The biggest news of the week in health is the emergence of a partially-effective vaccination against HIV, which in the study prevented an extra 30% of the treatment group from contracting the virus. There are many questions and caveats: the trial involved a population with a comparitively low incidence rate (relative to say, sub-Saharan Africa). It's also unclear why this vaccine would only be partially effective and whether or not it would work against different strains of the virus. One sentence I'm seeing quite often is jumping out at me:
The vaccine lowered the rate of infection by 31.2% — too little to be considered for licensing.That's from the Global Health Blog (with emphasis added).
My question is this: why is is 32% too low for a roll out of the vaccine? Is this purely the result of a cost-effectiveness argument, or is there some basic threshold a vaccine must pass to be considered viable? The global health community is in the process of pushing for a greater number of circumcisions in sub-Saharan Africa and now in the US, based on the similarly limited efficacy rate of 50% (something I've questioned before). Is the difference here just the cost, or is the context?
5 Comments
The fear is that the introduction of an AIDS vaccine will induce (negative) behavior change in immunized individuals, who will think they are protected and thus engage in riskier sex and other activities. At 32% efficacy, the AIDS community is legitimately worried that in this scenario the increase in transmission would offset any protective benefits of the vaccine.
That's a reasonable concern, but how are the possible problems with behavioural change any different than with circumcision, which has just a 50% efficacy rate for men (and offers no protection for women)?
I heard something on the news last night that says that with vaccines in general, you want something like an 80% threshold. (I think. My memory isn't great today.) It might be a purely scientific reason and not a behavior-based one, sort of how we look for a margin of error of +/-3% in a poll to indicate that it's reliable.
I don't know the answer to Matt's question. Who determines the threshold for vaccine effectiveness?
There's a comment on the Wisdom of Whores blog post on the vaccine (http://www.wisdomofwhores.com/2009/09/25/hiv-vaccines-the-ecstasy-and-the-agony/ ) that could be an answer: if effectiveness is low and the virus mutates fast (which we know HIV does), the vaccine's already-low effectiveness would become even less powerful in fighting new variants.
Since the comments bring it up, I don't think behavioral disinhibition is the reason why. Sure, in the US there was a reduction in the use of condoms following the availability of ART, but there is less convincing evidence of behavioral disinhibition elsewhere. In fact, a male circumcision cohort study in the Luo-dominated area of Kenya found:
During the first year postcircumcision, men did not engage in more risky sexual behaviors than uncircumcised men, suggesting that any protective effect of male circumcision on HIV acquisition is unlikely to be offset by an adverse behavioral impact.
I think we have to be cautious when considering the lack of behavioral change in the initial circumcision studies. The biggest difference between then and now: during the clinical trials, circumcision was not a proven treatment, where now it is being billed as a "must" in the fight against HIV/AIDS.
The subjects in the original trials weren't necessarily aware of its effectiveness, so there is little reason to change behaviour. Today, the average man going in for a circumcision does so because he knows it reduces his chance of contracting HIV/AIDS, and so I'd be very, very surprised if that awareness didn't aggravate bad behaviour.