So if I read you correctly, you're taking battlefield statistics from two 20th century wars, applying the ratio of wounded to dead soldiers to the latest estimates of overall mortality figures for Iraqis, and concluding yet again that one of the most prestigious medical journals in the world is off its nut -- all because of the inspiring words of a commenter named "Yancey Ward."

These marijuana cigarettes you're smoking -- where can I get some?

d - instead of being snarky, why don't you use your self-proclaimed intellectual superiority and address the subject of the wounded?

If there has been a excess of 650,000 deaths - caused by violence - then isn't it likely that there has been an excess of wounded? If fighting occurs in populated areas, then isn't it reasonable to expect a higher proportion of wounded among the civilian population?

Or are we to surmise that the coalition's plan is to leave no witnesses, therefore most casualties would be deaths, rather than wounded?

Enlighten us.

It keeps getting better with you folks. It's amazing.

First, the proprietor of this blog makes a stream of facially absurd remarks that indicate a basic misunderstanding of statistics, public health research, the academic peer review process, and any number of other related issues; he and his commenters (including you) go out of your way to mis-read the article under consideration, claiming that it argues something that it does not; and then, after it's pointed out over and over again (try here for starters) that criticism of the Lancet study is almost universally ignorant, CY (and now you) derive renewed inspiration from a commenter named "Yancey Ward" and promtly begin howling about a pile of utterly irrelevant, phantasmic statistics about wounded Iraqis.

There's no possible way to enlighten you. It's quite dark in there, I'm afraid to say.

In other words, we have absolutely no right to be astounded and skeptical of a death rate that is significantly higher (10x-13x) than what has been reported over the last 4 years?

We have no right to question the accuracy of the report?

We have no right to question the political motives of the authors of the report?

We have no right to ask for more than a statistical extrapolation as proof of the veracity of the report?

We must bow down and pledge unquestioning obedience to all (the right) scientists for they are pure, infallable, and apolitical?

Well, I am performing my patriotic duty by dissenting with your point of view.

The link you gave said (paraphrased) "Pay no attention the the really big number - it doesn't matter. What matters is that we have scientifically proven, beyond a shadow of any doubt, that Iraq is extremely more worse off since the start of the war than they were before the war started."

Until corroborated by other sources, I will remain skeptical.

I'm still waiting for Al-Jazeera to show the pics of these 600,000 graves. I won't be holding my breath, becasue if AJ can't can't find'em, nobody can.

Another example of a professor sucking more people into some fantasy so he can get attention. Like a spoiled brat they don't care what kind of attention just so it's attention. This bunch is so full of bull a kindergarden student would smell it out.

I have not read the study, and really do not intend to. But if more Iraqis died of US activity in a couple of years than Germans during four years of bombing not only factories but "population centers" (Berlin, Hamburg, Dresden and others) would not there be more than one study showing it?


Perhaps the statistical analysis of reported (to the studiers) deaths is good: can the same be said of their sampling methods? For example, asking interviewees to spread the word that they want to hear about deaths by violence, is it not likely that only people who had such deaths to report would then show up on succeeding days? Extrapolating from such a self-selected base would surely be akin to interviewing only people who claim success on the grapefruit diet and extrapolating that it must work...

Tequila Jack: I appreciate your willingness to publicly state your determination to know absolutely nothing about this study.

Scrapiron: You're absolutely right. I have no interest in this question except to draw people into some kind of "fantasy," and I am so desperate for attention that I can only receive it from people who denounce public health research they have publicly announced their refusal to read. Remarkable.

Southern Roots: I don't particularly give a squirt if you "question authority" or not. I'd remind you, though, that intelligent questions will always get you farther in life than the ones you knuckleheads are asking about the Lancet study.

Let me put this whole issue another way. Rather than focus on the estimated figure of 655,000 dead, I'd like someone -- why don't we start with you, Southern Roots? -- to explain to me why we should find it so implausible that the Iraqi mortality rate should rise from 5.5 per capita to 13.3 per capita over the course of three and a half years of warfare and insurgency, sectarian violence, infrastructural corrosion, massive unemployment, and degraded access to medical care throughout much of the country. The 655,000 figure you find so grotesquely overstated is actually derived from that spike in mortality.

I'm keen to hear your theories on this one.

d - We are having a very fine dance here. You are asked questions and you refuse to really answer them except to belittle and then fall back on the "read the report, believe the report, you no nothing of statistical analysis" theme.

In order to have an increase from 5.5 to 13.5, you are required to have over 600,000 more deaths than "normal" in a three year period.

There is no getting around that number. I find it very difficult to believe that it goes unnoticed and unreported.

If the death rate due to violence increases significantly, then there should be a corresponding increase in wounded. You have thus far refused to acknowledge this, let alone discuss it.

I also don't think anyone here has been arguing that there haven't been increase in civilian casualties due to the war, what we are disputing is the magnitude. Do you honestly believe that, if the casualties are as high as the Lancet report claims, that all those groups that detest and hate Bush (and America) are just sitting on their hands and not finding some way to use this information against Bush? How realistic is that?

When actions do not correspond the presumptions of statistical studies, I do get cynical about the theoretical side. And I don't "give a squirt" if that puts your panties in a bunch.

If the death rate due to violence increases significantly, then there should be a corresponding increase in wounded. You have thus far refused to acknowledge this, let alone discuss it.

No-one knows how many people have been wounded in Iraq over the last three years. You don't know, I don't know, The Lancet doesn't know, and George Bush doesn't know. There has never been a comprehensive study of this.

If the Lancet had tried to do the same survey of injuries as they did for deaths, the numbers would be practically useless. With a death, you're either living or dead, true or false, there's no middle ground. It's an easy thing to count because it doesn't rely on someone's interpretation. There's even official certificates. But for injuries, you'd need to send doctors out to assess every wound, which obviously isn't practical. Many of these people don't have access to medical care at all, so if you've got a few thousand doctors spare who are happy to be sent into dangerous territory, they really should be used to make people better, rather than collate statistics.

The comparison to war statistics is, as d says, a non sequitur. I think a lot of you are missing the point here: this number of 655,000 is not measuring war dead. That is, it is not measuring people who died in pitched battles. It is measuring excess deaths due to many causes, including sickness and accidents. It is measuring the number of extra dead people Iraq has as a result of the war, not the number of people killed directly in the war itself. Historically, it's perfectly normal for there to be a big difference between these two numbers.

As to how many people I think were wounded, I can only speculate. I might say that many of the dead bodies that have been turning up show signs of torture. That is, they weren't gunned down in a pitched battle, but abducted, tortured, and killed. If someone wants to do that to you, you either escape, or you're dead. There's no way to get injured. If that effect was statistically significant, it might produce a death/injury ratio higher than pitched battles would. But I should emphasise that that's just speculation: I have no idea what the injury rates should look like.

So we have no idea how many people we should expect to have been injured given 655,000 excess deaths, and we have no idea how many people were in fact injured. You are taking two numbers you know nothing about and concluding that one is less than the other. It's not very convincing.

To follow up on Mat's point in the last two paragraphs, I just typed "Iraq" into Google News. The top two headlines were:

Decapitated corpses found in Iraq
8 Females, 2 Teenagers Kidnapped In Iraq

I haven't done a scientific study that shows that the dead-to-wounded ration from decapitation and kidnap-murder is higher than the dead-to-wounded ratio of WWII-style combat, but it seems like a plausible hypothesis. (That is to say, freaking obvious.)

A couple of other points:
would not there be more than one study showing it?
Well, there was another study a couple of years ago. But these studies are very difficult to carry out -- they involve going around Iraq and interviewing families in large numbers, and Iraq is a dangerous place to do it. So no one else has actually attempted this kind of study, as far as I know.

I'm still waiting for Al-Jazeera to show the pics of these 600,000 graves.

There are some interesting articles lately about people fishing bodies out of the Tigris.

Rather than focus on the estimated figure of 655,000 dead...

Yes, lets ignore a patently absurd claim so we can bash others. Yea, that's the ticket.

Sorry jaggoff, the claim is so absurd its not even worth considering in serious company. Save this shit for you truther buddies who might listen.

You don't need to be a rocket scientist to figure this out, you just need to watch rabidly biased sources like Al-Jazzera to know its bogus.

Where did they grow you guys? I think I know what they used for fertilizer.

The staff of Iraq Body Count has now issued a press release accuses the Lancet of over-inflating the civilian body count in Iraq.

See the update to this post.

If there has been a excess of 650,000 deaths - caused by violence - then isn't it likely that there has been an excess of wounded?

Not particularly. Very few people who are kidnapped with the intent of being tortured to death and dumped as an object lesson to others wind up wounded instead of dead. Since deaths from violence would tend to involve victims rather than soldiers, I'd expect a far higher dead-to-wounded ratio.

In other words, we have absolutely no right to be astounded and skeptical of a death rate that is significantly higher (10x-13x) than what has been reported over the last 4 years?

From an interview with one of the authors (to which, strangely enough, this blog won't allow a link):

"We have gone and looked at every recent war we can find, and only in Bosnia did all governmental statistics add up to even one-fifth of the true death toll. And in Bosnia, the rate was 30 or 40 percent, with huge support for surveillance activities from the UN. So it’s normal in times of war that communications systems break down, systems for registering events break down.

And in Saddam’s last year of his reign, only about one-third of all deaths were captured at morgues and hospitals through the official government surveillance network. So, when things were good, if only a third of deaths were captured, what do you think it’s like now?"

You were saying?

Perhaps the statistical analysis of reported (to the studiers) deaths is good: can the same be said of their sampling methods? For example, asking interviewees to spread the word that they want to hear about deaths by violence, is it not likely that only people who had such deaths to report would then show up on succeeding days?

"And then, once we had picked that we were going to visit two or three neighborhoods in a certain governance or province, we would then make a list of all the villages and towns and cities, and again randomly pick one of those to visit, so that big places had a larger chance of being visited than smaller places. And then, finally, when we got down to the village level or to the section of a city, we would pick a house at random, visit it and the other 39 houses closest to it to grab a cluster of 40 houses."

You were saying?

Summary
Iraqi Body Count Press Release
October 16, 2006

"In the light of such extreme and improbable implications, a rational alternative conclusion to be considered is that the authors have drawn conclusions from unrepresentative data. In addition, totals of the magnitude generated by this study are unnecessary to brand the invasion and occupation of Iraq a human and strategic tragedy."
http://tinyurl.com/yd5o2j

In the light of such extreme and improbable implications, a rational alternative conclusion to be considered is that the authors have drawn conclusions from unrepresentative data.

Ooops - Appeal to Consequences of a Belief.

One notices that the IBC hasn't said anything about the statistical basis of the study, of which all informed experts have said is fairly robust. One also notes that the authors have addressed some of the points made, noting, for example:

"We have gone and looked at every recent war we can find, and only in Bosnia did all governmental statistics add up to even one-fifth of the true death toll. And in Bosnia, the rate was 30 or 40 percent, with huge support for surveillance activities from the UN. So it’s normal in times of war that communications systems break down, systems for registering events break down."

"And in Saddam’s last year of his reign, only about one-third of all deaths were captured at morgues and hospitals through the official government surveillance network. So, when things were good, if only a third of deaths were captured, what do you think it’s like now?"

So the IBC criticism is not only off focus, it is based on false assumptions.

Try again.

of which all informed experts have said is fairly robust.

Where have we heard the "all experts concur" line before?

Donald Berry, chairman of biostatistics at M. D. Anderson Cancer Center in Houston, was even more troubled by the study, which he said had “a tone of accuracy that’s just inappropriate.”

http://medpundit.blogspot.com/2006/10/lancet-strikes-again-i-admit-this.html

According to official reports , over 180,000 internally displaced refugees were reported just between the months of February and June of 2006. Undoubtedly those not registering pushes the number much higher. As I pointed out below, the survey methodology means that these displaced refugees had very little chance of being surveyed. But in addition to that, their migration is sure to skew the analysis of the data.

The authors acknowledge as much in their paper:

"The population data used for cluster selection were at least 2 years old, and if populations subsequently migrated from areas of high mortality to those with low mortality, the sample might have over-represented the high-mortality
areas."

http://notropis.blogspot.com/

"We have gone and looked at every recent war we can find, and only in Bosnia did all governmental statistics add up to even one-fifth of the true death toll. And in Bosnia, the rate was 30 or 40 percent, with huge support for surveillance activities from the UN. So it’s normal in times of war that communications systems break down, systems for registering events break down."

Even with such terrible registration rates for death certificates, isn't it just amazing how the Lancet group was able to get well over 80% of the households to produce a death certificate? This was a key element in their validation of interprolating the numbers out to >650,000.

Seems bogus to me.

To compare deaths-to-wounds ratios in combat with what's going on in Iraq is just pig-ignorant.

Then, to presume at least WWII-quality medical care is to ignore the fact that hospitals currently are serving as *bases* for some of the militias, so a lot of wounded probably aren't even seeking medical help.

Beyond that, sure, it's possible the Lancet study is flawed. But nothing to date proves it, and disinterested public-health professionals find the survey techniques generally unassailable.

As for IraqBodyCount, anyone who knows anything about statistics has known for some time that its own methodology was overwhelmingly likely to have been understating fatalities by anywhere from several multiples to an order of magnitude.

But don't let, you know, facts get in the way of a good story.

Where are the death certificates? If they could get them for 501 and of the 546 people they said were killed because of the war, where are the rest?

The methodology is sloppy, they virtually ignore rural areas, and the sample groups are too small and interrelated. Note how few real professionals are prepared to risk their reputations backing this absurd claim. It is like saying the earth is flat, it is ridiculous on its face.

It is political and it shows a cyncial disregard for the suffering of the Iraqi people. They are just cannon fodder for the people who did this report.

I wonder what the numbers would have been had they done a report like this when Saddam was around and used as a sample the Kurds and Shia? There were people in the Kurdish north whose entire villages were destroyed. I wonder how many millions they could have claimed as dead using this methodology? But then again, Saddam would not even have let them talk to people.

The Wall Street Journal says that the Lancet is a bogus report.

Specifically that the selection of only 47 clusters to represent 26 million people was bogus.

Also that no demographics were asked of the interviewees so that the samples could be compared with a census or other information.

According to the Lancet’s disputed study, 601,027 people—al Qaeda terrorists, insurgents, Iraqi soldiers, police, and true civilians—have been killed violently ("the most common cause being gunfire," says the summary) since the 2003 invasion of Iraq.

We also know that for every combat-related death, there are usually a far greater number of casualties.

"Violent" death and "combat-related death" are not synonyms.

A large number of targeted murders would yield a large number of bodies, and few wounded.

(By "targeted murder", I mean a murder where the murderer has selected a victim, and decided to kill the victim, attacked, and did not stop until certain that the victim was dead.)

However, if you can prove that there is damn good injury tracking in Iraq - a statement that we have no evidence for - and can show that nowhere near that level of injury has occurred, you might well have found a flaw in the report.

Uh, do you have any proof that the injury tracking in Iraq is so good that it calls the study results into question?

I mean, you haven't presented any; you've just insisted that we surely would have known if there were that many injuries. You're saying that one should trust an unknown system of tracking and reporting injuries, and use that trust to ignore solid research with a sound methodology.

It doesn't sound very sensible to me. My figuring is, if the study is done soundly (and it was), we should be digging to find out if it's true, not looking for self-serving reasons to declare that it's false.