In 2012, a new coronavirus emerged in Jordan and quickly spread to Saudi Arabia. Eventually, travelers to and from the Middle East spread the virus to 21 countries, resulting in over 2,000 infections and 600 deaths—a case fatality rate of over 30%. However, the transmissibility of Middle East respiratory syndrome, or MERS, turned out to be low, with almost all transmissions happening between people who were in prolonged close contact, such as family members caring for someone who was already ill. The basic reproduction number of the virus, called the R0 value—the number of people each infected person might be expected to infect—was set at less than 1. As terrible as the outcome of MERS was, by the end of that year, the outbreak was over.
Then it came back. In 2015, a traveler from Saudi Arabia visited South Korea. Within days, a burst of new cases—and over 100 deaths—resulted, before this smaller outbreak was brought under control. In this second instance, the R0 value for MERS was estimated at between 3.9 and 5.0. Why there was this difference between the first and second outbreaks remains unclear. Since then, MERS has simmered. It’s been found in a total of 27 countries, and has resulted in a total of 858 deaths. It’s been a constant, stark warning of what could happen with a coronavirus that paired high rates of transmission and mortality.
As the number of deaths from 2019-nCoV moves into four figures, there’s been continued concern that the publicized case fatality rate of around 2% may not capture the actual nature of what’s happening. So today we’re doing a deep dive … into death.
Before we get into the even darker numbers, here are the charts for Feb. 11.
And if it looks, once again, like tracing your finger along that curve suggests there’s an end in sight, that’s because the daily numbers can be given a somewhat hopeful spin when it comes to new confirmed cases.
What looks much less encouraging on this particular day are the outcome numbers. These are particularly thrown off because Hubei province in China, the area around Wuhan, reported about 30% fewer recovered cases than on Monday and a greater number of deaths. Most of the blob of new cases outside China is in the same place that it was on Monday—the poor Diamond Princess, where people are enjoying a cruise that has to be in the top ten on the list of awful.
Novel Coronavirus Outcomes
|
Recovered |
Died |
China |
4,049 (+380) |
1,016 (+106) |
Outside |
48 (+6) |
1(-) |
Within China, 7,333 cases are currently listed as serious or critical. Outside China, 12 are listed as serious.
Since I started tracking the 2019 novel coronavirus outbreak, I've gotten more questions about various measures of mortality than anything else. It’s also dominated the comments on several days. So with the grim milestone of 1,000 deaths now in the past, here are a couple of different ways of considering that issue.
The classic way of presenting this issue is case fatality rate (CFR). This is the number of deaths divided by the number of confirmed cases. This is the number you most often hear on the news, as well as the number that’s usually presented when looking at past outbreaks such as MERS and SARS. It is about the best thing available when looking at the question, “If I catch this thing, how likely am I to die?”
The dates on all these mortality-related charts are those for which I’ve been able to locate data that seems reliable—or at least, as reliable as anything else we know. In this case, after a bit of scrambling in the early days, the case fatality rate settles down around 2% and stays there. So when television news provides this number, it’s absolutely correct.
That value might not seem that awful until you consider that the comparable CFR for flu is around 0.1%. Now one day of reporting on 2019-nCoV has passed without several comments that the flu is already a much greater pandemic. After all, the flu killed 10,000 people in the U.S. in the last year alone. However, the number of confirmed cases of flu was around 20 million. If 2019-nCoV reached that kind of of spread in the U.S., the number of deaths wouldn’t be 10,000. It would be 400,000. That’s what makes people sweat.
But of course, some people are deeply concerned that a figure of 2% underestimates the potential threat. The itching worry for many is that the CFR, by using the total number of cases at a point when the number of cases is growing rapidly, isn’t accurately mapping outcomes. After all, think about a disease that is doubling every day: 10 cases on Monday, 20 cases on Tuesday, 40 on Wednesday, and so on. If everyone who got sick died on the third day, the reporting on Monday and Tuesday would be that CFR was zero, and on Wednesday that it was 20% … but all the while, everyone would be dying.
With the number of cases of 2019-nCoV still rising rapidly, could that number of confirmed cases be disguising the real chance of death? Instead, many have suggested looking at what we know of outcome mortality—that is, out of all the cases that have resolved in either recovery or death, what percentage of patients have died?
And when those are seen as a percentage:
This outcome mortality view is considerably more grim. In the first week of tracking it generates numbers even worse than the CFR of MERS, and even as time goes on it still produces a value 10 times that seen in the CFR values.
But what’s being seen here is largely a factor of time. What these numbers indicate is simply that it takes longer to fully recover from an infection of 2019-nCoV than it does to die, which is true of most infections. There are diseases from which you either recover quickly or progress toward a much bleaker outcome, but this doesn’t seem to be one of them. It looks like it takes somewhere over 10 days, maybe even 14 on average, to recover from a bout of 2019-nCoV—a recovery period similar to that of the flu. Those people who die don’t make it that long.
Think again about our doubling rate of disease in the last example, but now think that outcomes are divided equally between recovery and death—only, full recovery takes five days. So long as the rate of new cases keeps going up, neither CFR nor outcome mortality ever presents a good picture of what’s happening. Within three days, CFR locks in at 13% and stays there as long as the number of cases is going up. Too low. Meanwhile, by day 5, outcome mortality freezes at 80%. Too high. Neither gives a good picture of what’s really happening in terms of the prognosis for someone who comes down with this doubling fever—which is a 50% chance of death.
So what’s the truth about 2019-nCoV? When the numbers are all tallied and this outbreak is in the history books (the hopefully very-near-history books), the number is likely to be greater than 2%. It’s also going to be considerably less than 20%. And there’s a particular factor that may be shaping the outcomes at the moment—one that suggests that the key to reducing the mortality rate for 2019-nCoV is tied directly to getting down the number of cases.
In the 2012 MERS coronavirus outbreak, 80% of patients needed respiratory assistance at some point in their treatment. That included the majority of patients who survived the infection. Right now, of the cases of 2019-nCoV that have occurred outside China, only one has ended in death. That’s a CFR of 0.02%.
What this suggests is that a large number of the deaths now happening in China, and in Hubei province in particular, simply represents a healthcare system that’s overwhelmed by the number of cases. So long as this novel coronavirus is held in check at the point where cases can be dealt with with a high level of care, the survival rate would seem to be very good—much more like the flu. But when the system is simply swamped with sick people, the rate of deaths goes up enormously.
So think about it this way: Right now, there are people in China who are being held in quarantine conditions, without the kind of care this disease requires, in the name of preventing the outbreak from becoming a sweeping epidemic within China and a nightmare pandemic around the world. Many of those people are dying—hundreds of them, almost certainly thousands before this is over—so that you, and your family, will not face a system overwhelmed by the kind of disaster it would represent for novel coronavirus to emerge as a pandemic. That may not be the intention of the men who put those ill people into converted gyms or warehouses, but it will be the legacy of those who are dying.