Last week was not a good one in the Coronavirus department. For some of the smaller or more rural counties, the week was frankly terrible. Because of the highly variable day-to-day numbers of new cases reported, it is usually dicey to make predictions about what tomorrow will be like. Even the use of 7-day averages does not “smooth out the curve” as much as one might like. Some analysists are shifting to 14-day averages for this reason. Looking at weekly totals is one easily understandable version of a 7-day average. It provides a rational and understandable way to minimize the distortions of using a 5- or 6-day work week to describe a 7-day virus. KHPI has been using a Sunday to Saturday weekly interval to aggregate data. Last Saturday, August 22, closed out the most recent full week. The results are displayed below. (The full panel of fully updated interactive data visualizations remain available on the KHPI Tableau Public website.)
The following graphic displays the daily counts of new cases (less duplicates and the like) current through August 22d. New cases remain high, with both high and low days dwarfing the counts of the first 3 months of our epidemic.
The next graphic displays the full weekly totals of new cases. This week gave us the third highest total of new cases in Kentucky ever. I am unaware of any “special” boluses of new cases such as from prison or nursing home populations that would have caused a big jump as we have seen for other new highs. We certainly cannot yet say that the number of new cases is trending down.
On the other hand, the number of deaths last week was the highest ever for Kentucky. We knew this was coming. All indications are that this next week will be worse.
Hospital and ICU utilization remain high. Hospital data has been tricky to gather nationally and in Kentucky. It would appear that the number of current patients in ICUs is trending upwards. That is likely a predictor of any future increase in deaths.
Today (Monday) is traditionally a low reporting day, following as it does Sunday, a day of partial epidemiologic rest. With 463 new cases and 9 deaths, neither were demonstrably low on Sunday. We will see where we are this evening (Monday) but it is Tuesday’s report that may is more likely to provide a hint of what is to come.
Covid-19 in individual Kentucky Counties.
The number of cases in the big-four counties continues to increase steadily. I am now able to display the epidemiologic curves of each of Kentucky’s 120 counties individually. I believe there are many lessons to be learned if analysists and other interested pesons on the local ground can interpret them for us.
For example, above is the case curve for Green County. I read in the papers that the county opened and then quickly closed its public schools because of the identification of Covid-19 infections. I think it is fair to say that Covid-19 infections are exploding in Green County. What I do not have any feel for is how the timeline of reported new cases compares to the school calendar. Was the virus well intrenched before schools opened, or did opening schools accelerate the expanding number of daily cases beginning in July, or both? Remember that it takes a few days after exposure for the disease to be symptomatic, then diagnosed, and then reported. The longish interval between contact and report can be significant practically. For what it is worth, as of August 22, as many at 1% of the residents of Green County are already infected. Deaths in Green county are rising too. Of the 5 deaths so far, three occurred in the last 2 days displayed in this graphic.
I am requesting the help of my readers. Click through the individual counties in KHPI’s interactive county visualizations. What stories can you extract? What lessons are there for us to learn. How can our home-grown experiences inform the inevitable necessary public policy, educational, and business decisions that confront us?
Peter Hasselbacher, MD
Kentucky Health Policy Institute
August 24, 2:00 pm.