Condom use and the risk of HIV infection: who is being protected?

Descriptive baseline data at enrolment into a cohort of male factory workers who were tested for HIV serology and monitored for sero-con version over time, were analysed for condom use. At recruitment, the 1 146 men were asked about their sexual behaviour, history of sexually transmitted diseases (STDs), condom use and circumstances under which condoms were used. HIV seroprevalence in the cohort was 18,2 pc. Self reported use of condoms was low, with only 5 pc of the men reporting using them all the time. Forty four pc reported that they had never used a condom, 11,5 pc tried a condom only once, and 30,5 pc used condoms less than half the time. HIV positive men were more likely (Odds Ratio [OR]= 2,2 95 pc Cl: 1,3 — 3,3) to use condoms than those who tested negative. Men using a condom more than once were younger and had more education (p values < 0,0005). Univariate analysis showed that men with self reported risk factors for HIV infection were more likely to use condoms. Significantly more condom users reported paying for sex, multiple sex partners or (for married men) a girlfriend (p < 0,005). Condom users also more often had a history of genital ulcers, urethral discharge or other STDs. Few married men (24 pc) reported using a condom with their wives. Condom use was more commonly reported with commercial sex workers (44 pc) or other extramarital partners (36 pc). Some risk factors for HIV infection were also present amongst men who reported that they did not use condoms. Independent determinants of condom use identified by stepwise logistic regression analysis included young age, having a girlfriend (OR = 2,2; 95 pc Cl: 1,47 — 3,3), number of sex partners in the last year (OR = 1,27; 95 pc Cl: 1,06 —1,51 per partner), and paying for sex in the preceding year (OR = 1,74; 95 pc Cl: 1,06 — 2,83). The results show that men use condoms with partners considered risky, such as prostitutes or girlfriends but use condoms less often with their wives. The results underscore theneedfor health education for behavioural change that promotes universal, consistent use of condoms or monogamous partnership