CROI: ARV provision in Africa could cut HIV transmission by 90 per cent

This article is more than 16 years old. Click here for more recent articles on this topic

The provision of antiretrovirals (ARVs), along with comprehensive sexual risk behaviour and ARV adherence support programmes, cut the risk of HIV transmission by 91% over a three year period in a study from eastern Uganda, the Fifteenth Conference on Retroviruses and Opportunistic Infections heard today in Boston.

There was some evidence of an increase in risky sexual behaviour in people receiving ARV treatment, the conference heard. But the decline in the risk of HIV transmission due to the reduced viral load in people on ARVs far outweighed any effect of increases in sexual activity and risky sex.

Rebecca Bunnell of the US Centers for Disease Control’s programmes in Kenya and Uganda presented findings from a study of 926 HIV positive people who initiated antiretroviral therapy (ART) in Tororo province in south-eastern Uganda, funded by the US PEPFAR programme.



Choosing sexual partners of the same HIV status, or restricting condomless sex to partners of the same HIV status. As a risk reduction strategy, the drawback for HIV-negative people is that they can only be certain of their HIV status when they last took a test, whereas HIV-positive people can be confident they know their status


Short for voluntary counselling and testing.


The transition period from infection with HIV to the detectable presence of HIV antibodies in the blood. When seroconversion occurs (usually within a few weeks of infection), the result of an HIV antibody test changes from HIV negative to HIV positive. Seroconversion may be accompanied with flu-like symptoms.



A serodiscordant couple is one in which one partner has HIV and the other has not. Many people dislike this word as it implies disagreement or conflict. Alternative terms include mixed status, magnetic or serodifferent.


A serodifferent couple is one in which one partner has HIV and the other has not. 

The treatment programme also offered an integrated TB and preventative care package for people with HIV and involved their families, with home-based voluntary counselling and testing (VCT) offered to all family members including children. All tests were conducted by lay workers at home; there were no clinic visits after study enrolment.

Annual VCT was offered to HIV negative sexual partners, couples drew up ‘sexual behaviour plans’, and individual and couple counselling and condoms were provided on request.

Participants’ sexual behaviours and HIV viral load were assessed at baseline and every six months. ‘Risky sex’ was defined as inconsistent or no condom use with HIV-negative or unknown-status partners in the previous three months.

Of the 926 people starting ART, 691 (75%) were women. There were few differences between the men and the women except that considerably fewer of the women were married. This was because 60% had already been widowed, presumably in many cases due to AIDS, compared with only 15% of men. Only a quarter of the women were currently married compared with 68% of men. The only other difference was that 30% of the men drank alcohol regularly compared with 10% of women. The average age of the cohort was 39.

During the three-year follow-up, 10% of ART recipients died. The proportion who either died or were lost to follow-up after three years was 18 per cent.

There was strong evidence of both increased sexual activity and increased unprotected sex – and also of serosorting. Rates of sexual activity (any sex in the last three months) increased from 48% to 72% in men and from 21% to 31% in women. Rebecca Bunnell commented that these increases went hand-in-hand with increases in reported quality of

life and decreases in depression, as measured by standardised questionnaires.

The vast majority of sexually active participants were monogamous, with 94% only having had one partner in the previous three months. But the proportion of partners who had risky sex, as defined above, decreased from baseline but then started increasing, especially after two years, and by three years was almost back to baseline levels in men and women.

The proportion of men who had risky sex declined from 9% at baseline to 3.8% at 24 months but then returned to nearly 9% at 36 months. In women 4.5% were having risky sex at baseline; this decreased to less than 2% at 6 months but by 24 months had returned to baseline levels.

The mean number of unprotected sex acts over the previous three months decreased from 4.7 to almost zero in men, but then increased to 2.6 by 36 months; in women the number of risky sex acts decreased from 5.6 at baseline to 2.1 at 24 months but then increased to 5.2 at 36 months.

Factors associated with risky sex included, at baseline, being married; married participants were, not very surprisingly, nearly seven times more likely to have unprotected sex with a sero-different partner. Prosperity was associated with risky sex at baseline too, with those having a trade or profession3.1 times more likely to be associated with risky sex. Alcohol drinking was associated with risky sex both at baseline (odds ratio 3.0) and follow-up (OR = 3.7), so remained a risk factor throughout the study.

Adherence amongst patients given ARVs was extremely high. Over 98% of patients took over 95% of their doses, as measured by pill count. As a result, the median viral load in those reporting risky sex declined from 150,000 copies/ml to below 50 copies/ml.

Based on these figures, and using estimates of infectivity and viral load derived from previous studies, Bunnell and colleagues calculated that there had been an overall reduction of 91% in the chance of HIV transmission from the people on ART during the three years of follow-up. At baseline it was estimated that their risk of transmitting HIV was just under 5% a year. By the ninth month of the study this had fallen to 0.1%. By month 30, because of the increases in risky sex, this had risen to just under 0.5%.

There was only one seroconversion amongst negative partners. Among 62 serodiscordant couples, one husband of a female participant became HIV-positive, within the first year of the study, with no further seroconversions after that. Bunnell said that the man’s wife had responded slowly to ART and had taken six months to achieve undetectability; however, at present, it was not known exactly when her husband acquired his infection.

Bunnell drew a number of conclusions from the study. The observed waning of behavioural impact over time was of concern, and pointed towards intensified prevention messages being needed early in the ART experience. Partner testing was important, especially as there was evidence of serosorting. And in this population, interventions targeted at drinkers were also needed.

However, she added, integrated ART and prevention programmes based on the PEPFAR model had the potential to greatly reduce HIV transmission in Africa. “The important thing now,” she added at a press conference, “is to target the 75% of people in Africa who have never had an HIV test.”


Bunnell R et al. 3-year follow-up of sexual behavior and HIV transmission risk of persons taking ART in rural Uganda. Fifteenth Conference on Retroviruses and Opportunistic Infections, Boston, abstract 29, 2008.