Ah, well then. My mistake.

2/4 -This is not the case. Indeed, if anything, our latest estimates suggest that the virus is slightly more transmissible than we previously thought. Our lethality estimates remain unchanged.

— neil_ferguson (@neil_ferguson) March 26, 2020

4/4 – Without those controls, our assessment remains that the UK would see the scale of deaths reported in our study (namely, up to approximately 500 thousand).

— neil_ferguson (@neil_ferguson) March 26, 2020

Reading this New Scientist piece, it sounded like Ferguson had lowered his estimate of the death toll based on news from Europe that the virus was more easily transmitted than thought. That fit with the assumption in the Oxford model of the disease that many more of us are infected than anyone realizes, which means the number of people hospitalized right now is just a tiny fraction of the total infected population. The severity of the disease has been wildly overestimated. We might be well on our way to herd immunity and putting the worst of this crisis behind us in the next few months.

That’s not what Ferguson is saying, per his clarification. His opinion of the disease’s severity hasn’t changed since the dire Imperial College study. What’s changed is people’s behavior in the U.S. and UK since it was published. The only way to slow down infections and minimize deaths, the study claimed, is to implement strict social distancing measures. The Oxford model’s not going to save us; we’re in this for the long haul. Either tamp down the spread of the virus immediately or pay a supreme price.

Well, Boris Johnson and Trump came around to that view. (Whether Trump sticks with it is a separate question.) And so, Ferguson is happy to report, Britain and the U.S. are on a better path now. Here’s a table from the original Imperial College study projecting deaths based on a variety of social distancing measures. “CI” is “case isolation,” “HQ” is “home quarantine,” “SD” is “social distancing,” and “PC” is “school/university closure.” R0 is the assumed rate of infection:

If all four strategies are being used simultaneously, as they are at the moment, you see how low the death total can drop in the fourth column.

Ferguson imagines those social-distancing strategies being turned “on” and “off” as needed during the 12-18 months he expects the outbreak to last, believing that the virus will inevitably begin to spread again during periods when we’ve relaxed our measures to contain it. Researchers at Harvard came to the same conclusion. The good news is that we shouldn’t need to be as draconian with social distancing in future periods when we’re forced to turn it back “on” to contain the spread. That’s because people who’ve been infected in prior outbreaks presumably will acquire immunity; the disease might not spread as rapidly in each successive outbreak as it did in previous ones.

The bad news is that we may have to go through several periods where, say, schools close and mass gatherings are canceled. That’s what the “trigger” column is about in the Imperial College graph. Once a critical threshold of severely ill patients is reached, that would be the “trigger” for a new round of heightened social distancing until the load on hospitals lifts. Or, of course, we could all just cut our number of daily contacts waaaaaay back for many months and produce a graph like this, which researchers in Texas presented today as a model for how things might develop in Austin:

You can find the full UT Austin report here: https://t.co/lqHCVMnC8E

— Alyssa Goard (@AlyssaMGoard) March 26, 2020

Eliminate 90 percent of their daily contacts — no sweat! — and Austinites can prevent overloading local hospitals well into late summer. Anything less strict than that and hospitals will reach capacity sometime soon-ish, if, that is, the model’s assumptions about the virus (e.g., one in 20 patients requires hospitalization) are correct. The point is, for both Ferguson and the Austin researchers, disciplined social distancing is the only game in town when it comes to averting catastrophe. (Especially if the transmission rate is higher than we thought, per the new European data.) We’re not going to be locked away in our homes forever, but we at least have to be prepared for major closures and disruptions of public spaces sporadically for the next year or more in order to douse the latest outbreak once it begins to burn again aggressively. There’s no easy way out.

But if you’re clinging to the Oxford model for hope, as we all are, here’s a new argument in its favor from a retired British pathologist. His logic will be familiar to you by now: There’s circumstantial evidence that the universe of infected people is much bigger than we believe, which means the rate of hospitalization and death is much smaller than we believe. In relatively rare cases people will suffer and die; an extremely vast majority of us, if not most, will have mild symptoms or none at all. The thorny question that may be on the horizon if antibody testing confirms that the death rate is low is how much social distancing we should apply to try to lighten the load on hospitals anyway. If “only” 30,000 people are expected to die under the Oxford model but American doctors and nurses face being besieged for, say, three months by the inundation of patients as we reach peak infection, what’s to be done? We’re not going to shut down the whole economy for a disease that kills about as many as the flu, but we’ll have to shut down some things for a disease that causes way more pressure on the U.S. health-care system than the flu does, and more than it can bear. Which things do we close?