It may be of some interest to people in this thread that in this NYT opinion piece, the author/doctor claims the fatality rate is unknown.
This is a known problem with a ‘fatelity rate’ statistic. The people calculating the value understand the problem. Presumably the statistic is called something else in French or German, but it is what it is.
On the other hand, there is no reason at this stage to believe that ‘fatality rate’ bias makes it inaccurate. Having observed the course of the disease over 100+ fatal cases, there is no reason to believe that the existing cases will suddenly convert to late-stage deaths.
Morbidity and Co=infection rates?
Have we seen any estimates for what % of confirmed infection will require hospitalization (other than just for isolation)?
And presumably a large % of the deaths are from bacterial pneumonia. That’s the way it normally happens with flue or other respiratory problems. Have we seen any numbers on that?
(Thought: Wuhan is at around 150 on the air pollution list? Bit it’s not a good number. That has to be affecting fatality rates. Expect fatality rates to be even worse if it gets into a city with much worse air pollution).
In the initial stages the fatality rate for SARS (severe acute respiratory syndrome) was estimated to be 4%. Later revisions brought it to 14% to 15%. From the same cited article the fatality rate for patients aged 60 and older was 55%. It’s a little hard to understand how they could have been so far off with the initial estimates. If more than half the people older than 59 died from it, how could they ever have come up with a 4% rate?
It makes perfect sense if you don’t wait for recovery before coming up with your rate. That’s the point I’m making. If we have 100 infected and 2 dead, 2 recovered, they’ll claim the fatality rate is 2%. But we don’t really know that yet because 96 people are still sick! People can’t come back to life, so the floor is 2% (assuming no more infections), but the ceiling is a 98% fatality rate. It can only go up.
Imagine a disease that is spreading through a population such that the number of cases doubles each week. Imagine that this disease takes exactly one week to kill you - if you don’t die after one week you will recover. A fatality rate calculated as deaths/infections will understate the actual risk of death by a factor of 2.
But, again, what better measure is there? 1 minus the recovery rate would imply that herpes is 100% lethal, which is certainly badly misleading.
I think the numbers being quoted are actually the case fatality rate. It’s not intended to show the actual fatality of the disease. It’s just the number of fatalities divided by the number of diagnosed cases. And:
Numbers today looked more like 1.4% mortality. 14K cases, 200 some deaths. Flu is running far higher. I’d like to see some data behind the claims of a “high mutation rate”.
Deaths/ Recovered or high likelihood of survival. It’s not perfect but better than deaths/infected. Bottom line is if doctors are using deaths/infected to compare it to threat of SARS, then at this stage they could be making terrible threat response decisions. Better to say “we don’t know” than use a bad metric.
just two nitpicks: norovirus is nasty to have, and extremely contagious but the body “clears” it faster than a cold. People usually recover within three days:
And the common cold is a form of coronavirus. So are SARS and MERS. I’ve been calling this one Wuhan.
I think it’s just been misused by the media. I think doctors know it’s just a very general estimate, and that the numbers will change. Basically, the information is that it’s deadly enough to be a big deal, but not as deadly as Ebola.
The numbers are biased in both directions. On the one hand, most of the infected people are neither dead nor recovered, and I think a lot are still in that limbo where nobody knows how they will do. On the other hand, there may be tons of people with minor infections who haven’t sought medical care. Last I looked there weren’t any people under 15 who were infected. That seems implausible. Far more likely is that young people who contracted it haven’t gotten very sick.
In most cases, the fatality rate is by far the most reliable indicator of how dangerous a disease is.
If you’re going to ignore fatality rates, then you might as well claim that the flu is as dangerous as smallpox, which is insane.
“How dangerous” needs not only the fatality rate, but also how contagious it is.
I think you’re right that, in general, it’s the media that’s misusing it and that legit epidemiologists know well that stat’s limitations. I did find it concerning however to see some doctors quoted as making direct comparisons using the nascent Coronavirus stat with the mature SARS stat. I like my scientists precise and rational.