In Sub-Saharan Africa, HIV-Serodiscordant Couples Often Use Condoms—But Not Other Forms of Contraceptives

A. Kott

First published online:

While African women whose partner's HIV status does not match their own generally use condoms, they do not typically use other forms of contraceptives consistently, according to data from a multinational trial testing an antiviral drug.1 By the end of the two-year study, more than 90% of women were consistently using condoms. However, only 30% of HIV-positive women with uninfected male partners and 14% of HIV-negative women with infected male partners were consistently using a form of contraception other than condoms. Consistent dual contraceptive use was reported at 16–24% of study visits. Women who had had unprotected sex (i.e., without condoms) in the past month were more likely than those who had consistently used condoms to have used another form of contraception every time they had sex (odds ratios, 1.3–1.4).

Most cases of HIV transmission in Africa involve couples in stable relationships. Sero-discordant couples (those in which one partner has HIV and the other does not) in such relationships may be concerned about preventing pregnancy, preventing HIV transmission or both. Dual contraceptive use, which combines a condom with a highly effective form of contraception, is an effective strategy for addressing both issues.

To explore use of contraceptives in HIV- serodiscordant relationships, investigators conducted a secondary analysis of data from 3,407 heterosexual serodiscordant couples in East and southern Africa who had participated in a randomized trial examining whether the antiviral drug acyclovir reduces HIV transmission in couples affected by herpes simplex virus type 2. In 67% of couples, the woman had HIV; in 33%, the man did. Most of the women were married, were younger than 35, had completed no more than eight years of school, earned no income and had at least two children. All participants received risk reduction counseling, free condoms and treatment for STIs during the two-year trial. HIV-positive participants were monitored monthly; visits for HIV-negative participants were quarterly. At each study visit, women reported their sexual behavior and condom and contraceptive use with their partner during the past month. Most participants—92% of HIV-positive and 84% of HIV-negative women—remained in the study for two years.

For some analyses, researchers combined data from all follow-up visits; data from visits when the woman was pregnant and those that followed seroconversion were excluded. The investigators used generalized estimating equations to analyze, separately by women's HIV status, the relationships between contraceptive use and demographic and behavioral characteristics, as well as between unprotected sex and contraceptive use.

At enrollment, more than two-thirds of women reported having consistently (always) used condoms in the past month—71% of HIV-positive women and 73% of women without HIV. Twenty-four percent of HIV-positive women and 21% of HIV-negative women reported having consistently used some type of contraceptive besides condoms (IUD, tubal ligation, hysterectomy, or an oral, injectable or implanted hormonal method). The prevalence of consistent contraceptive use was higher at baseline among women from southern Africa (30% of HIV-positive and 29% of HIV-negative women) than among those from East Africa (20% of HIV-positive and 18% of HIV-negative women).

By the final study visit, two years later, the prevalence of consistent condom use had increased substantially, to 93–96%. However, only 30% of HIV-positive women and 14% of HIV-negative women reported consistent use of other contraceptives. Twelve percent of women with HIV and 7% of those without reported consistent use of contraceptives (excluding condoms) at every study visit, and an additional 34% of HIV-positive women and 28% of HIV-negative women reported consistent contraceptive use at least once.

Injectable contraceptives were the most popular method, reported by HIV-positive and negative women at 17% and 9%, respectively, of all study visits. Consistent use of oral contraceptives was reported at 5% of visits by HIV-positive women and 3% of those by HIV-negative women, while female sterilization, implants and IUDs were used even more infrequently (1–4% of visits each). At most visits, HIV-positive and negative women reported having had sex with their partner in the past month (81% and 84%, respectively). Although reports of unprotected sex were relatively uncommon, 26% of women with HIV and 25% without reported having had unprotected sex at least once after enrollment.

Consistent dual contraceptive use in the past month was reported at 24% of visits by HIV-positive women and 16% of those by HIV-negative women. While dual contraceptive use increased among HIV-positive women during the study period, no change occurred among uninfected women.

In multivariate analyses that controlled for demographic and socioeconomic variables, the odds that an HIV-positive woman reported consistent contraceptive use at a visit were lower if the woman was 45 or older than if she was aged 18–24 (odds ratio, 0.3); the odds were also reduced among women who were from East Africa (0.6) or married (0.8) or had had a live birth since their last study visit (0.6). The odds of contraceptive use were elevated for women who had at least one living child (2.4–4.6, depending on parity), had had unprotected sex with their partner during the previous month (1.3) or had had sex with an outside partner during the past month (2.0). For visits by HIV-negative women, the odds of consistent contraceptive use were reduced among those who were from East Africa (0.6) or had had a live birth since their last study visit (0.1), but elevated among women with one or more living children (2.3–3.3) and those who had had unprotected sex with their partner during the prior month (1.4).

The odds that a woman reported having had unprotected sex with her partner were elevated at visits where use of a contraceptive method other than condoms was reported (odds ratios, 1.3). No increase was observed, however, among women who had used contraceptives inconsistently.

According to the authors, the study's limitations include its lack of measures of fertility goals and views on contraception, and its reliance on self-reports of contraceptive use; however, among its strengths are its large, diverse sample and high retention rate during follow-up. The investigators endorse ongoing efforts to understand fertility desires and behaviors, increase use of dual protection and reduce unintended pregnancies among high-risk couples. "As dual contraceptive use has great potential to prevent unintended pregnancy and avert new HIV infections, it should be a high priority for programs counseling women in HIV serodiscordant partnerships," they contend.—A. Kott


1. Heffron R et al., A prospective study of contraceptive use among African women in HIV-1 serodiscordant partnerships, Sexually Transmitted Diseases, 2010, 37(10):621–628.