Human and Social Development, Human Sciences Research Council, Pietermaritzburg, KwaZulu Natal, South Africa.
Developmental Pathways for Health Research Unit, University of the Witwatersrand, Johannesburg, Gauteng, South Africa.
PLoS One. 2020 Oct 2;15(10):e0239304. doi: 10.1371/journal.pone.0239304. eCollection 2020.
Sub-Saharan Africa has been heavily impacted by the HIV/AIDS epidemic. Social data (e.g., social media, internet search, wearable device, etc) show great promise assisting in public health and HIV surveillance. However, research on this topic has primarily focused in higher resource settings, such as the United States. It is especially important to study the prevalence and potential use of these data sources and tools in low- and middle-income countries (LMIC), such as Sub-Saharan Africa, which have been heavily impacted by the HIV epidemic, to determine the feasibility of using these technologies as surveillance and intervention tools. Accordingly, we 1) described the prevalence and characteristics of various social technologies within South Africa, 2) using Twitter, Instagram, and YouTube as a case study, analyzed the prevalence and patterns of social media use related to HIV risk in South Africa, and 3) mapped and statistically tested differences in HIV-related social media posts within regions of South Africa. Geocoded data were collected over a three-week period in 2018 (654,373 tweets, 90,410 Instagram posts and 14,133 YouTube videos with 1,121 comments). Of all tweets, 4,524 (0.7%) were found to related to HIV and AIDS. The percentage was similar for Instagram 95 (0.7%) but significantly lower for YouTube 18 (0.1%). We found regional differences in prevalence and use of social media related to HIV. We discuss the implication of data from these technologies in surveillance and interventions within South Africa and other LMICs.
撒哈拉以南非洲深受艾滋病毒/艾滋病的影响。社会数据（例如社交媒体、互联网搜索、可穿戴设备等）在公共卫生和艾滋病毒监测方面具有很大的应用潜力。然而，这一主题的研究主要集中在资源较丰富的环境中，例如美国。在受艾滋病毒流行影响严重的低收入和中等收入国家（LMIC），如撒哈拉以南非洲，研究这些数据源和工具的流行程度及其潜在用途尤为重要，以确定这些技术作为监测和干预工具的可行性。因此，我们 1）描述了南非各种社会技术的流行程度和特征，2）以 Twitter、Instagram 和 YouTube 为例，分析了与南非艾滋病毒风险相关的社交媒体使用的流行程度和模式，3）绘制并统计测试了南非各地区与艾滋病毒相关的社交媒体帖子的差异。2018 年的三周时间内收集了地理标记数据（654373 条推文、90410 条 Instagram 帖子和 14133 条 YouTube 视频，其中有 1121 条评论）。在所有推文中，有 4524 条（0.7%）与艾滋病毒和艾滋病有关。Instagram 的比例相似，为 95（0.7%），但 YouTube 的比例明显较低，为 18（0.1%）。我们发现与艾滋病毒相关的社交媒体的流行程度和使用存在区域差异。我们讨论了这些技术的数据在南非和其他 LMIC 中进行监测和干预的意义。
