Coronavirus in Oklahoma: some data from the first week of April

As we have now left March 2020 in the rear-view mirror, I thought it might be a good idea to adjust my semi-weekly interpretation of national comparisons on #COVID-19 deaths and drill drown into some data from US states. Apropos of, well, everything to our current state of residence, I’ve focused here on our adopted home of Oklahoma.

Data herein come from Oklahoma’s State Department of Health and the Our World in Data Coronavirus Disease page (compiled from the European Center for Disease Prevention and Control by Roser et al. 2020).

Incomplete testing for novel coronavirus makes the estimation of death rates, case fatality rates, and similar measures highly unreliable, especially early in a phase of community spread. This is certainly the case in Oklahoma, where we were ranked 51st in COVID19 testing last week (just 541.8 tests per million residents). Death rates are inflated without good data on infection rates, and we might not have good infection rate data for months.

See how Oklahoma’s red dot is not only small, it’s close in size to the gray dot containing it. This means that we’re not testing many people, and the ones we are testing are the most gravely ill. Of 1159 positive tests, 316 were sick enough to require hospitalization (that’s 27%!) and 42 died (3.6% case fatality rate).

To avoid some of that uncertainty, I prefer to examine those data for which we have the greatest confidence: confirmed deaths from COVID-19. These are undercounted too, but much less-so that the undercounted number of total cases. We can track the number of deaths over time to examine the doubling rate, i.e., the number of days it takes for the death toll to double as an indication of the speed of spread through a population. Note that if our observed data indicate any kind of predictable pattern, the number of deaths will be ~13% of the number of people requiring hospitalization to recover, and that helps us examine and predict overload capacity at our hospitals. For example, when we hit 100 deaths in Oklahoma, we will have have needed to accommodate 769 people in hospital.

…when we hit 100 deaths in Oklahoma, we will have have needed to accommodate 769 people in hospital.

So here are some data on deaths in Oklahoma:

Our first confirmed death from COVID-19 was on 19 March. As of 4 April, 42 Oklahomans have died from coronavirus.

Because raw data can jump around so much from day to day, here I’ve plotted deaths and a rolling average number of deaths. The rolling average was taken from the day listed plus the number of deaths on the previous two days. This 3-day average smooths the variability a bit, revealing a fairly steady increase in deaths.

Doubling time is the number of days it takes to double the number of deaths. Early on in an epidemic there is a lot of variability. The US doubling rate is currently 4 days; global deaths are doubling every 7 days. So far we started at about 5 days, got down to 1–2 days (very rapid increase) and are maybe stabilizing now at 2–3 days. Our overall number of deaths is not comparatively large just yet, but the rate at which they are accumulating is fast.

So how early are we in this fight? Just for my own edification, I’ve considered the day each country hit double-digit deaths as the real start of the epidemic in those countries. This is just to trim off some of the first data points to help see patterns a little better.

I see two interesting things in these data. First, the 9 days it took for Oklahoma to register double-digit deaths is more similar to the 11 days for China than for the 3–4 days for the other countries (and Washington State). This tells me that we’re about as in the dark on our outbreak now as China was way back in mid-January. Next, we are just beginning here in Oklahoma. It was just March 28th that we hit double digits – just 8 days prior to today that I’m writing this (5 April).

That last point bears repeating. Here in Oklahoma we hit double-digit deaths on 28 March, 8 days ago. We are currently under the Safer At Home ordinance in Oklahoma, and will be until 30 April. This is not nearly as stringent a requirement nor as clear a message to citizens as the Shelter in Place order that most other states have adopted. China has fared much better than we have in the US or in Italy or Spain, precisely because they rapidly moved to confine people to their homes and enforced that social isolation. For example, the Wuhan (a city of 11 million people) lockdown began on 23 January – a mere 1 day after China’s double-digit death occurred. In contrast, it is EIGHT DAYS past our double-digit date in Oklahoma, we are still nowhere near as tightly confined to our homes as were the people of Wuhan.

The longer we put off serious social distancing, the longer we will have to do it. It is not enough for those of us voluntarily sheltering in place in Oklahoma when so many still are not. I call on Governor Kevin Stitt to immediately issue a shelter in place order for the Sooner State. Every day we delay will result in needless deaths for our state.