It is mathematically, and morally indefensible to continue to hold up New York, either as a model for handling coronavirus, or as a bogeyman for the rest of the country if we as citizens fail to "contain" the virus. This has not stopped multiple U.S. public health officials proffering NY as a model of "crushing the curve." But let's take them at their word. What would it look like if the whole U.S. had been as "successful" as New York in crushing the curve? Well, as you guessed from the title, we'd have 565,800 dead--and 330,000 people hospitalized at the peak.
As of October 26, 2020, New York has the country's second worst death rate after NJ, at 1725 deaths/million. The country as a whole stands at 697. New York's death rate is more than 2.5x that of the country at large. Figure 1 below shows what the country would have looked like as a whole, had we "succeeded" like New York in crushing the curve.
Figure 1: Hospitalizations & Deaths in NY, All of the U.S. excluding NY, and Deaths & Hospitalizations Nationally, if the rest of the U.S. had "Succeeded" like NY
Source: COVID tracking project. Analysis by Emily Burns
If the entire U.S. "had succeeded" like New York, in addition to having more than 350,000 additional dead, we would have seen more than 13,000 dying daily at our peak. This is in contrast to the than 2100 we actually saw, 700 of which (or 1/3) were contributed by New York alone. But more to the point is the number of hospitalizations. New York, at its peak, had nearly 19,000 people hospitalized. This is almost half of the number that are hospitalized nationwide now. If the entire country had "crushed the curve" by allowing a large spike like New York, we would have had 330,000 people simultaneously hospitalized--35% of all U.S. hospital capacity (925K staffed beds). As it stands, we have never had more than 65,000 people hospitalized (8%), and when we did, nearly a third of those people were in New York. Now, Governor Cuomo has stated recently that New York's hospitals were never overwhelmed. Perhaps that is true. Even if we were to reach this catastrophic 300,000+ number, we would still be using "only" 1/3 of our hospital capacity, and might, like New York, not have been overwhelmed (we currently have around 275K hospital beds free in the U.S., and are using 43,000 for COVID patients).
As for ICU usage, at the moment, COVID ICU usage runs at roughly 20% of all COVID-hospitalized patients. If such a surge had materialized it would mean around 60K ICU beds in use by COVID patients nationwide, or 60% of the 100K ICU beds available. This is undoubtedly less than the percentage of ICU beds that was in use in New York, given that in New York, 33% of COVID hospitalizations were in the ICU at the peak. This suggests, contrary to governor Cuomo's claims, that hospitals in New York were overwhelmed, and were likely turning away many patients who ought to have been hospitalized--which may explain at least some of their elevated death rate.
In a "by the numbers" analysis, had the entire country "crushed the curve" New York fashion, we, like New York, might not have overwhelmed the hospitals. But the question remains, why would we want to? Why is this being pointed to as success, when the country as a whole has 40% fewer deaths per capita, and, at our very peak, 80% fewer hospitalizations (again, per capita).
The reason of course, is that our epidemiological overlords are simply enamored of just how "flat" New York's curve is now. This deliberately ignores the likely impact of acquired immunity being at least in some measure responsible for this flatness. To test this assumption, let's take a look at Sweden, which famously never locked down, and never had a mask mandate. Sweden's population is 10 million, New York State's 19.5 million. Stockholm has roughly 20% of its population showing antibodies (evidence of prior infection from coronavirus) to New York City's 25%. Below are the death curves for both areas.
They look pretty similar, don't they? There are four aspects of Figures 2 and 3 that bear highlighting. 1) Sweden's curve is actually "flatter," suggesting, along with their lower antibody counts, that their mild mitigation measures worked better than New York's more draconian measures to flatten the curve. 2) New York's curve looks far more like a natural flu/pneumonia curve (shown below, in Figure 4), adding credence to the theory that New York's curve was flattened not due to lockdowns, but due to the disease burning out. 3) At only twice the population, New York has 10x the number of deaths on its deadliest day (and more than 3x Sweden's deaths/capita total)--this doesn't seem like an argument in favor of the NY approach 4) Neither locale has registered any discernible increase in deaths since returning to baseline, this despite Sweden and New York both having a case "surges," Sweden's being equivalent to its "high" back on June 24th ("high" because Sweden's actual peak cases were clearly in the beginning of April, as deaths are the only reliable measure of cases, giving wildly variable testing regimes throughout the pandemic).
I believe that these four things point strongly to New York's current "flatness" (as well as all other hard-hit northeast states) not being due to the success of their mitigation measures, but actually to their having achieved whatever level of population immunity is necessary (which may have far more to do with strong T-cell responses in younger populations than antibody responses) for the virus to cease being a major health concern in a given area. Of course, achieving some level of immunity is not the reason that New York is being touted as a success story--in fact, it's never even discussed, because antibody levels are not at the 60% that epidemiologists estimated at the beginning of the epidemic. This ought to be part of our discussions (which we could ascertain with T-cell studies, that have, pardon the pun, been studiously avoided in the U.S., as well as simply opening up New York's society). To the extent that this is true, identifying the causes for New York's stratospheric mortality rate versus other countries and states that took a lighter touch for the same immune benefit must be part of the conversation.
Figure 4: Monthly Flu and Pneumonia Deaths
At the beginning of this post I stated that holding up New York as a model was both morally, as well as mathematically indefensible. I think I have demonstrated the truth of the mathematical indefensibility, now I will turn to the moral indefensibility. I believe that New York has built whatever level of herd immunity it has on the backs of the most vulnerable in its population--minorities, the poor, the elderly. Let's return to the rates of death in Sweden and New York. Why is it that Sweden, without locking down has a 3x lower mortality rate than New York? The standard answer is, "well, the nursing homes, and the minorities, because minorities die at higher rates from COVID, therefore, naturally New York would have a higher death rate than, say, Sweden." Yes to nursing homes, but let's look closer at minority deaths. If, instead of looking just at deaths of minorities, you look at cases AND deaths, you see very quickly that minorities not only die at higher rates, they are also INFECTED at higher rates. See the chart from the CDC below.
The discrepancy in deaths between minorities and whites from COVID-19 has been portrayed as an example of "systemic racism" which over decades has created generally worse overall health for minorities. The chart above provides a powerful argument that minorities are dying not because they have worse health, but because they are exposed to the virus at higher rates. I believe their is a causal relationship between their excess exposure and our lockdown policies. Indeed, our public health officials are not just aware of this, but openly admit it. And yet, despite a policy which is blatantly classist, if not outright racist, they persist in pushing lockdown, or lockdown light policies, the result of which is to infect the poor and protect the the rich.
Perhaps, in the most cynical of worlds, it would be acceptable to sacrifice the most vulnerable in your population in service of the greater good of "stopping the virus". But when it becomes clear that you are not stopping the virus, that the virus cannot be stopped, that whether by design or by accident, you are moving towards some kind of population immunity, such a policy is simply barbaric--well beyond systemically racist. New York is not an example of coronavirus success, it is an example of the grossest failure of accidental herd immunity--literally a "survival of the fittest" perversion of herd immunity. The death rate in New York is what happens when, instead of protecting the vulnerable and using the acquired immunity of the strong to protect the weak, you use the corpses of the poor to build an immunological island on which the rich can ride out the viral storm. It is a model for nothing except how to exploit deaths of the most vulnerable for political gain.
The Next "Wave"
There are many epidemiologists and public health officials sounding the alarm that if we don't take drastic measures, we will, as a country look like New York, that we will have our deaths go up to over 500,000 by the end of February. They are basing this on the current increase in cases, which they cite as doubling of the 7 day average cases from 34,000 on 9/12 to more than 68,000 as of yesterday. However, testing has also increased more than 60%. If testing were held steady at its rate on 9/15 (shown by the green line in Figure 6), cases track exactly with hospitalizations. Further, we see that we are not indeed at our highest number of cases yet, we are, as deaths would tell us, approaching another hump, but one that is much smaller than the first two. While hospitaliaations are up, they are up by half, not double, and still 25% below the July peak. The end of February is approximately 120 days away. To reach this 500,000 number, we would have to AVERAGE 2500 deaths/day for every day between now and then. We have never gone above 2100 for a 7-day average, and we needed the disaster of NY to "achieve" that. We are currently at 800 deaths/day. Hospitalizations would need to surge to 115,00 (nearly 3x where they are now), and remain there from tomorrow until the end of February.
Figure 6: U.S. Cases, Deaths, Hospitalizations, Testing and Estimated Cases based on Consistent Testing through 10/26/20
Source: Source: COVID tracking project. Analysis by Emily Burns. Estimated cases based on consistent testing based on reference date of 9/15 (could chose any date): Estimated Cases on Date = 7-Day Avg Observed Cases on Date/(7-Day Avg Number of Tests Performed on Date/Testing Baseline on 9/15). On 9/15 there were 700,000 tests performed.
Saying that we are headed into the darkest days of the pandemic ignores all other data beyond cases, which is the most fickle of our data points. This is pure fear-mongering that bears no resemblance to what the data show us. It is based on extrapolating from the European second wave, which ought rather to be understood as more similar to our summer "second wave," but pushed further out due to Europe's harder initial lockdown. It is worth noting though, even if we were to see this step function increase in hospitalizations and deaths, that level of hospitalizations would STILL be 1/3 the level of hospitalizations seen when New York was "crushing its curve". Similarly, if through some sort of dark magic we were able to "achieve" this step function increase, the country as a whole would STILL have a lower death rate than New York's. But this is not going to happen. The rest of the country flattened the curve, we didn't crush it. We will continue along our gradual curve, likely with a slight bump here and there, then gradually resume our trend towards a low-level endemic disease with no excess mortality, until we have a vaccine. At which point those people who might have been taken off by coronavirus, will instead be escorted out of this world by flu.