View Full Version : Mortality stats

Brian Hanley
11-05-2007, 01:18 AM
Something I have toyed with from time to time is the thought of an epidemiology type article on mortality and injury statistics in Iraq versus the USA. This caught my attention some time ago when I was looking at the Lancet article that caused so much flap. (I did, by the way, write extensively to both the author of the study and the Lancet, about major deficiencies in that study's methodology. Might try getting that through someday too. But I need more clout developed first.)

I thought I would post this here and see what people might have to say, collect responses. It's not quite an RFI, it's something that you all could comment on and perhaps help make such an article happen.

I'll open it by noting these statistics taken from CDC report. http://www.cdc.gov/nchs/data/nvsr/nvsr55/nvsr55_19.pdf
I've attached a basic graph that shows deaths per 1,000 from all causes by age group. It gives a range from 110 to about 400 per 100,000 for men in the service age range, deaths per 1,000 from all causes. Table 9 shows rates of 32 to 40 per 100,000 for accidental death for ages 17-40.

Up against that, from Iraq's theater of war, 3913 over 5.5 years with a fluctuating number of troops in country. Best average I can come up with is around 120,000. That gives a rough figure of 592 deaths per 100,000 per year from all causes in Iraq. That is somewhere around roughly 250 per 100,000 per year above background. But, in other sources I have seen reference to statistics that show that some ethnic groups and locations in the USA have death rates fo 1000 and up per 100,000 per year.

My general observations are two:
First, that casualty rates for US forces in Iraq are really quite low. Not very far above background. I suggest that this might cast the use of suicide bombers in a different light, because for commanders of troops in the insurgency, suicide bombers have a positive kill ratio when other methods don't.
Second, that for a fair number in the USA, being in Iraq is sight safer than staying home.

The matter of injuries is another matter that should be treated properly since it is a major element.

I'm looking for comments and ideas on sources of data. I imagine that DoD has stats, but may be reticent to give access to them. But it might be interesting to do a study and perhaps a more popular article, and take people through this in detail grounded against the USA.

11-05-2007, 01:48 AM

Trying a general population in the US vs. soldier population in Iraq comparison is riddled with selection bias to the point that any study would be worthless IMO. Without controlling for health; intelligence; freedom to make dumb decisions every night, all the time; etc.; you're left with an apples and oranges comparison.

I don't have any stats in front of me, but I'd imagine that within the age group that comprises the bulk of your distribution of soldiers, the cause of death in the civilian population would be car accidents, with most of them involving alcohol, and with either smaller cars or else SUVs that are likely to roll. When you compare this Iraq, you now insert General Order #1, which means no drinking, and so you remove the major cause of deaths in the US population from your Iraq population. Also, if you are now in the bigger vehicle in any accident, your chances of survival in most cases goes up.

What you're left with is not that Iraq is safer, but that you are comparing populations that are apples and oranges and scenarios that are also apples and oranges.

Brian Hanley
11-05-2007, 02:24 AM
Valid thoughts. However, the bottom line is valid in the end. If you take away alcohol, that's not a reason to say that the overall safety level for a demographic slice is not valid. Otherwise you could not say that it was safer in the Hamptons than it is in the Bronx. Those factors are reasons that influence the overall result, but they don't invalidate a comparison, they are the "Why is this?" discussion.

Demographic slicing is valid, and should be done, age, comparable health, education, pay scale are more important. But even there, using the example of death rates in the Hamptons versus the Bronx, all of those introduce skews of their own.

Good comment. Would need to be addressed head on in the discussion. :)