Impacts of four communication programs on HIV testing behavior in South Africa

Impacts of four communication programs on HIV testing behavior in South Africa
复制标题

DOI:
10.1080/09540121.2014.901487
复制
发表时间:
2014-01-01
影响因子:
1.7
通讯作者:
Figueroa, Maria Elena
Figueroa, Maria Elena
中科院分区:
医学4区
文献类型:
--
作者:
Do, Mai;Kincaid, D. Lawrence;Figueroa, Maria Elena

文献摘要

被引文献

相似文献

本文旨在评估四种传播计划在促进南非性行为活跃的个人艾滋病毒检测行为方面的影响。这四个项目由南非约翰·霍普金斯健康和教育部门实施,旨在促进艾滋病毒预防行为,以及基于性别的暴力预防、结核病筛查和治疗以及减少饮酒。它们在2009至2010年间推出,都在推广艾滋病毒检测。数据来自以人口为基础的第三次全国艾滋病传播调查2012;6004名在过去12个月内发生性行为的男性和女性被纳入分析范围。多重因果归因分析被用来证明因果参照是合理的,并估计沟通计划的影响。研究结果表明,这些项目对艾滋病毒检测行为有显著的直接和间接影响。间接影响通过增加一个人察觉到他们的朋友被检测的可能性,以及与性伴侣和朋友谈论艾滋病毒检测的可能性,这反过来又增加了艾滋病毒检测的可能性。研究结果表明,沟通项目可以从多个角度促进艾滋病毒检测。这项研究还展示了在不可能随机化的仅后置设计中,因果归因的多种统计技术的使用。
This paper aims to evaluate the impacts of four communication programs on promoting HIV testing behavior among sexually active individuals in South Africa. The four programs, implemented by Johns Hopkins Health and Education in South Africa, are aimed to promote HIV prevention behaviors, as well as gender-based violence prevention, tuberculosis screening and treatment, and reduction of alcohol consumption. Launched between 2009 and 2010, they all promoted HIV testing. Data came from the population-based Third National AIDS Communication Survey 2012; 6004 men and women who had sex within the last 12 months were included in the analysis. Multiple causal attribution analysis is used to justify causal reference and estimate the impact of communication programs. Findings indicate significant direct and indirect effects of the programs on HIV testing behavior. Indirect effects worked through increasing one's likelihood of perceiving that their friends were tested and the probability of talking about HIV testing with sex partners and friends, which in turn increased the likelihood of HIV testing. Findings suggest multiple angles from which communication programs can promote HIV testing. The study also demonstrates the use of multiple statistical techniques for causal attribution in a post-only design, where randomization is not possible.