I have crunched numbers, discussed them with advisers who are good with statistical models and observed the modeling done by national healthcare officials.
In this article, I try to parse the arguments and data to explain how there can be such discrepancy – to the point that most mainstream experts are convinced that Florida (and presumably, Miami) is “behind the curve” and likely to become a “hotspot” along the lines of New York state and/or New York City.
After looking at data for a week, I argued my case (as I do once or twice a day) to my Boston-based brother, Mel, who is an M.I.T. engineer and professionally a systems analyst. We both studied mechanical engineering at Villanova and are reasonably fluent in statistical models. We both also have some familiarity, for professional reasons, with healthcare demographics.
So here are the reasons for the discrepancy that leads, for example, the IMHE model (on which my brother greatly relies) to predict that Florida is not yet anywhere close to reaching a peak in infected subjects, hospital admissions and deaths.
I will take them in no particular order.
First Assumption: States Outside New York/New Jersey Are Lagging in the Cycle of Contagion.
It took me a whole week to figure this out. Now I think I know but I am still not sure I can explain.
I remember hearing Governor Cuomo say that his state would likely experience 40% to 60% spread of the virus to the overall population. I wondered, at the time, where he got that figure.
As I have found subsequently, many models make that assumption – including the aforementioned IMHE model that is widely quoted by the president’s task force.
I have not been able to pin down where those assumed rates of contagion come from. I know this: the only randomized testing so far reported publicly was done in San Miguel County, Colorado. It involved about 650 subjects, of which only 1% tested positive; another 1-2% were inconclusive; 97% tested negative.
In Miami Dade, we have tested some of our first responders using the same technique that we plan to use for a random test of 7,000 residents, using FPL data banks and methodology designed by a University of Miami public-health expert. About 5% tested positive, and it is entirely possible that those are people who have been exposed to strong viral doses.
Again, I have no idea where the models being touted base their assumptions as to the overall spread of the virus. We do know that about 10% of those with symptoms test positive. And we have very recent figures from the City of Miami that show only 7% of those tested with the contagion, but that reflect some self-selection, as some of those tests were done in the convenience of the subject’s home. (And it may also reflect some late-reporting results.)
Are our low rates of contagion, or of mortality, due to an assumption that those infected, presumably overseas, entered the United States through New York, rather than through Miami? If so, I have seen no data whatsoever to back that model of geographic dispersion.
Second Assumption: Governmental Actions Are Not a Big Factor in the Initial Spread of the Virus
In the United States and Europe, the decision to stop all human interaction – or to implement mandatory “social distancing” came perhaps a bit late, in the opinion of the experts, to stop the virus cold.
On this issue, we can certainly agree that New York should have stopped all subway transit, New Orleans should have canceled Mardi Gras, and Washington State and Broward should have stopped visitors at nursing homes.
Perhaps New York City should have required all persons using elevators to wear masks. (If not, maybe it’s not too late to implement that measure.)
Miami, as stated in my prior article, took prompt, and probably effective, steps in canceling Ultra and Calle Ocho. As a very horizontal metropolis (2,400 square miles) compared to Manhattan (22 square miles), which implemented social distancing, shut down hotels and restaurants, and prohibited large gatherings, it is entirely possible that we will never follow the New York model of either infection or mortality.
Lately, we are maintaining a constant rate of deaths to confirmed cases (about 0.8% in the county), which is in stark contrast with New York and New Jersey, which are closer to the worldwide norm of 5%.
Third Assumption: The Virus Spreads at the Same Rate in the North as in the South
This assumption is one I just don’t understand. Almost universally, the southern states, from California to Florida, are experiencing much lower fatality rates, in proportion to confirmed cases, than the northern, urban states. Typical is Florida, which has consistently stayed at about 1.7%; California is at 2.3% and Pennsylvania, which is much colder, is even lower at 1.3%.
In conclusion the models being thrown around all seem to suggest that Florida and California are somehow “behind the curve” in relation to New York and New Jersey.
But that may not be the case at all – particularly if those spreading the virus are limited in their travel, or at least required to prove that they are contagion-free.
In that regard, I must, once again, give kudos to the mayor of Miami, who recently asked the president to restrict passenger travel from northern “hotspots” to our fine city, which has no observable reason to become a “hotspot,” except in the meteorological sense…