Funny question about death tolls

[Jens]

This is something I've sometimes wondered about; maybe it's a bit insensitive. But anyway.

Whenever there's a natural disaster like a hurricane, death tolls are mentioned; 1,000 or whatever. Sometimes during even a minor storm, the news says, "as a result of the storm, three people slipped and were injured" or whatever. But what I always wonder is, a certain number of people are going to slip and injure themselves on every single day, regardless of whether it's a storm or not. It could be clumsiness, or a broken heel, or whatever. And in the case of the hurricane, many of the victims were (I think) elderly people, and elderly people tend to die at a higher rate than younger people. There was one instance in New Orleans where elderly patients were found dead in a hospital. I assume that they were fairly sick patients, and so you'd expect that some would die whether there was a storm or not. So what I'm wondering is, when they announce statistics on death rates, are they assuming that all the deaths are a direct result, or do they calculate it above the "background death rate" that you would expect for people in a normal situation.

Again, no offense intended, it's just a question about the use of figures.

[Frog march]

Death figures are only estimates anyway. I wouldn't have thought that they would bother with adjusting for natural deaths. Old people don't die that often.

[Candy]

I want to witness this future discussion. :)

[Lianachan]

Funny you should mention that - I'd just been looking at the UK earthquake death tolls (http://www.quakes.bgs.ac.uk/hazard/faq1.htm#FAQ14) and was wondering if they are all a direct consequence of the earthquakes.

[Fram]

One can wonder indeed how a suicide can be attributed to an earthquake. That's like saying Marilyn Manson (or Mariah Carey or whoever, just an example) kills five people. Falling rocks and masonry is of course something more of a direct consequence...

[farmerjumperdon]

Incredible as it may seem, the Probability of Dying (http://www.cdc.gov/nchs/deaths.htm) of a person between 80 and 81 years of age is only 7% (besides, in the US, an 80 year-old person can expect to live 88 years, while a newly born can only expect to live 77 years). Thus, the high death toll among the elderly after a disaster can be easily credited to the disaster itself.

Interesting how those stats come out. Which would you rather be?

1 - A newborn being told you can expect to live 77 more years, based on averages, but with no promises you will even see the next day, or . . .

2 - An 80 year old being told you can expect to live 7 more years, based on averages , but with no promises you will even see the next day.

[Argos]

Oops. I accidentally deleted my previous post.

I´m not sure if the "probability of dying" stats include deaths by disasters (I suppose it does). So, my reasoning above must be incorrect.

[gethen]

Perhaps I'm wrong, but I always thought that those estimates were billed as "Hurricane related" or "attributed to the tornado," and that deaths from old age or such were not counted.

[HenrikOlsen]

Based on numbers available on the net (National Center for Health Statistics (http://www.cdc.gov/nchs/)):
In Orleans, coext. with New Orleans City, 5440 deaths in 1999
this would mean about 15 deaths a day normally.
They do have an additional classification saying hurricane related, but it's very restrictive.

I just checked their guidelines, they have a lot to say about how they classify things, many deaths and injuries we would call hurricane related will not be classified as such.
Examples explicitely mentioned in the guidelines I found are: a police officer injured trying to stop looting, and people getting hurt in an evacuation where neither will go in the statistics a hurricane related.

[Gillianren]

as for the hospital, I would find it unlikely that, what, forty-five people would've died in it during the few days involved. so did the people who worked there.

I know heart attacks get attributed to earthquakes, given that, well, they're scary and can apparently trigger them. but it would take some pretty interesting maneuvering to get me to accept suicides attributed to earthquakes. (though the police psychiatrist I saw on Court TV last week said the police suicides following Katrina could be chalked up to the earthquake.)

[CalabashCorolla]

Many in the meteorology community, at least, tend to define weather-related deaths as those that took place during the disaster, not in the aftermath. So a person who has a heart attack and dies upon viewing their destroyed home would not be counted under this reasoning. There's some reason to this method, which is if one wants to determine if deaths from a particular disaster are increasing/decreasing over time, then they would want to know how many people that disaster actually killed on its own, since that kind of number is a better indication of how prepared and warned the potential victims were for the disaster.

Katrina is a very sticky situation, since in the aftermath many very vulnerable people were left in an extraordinarily bad situation, which undoubtedly resulted in several deaths. Probably these deaths are being counted in the total, but if someone, for instance, is electrocuted by a downed wire during cleanup, then that sort of death may NOT be counted at this point.

[zebo-the-fat]

Death figures are only estimates anyway. I wouldn't have thought that they would bother with adjusting for natural deaths. Old people don't die that often.

No, usualy it's just once!