Bloodborne HIV: Don't Get Stuck!

There’s a recent article in the English Guardian about the US phenomenon of ‘purity balls‘, where daughters take a pledge to ‘remain pure’ (which clearly means they must ‘abstain’ from sex before marriage) and their father pledges to guard their virginity. Whatever people may feel about creepy ideas that some people in the most democratic nation in the universe may have, I find it even more objectionalbe that these ideas are imposed on HIV policies in countries with the worst epidemics in the world, all in sub-Saharan Africa.

The Zambian Demographic and Health Survey (DHS) for 2007 asserts on page 25: “Given that most HIV infections in Zambia are contracted through heterosexual contact…[etc]”. That’s no different from the implicit or (more frequently) explicit assumption that almost all HIV transmission in high HIV prevalence countries is a result of heterosexual sex, which you will find in DHSs for most (if not all) other African countries and in the bulk of publications on the subject. Never mind that most of these publications themselves present findings making it clear that not all HIV transmission is a result of heterosexual sex.

There’s an interesting video on YouTube demonstrating how people like to try to confirm what they already believe and it takes some coaxing to get them to change their protohypothesis, it’s worth watching. Similarly, many scientists who write about HIV seem so reluctant to change their (proto?)hypothesis that they end up making statements that are flatly contradicted by their evidence. The Zambian DHS is a good example because the authors find that 3.5% of people who never had sex are HIV positive (3.7% of women and 3.5% of men). Enter an alternative hypothesis?

Not quite. The authors say on page 261: “This suggests either that some women and men incorrectly reported that they were not sexually active or that there is some degree of non-sexual HIV transmission occurring (e.g., through blood transfusions or non-sterile injections, etc.).” You might expect this to lead to some kind of investigation of the possibility of such non-sexual transmission, because if it happened to non-sexually active people, it could also happen to sexually active people. Evidence of sexual activity is not evidence of the route of transmission of HIV, which is difficult to transmit through heterosexual sex but easy through contaminated blood.

The authors seem unaware of a Zambian study that came out the year before (perhaps because they, like me, do not have access to Sage Publications). The paper finds that “Medically administered intramuscular or intravenous injections in the past five years (but not blood transfusions) were overwhelmingly correlated with HIV prevalence, exceeding the contribution of sexual behaviours in a multivariable logistic regression.” They conclude “the disproportionate association of medical injection history with HIV highlights the need to investigate further and prospectively the role of health-care injection in sub-Saharan Africa’s HIV epidemic.”

I am not aware of the results of any such investigation for Zambia, nor for any other country. But there are numerous examples of HIV probably being transmitted through non-sexual routes, often discovered because infants are found to be HIV positive although their mothers are HIV negative. We have a collection of similar cases, and investigations that have taken place in non-African countries, where the possibility of non-sexually transmitted HIV is sometimes (thought not always) taken more seriously; hence the long list of cases compared to the short list of investigations.

The above paper was published 8 years ago and the DHS was published 7 years ago and another Zambian DHS will appear in the next year or two, following the collection of data that started in early 2013. In the meantime, Kenya has also carried out a DHS in 2012 and the report may appear as early as this year. But already, a number of papers has appeared which are based on the collected data. One of them was published this month, and it is available free of charge, so concerned researchers have no excuse for not examining the complete findings.

The Kenyan paper finds that there is a strong preference for injections over pills, a phenomenon that is very common in developing countries, many of which have poor quality health facilities with shortages of skilled personnel, equipment and supplies. While 95.9% of people were said to have observed the health practitioner opening a new injection pack, men and women who had received an injection in the previous 12 months were “significantly more likely to be HIV infected compared with those who had not received medical injection in the past 12 months”. Hardly a smoking gun, but this is a step in the right direction. People who don’t give the expected answer to questions are not necessarily mistaken, lying or deluded.

But we have a long way to go. A recent paper on students in Uganda notes that many students are sexually active but that few studies have measured HIV prevalence, or even prevalence of various sexually transmitted infections (STI). They find that “there is little evidence of substantially increased HIV risk among” 640 students in five large universities in Kampala, the capital. HIV prevalence was .4% among males and .9% among females. Rates for some STIs were higher, especially among females, but the question is, did the researchers start to wonder if sexual transmission may not account for all HIV transmission among sexually active people?

The Ugandan study was a ‘Crane Survey’, which seem to take an interest in sexual behavior and people thought to be most at risk of HIV infection, as a result of their sexual behavior of course. Many studies and many sources of funding concentrate on sexual risk and sexual behavior but very few concentrate on non-sexual risk or behavior. People in high HIV prevalence countries face serious risks through unsafe healthcare, cosmetic and traditional practices, but these are rarely alluded to and even less frequently the subject of thorough inquiry.

If you start with a hypothesis and aim to collect data that you believe supports the hypothesis, while systematically ignoring, denying or otherwise belittling any evidence that does not fit with your hypothesis, you will end up with a very biased view of your research field. Yet that is how much of the research into HIV transmission appears to proceed, as if we already know that almost all transmission is a result of sexual behavior, but we just have to find out what kind of sexual behavior. It’s time to change the hypothesis: some HIV transmission is not a result of sexual behavior, so let’s deal with it.