Sunday, October 2, 2011

Fruitful Discord At Lomborg's Copenhagen Consensus Center?

Up to now, UNAIDS' method of verifying their data has been the equivalent of printing out lots of copies of their reports and concluding that, because they all say the same thing, they must be true. First on the chopping block should be the HIV Modes of Transmission analyses (MoT), which purport to estimate the relative contribution of various routes of infection, sexual and non-sexual. In reality, the data used is a mishmash of guesswork and hot air.

The first paper to be published for RethinkHIV is by Lori Bollinger, who considers the cost effectiveness of non-sexual HIV transmission interventions. It's long and boring. But it's based on data such as that from MoT reports. Much though I'd like to criticize Bollinger's offering, I'll leave that to the authors of the second paper, Rob Baltussen and Jan Hontelez.

They are mercifully brief in their analysis, though they are far from merciful to Bollinger. They question if analyses such as Bollinger's can "really provide estimates that are sufficiently transparent, valid and reliable at the country level", and express serious doubts about the value of her work.

Baltussen and Hontelez don't feel that the models Bollinger uses "reflect the actual epidemiology" in the countries in question. Thankfully, they also doubt the adequacy of the "estimates of the relative contribution of each transmission route to the overall epidemic". In particular, they question the validity of the MoT reports used.

These researchers also examine the claimed impact of various HIV prevention interventions and remain unconvinced. They even have doubts about the costing data used for Bollinger's cost/benefit analysis. They point out that all these limitations are "inherent to the task at hand and therefore virtually inevitable". Baltussen and Hontelez do well to raise the issue of the usefulness of such data; yet much of UNAIDS' HIV policy is based on it.

The list of limitations goes on and on. While it is not one of the authors' conclusions, every criticism of Bollinger's offering is a criticism of UNAIDS, the HIV industry as a whole, and much of the HIV literature that has launched a thousand failed interventions per year for the last 20 years or so. The 'successes' among these interventions are based on grotesque overestimations that remain unquestioned even when program after program has failed to deliver the goods.

Despite unearthing all these limitations in the work of Bollinger, and much of the work of the HIV orthodoxy, the authors agree with Bollinger's conclusion: "that interventions to reduce non-sexual transmission of HIV are generally economically attractive". Which is great, as long as the "absence of comprehensive data" that Baltussen and Honetlez note is also rectified.

I'm still mystified as to the lack of consensus between the first and second paper commissioned by Lomborg and his gang. However, I'm not complaining. It makes a pleasant change to see researchers challenging each other rather than patting each other on the back. A few more papers like this and UNAIDS may even have to revise their lynchpin: the assumption that most HIV transmission in African countries is through heterosexual sex.

At less than five pages, Baltussen and Honetlez's paper is highly significant. But what influence will RethinkHIV have on the HIV orthodoxy? UNAIDS has discredited and branded anyone who has dared to challenge their racist, sexist and highly destructive stance. Will they do the same to these authors, or even to RethinkHIV? Perhaps Lomborg has miscalculated his credibility; HIV celebs gain their strength by supporting the orthodoxy, not by challenging it.

Incidentally, Lori Bollinger says elsewhere in a throwaway remark "the work we do is not about numbers and equations, but about people." But it is about numbers and equations and it is not about people. When Africans are asked about their sexual behavior, their answers show that they are human beings. They have similar sexual behaviors to other human beings. But researchers conclude that Africans 'underestimate' and/or 'overestimate' in their answers, effectively calling them all liars. (Thank you to Dr David Gisselquist for the Bollinger citation.)

Researchers tend to assume as a starting point that African sexual behavior (yes, apparently Africans are all the same) explains extremely high prevalence of a virus that is difficult to transmit sexually. Without this assumption their research is unlikely to be funded or published. But this has resulted in the current impasse in HIV prevention. The way forward is to investigate non-sexual HIV transmission but to use empirical data, not the stuff UNAIDS calls data.

[For more about non-sexual HIV transmission modes, such as unsafe health care and cosmetic services, visit the Don't Get Stuck With HIV website]


No comments: