4.10.2006
Less AIDS in Africa?
An article in the Washington Post from Thursday, April 6, talks about how estimates of HIV and AIDS burdens in African countries have been gradually revised down as better data becomes available. Now, to estimate prevalence, researchers conduct random testing, and verify positive tests. But, in the early days of the epidemic, researchers estimated prevalence using blood samples that they took from women attending prenatal clinics in urban areas, assuming that these women represented the general population. The tests they used to determine HIV status also returned a lot of false positives. So the numbers were higher than they should have been.
The article implies at points that UNAIDS may have issued extremely large estimates in order to mobilize funding. But accidental systematic overestimation makes sense—with limited data, the assumptions were flawed and the models were flawed, so even as more data was gathered, the estimates were still too high. Plus, the downward revisions may affect national prevalence estimates, but there are still extremely high rates in sub-populations, including people in most countries’ most productive age groups. And even if the UN had to revise its estimates for Africa, the revision was from 30 million to only 25 million.
25 million people are living with HIV in Africa.
The article implies at points that UNAIDS may have issued extremely large estimates in order to mobilize funding. But accidental systematic overestimation makes sense—with limited data, the assumptions were flawed and the models were flawed, so even as more data was gathered, the estimates were still too high. Plus, the downward revisions may affect national prevalence estimates, but there are still extremely high rates in sub-populations, including people in most countries’ most productive age groups. And even if the UN had to revise its estimates for Africa, the revision was from 30 million to only 25 million.
25 million people are living with HIV in Africa.
Hit Counter
