in daily COVID-19 growth rates after issuing these mandates compared with states that did not issue mandates." The University of Kansas Medical Center's COVID-19 Daily Digest (which I highly recommend, even though it's not actually updated daily), includes an entire section on masks, with links to several more studies. The data from these studies is quite clear: masks are a good way to reduce the spread of the novel coronavirus, even among people who live in the same house.

The question of whether you HAVE TO wear a mask in indoor public places depends on who is in office where you live, so it is a political question, by nature. But the question of whether you SHOULD wear a mask is not actually political at all. It's a question about your health, as well as the health of your family, your community and your state. 

Kansas COVID-19 Update, Week 3

Updated Aug 3: The test positivity rate cited below (18.6%) appears to be an error. After taking a deeper look at the Johns Hopkins data, it includes one day (July 27) in which test positivity was reported to be 109.7%. This is obviously impossible (you can't have more than 100% of tests come back positive), so there's something wrong with the data on that day. Test positivity on the other days in the rolling average are 11.2%, 10.8%, 10.8%, 11.6% and 12.5%. So the true number is probably somewhere between 10.8% and 12.5%, which is closer to what it was last week. Still not good, but much better than 18.6%.

The format has changed a bit with this week's Kansas COVID-19 update. Only one of my key stats was "bad" compared to the week before, but it was SO bad that it has skipped right to the "ugly" category. It is a truly shocking number, and one I hope is a fluke, or a data error. As a palate cleanser, I've included another "bonus" section on a single county.

The Good: Once again, Kansas' reproduction rate, or Rt, has ticked down a bit, now to 1.07. Until there is a vaccine or an effective treatment, getting that Rt below 1.0 remains our best chance of containing the virus. (Rt below 1.0 means every infected person spreads the virus to fewer than one other person, on average, which would cause COVID-19 to slowly die out for lack of new hosts. Rt above 1.0 means the virus' spread is increasing). If Rt is decreasing, that probably means more people are wearing masks and social distancing. 

The Neutral: ICU admissions were stable this week, with at least 98 COVID-19 patients in intensive care and 36% of beds available on July 30 (the same numbers as on July 23).

The Ugly: According to Johns Hopkins, the percentage of COVID-19 tests that came back positive in Kansas jumped from 11% last week to 18.6% this week. If that's accurate it means Kansas is not only far from the World Health Organization's standard of 5% or less, but it's tied with Arizona for the fifth-highest test positivity rate in the country. It's also disheartening in that it means we're probably missing a lot of cases and that drop in the Rt might be a mirage caused by hidden infections. 

A quick clarifying note: many Kansas media outlets, when they report test positivity rate, use the figure reported by the Kansas Department of Health and Environment, which is much lower (9.5%). But KDHE appears to be calculating test positivity based on the cumulative number of all tests in the state since the beginning of the pandemic (now almost 300,000). That's not wrong, but it does mean that the percentage isn't going to move much from week-to-week. I prefer to look at Johns Hopkins' number because they're calculating test positivity based on a rolling average of the last seven days of testing. That provides a more current snapshot of the situation, but it also makes the positivity rate more susceptible to weekly fluctuations in testing. 

For example, there has been a lot of press lately about long wait times for outpatient testing done at private labs (mostly Quest and LabCorp). Kansas depends heavily on those labs (they've recorded 70% of the state's positive cases, so far). If their results were really backed up last week and most of what was actually reported were results from hospital inpatient labs, that might skew the test positivity number higher (because patients sick enough to be hospitalized are more likely to have COVID than outpatients). But still, 18.6% is a number that should cause all of us to sit up and take notice. We need more testing in Kansas.

Bonus: At the end of May, Sedgwick County had its test positivity rate down to about 1%. But then the shops, restaurants and bars opened back up and the rate quickly ballooned to about 14% by July 23. Things were going the wrong direction. At one point in mid-July, Wichita's hospitals estimated they had only eight ICU beds available. But since July 23, the test positivity rate has made an abrupt downward turn and is below 12% and falling. What happened? Here's one theory: on July 8, the Sedgwick County health officer ordered residents to wear masks in public areas. Almost exactly two weeks later (the maximum incubation period of the novel coronavirus) the county's test positivity rate began to go down. It's impossible to prove that the mask order caused the downturn, but the timing matches up. Sedgwick County still has a long ways to go, but it's now trending in the right direction, and, as I wrote last week, trends are everything with outbreaks of infectious disease.