Bad statistics from HIV/AIDS "skeptic" Peter Duesberg debunked

My favorite part is where the defender argues that Duesberg is really saying the average person who contracts HIV has had unprotected sex with an infected person 1000 times. Besides that this is no less fallacious than what Duesberg actually said--Duesberg never once used the word "average" and in fact explicitly said "all positive teenagers would have had to achieve..."

Where on earth was this paper from anyway? Was this actually published in a peer-reviewed journal?

Katie said:

Where on earth was this paper from anyway? Was this actually published in a peer-reviewed journal?

Ha! There's the problem in a nutshell.

Generally speaking though, I think the scienceblogs should recognize a commonality between these AIDS denialists and global warming/evolution denialists.

Again and again the discussions end up being the same, it's like a syndrome, and I think we should refer to all these guys as a group. They always use the same tactics, they pull quotes out of context, they fudge math/stats, they outright refuse to acknowledge contradictory data, and they never have any evidence of their ideas in peer-revied journals. They should be called "denialists" and we should treat all their arguments in essentially the same way.

I guess I'm saying I don't understand why when these trolls visit the scienceblogs we don't just clump them into this denialist group, and dismiss them out of hand. We could say that the scienceblogs exist to inform the public by debate and shouldn't be elitist or exclusionary, but do debates with denialists actually accomplish anything? It's not like there is any piece of evidence that will convince them to change their minds, that is part of the damned syndrome. In fact, a great starting point with the denialists is to demand from the outset exactly what data it is that will convince them they are incorrect, because if they truly are arriving at their opinions based on facts, if the facts contradict them they should be willing to change their minds.

I get the feeling that most denialists when posed this question, no matter what they are refusing to believe, will either be unable to answer it, or will suggest an unverifiable experiment or impossible fact. At that point, the argument is already over, and they've lost.

Where on earth was this paper from anyway? Was this actually published in a peer-reviewed journal?

This article was published in Pharmacology & Therapeutics which is a peer-reviewed journal with a respectable impact factor of 8.357.

It's hard to understand how a lot of it got through peer-review. In reality peer-review is not perfect.

In a completely amazing coincidence Phillip Johnson the founder of the Intelligent Design/Creation movement was involved in the submission of this article. Skeptics would argue that this crossover was so improbable that it could not possibly have occurred by chance. Evidence of design?

The second experience that brought me to a decision was that of participating in the pre-publication review of Duesberg's major paper, "AIDS Acquired by Drug Consumption and Other Noncontagious Risk Factors," which appeared in the international journal Pharmacology and Therapeutics in 1992. Duesberg had asked me to give a critical reading to an early draft of the paper, and I was sufficiently interested to go through it line by line and source by source. By accident I had an unusual additional role, because the supervising editor for Duesberg's paper was Professor David Shugar of the University of Warsaw, who could communicate efficiently with Berkeley only by email. I was just then learning the ways of the internet, and Duesberg hardly knew what email was, so I became by default the go-between for the author and editor. Many criticisms came from multiple editorial consultants, and all had to be considered. Through this process I became intimately familiar with the jots and tittles of the HIV/AIDS controversy, and I became convinced that Duesberg was practicing honest science and the HIV establishment was not.

So there you go a Creationist lawyer couldn't find any errors in Duesberg's article so it must be true.

I can only conclude that the original draft must have been even worse than the version that ended up in print and that worse errors were weeded out in the process.

The supervising editor at the time David Shugar later signed the list of HIV "rethinkers" so it seems that he was sympathetic to Duesberg's claims.

http://www.virusmyth.net/aids/group.htm

The wrost thing is that Duesberg's supporters argue that his articles must be correct because they have been published in peer-reviewed journals. The same argument apparently doesn't follow for the vast majority of peer-reviewed literature that contradicts their beliefs.

The reason this paper got accepted could also be due to Duesberg's reputation, which hadn't yet plummeted as a result of his HIV/AIDS "dissident" stand yet. Like it or not, sometimes peer reviewers will let stuff from prominent, well-published scientists slide that they wouldn't let in if it came from a less prominent scientist. Sometimes it's because of admiration; sometimes it's because they don't want to cause too much trouble with such a figure. Also, prominent scientists also often know the editors personally and will sometimes call to lobby for certain papers. I have no idea if that's what happened here or if it was something else, but it's one possible explanation for how this piece of dreck ended up getting published in a pretty good journal.

But but but...
Am I missing something, or is his logic incredibly wrong, and in a totally fundamental way? I have surely read a lot of peer-reviewed papers with jumps in reasoning and overstated points, but I don't think I've ever read something so transparently wrong. Maybe I am just misinterpreting it and it's actually much less obvious than it seems. Or maybe I just haven't read enough. Granted it seems like this is a minor point in the paper--but it's so wrong!

Also, wouldn't this logic lead to the conclusion that HIV actually could not have infected so many people, since by his reasoning they would all have to have had random unprotected sex a quarter of a million times? Doesn't this lead you into the "virus doesn't even exist" camp?

Well, even Jon Wells of the discover institute has a "peer reviewed" publication in MCB. So this guy snuck one in a couple of decades ago, but it's one paper vs thousands upon thousands that contradict it. No peer reviewed articles are getting published these days promoting this claptrap, and while science does self-correct, it doesn't exactly purge the literature of the crap, so anyone with an internet connection can find the junk science at any time, without seeing any subsequent correction to the record (so it seems like it's still true). I guess we need to make science a little bit more like wikipedia, not that just anyone should be allowed to edit the articles, but it would be nice if we had a central repository summarizing findings in fields, edited by scientists working in that field and updating it with additional papers being published and reversals of previous hypotheses/theories. Kind of like a wiki-review article for each relevant topic.

Without that kind of transparent correction denialism seems to be perpetually encouraged since they can take the one paper that confirms their view (even if debunked), over the entire field which contradicts it which they don't see or care to see.

I'm just becoming more familiar with the AIDS/HIV denialist movement since I first started reading about them on Aetiology. The similarities to the creationists/IDers/anti-global warming crowds are striking.

Am I missing something, or is his logic incredibly wrong, and in a totally fundamental way? I have surely read a lot of peer-reviewed papers with jumps in reasoning and overstated points, but I don't think I've ever read something so transparently wrong. Maybe I am just misinterpreting it and it's actually much less obvious than it seems. Or maybe I just haven't read enough. Granted it seems like this is a minor point in the paper--but it's so wrong!

No, you are correct. In this instance Duesberg's logic is spectacularly wrong. The logical fallacy that he falls victim to is a relatively common misunderstanding of probability. It is surprising that this got through peer review. We don't know who reviewed the paper and what field they had expertise in.

Most of his peer-reviewed articles have few whoppers like this. Duesberg often selectively cites the literature, chooses interpretations that fit his theories and makes conclusions that are not based on the data etc. All of these are more difficult to recognise and can be "debated" so they do get through peer-review.

With some papers like his latest which got published in the Journal of Biosciences it seems likely that the article was rejected several times before Duesberg found an editor that was sympathetic or got reviewers that were not experts in the field.

Also, wouldn't this logic lead to the conclusion that HIV actually could not have infected so many people, since by his reasoning they would all have to have had random unprotected sex a quarter of a million times? Doesn't this lead you into the "virus doesn't even exist" camp?

It is difficult to work out a coherent model for what Duesberg really believes. He seems to make different assumptions when he makes different arguments for why HIV cannot cause AIDS. The only common pattern is that he argues from the start that HIV cannot cause AIDS. Everything such as his assertion that HIV depends on perinatal transmission for survival is just an ad hoc argument to support his dogmatic inistence that HIV cannot cause AIDS.

From the 80s it has been clear that the majority of HIV infections are in specific high risk groups and are therefore not acquired perinatally.

Perhaps he believes that before the gay revolution and the rise of injecting drug lead to rise in prevalence HIV was endemic and predominantly transmitted perinatally. However, at other times he argues that HIV has been at a constant prevalence of 1 million for the past two hundred years.

I can't find a coherent model in Duesberg's articles. The only continuos idea is that HIV cannot cause AIDS.

The reason this paper got accepted could also be due to Duesberg's reputation, which hadn't yet plummeted as a result of his HIV/AIDS "dissident" stand yet. Like it or not, sometimes peer reviewers will let stuff from prominent, well-published scientists slide that they wouldn't let in if it came from a less prominent scientist. Sometimes it's because of admiration; sometimes it's because they don't want to cause too much trouble with such a figure. Also, prominent scientists also often know the editors personally and will sometimes call to lobby for certain papers. I have no idea if that's what happened here or if it was something else, but it's one possible explanation for how this piece of dreck ended up getting published in a pretty good journal.

From personal experience there is also a large distribution in how rigorously reviewers actually review articles. Sometimes one reviewer will find a missing apostrophe and another reviewer will find a major flaw. If you submit the same paper several times you might strike it lucky.

Some of Duesberg's papers have hundreds of citations. It is naive to expect that a reviewer will each and every single citation to check that Duesberg accurately and honestly presents the findings of these papers.

Duesberg's friendship with Harvey Bialy does explain his Bio-Technology paper getting through.

The editor-in-chief of Genetica John Mcdonald also gave the dissidents an entire issue of this journal to present their arguments in an effort to counteract the claim that the "rethinkers" were being censored.

Duesberg was the editor of this edition so again the editor was a friend.

As a member of the NAS Duesberg had automatic rights to publish what he wanted in the PNAS. There is no peer-review for members. The NAS changed the rules after Duesberg and Linus Pauling abused the privilege by promoting their respective pseudoscientific nonsense. The PNAS papers that got through were rejected by scientists that reviewed them but were published anyway.