Dual protection, defined as any strategy that prevents both unwanted pregnancy and sexually transmitted infections (STIs), including HIV, is emerging as an important preventive approach in reproductive health.1 It may include various combinations of pregnancy and STI prevention, such as the use of condoms with hormonal contraceptives, or it may consist of other risk reduction behaviors such as nonpenetrative sex or abstinence.

In South Africa, the predominance of hormonal contraceptives, particularly injectables, means that the simplest approach to dual protection is through the combination of a barrier method, most commonly the male condom, with a nonbarrier (primarily hormonal) contraceptive. South Africa's new reproductive health policies promote dual method use as an important means of preventing both unwanted conceptions and transmission of HIV and other STIs.2 Nevertheless, little is known about the current levels of, and barriers to, dual method use in sexually active South African populations. In this article, we explore the current levels and predictors of dual method use among condom users in South Africa and consider potential interventions to promote dual protection.


The study was undertaken during 1998-1999 in 12 public health facilities from four health regions of South Africa. (Detailed data collection methods have been published previously.3) Consecutive individuals leaving these sites with condoms were invited to participate in the study. Eligible respondents included clinic clients who had asked for condoms as well as family planning and STI patients who had obtained condoms from health care providers during their clinic visit. Of the 594 individuals approached, 554 men and women (93%) completed a baseline interview, which used a semistructured questionnaire to collect demographic data and information on recent sexual behavior and on condom-related knowledge and attitudes. At the end of this interview, the respondents were scheduled for two follow-up interviews over the next five weeks. At each follow-up, the respondents participated in loosely structured in-depth interviews exploring sexual behaviors; condom and contraceptive use; and their attitudes and those of their partners toward HIV/AIDS, other STIs and pregnancy. All interviews took place in a private room at the health care facilities and were conducted in the respondents' own language by trained fieldworkers.

Statistical analysis used chi-square tests and t-tests to identify the factors associated with dual method use, defined as the use of condoms with another form of contraception (including sterilization, the IUD or the diaphragm) during their most recent sexual encounter. Factors shown to have crude associations with dual method use were entered as independent variables into a logistic regression model that adjusted for participant age and gender; variables were removed if they did not appear to be associated with dual method use in the model and if their presence did not significantly influence the associations between the other covariates and dual method use. For qualitative analysis, interviews were sorted by gender and by condom and contraceptive practices, and then individually coded and grouped into themes.


The respondents' mean age was 26 years (range 14-63), and 43% were women (Table 1). About half the participants (51%) reported having used a condom during their most recent sexual encounter, and 34% reported having used a contraceptive other than a condom. Of that 34%, the vast majority (94%) reported having used the pill or the injectable. The 189 participants who had used a contraceptive (with or without a condom) at last intercourse were slightly less likely to have used a condom during that episode than were the 365 participants who had used only a condom or no method (47% vs. 53%); the difference between the two groups, however, was not statistically significant (p=.09). Eighty-eight sexually active participants (16%) had used both a condom and another method of contraception during their last sexual encounter. In a logistic regression model adjusted for age and gender, dual method use at most recent sexual encounter was associated with increased schooling (odds ratio of 1.1 for each additional year; 95% confidence interval, 1.1-1.2) and previous instruction on condom use from health care providers (1.7; 95% confidence interval, 1.1-2.8 ).

Three hundred and eighty-four (69%) of those surveyed at baseline also completed the qualitative interviews. Respondents who were successfully followed up were more likely to be male and were more likely than those lost to follow-up to report having used a condom during their last sexual encounter before the study.4 In follow-up interviews, male and female participants agreed that the primary function of condoms was to protect against HIV and other STIs, but were less comfortable with the use of condoms as contraceptives.

Although the protection condoms provide against HIV and other STIs benefits both partners equally, study participants perceived condoms primarily as a means of protecting men from infection by their female partners. Although many female respondents had an opinion on condom use, both males and females felt that the final decision on whether to use condoms as STI prophylaxis was the man's. Conversely, men and women concurred that although men usually expressed a preference about the use of methods to prevent pregnancy in their relationship, the decision was ultimately up to the female partner. These and other qualitative findings from participants across all sites suggest a framework in which decisions about condoms and other contraceptives are considered the domain of either the male or the female partner, but not of both. Dual method use, rather than being a consensual choice, generally occurs only when a man's aim of protecting himself from STIs coincides with his female partner's goal of preventing unwanted pregnancy.


To our knowledge, this is the first study of dual method use in South Africa. Although most previous dual method use research has focused on female family planning clients,5 this study interviewed both men and women who had obtained condoms about their condom and contraceptive use. Despite differences in study populations and locations, our finding of a low level of dual method use is broadly consistent with previous studies, which have found levels of dual method use ranging from 13% among family planning clients in Kenya6 to 17-20% in various U.S. populations.7

All of the participants in this study had access to primary health care facilities; dual method use may be lower in the general population. Although further research is needed to evaluate the levels of, and barriers to, dual method use in different South African populations, the next step in national and local policy-making is to develop concrete guidelines for the promotion of dual protection as part of primary care services. Such guidelines could focus on the integration of family planning and STI services (which remain separate in many South African health care settings), and seek to implement strategies for primary care providers to provide individuals with risk assessment and counseling on HIV and other STIs as well as on unwanted pregnancy.

This research has highlighted the gender-specific nature of existing contraceptive and prophylactic options. New reproductive health choices, such as the female condom, vaginal microbicides and male hormonal contraception, have the potential to rearrange the gender-specific obstacles to use of contraceptives and barrier methods by giving partners of both sexes a range of options through which to achieve dual protection.

Landon Myer is a senior scientist in the HIV Prevention and Vaccine Research Unit, South African Medical Research Council, Hlabisa, South Africa, and fellow, Fogarty AIDS Information, Training and Research Program, Mailman School of Public Health, Columbia University, New York. Chelsea Morroni is a scientific officer in the Department of Public Health and Primary Health Care, University of Cape Town, South Africa. Catherine Mathews is a senior scientist in the Health Systems Research Unit, South African Medical Research Council, and an honorary lecturer in the Department of Public Health and Primary Health Care, University of Cape Town. Francesca Little is a lecturer in the Department of Statistical Sciences, University of Cape Town.