Science Spits on Death Rates

According to a new study released yesterday by the British medical journal, the Lancet, the risk of mortality in Iraq has increased by 1.5% in the year and a half since the invasion. The chance of dying is up to 7.9%, which the researchers estimate has led to the deaths of more than 100,000 Iraquis who would otherwise still be alive.

Some numbers to consider:
Mortality in Iraq: 7.9%
The overal mortality rate in the U.S.: 0.84%* (847 per 100,000)
The risk of being a dwarf: 0.25% (1 in 40,000)
The risk of being the decendant of murderous dwarves: 100%

Does that put things in context?

The report was "early released" (published online prior to print publication) -- a practice that is by now relatively common in the medical publishing world. It is supposedly reserved for the findings of studies that are of the utmost importance to public health or clinical practice. The editor, Richard Horton, may have overstepped his bounds, however, with his strong words for the "democratic imperialists." From the editor's commentary:

With the admitted benefit of hindsight and from a purely public health perspective, it is clear that whatever planning did take place was grievously in error. The invasion of Iraq, the displacement of a cruel dictator, and the attempt to impose a liberal democracy by force have, by themselves, been insufficient to bring peace and security to the civilian population. Democratic imperialism has led to more deaths not fewer.

This, with some qualification, appears to reflect the wishes of the head researcher, who said to the AP:

I emailed [the report] in on Sept. 30 under the condition that it came out before the
election. My motive in doing that was not to skew the election. My motive was that if this came out during the campaign, both candidates would be forced to pledge to protect civilian lives in Iraq.

This statement strikes me as unbelievably sexy, but realistically -- especially when you have Mr. Horton slathering on the New World Order epithets -- it's a bit biased. This might (reasonably) spark a debate about the role of politics in medical publishing, but it shouldn't reasonably spark a debate in politics about the reasonableness of medical publishing. I say this only because it looks like we'll see more of the following (from this Post article):

"The methods that they used are certainly prone to inflation due to overcounting," said Marc E. Garlasco, senior military analyst for Human Rights Watch, which investigated the number of civilian deaths that occurred during the invasion. "These numbers seem to be inflated."

That's nothing compared to what a campaign spinner could do with a similar line of thought. Basically, though, what Mr. Garlasco is indicating is that the study samples (33 clusters of 30 households in different areas of the country) could have been concentrated in areas that were more prone to violence than others (in other words, that they weren't representative). This is a reasonable critique, but has little bearing on the overall significance of the report. Medical studies operate on certain scientific principles meant to minimize statistical error. Studies that pass muster (as determined by the editorial board) are judged to have statistically significant findings and an insulated amount of error. There will of course *always* be error, but by and large, if you see people trying to poke holes in these findings they're just obfuscating. The number may be closer to 80,000 additional deaths than 100,000 -- but that doesn't change either the science or the politics: the real number (which we'll probably never know with exactitude) is still bound to be statistically signficigant - tens of thousands of Iraqis have died.

So, the moral of the story: we now have independent, scientific proof of a substantial rise in the death rate that is strongly correlated to the war. So what? War always causes death, right? Well, aside from the fact that having a cavalier attitude to death on that scale is repugnant, Americans are allergic to high civilian casualties. Does the almost too obvious to mention connection between death rates and missing explosives push us over the hypoallergenic tipping point?

Damn the dwarves and everthing they started.