A Multilevel Analysis of Dual-Method Contraception Use Among Women in Sub-Saharan Africa

Background

Although modern contraception is increasingly available and affordable, many women remain at-risk for unintended pregnancy and HIV throughout sub-Saharan Africa (SSA) due to non-use of modern contraception. Dual-method contraception (use of both condoms and non-barrier contraception including pills, injectable, intrauterine device, implant, or sterilization) is recommended to prevent both unintended pregnancy and HIV transmission. Despite promotion efforts, as women gain familiarity with non-barrier methods, they may reduce their use of condoms.

Educational attainment may directly and indirectly prevent HIV transmission; however, educational standards differ throughout SSA. In addition, human developmental progress, such as better labor-market conditions, may influence the use of condoms. However, macro-level factors are often neglected in the health behavior literature; therefore, we do not know if these factors impact dual-use across the region. Social-ecological approaches, such as multilevel analysis, can identify how contextual structures affect individual-level behavior. The purpose of this study was to examine the magnitude of the relationship between dual-method use and education across SSA countries and determine if national-level human development moderates this effect among current users of non-barrier contraception.

Methods

We used individual-level data (N=87,276) from the Demographic and Health Surveys and national-level data (N=30) from the World Bank and UN Statistics Division. Data analysis was conducted using STATA 14.0 and HLM 6. We ran multilevel logit models of the association between the dichotomous outcome (dual-method use versus no dual-method use) and individual-level education and national-level human development (human development index score). Based on prior literature, we decided a priori to retain individual-level age, parity, residence type, and marital status and national-level HIV exposure (HIV prevalence rate) and socioeconomic status (per capita gross domestic product) as controls. Analysis was restricted to women of reproductive age (15-44 years) who were not pregnant.

Results

Eight percent of the 87,276 women were dual-method users. Fourteen percent of the true between-country variance in dual-method use was explained by individual-level age, residence type, parity, marital status, and education, and 58% was explained by HIV exposure, human development, and national socio-economic level. The magnitude of the relationship between dual-method use and educational attainment varied across countries (𝑋² = 149.18, df = 29, p = 0.020). Human development accounted for the country-to-country variation in the effect associated with education (OR 0.74; 95% CI: 0.58, 0.95).

Conclusions

Among current users of non-barrier contraceptive methods, few continue to use condoms. Although, formal educational attainment increased the likelihood of condom use. This effect was moderated by achievements associated with human development. Specifically, in more developed countries, formal education did not increase the odds of condoms use, but in less developed countries, it did.

Implications

Countries across SSA should focus capacity building efforts on social and economic development. Formal educational opportunities are continued to be of need in less developed countries and can act as a mechanism to increase human development. Increases in HIV prevalence and unintended pregnancy, however, work against development and need to be addressed by policy makers on a macro-level. Policies should be enacted to spur social and economic growth.