MrK - HIV/AIDS QuestionsHow could a contraceptive test positive for HIV/AIDS, when HIV/AIDS tests are 99% accurate (in the lingo, specific and sensitive)?
This is a very interesting question. There are many questions about HIV/AIDS that have never been addressed. Here are few of them:
- Why are different tests or combinations of tests used on different continents?
- Why is the same test, Western Blot, used as a confirmation test in the USA, but not in the UK?
- Why is the highly sensitive but less specific screening test, ELISA or EIA, used as both a screening test and a confirmation test in the UK, but only used as a screening test in the USA, where it is considered too sensitive for diagnosis?
- Why is Western Blot interpreted differently in pretty much every country it is used in?
However, to come to the substance of HIV testing:
- Why is there not a single test in use, that tests for the whole HIV virus?
- Why do the different tests, test for different parts of the virus?
As to the epidemiology of AIDS:
- Why do surveys based around testing pregnant women at antenatal clinics, give so much higher positive rates, than tests that use representative samples of the general population? Is it just that pregant women are younger and more sexually active, or is there something about pregnancy that causes false positive test results to occur much more frequently?
- Why did it take so long for UNAIDS, the UN's official AIDS watch dog and advocacy body, to give up on using these antenatal clinic surveys (ANCs), for representative surveys (called DHS or Demographic and Health Surveys)?
- Why was there so much resistance to doing so?
- Did UNAIDS have a vested interest in keeping estimates high?
To read a complete review of the revision of worldwide HIV/AIDS infection rates, specifically in Africa, see these articles from The Boston Globe (2004) and The Washington Post (2006).
This followed on the heels of similar DHS versus ANC results from Swaziland. The most remarkable downward revision of HIV prevalence in Africa came for Sierra Leone, which went from 7% to 1% (see chart in the Washinton Post article). Rwanda went from 11.21% in 2000 to 3% in 2005. Keep in mind, the only thing that changed, was the type of survey conducted (DHS instead of ANC), although there are plenty of organisations that stepped forward to claim credit (there was as brief scuffle between the 'abstinence only' and 'condom for all' crowds).
This should make anyone urgently question the seriousness of the data, how it is arrived at, and in fact the tests themselves. Although officials acknowledged how far off their predictions were, they still maintain that their projections for Southern Africa are correct. In fact, they now make Southern Africa the focal point of HIV/AIDS worldwide.
Lastly, I question:
- Have population projections that take into account mortality from HIV/AIDS turned out correctly?
For that I can provide an answer: no. There is one country in Africa that we have correct population data from, even if it is only from 1998 onwards, and that is South Africa. Again, HIV/AIDS as a model to project population size has failed.
The US Census Bureau runs two models: one that takes into account mortality from HIV/AIDS, one that does not. Here are the date from both:
(Pop. of South Africa x 1 million)
With mortality from HIV/AIDS
Without mortality from HIV/AIDS
(Source: personal correspondence with the US Census Bureau, but the with aids number is available on their website - also note that on this website, they project South Africa's population to decline to 39 million by 2025).
Last year, Statistics South Africa did a community survey, and they put the mid-year population of South Africa in 2007 at 48.5 million.
Obviously, the 2007 population number from the actual community survey was a lot closer to the without HIV/AIDS model (49.28 million), than the with HIV/AIDS model (43.99 million) projected for the same year.
Not only that, but the with aids model shows a declining population, while both the without aids model and the household survey show a growing population.
This should raise serious doubts about the presence of HIV/AIDS in South Africa, it's extent, and/or it's impact on mortality.
If HIV/AIDS prevalence in Sierra Leone can be downwardly revised from 7% of the population to 1% of the population, how much room is there for a downward revision in South Africa, where population growth puts a lie to the notion of a population declining because of widespread mortality from HIV/AIDS infection?