Skills-building behavioural interventions advocated for reducing alcohol-related HIV risk in South Africa

Gus Cairns, Kelly Safreed-Harmon
Published: 07 August 2008

Findings from a South African study presented on Wednesday at the XVII International AIDS Conference suggest that people whose alcohol consumption puts them at higher risk of HIV infection may benefit from skills-building risk reduction interventions.

The study was a successful transfer of a community based alcohol-related risk intervention from the USA, adapted for local needs in South Africa. A previous study from the same team using a single 60-minute counselling session produced a 25% increase in condom use and a 65% reduction in unsafe sex sustained over six months among participants.

The present study used a three-hour group workshop. This produced a decline of two-thirds (67%) in unsafe sex sustained over three months compared with a one-sixth (17%) decline in a control group in lighter drinkers. However in heavier drinkers the effect was less marked, producing only a 25% decline in unsafe sex, which was similar to the control group. At six months the effect of the intervention in lighter drinkers had somewhat dissipated, although they were still having less unsafe sex than at baseline. The study had an 89% retention rate.

In this study, 353 men and women targeted at local shebeens (informal bars) and through street-intercept surveys were randomised either to receive, at a local community centre, a single three-hour workshop on alcohol use, HIV risk and the building of behavioural skills to reduce risk, or a one-hour control intervention which consisted of the first hour of the workshop. This had a purely educational content giving information about the link between alcohol and unsafe sex. The full workshop added discussion of what triggered unsafe sex, role-plays, goal-setting and teaching safer-sex negotiation skills.

The study recruited 117 men and 236 women. The average age was 34, and mean length of HIV diagnosis was nine years. Not all the group were black African; 23% were from other racial groups (coloured, Asian, white being the categories used in South Africa). The vast majority were unemployed and 18% had had a sexually transmitted infection in the last year. Their AUDIT score was also measured; this stands for Alcohol Use Disorders Identification Test and is a World Health Organization-accredited standardised questionnaire-based measure of alcohol use.

Seventy-one per cent of participants had an AUDIT score over eight (out of a maximum of 20), which indicates problem drinking; the average score was 14.5.

At baseline, light drinkers in the intervention group reported an average of 3.5 and heavy drinkers four episodes of unprotected sex in the previous month. At three month follow-up, this had declined to 1.2 episodes in lighter drinkers but only to three episodes in heavier drinkers. In contrast, unsafe sex episodes in lighter drinkers receiving the control intervention only declined from four to 3.5 episodes in the previous month.

There were also declines in the involvement of alcohol in unsafe sex. In lighter drinkers, the number of episodes of unsafe sex which involved getting drunk declined from 2.3 to 0.7 per month in the intervention group but only from three to 2.5 in the control group. Again, heavier drinkers benefited less from the intervention, with alcohol use preceding unsafe sex only declining from four to three episodes a month.

The experimental cohort also reported a 74% increase in condom use at three-month follow-up (from 43% to 75% of sex acts protected), fewer sex partners, and fewer partners met at shebeens. The researchers suggest initiating HIV prevention efforts in alcohol establishments in South Africa and presenter Professor Leickness Simbayi of Cape Town University said the research team was now testing a multi-component intervention for men who drink at shebeens and members of their social network.