June 29, 2020: EPI's COVID-19 situation analysis (newest)
A new EPI analysis examines COVID-19 statewide cases by age, plus health care trends at the state and county level.
- Thomas Hladish, Ph.D., Department of Biology, and the Emerging Pathogens Institute, University of Florida, Gainesville, USA.
- Arlin Stoltzfus, Ph.D., National Institute of Standards and Technology.
- Kok Ben Toh, M.Sc., School of Natural Resources and Environment, University of Florida, Gainesville, USA.
- Sanjana Bhargava, B.S., University of Florida, Gainesville, USA.
- Alexander Pillai, Microbiology major, University of Florida, Gainesville, USA.
- Dianela Perdomo, Biology major, University of Florida, Gainesville, USA.
Florida is beginning to approach health care capacity in some places, but not yet in any of the state's larger counties. Our concern at the moment is for small hospitals in sparsely populated areas that could become overwhelmed. We estimate that reaching health care capacity at a large scale is more than a week away. Here, we document a wave of relatively young people testing positive, and we predict that this wave will cause "spillover" events into older age groups that will then have a disproportionately larger need for health care services in the coming weeks. We also analyze health care trends at the state and county level related to hospitalizations and use of intensive care units.
Case trends by age
The collection of four graphs below show positive COVID-19 cases reported by age group from March 2 to June 25. (Top two rows, left to right: 0-24 years and 25-44 years; bottom two rows: 45-65 years and 65 years and older.)
Cases are rising fastest in the 25- to 44-year-old age group compared with all others. Cases in the 65 and older age group are rising the slowest. (The shaded region represents June 11 to June 25, and our graph may potentially be revised for these dates as information is refined or made newly available.)
To further explore this relationship, we compared the daily reported cases with ages of those testing positive for COVID-19. The top graph below shows the total cases reported each day, while the bottom graph depicts trendlines for the mean age (or average age) in blue and the median age in red.
The mean is a calculation showing the average of all ages. As the graph shows, in this case, the mean skews in recent days to people in their late 30s. The median, on the other hand, calculates the point in the ages where roughly half of the cases fall below, and the other half fall above. The median in our graph began to skew much younger compared with the mean starting in May. After a long downward trend of mostly younger people testing positive, we are just beginning to see the median tick up in recent weeks, from people in their early 30s to those older than 35 in recent days. This reinforces our prediction that mass cases in younger people will "spillover" into older age groups that will likely cause a greater burden on our health care system.
Statewide health care trends
The graphs below show statewide trends in health care between March and late June.
Columns on the left show trends in hospitalization, and columns on the right show trends in the use of intensive care units (ICU).
- Row 1 displays identical information to reinforce time frames in the columns below
- Row 2 displays information about occupancy rates.
- Row 3 displays information about patients hospitalized with COVID-19.
- Row 4 displays information about the proportion of patients hospitalized with COVID-19 out of all hospitalized patients.
(Note: these graphs are available at full size in a linked PDF at the end of this post.)
Select county health care trends
We evaluated the same criteria shown above for all 40 Florida counties that have at least one ICU bed. Below, we show select graphs for five counties:
All of our data is available in a PDF that includes graphs for all 40 counties included in our analysis. If a county is not listed, it's because it lacks an ICU bed and did not fit our evaluation criteria. We emphasize again that our concern at this time is for small hospitals in sparsely populated areas that could become overwhelmed.
For questions about this data, please contact Tom Hladish.