Wednesday, February 11, 2009

The Lancet and the Iraqi Studies

This is a response to Myk's comment, I wanted to touch on both things included in the comment so I decided to make a new post as this no longer pertains to the marijuana commercial post. I probable should have saved this for Grandma's B-day, oh well.
"Peter:"That article I posted in my comment is from the Lancet a really good medical journal, the British version of the New England Medical Journal."

My public health expert I used some years back in an AIDS case (from of all places Johns Hopkins), said that Lancet, while commanding a certainly level of respect, tended to be out there somewhat and not trusted by everyone. Lancet, if you recall, published the two fairly heavily criticized Iraq mortality studies."

I wonder if the health expert knows the authors of the Iraqi study-it was a Johns Hopkins study. Also "not trusted by everyone" not sure about the Lancet but for the Iraqi study, the non trusting groups are mainly the U.S, British, Iraqi government and journalists. Large numbers of statisticians, epidemiologist and pollsters agree with methodology and results. Of course there is always debate around the imprecision that results from all cluster polling however I would not go so far as to say it heavily criticized by the scientific community, opposite in fact. It was lambasted in the media (The study needed Nate Silver as PR, he would have won over the media like he did my heart with his savvy polling). I do not take the study as the complete truth (though very objective in carrying out standard population polling methods), however I do believe the Iraqi study is very important as it was the only survey to use active population polling versus every other method that used passive surveillance (news reports and hospital records etc). Interestingly Les Roberts (one of the authors) did a very similar study (he has done countless) on Rwandan refugees-1.7 million had been killed. The study is widely accepted and resulted in the UN requesting 140 billion in aid, the withdrawn of foreign armies from Congo and the US state department pledge an addition 10 billion in aid.

As for the Lancet, I was a little off in my comparison to the NEMJ. It is more like NEMJ=Princeton University, the Lancet=lowly Duke or Columbia University. The impact factor (scientific study on validity of Scientific Journals) of the Lancet is extremely high, meaning high number of scientists use its resources, ranking it in the top ten of hundreds of journal publications (#10 versus NEMJ #2). Three docs I talked to this morning (anecdotal evidence) would have no reservation in using journal articles from the Lancet. All scientific journal articles are debatable (some more than other) hence peer reviewing. The Iraqi study was no different, garnering debate (though not any more extensive versus other polling studies) over its shortcomings-heavily criticized though, not really.


Big Myk said...

I have had exactly zero training in epidemiology, so I don't pretend to be able to pass judgment on the Iraqi death studies published in Lancet. My only point was that Lancet is a bit more willing to court controversy than other medical journals. And, by the way, the reference to the death studies was my comment and not my expert's.

In fact, my contact with this expert occurred many years before the Iraq war. The issue was the value of screening state prisoners for HIV infection. One of my expert's main points was the problem of false positives (at least at that time), and one of the strongest studies we had on this point was published in Lancet. His comment to me was that our reliance solely on a Lancet article might expose him to additional cross examination -- since there is a perception that Lancet is willing to publish studies other journals wouldn't touch. He suggested that we keep looking for additional articles to back up the Lancet piece.

I was only sharing this little tidbit for what is was worth. I didn't exactly expect the Spanish Inquisition.

Peter H of Lebo said...

Sorry I went overboard, my long rant was that the level of credibility/respect of the Lancet has no relation with the controversy/debate Iraqi study. I mistakenly thought you implied, when you stated, "Lancet, if you recall, published the two fairly heavily criticized Iraq mortality studies." that the Lancet and the Iraq study as well as the drug (marijuana) article was somehow less scientific (or to be taken with a grain of salt) because it publishes controversial articles. Your comment was a response to my statement "That article I posted in my comment is from the Lancet, a really good medical journal." A good article in a good journal. I completely agree that one must take gather evidence from a wide variety of sources but that has no relevance to my comment that the Lancet is a good journal. I applaud the Lancet for publishing the Iraqi study (though it makes sense that a science journal publishes a science article). I think that the Lancet helped facilitate discussions on science in politics and methodology practices and precision in extreme environments. Also if you were going to get Spanish Inquisition I would have attacked you in a catchy musical number if I recall correctly the methods from my Hopkins history education.