Florida: Second Wave Analysis
[07/29/20 NOTE: Read our updated analysis for a more up to date recovery assessment: Florida: Summer Analysis Update]
[07/23/20 NOTE: This Analysis Will be Refreshed. Intro Graphs Now Include All Cases]
Current trends in Florida show a significant increase in percent positive test results in recent weeks. Florida has significantly increased the amount of testing. The average age of infection has dropped substantially and there is clear evidence that younger people who are much lower risk of death from COVID-19 are out and about.
This second wave of infections is likely similar in magnitude to the initial outbreak in March, but early testing was limited, so this can be inferred by secondary analysis using our COVID Decision Model. We take close look at the current data and evaluate outcome trends in Florida. Our model is well suited for modeling dynamic heterogeneous effects, especially those based on age demographics. Note that data reporting is delayed, and as such, the model may deviate somewhat from updated data. We have modeled a range of scenarios to assess long term outcomes.
We reference the Florida Department of Public Health for statistics. We present the model results for five scenarios. The details of each case study are presented at the bottom of this post.
We make several observations:
- People are modulating their connectivity in proportion to their risk.
- Infections are up, but deaths will continue to decline.
- The Infection Fatality Rate is declining rapidly and approaching flu like levels.
- The second wave will rapidly decline over the summer and the virus is likely to die out without the need for vaccine if behavior continues to be modulated by risk.
- Florida is doing a good job protecting a large population of elderly vulnerable people.
Despite the criticisms from many seeking to create panic, the people of Florida offer a great example the path forward for COVID-19 mitigation. As the population organically loosens up by age demographic where risks are low, the infection is essentially working through those populations and burning itself out.
From a public policy perspective, it is imperative to educate people to their actual risks so that those who are vulnerable can minimize risky interactions. It is important to continue to protect the vulnerable people who can’t protect themselves in institutions or are under the care. Younger people should continue to minimize potentially infectious contact with vulnerable populations. Good hygiene should continue to be encouraged, including hand washing and wearing masks in close quarter indoors.
This recovery approach provides the best balance of economic recovery and public health.
The Florida Second Wave Model
We have an increase in infections that are likely resulting from an overall increase in connectivity proportional to age (younger people dropping social distancing norms). The loosening of stay at home orders in mid May and widespread civil protests after Memorial Day both contributed to this increased connectivity. This increased social contact for younger demographics has caused a large spike in infections. Death rates resulting from these infections will be very low.
We have modeled a heterogeneous shift in behavior by age group, proportional to actual individual risk. We assess five scenarios to understand the nature of this second infection wave in Florida and a range of outcomes over time. A simulation population of 2 million agents is scaled to the 21.4 million in presented results.
- Case 1 assumes modest social distancing and reduced connectivity in proportion to risk (age). This case best matches the current data trends.
- Case 2 is the same as Case 1 with incrementally more increased contact.
- Case 3 is the same as Case 2 with significantly increased contact going forward. This is not likely a realistic scenario.
- Case 4 is similar to Case 1, with reduced contact in July and then increased contact similar to Case 2 from August onward.
- Case 5 is a light switch return to normal to show a worst case and is presented for reference. This is not a realistic scenario.
Case 1, 2 and 3 are representative of likely scenarios. Death and infection trends for these scenarios are presented below. Death curves are well matched to the existing trends. The infections rate shows a clear second hump due to the age stratified increase in mobility (younger cohorts are more mobile) and matches current % positive test trends. We also show the due to the younger distribution of infections, the Infection Fatality Rate (IFR) has dropped from an initial value of 0.5% to less than 0.2%.
Dynamic Model Input Parameters
The dynamic inputs for each simulation run are shown below. Age cohorts are divided by decade and mobility is inversely proportional to risk. Vulnerable individuals are protected in all cases. We assume an decrease in mortality due to therapeutic advances. We also assume a mild seasonal effect to reduce transmission and mortality.
Florida Death Trends
The model is well matched to current trends. Case 1,2 and 4 are likely scenarios, as behavior will be modulated by increases in infections (a natural response).
Florida Infection Trends
Summary Outcome Tables