Predictors of condom use and refusal among the population of Free State province in South Africa
1 Department of Health, Free State, Bloemfontein, 9300, South Africa
2 Institute of Tropical Medicine, Antwerp, Belgium
3 Department of Epidemiology and Social Medicine, University of Antwerp, Antwerp, Belgium
4 South African Centre for Epidemiological Modeling and Analysis, University of Stellenbosch, Stellenbosch, South Africa
5 Institut de Recherche Santé et Societé (IRSS), Université catholique de Louvain, Brussels, Belgium
Citation and License
BMC Public Health 2012, 12:381 doi:10.1186/1471-2458-12-381Published: 28 May 2012
This study investigated the extent and predictors of condom use and condom refusal in the Free State province in South Africa.
Through a household survey conducted in the Free Sate province of South Africa, 5,837 adults were interviewed. Univariate and multivariate survey logistic regressions and classification trees (CT) were used for analysing two response variables ‘ever used condom’ and ‘ever refused condom’.
Eighty-three per cent of the respondents had ever used condoms, of which 38% always used them; 61% used them during the last sexual intercourse and 9% had ever refused to use them. The univariate logistic regression models and CT analysis indicated that a strong predictor of condom use was its perceived need. In the CT analysis, this variable was followed in importance by ‘knowledge of correct use of condom’, condom availability, young age, being single and higher education. ‘Perceived need’ for condoms did not remain significant in the multivariate analysis after controlling for other variables. The strongest predictor of condom refusal, as shown by the CT, was shame associated with condoms followed by the presence of sexual risk behaviour, knowing one’s HIV status, older age and lacking knowledge of condoms (i.e., ability to prevent sexually transmitted diseases and pregnancy, availability, correct and consistent use and existence of female condoms). In the multivariate logistic regression, age was not significant for condom refusal while affordability and perceived need were additional significant variables.
The use of complementary modelling techniques such as CT in addition to logistic regressions adds to a better understanding of condom use and refusal. Further improvement in correct and consistent use of condoms will require targeted interventions. In addition to existing social marketing campaigns, tailored approaches should focus on establishing the perceived need for condom-use and improving skills for correct use. They should also incorporate interventions to reduce the shame associated with condoms and individual counselling of those likely to refuse condoms.