We are in an upwardly creeping plateau even before the recent event-filled holiday week and school openings.
As had already been reported, both new cases and deaths during the week of August 30 through September 5 were the highest since Covid-19 arrived in Kentucky– even without the major outbreaks in prisons or long-term living facilities that boosted earlier bad weeks. Day-to-day variation of reported numbers due to lack of capture of data over the weekends grossly distorts epidemic curves, confounds the identification of trends, and reduces considerably the public health effectiveness of case identification and contact tracking. The resulting volatility is evident in the figure below. (See also the interactive data visualizations of all the figures below here.)
The saw-tooth shape of such curves is also evident in tallies of national data and has lead those tracking the epidemic (including KHPI) to use curve-smoothing averaging methods such as reporting rolling 7-day or even 14-day averages or, as illustrated below, comparing the number of cases or deaths per week.
Comparison of 7- and 14-day rolling averages of new cases.
When the red 7-day line is above the blue 14-day rolling average, the overall rate of new cases is increasing. The converse situation of 14-day average being greater has happened infrequently for short periods. As we emerge from an action-packed Labor Day season punctuated by the Kentucky Derby; the opening of schools, colleges and universities; and the promise of continuing large-scale sporting events; in my opinion we will be seeing a much more dramatic red-over-blue scenario in the next few weeks.
The number of deaths last week leapfrogged over even the bad weeks of April and early May. If the accuracy of hospital utilization reporting is to be believed, we are busy but not clinically overwhelmed.
It is clear from multiple reports across the country that clinical laboratories are becoming overwhelmed again because of both high volumes and ongoing limitation of supplies. Less-accurate (more false positive and negative) antibody and antigen tests are now counted along with the definitive viral RNA tests. It goes without saying that management of our epidemic, including case identification and contact tracing, depends critically on timely testing and reporting.
What is happening around the state?
I do not see things slowing down. Even with data current only to September 4, multiple Kentucky counties are considered to be “in the red,” indicating a rapidly rising number of new cases. The epidemic curve for Jefferson county below is compatible with “catching fire.”
Green and Todd counties made last weekend’s short list of national hotspots for new cases per 100,000 inhabitants. Both these counties have added new cases since the date of these plots. I believe the use of the term “exploding” is not unreasonable. Can anyone suggest what is going on in those two counties, or in any of the other counties with similar expanding epidemic curves that can be identified in KHPI’s interactive data visualization?
What is to come?
Even as a “glass-half-full” kind of guy, I am more than worried about what is poised to occur. Major stadium events and openings are planned even as we wait to see the outcomes of recent Labor Day week events. (We will not catch up with reporting and analysis for several days yet as the usual weekend pause is longer.) Surges of cases followed Memorial Day and Fourth of July holidays and concurrent openings. School districts across the state have already reported hundreds of cases in students and staff. As I bicycled past a school near Rock Creek here in Louisville a week ago, students where being let out to the adjacent public park for athletic activities– including huddles with coaches– but were largely unmasked. The parking lot of a small bar on River Road was packed. As of September 9, the University of Kentucky reported to the state that it had identified 991 cases in students. For all colleges and universities that number was 1469. For the University of Louisville, the state’s public health report for colleges and universities lists only 8 cases. (The list of colleges and Universities reporting is grossly incomplete. UofL is currently the only school in Jefferson County included on the list!) Clearly there are more cases at UofL than 8 as they prepare to open their football season next week. Which is more important, keeping open the option for in-person classes or hosting major stadium sporting events? At UofL or anywhere? Which is the tail and which is the dog? A New York Times analysis recently identified college and university towns as key drivers of Covid-19 infections in their communities. It can happen in our state too. Should not our priorities be in taking care of our students of all ages, and protecting the health and financial survival of individuals in the community? None of us will be either safe or healthy until the least and most vulnerable of us are.
Peter Hasselbacher, MD
Emeritus Professor of Medicine, UofL
September 8, 2020