Untested HIV-positive individuals more than twice as likely to engage in high-risk sex than those aware of their HIV-positive status

A meta-analysis of eleven studies has found that the prevalence of high-risk sexual behaviour is between 53-68% lower in HIV-positive individuals aware of their status than in HIV-positive individuals unaware of their status. The study, from the US Centers for Disease Control and Prevention (CDC) in Atlanta, was published electronically ahead of print in the Journal of AIDS at JAIDS online.

The US CDC estimates that of the 850,000 – 900,000 individuals currently living with HIV in the United States, 250,000 are unaware of their infection. However, the proportion of individuals unaware of their HIV infection varies substantially by risk group. Earlier this year, the CDC published data which found that 77% of HIV-positive young gay urban men were unaware of their HIV status .

Few data exist that compare the relative differences in high-risk sexual behaviour between those aware and unaware that they are HIV-positive. To this end, researchers from the CDC conducted a meta-analysis of published articles and conference abstracts that fulfilled the following criteria:

• Data were from the United States.
• Comparing a group of HIV-positive-aware individuals with an independent group of HIV-positive-unaware individuals.
• Alternatively, measuring seroconverting individuals before and after receiving an HIV-positive diagnosis.
• Measuring either unprotected insertive or receptive anal intercourse, unprotected vaginal intercourse and/or consistency of condom use during sexual intercourse.

Of 620 abstracts screened, eleven independent findings from eight studies conducted between 1988 and 2003 were included in the meta-analysis. These included four studies published in peer-reviewed journals and four multi-site data sets from the Multicenter AIDS Cohort Study (MACS), the HIV Epidemiology Research Study (HERS), the Supplement to HIV/AIDS Surveillance (SHAS phase 1, 1995-2000) and SHAS phase 2, 2000-2003. Six of the findings compared the high-risk sexual behaviour of individuals who were aware that they were HIV-positive with HIV-positive-unaware individuals (between-group comparisons) and five compared seroconverting individuals before and after being notified of their HIV-positive status (within-subject comparisons).

McCusker et al, Am J Public Health

This study from 1988 was a between-group comparison of HIV-positive gay men recruited at a Boston health clinic who reported engaging in unprotected anal intercourse (UAI) before getting HIV test results. The CDC calculated a 59% reduction in prevalence of high-risk sexual behaviour between HIV-positive-aware individuals and HIV-positive-unaware individuals.

Valleroy et al, JAMA

This between-group study, from 2000, of HIV-positive gay and bisexual men recruited at seven urban sites found that 37% of HIV-positive-unaware men engaged in UAI with any partner in the previous six months, compared with 13% of the HIV-positive-aware men. This was calculated as a 65% reduction in prevalence of high-risk sexual behaviour between groups.

SHAS, phase 1

The CDC calculated that the between-group reduction in prevalence of high-risk sexual behaviour with any partner in this study of HIV-positive men and women surveyed in twelve cities or states between 1995-2000 was 63% for men and 59% for women.

SHAS, phase 2

The CDC calculated that the between-group reduction in prevalence of high-risk sexual behaviour with at-risk partners in this study of HIV-positive men and women surveyed in 19 cities or states between 2000-2003 was 59% for men and 56% for women.

Cleary et al, Am J Public Health

This study from 1991 was a within-subject comparison of 153 HIV-positive male and 43 HIV-positive female blood donors from New York between 1986 and 1988. It reported on the difference in sexual behaviour (unprotected anal or vaginal intercourse, or unprotected oral sex) in the week before and up to two weeks after diagnosis. The CDC calculated a 41% reduction in these sexual behaviours for men, and a 34% reduction for women, after an HIV-positive diagnosis.

HERS

Fourteen women (79% injecting drug users) from multiple urban sites provided a within-subject comparison of unprotected anal or vaginal intercourse with any “casual” sex partner in the twelve months prior to, and 18 months after an HIV-positive diagnosis. The CDC calculated a 52% reduction in high-risk sexual behaviour after an HIV-positive diagnosis.

MACS

This within-subject study of 90 gay and bisexual men recruited at urban locations who seroconverted between 1988 and 1999 compared unprotected anal intercourse with any sex partner in the twelve months prior to, and 18 months after an HIV-positive diagnosis. The CDC calculated a 25% reduction in high-risk sexual behaviour after an HIV-positive diagnosis.

Colfax et al, AIDS

Forty-three gay and bisexual men from urban locations provided a within-subject comparison of unprotected insertive anal intercourse with a partner reported to be HIV-negative or of unknown HIV status in the six months prior to, and six to twelve months after an HIV-positive diagnosis. The CDC calculated a 59% reduction in high-risk sexual behaviour after an HIV-positive diagnosis.

The CDC researchers then calculated the effect size (i.e. the likelihood that knowledge of HIV-status resulted in a reduced prevalence of high-risk sex) of these studies. They found that, using an unadjusted model, the combined effect size for all eleven findings was an average of 53% (95% CI: 45-60%) lower in HIV-positive-aware individuals relative to HIV-positive-unaware individuals.

After adjusting the data to focus on unprotected anal or vaginal intercourse with partners who were not already HIV-positive, the researchers calculated that being aware of one’s HIV-positive status resulted in a 68% (95% CI: 59-76%) reduction in high-risk sex. The reductions were larger in between-group comparisons than in within-subject comparisons (p 0.10).

The researchers point out several limitations with the studies examined in the meta-analysis, in particular the use of self-reported sexual behaviour, which tends to lead to under-reporting of high-risk sex. The studies could also not ascertain HIV transmission risk, which is subject to many variables, including stage of HIV infection, viral load, and other concurrent sexually transmitted infections.

They argue, however, that their findings “reinforce the need for a multidimensional approach to HIV prevention”. This would include scaling-up resources and efforts “to make HIV testing opportunities more accessible and to reduce barriers to testing so that infected persons learn their status”.

They conclude by suggesting that HIV counselling and testing is not enough to control the HIV epidemic, and that the continued high-risk sexual behaviour of people aware of their HIV-status can also be modified through behavioural interventions, although “the challenge is to find settings and approaches for delivering prevention programs to this population over time”. They suggest a multi-factorial approach that includes counselling in HIV clinics, peer intervention and “assisting HIV-positive persons to establish social networks that encourage risk reduction and provide social support for seeking medical care and adhering to treatment regimens”.