The Orlando sentinel headline today reads “Florida breaks state, US record with 15k Covid coronavirus cases reported in a single day, an all-time high for a state”. Let’s take that headline and unpack that further.
As I have said previously, it is nearly impossible for the public to get a true understanding of what is happening with Covid by relying on the media with their focus on headlines and the popular narrative.
We analyzed publicly available data through the Florida Department of Health and the Covid Tracker project to look at aggregate data trends which are summarized in Tables 1 and 2 below. Here are eight insights below the headlines.
- Cases and positivity rates are both increasing. There is no doubt that the higher cases are not just from increased testing alone. Total new cases for the first half of July are 102 thousand with a positivity rate of 19%. This is dramatically higher than the first half of June where new cases were 21 thousand with a positivity rate of 5%.
- Even so, exposure to Covid may be greatly underreported, as much as 11-fold. In published data last week, the CDC showed that official cases reported in South Florida from April 6 to April 10 was less than the actual prevalence by 11-fold. At the time the actual case rate determined by random sampling was 1.65%. The CDC found similar findings in other cities where undercounting varied from 6-fold to 24-fold. Since that time Florida testing has increased 4x. Since the current case report rates are now 1.13% it could be that the current Florida prevalence may be over 7%.
- Exposure to Covid measured by T-cell response could be 5x higher than antibody testing. A study this week from the prestigious Karolinska Institute reported that T-cell immune reactions were present in 30% of blood donors in Sweden which was 5 x HIGHER than other reported antibody prevalence of 6.3% of Swedish blood donors.
- T-cell response may be more important than antibodies when looking at immunity for Covid. Prior SARS-COV-1 research has shown that memory T-cells persist for many years while B cells (antibodies) were more short-lived. Lack of measurement of T-cell response is a serious deficiency in our current testing strategy. While measuring T-cell responses are just starting to move out of the research setting into broader clinical care, we must follow up on the Swedish data and start to investigate the degree of T-cell responses in the US population. If numbers are in the Swedish range, this has major implications for public health strategy (e.g., we start to get close to herd immunity).
- Increases in case rates are being driven by higher case rates in the young. The 25-34-year population has an infection rate that is 2.5 x higher than the 65-74-year population. Older people are protecting themselves from the virus, which is critical given the higher risks that they have. Clearly there is a risk of older people contracting the virus from the young, but this is not showing up in the data yet.
- Covid deaths in people less than 55 constitute less than 7% of the total deaths from Covid. Increased infection rates in this population will likely lead to lower death rates.
- Hospitalization rates and death rates are plummeting, with a reduction of over 80% from the peak. The peak hospitalization rate was 23% in the first half of May and has now plummeted to 3.4%; a reduction of seven-fold. The case fatality rate was 7.4% in the first half of May and has now plummeted to 0.7%, a reduction of ten-fold. These reductions should be making headlines. While the virus is spreading rapidly the amount of harm it is doing has decreased dramatically. There has been a great deal of online commentary that the increased death rates are “two weeks away” but the drop in death rates has been taking place over the last 2 months and is being replicated in other states. We have no reason to believe that death rates from Florida are collected in a different way than other states or have changed over time, but this should be closely reviewed.
- The death rate from Covid for the entire pandemic is 1/20th of the expected “all-cause” death rate. While no death is good, we need to understand death rates from Covid compared to the expected death rate in the population. We calculated the entire death rate from Covid and compared it to the expected death for the time from March 1st through July 11th. For the entire population, the death rate from Covid was 4.6% of the expected “all cause” death rate. For the 25-34 year-old population the death rate from Covid was 1.84% of the expected “all cause” death rate. This is the maximum “excess deaths” since we assume that 100% of deaths from Covid wouldn’t happen if there was no Covid. This certainly is not the case in older adults.
A recent JAMA study looked at excess mortality rates for an earlier time period (March through May) and concluded that in COVID contributed to an extra 6% of total deaths in Florida. So, the excess death rate may be dropping as younger patients getting infected.
To conclude in Florida, the data is taking us in two different directions:
- Cases are significantly increasing and may reflect a reality where the rates are even higher: 7-10%+ if we look at antibodies and 10-30% if we assume that the T cell immune response is similar to Sweden.
- Death rates and hospitalizations are plummeting which is welcome news. If current trends hold excess death rates from Covid will likely contribute to less than 6% of total expected deaths by the end of the year.
Don’t just look at the headlines — think for yourself!
|Dates||New Cases||New Tests||New Hospitalizations||New Deaths||Positivity Rate||Hospitalization Rate||Case Fatality|
|Ju1y 1 - July 11||102,077||530,304||3,462||697||19.2%||3.4%||0.7%|
|June 16 - June 30||75,108||514,811||2,581||574||14.6%||3.4%||0.8%|
|June 1 - June 15||21,163||408,835||1,845||496||5.2%||8.7%||2.3%|
|May 16 - May 31||12,953||412,512||2,460||543||3.1%||19.0%||4.2%|
|May 1 - May 15||9,520||226,432||2,198||701||4.2%||23.1%||7.4%|
|April 16 - April 30||11,179||168,896||2,451||681||6.6%||21.9%||6.1%|
|April 1 - April 15||16,173||152,886||2,521||532||10.6%||15.6%||3.3%|
|March 1 - March 31||6,338||60,597||823||77||10.5%||13.0%||1.2%|
|Covid Cases||Case Rate||Covid Hospitalizations||Covid Deaths||Expected Total Deaths||COVID Expected Deaths|
|0 - 4||4,211||0.36%||91||0||205||0.00%|
|5 - 14||9,026||0.38%||69||2||192||1.04%|
|15 - 24||43,166||1.69%||446||11||672||1.64%|
|25 - 34||55,097||2.03%||1,143||24||1,301||1.84%|
|35 - 44||43,336||1.73%||1,714||86||1,780||4.83%|
|45 - 54||41,085||1.61%||2,458||178||3,695||4.82%|
|55 - 64||32,712||1.12%||3,206||398||9,508||4.19%|
|65 - 74||19.058||0.78%||3,511||868||16,858||5.15%|
|75 - 84||11,148||0.78%||3,188||1,186||25,965||4.57%|
- Data accessed from the Covid Tracking project on July 12th
- State population data from Wikipedia
- Florida Department of Health Covid Report . Accessed July 12th showing July 11th data.
- Social Security Actuarial Table for 2016
- Census projections for 2020, Florida Government
- https://doi.org/10.1101/2020.06.29.174888 Sekine et al. Robust T cell immunity in convalescent individuals with asymptomatic or mild COVID-19
- Weinberger D et al Estimation of Excess Deaths Associated With the COVID-19 Pandemic in the United States, March to May 2020 JAMA JAMA Intern Med. doi:10.1001/jamainternmed.2020.3391