A deeper look at local COVID-19 numbers shows a 5,000% increase, but also some encouragement
photo by: Associated Press
Douglas County is living on the edge these days.
Last week, the health department announced the county’s overall rate of COVID-19 tests coming back positive was up to 9.7% over the last 14 days. If the county gets to a 10% average, new restrictions — especially for schools and sports — could be mandated by health leaders. A new positivity rate is scheduled to be released Thursday.
In other words, it is a big week on the local COVID-19 front.
Given that, I thought it would be worth taking a deeper look at our local COVID-19 numbers, something we haven’t done for awhile.
I’ll get to the details in a moment, but here’s the big broad-brush summary: 1. New cases numbers are discouraging. Compared from June to September, the number of active cases we have in the community on any given day has increased by a little more than 5,000%. 2. The positivity rate mentioned above is a mess of a metric. It can lead you in a lot of false directions if that is the only metric you are watching. 3. One decent local number hasn’t gotten much attention: hospitalizations. Those statistics aren’t universally heading in the right direction, but they are much better than where they easily could have been.
There’s no way to sugarcoat this one. The number of active cases we have in the community has exploded. The numbers really are up 5,000% from June. That’s not hyperbole.
Here’s what I did with those numbers. We hear all the time about the number of new cases that we are adding. What we hear less of is about the number of active cases that exist on a daily basis. In other words, the number of new cases we have in the community, minus the number of cases that are past the incubation period or cases that people already have recovered from.
The health department has been providing the data since about the second week of April that is needed to compute the number of active cases we have on each day. So, I did those computations.
They show that we are in at least our third stage of COVID-19. They are:
• Stage 1: More Concern than Cases: In April, May and June, the community implemented several restrictions and new health guidelines as we watched what was happening in New York and other hard-hit areas. But, the numbers show that COVID-19 hadn’t really arrived in Douglas County in a significant way. On any given day in April, our number of active cases was just over 10. In May, it was just over nine. In June it was just over 12.
• Stage 2: COVID Arrives: In July the situation really changed. Notably, that is before students started arriving en masse. Some certainly had returned, but perhaps July is when people decided to get more active after seeing three months of very little COVID-19 activity. Or, maybe it is just when the virus really made its way to Douglas County from other regions. I don’t know, but the numbers show the progression. On June 15, we had 10 active cases in Douglas County. On June 22, we had 23. On June 29, we had 67. On July 1, we had 92. We hit our peak in July on July 10 at 194. (Well, based on the numbers I have. The health department data doesn’t really cover weekends or holidays, and there are a couple of other days when the department didn’t issue reports). Our overall average for July was about 138 active cases per day. That was up from 12 in June, an increase of a little more than 1,000% in one month’s time. Remember that number when trying to determine how quickly situations can change.
• Stage 3: Students Arrive (Also known as, if you thought July was bad, you’re really going to hate September.): Students started arriving to move into dorms in late August. KU, to its credit, provided testing for every student enrolled. Predictably, the number of cases soared. While it was predictable, the numbers are still shocking from a statistical standpoint. Through Sept. 14, the average number of active cases we have on any given day was a little more than 638. And that number has been on a general upward trend. On Sept. 14, for instance, it was at 752 active cases. So, from June, we went from an active case average of 12 to, thus far in September, 638. That’s your 5,000% increase.
It is easy to be negative about our positivity rate in Douglas County. After all, we are on the verge of crossing that 10% threshold. But, if you like statistics, it also is easy to be negative about the positivity rate in general. It is a pretty flawed statistic at the moment.
The reason is because the testing program keeps changing. If you test only people who have symptoms, then you are going to get a higher positivity rate than if you test people who are asymptomatic. In Douglas County, we have taken both approaches at various times. As a result, you can really draw some wrong conclusions if you focus too much on this metric in isolation.
For example, look at the chart below. It would suggest that our COVID-19 environment was its safest at the same time thousands of KU students came to town. That, of course, isn’t the case. The positivity rate was low because we started testing asymptomatic people, most of which came back negative for having the virus. Thus the lower positivity rate. But we weren’t safer as a community because we still had hundreds of new people in the community who had the virus.
photo by: Courtesy: Lawrence-Douglas County Health Department
Health department leaders understand this just fine. They know you can’t look at the positivity rate in isolation. They also know that all of this would be more effective if we had a broader testing program. Yet — other than KU — we’ve really struggled to test broad swaths of the Douglas County population. At some point community leaders need to figure out whether broad-based testing in the county is a priority. Or maybe the governor’s recent comments will take care of that for us. But, I digress. Here’s what Dan Partridge, director of the Lawrence-Douglas County Public Health, said when I asked him about the drawback of the positivity rate as a measuring stick.
“The percent positivity rate is one of several measures used to estimate the rate of transmission of COVID-19 in the community. It’s value as a metric becomes stronger as the level of testing in the community increases. Because of these limitations we encourage the public to not look at our percent positive rate in isolation but as a part of the whole.”
So, that’s the public service announcement for today. Don’t focus only on the positivity rate. However, I can envision some members of the community beating public officials over the head with that number as groups like the Lawrence school board in the coming weeks decide whether to reopen for in-person learning.
Simply put, if you focus too much on this number, you are at risk of making bad decisions.
While crunching numbers, I also spent time compiling statistics on the number of people who have been hospitalized at LMH as a result of the virus. They are reported every day by the health department, but they get less attention from the public.
However, I wonder if they can’t be more useful at actually tracking the severity of our COVID-19 environment. The thing that hospitalization numbers have going for them as a statistic is they avoid the problem the positivity rate has. The pool isn’t changing all the time. Unlike the testing program, the criteria for getting admitted into the hospital with COVID-19 hasn’t dramatically changed. Plus, it measures our greatest fear. What we really are trying to prevent here is people getting so sick that they greatly suffer or die. The hospitalization rate is a metric that measures that. Plus, we are trying to prevent our hospitals from becoming overwhelmed to the point that they can’t provide other needed care. This metric best measures that.
The numbers in Douglas County aren’t universally positive, but there’s some good news in them. The best news is this: We’ve seen a 5,000% increase in average active increases since June, but we have come nowhere close to overwhelming our hospital with cases. Sometimes that gets glossed over, but that is really important.
Here’s a look at the average number of COVID-19 patients hospitalized on any given day at LMH during the following months:
• April: 0.96 patients per day.
• May: 0.29 patients per day
• June: 0.16 patients per day
• July: 3.03 patients per day
• August: 6.48 patients per day
• September: 5.75 patients per day
Again, those numbers aren’t universally good news. The June to July jump in patients was concerning. That’s an increase of about 1,800%. But when you look at September, that clearly is good news. The patients per day have dropped thus far, even though our number of active cases in the community is at a new high. That seems significant.
But again, the most important thing with these numbers is that 5.75 patients on a given day does not overwhelm LMH. Even 6.48 patients per day doesn’t overwhelm LMH.
But health department leaders are appropriately highlighting one caveat: We’re not at the end of the race. The situation could still change. But, still, the numbers offer some encouragement.
“We should be proud of the work we’ve done to protect the hospital, but we should not get complacent or rest on our laurels,” Sonia Jordan, director of informatics for the health department, said via email. “We have wonderful mask compliance, and we should keep that up. We have invested community members that are trying to do the right thing to protect themselves and others, and we should commend their efforts.”
But it is still too early to say that the high number of KU cases is unlikely to produce hospitalizations. Even if they do not produce hospitalizations of actual students, Jordan said there is concern the KU cases will spread into other parts of the community. Time will tell whether those KU cases infect other people who later need to be hospitalized.
“It will take some time to move from one community to another,” Jordan said via email. “We do still have some concerns that the increase in transmission at KU will have negative effects on other areas of the community, such as long term care facilities or high-risk congregate care settings, which will negatively affect hospitalization rates. We should remain vigilant to this.”
There’s also one other way to look at these hospitalization numbers. Maybe some won’t find it useful, but I believe it may be important as time passes. I calculated the hospitalization rate per 100 active cases. Here’s what those averages showed for the following months:
• April: 10.9 hospitalizations per every 100 active cases in Douglas County.
• May: 3.3 hospitalizations per every 100 active cases in Douglas County.
• June: 0.05 hospitalizations per every 100 active cases in Douglas County.
• July: 2.44 hospitalizations per every 100 active cases in Douglas County.
• August: 4.7 hospitalizations per every 100 active cases in Douglas County
• September: 0.9 hospitalizations per every 100 active cases in Douglas County.
How might these numbers be useful? They’ll be useful if the community decides that not all COVID-19 cases are equal in their importance. That may be a controversial view, though, and community leaders will have to decide whether to adopt it. But, if your main goals are to prevent great suffering, death and overwhelming our hospital, then simply looking at the total number of cases in the community might not be the best metric. You are going to want to know the likelihood of new cases producing hospitalizations. This type of statistic can help with that.
But again, is that the approach the community is going to take? I don’t know — and I’m not weighing in on what approach it should take — but I do think it will be an important issue to sort out. I think you are seeing some of it unfold on the KU campus. Chancellor Douglas Girod last week made an important statement when releasing new COVID-19 numbers from campus.
“Our goal has never been to eradicate COVID-19,” Girod said. “Rather, our goal has been to take science-based steps to co-exist with it while fulfilling our mission of education, service and research.”
If the national health leaders are right, and we should expect restriction on our life until at least late 2021, we all may have to give new thought to what “co-existing” means.
Regardless, stay safe, be smart, wear your mask.