Friday, September 25th, 2009

Why We've Been Ignoring the AIDS Vaccine Story

· AIDS Vaccine "Important Step" Against Disease (Reuters)
· AIDS Vaccine Trial Offers Hope (The Advocate)
· For First Time, AIDS Vaccine Shows Some Success (New York Times)
· HIV vaccine breakthrough (Sydney Herald)
· Vaccine Shows Promise in Preventing HIV Infection (Wall Street Journal)
· Scientists Puzzle Over Minor Success Of AIDS Vaccine (Washington Post)

Oh wait-sorry, what's that, Washington Post? "The results were barely significant on statistical grounds, perplexing for scientific reasons and unanticipated by most researchers."

10 Comments / Post A Comment

RocketSurgeon (#1,632)

There are lies, damn lies and statistics. It's tempting to make a correlation vs. causation error in this case As the Post says, the difference could have been accounted for by chance. Also, there's the possibility that the HIV-prevention counseling bundled with the intervention had the intended effect. While this is a step forward, the fact that they're not offering it to the placebo group is telling. Good news does not equal efficacy.

Is this the appropriate venue for a Ronald Reagan joke?

HiredGoons (#603)

is there ever an inappropriate venue for one?

Rod T (#33)

The combination of Magnums and Eros Light continues to burn.


Abe Sauer (#148)

I hate humans for being dismissive or cynics or uninterested in any way about this.

smapdi (#1,306)

We went to bed and the moon was a waterless desert, an hiv vaccine would probably never be found, and
we could not draw a straight line from dinosaur fossils to birds.

ShanghaiLil (#260)

Sorry, guys, but this has been my area of professional expertise for >20 years or so, and for once, the WaPo is right. The likelihood that this study result would occur purely from chance is just under 4% — the limit for declaring a study "statistically significant" is >5%, and the comparable likelihood for something like the effectiveness of aspirin in preventing heart attacks in people who've already had one is something like 1/0.00000009. I'm not saying that the study result is false, but the probability is much higher than is true for other medications of comparable effectiveness, and no confirmatory studies exist. More caution about these findings is called for than has, by and large, been shown by major media outlets.

Post a Comment