Determinants of condom use among migrant farm workers in two South African provinces.

ABSTRACT Background: Previous studies have shown that HIV prevalence rates are relatively high while condom use is low in migrant communities in South Africa (Zuma 2003, Weltz 2007, IOM 2008). This study seeks to investigate factors associated with condom use in selected commercial farms in Limpopo and Mpumalanga provinces of South Africa. Methodology: This study is a secondary data analysis of the Integrated Biological and Behavioural Surveillance Survey (IBBSS) implemented by the International Organisation for Migration (IOM) in 2010 among 2,810 farm workers. The study population comprises all sexually active non-South African nationals who have worked on the farm from a period of less than one year to over ten years. Majority of the migrants are from Zimbabwe, Mozambique and Swaziland. The outcome variable is condom use at the last intercourse while the explanatory variables include sex, age, marital status, financial stability, sexual abuse, condom availability, transactional sex, attending an HIV function and period worked on the farm. Data analysis was undertaken in three stages. Firstly, univariate analysis of the variables was done to provide descriptive statistics of the study population. The second stage was bivariate analysis producing unadjusted odds ratios to examine the association between each of the predictor variables and the outcome variable. The final stage was multivariate analysis using logistic regression and producing odds ratios to examine the association of more than one predictor variable with the outcome variable. Results: The results show that access to free condoms, having sex while drunk, financial stability and living arrangements with spouse are the factors associated with condom use among migrant farm workers in Limpopo and Mpumalanga. Among migrant men financial stability and access to free condoms were significant while among women, marital status, having attended an HIV function, having sex while drunk and living arrangements with spouse are the significant factors associated with condom use. Unexpectedly, socioeconomic factors such as transactional sex, forced sex and demographic factors such as age, proved to be insignificantly associated with condom use. Conclusion: It was noted that condom use levels and determinants vary between male and female migrants. Although men portrayed exhibiting more high risk sexual behavior, women proved to be more vulnerable. Policies and programmes targeting migrant farm workers should be gender sensitive