Am J Epidemiol. 2015 May 15;181(10):747-9. doi: 10.1093/aje/kwv007. Epub 2015 Apr 15.
Social media-based recruitment for epidemiologic studies has the potential to expand the demographic and geographic reach of investigators and identify potential participants more cost-effectively than traditional approaches. In fact, social media are particularly appealing for their ability to engage traditionally "hard-to-reach" populations, including young adults and low-income populations. Despite their great promise as a tool for epidemiologists, social media-based recruitment approaches do not currently compare favorably with gold-standard probability-based sampling approaches. Sparse data on the demographic characteristics of social media users, patterns of social media use, and appropriate sampling frames limit our ability to implement probability-based sampling strategies. In a well-conducted study, Harris et al. (Am J Epidemiol. 2015;181(10):737-746) examined the cost-effectiveness of social media-based recruitment (advertisements and promotion) in the Contraceptive Use, Pregnancy Intention, and Decisions (CUPID) Study, a cohort study of 3,799 young adult Australian women, and the approximate representativeness of the CUPID cohort. Implications for social media-based recruitment strategies for cohort assembly, data accuracy, implementation, and human subjects concerns are discussed.
社交媒体招募在流行病学研究中具有扩大调查人员的人口统计学和地理范围的潜力，并比传统方法更有效地确定潜在参与者。事实上，社交媒体特别吸引人之处在于它们能够吸引传统的“难以接触”的人群，包括年轻人和低收入人群。尽管它们作为一种流行病学家的工具具有巨大的潜力，但基于社交媒体的招募方法目前并不比基于金标准概率抽样的方法更有利。社交媒体用户的人口统计学特征、社交媒体使用模式以及适当的抽样框架的数据稀疏限制了我们实施基于概率的抽样策略的能力。在一项精心设计的研究中，Harris 等人（Am J Epidemiol. 2015;181(10):737-746）研究了基于社交媒体的招募（广告和推广）在避孕使用、怀孕意向和决策（CUPID）研究中的成本效益，该研究是一项对 3799 名澳大利亚年轻女性的队列研究，以及 CUPID 队列的近似代表性。讨论了基于社交媒体的招募策略在队列组建、数据准确性、实施和人类受试者关注方面的影响。
