Correlation of 'Unsafe' Sex With HIV Transmission Does Not Prove Causation

A couple of HIV researchers have come up with a 'plan' to reduce HIV transmission: no sex for a month, or 100% condom use for a month. This plan hit the media some months ago because I discussed it in July, but it's definitely media friendly. The argument runs that a month of safe sex, or of no sex at all, could "help reduce the spread of HIV by skipping the period immediately after an individual acquires the virus when they are most infectious".

The researchers, Alan Whiteside and Justin Parkhurst, reckon 10-45% of infections result from sex with people during this brief period, when infected people are most infectious. Apparently these estimates are based on 'models'.

They bemoan the fact that nearly 50% of Swazi women aged 25-29 are HIV positive, despite past prevention efforts. However, these past prevention efforts included abstaining, being faithful to one partner and using condoms (without too much emphasis on the condoms in most instances). This was the much vaunted 'ABC' strategy, also considered media friendly in its day.

UNAIDS have coined the term 'Treatment 2.0' to refer to the sad bunch of 'strategies' they expect to use to reduce HIV transmission in the future. And, like 'Web 2.0' that inspired the name, there isn't really anything new about Treatment 2.0. It's just the same old stuff in slightly different packaging. This also seems to be true of the plan to advocate no sex for a month or 100% condom use for a month.

I accept that 'abstain for a month' sounds far more achievable than just 'abstain' and that 'use condoms for a month' sounds far more achievable than 'use condoms'. But Demographic and Health Surveys don't show that ABC and similar strategies didn't work very well: they suggest that there is little correlation between knowledge about safe sex, safe sex behavior and HIV transmission. In other words, behavior change communication (BCC) has been a total failure, in every country where it was implemented.

In common with the rest of the HIV prevention rhetoric, the insistence that transmission in African countries is all about sex seems to be mistaken. Some HIV transmission may be sexual, but not all. Therefore, HIV prevention strategies could include some that target sexual behavior. We would still be left with the problem of which ones, given that none have been particularly successful. But it's a start.

More importantly, though, a lot more attention needs to be paid to the question of how much HIV transmission is down to sexual behavior and how much is not. What are the other drivers of HIV transmission? There's clearly an issue with transmission through unsafe medical procedures, such as injections, because the WHO has reported that 70% of injections are unnecessary and 17-19% of them are unsafe in sub-Saharan Africa.

What about other medical procedures? Service Provision Assessments for African countries show that many health facilities lack trained staff, equipment, safety guidelines and the ability to carry out infection control measures. Many non-African countries have had huge levels of nosocomial HIV transmission, transmission as a result of health procedures. If it can happen in Western countries and Asian countries, why not in African countries?

Perhaps Whiteside and Parkhurst are right about selecting a specific time period, such as one month, but wrong about confining it to sex. Perhaps we could have a medical safety month, as well. This would be a good opportunity to estimate the effect such a measure would have on HIV transmission. The only problem is that health facilities don't have the capacity to provide safe healthcare for a day, let alone a month.

And that's a problem for a no-sex/safe-sex month, too. African countries don't have much capacity when it comes to national campaigns. They don't have the health facilities, the educational capacity, the infrastructure or anything else that would be required (though it has alway been a bit of a mystery as to what could ensure 'safe sex' in any country, for any period of time). There are many indications of how poor health services are, such as the number of people who suffer from and die from preventable and curable diseases, especially water borne and respiratory diseases.

Another indication, though, is the high rates of preventable and curable sexually transmitted infections. It is known that these can increase HIV transmission, in addition to causing a lot of poor health, misery and even death.

Perhaps Whiteside and Parkhurst are having a bit of a joke when they refer to Muslims 'abstaining' from sex during Ramadan, even though they only abstain during daylight hours. Does avoiding sex during the day constitute 'reducing risky sexual behavior'? HIV prevalence is often lower among Muslims than among non-Muslims and this may be related to sexual behavior, but it's unlikely to be related to sexual behavior alone. HIV, unlike many other sexually transmitted infections (STI), is difficult to transmit sexually but relatively easy to transmit through unsafe medical procedures.

Imagine this scenario: if high levels of unprotected sex were to give rise to high levels of sexually transmitted infections aside from HIV, this might result in a lot of people visiting hospitals and clinics for diagnosis and treatment. If those health facilities were inadvertently transmitting HIV, it would appear to be the sexual behavior that was driving the epidemic when, in reality, it was also unsafe medical procedures.

Populations that engage in unsafe sex are doubly at risk of being infected with HIV if transmission of other STIs is also high, or is thought to be high. Sex workers, men who have sex with men, intravenous drug users, perhaps even truck drivers, who have also been branded as 'risk' groups, pay regular visits to STI clinics to receive both preventive and curative measures.

So there's a bit of homework for Whiteside, Parkhurst and the extremely well funded bunch that make up UNAIDS and the HIV industry: are we assuming that correlation (of sexual behavior with HIV transmission) is equal to causation? When people are clearly at risk of being infected with HIV through sexual behavior, we are assuming that that's always how they are infected, so we are not looking for any other explanation. Perhaps it's time we looked at other possibilities.

If we don't look at other possibilities for HIV transmission, we may continue to think that it's all to do with sex. As a result, we may continue to fail to prevent the bulk of infections while patting ourselves on the back for appearing to prevent some. Ultimately, UNAIDS and the rest of them want to know the truth, right? Aids is not just about making money or careers, is it? It's unlikely that large numbers of people will abstain from sex for a month, or even use condoms consistently for a month. But even if they do, HIV transmission will remain high enough to maintain a serious epidemic.