指南与共识

欺诈、欺骗与颠覆:维护数据完整性的建议

Fraud, Deception, and Subversion: Recommendations for Maintaining Data Integrity.
指南 / 共识 · Arch Phys Med Rehabil · 2025-08-07

该特别通讯描述加拿大和美国两个研究组在随机对照试验和观察性研究中遭遇欺诈者的真实经历,并通过文献综述识别32篇同行评议文章,综合提出在研究设计、广告、招募筛选、评估、数据分析及报酬各环节预防、检测和管理欺诈的可操作策略。

要点

  • 在线研究易受欺诈活动影响,可损害研究结果有效性并消耗资金和人力资源。
  • 研究者应在研究设计、招募筛选、评估和数据分析等环节采取针对性防欺诈策略。
  • 并非所有策略普遍适用,需根据研究设计的具体风险评估选择匹配策略。

核心结论

  • 在线研究虽具优势但易受欺诈影响,研究者需评估研究设计特定风险并选择匹配策略以平衡减少欺诈、招募真实参与者和维护数据完整性。

研究设计与样本

  • 特别通讯,描述加拿大和美国各1个研究组在随机对照试验和观察性研究中遭遇欺诈者的真实经历,并进行文献综述识别32篇同行评议文章。

主要发现

  • 两个研究组在随机对照试验和观察性研究中均遭遇欺诈者。
  • 文献综述识别出32篇同行评议文章,涉及在线研究中的欺诈性参与和欺骗行为。

关键数值与推荐

  • 研究设计策略包括社区参与。
  • 广告策略包括限制信息。
  • 招募和筛选策略包括基于知识的资格问题。
  • 评估策略包括嵌入重复问题。
  • 数据分析策略包括移除欺诈数据。

临床怎么用

  • 研究者应评估其研究设计相关的特定风险,并选择与研究目标一致的战略,以平衡最小化欺诈、招募真实参与者和维护数据完整性。

局限性

  • 每种研究设计都有独特的脆弱性,并非所有方法都普遍适用。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

Researchers increasingly use online tools to advertise studies, recruit participants, and collect data, enabling access to individuals in remote or rural areas, those with limited mobility, and participants facing time constraints. While online platforms offer significant advantages, online research remains vulnerable to fraudulent activity, which can compromise the validity of study findings and deplete both funding and human resources. This special communication describes the real-world experiences of 2 research groups, 1 in Canada and 1 in the United States, who encountered fraudsters in both randomized controlled trials and observational studies. We conducted a literature review and identified 32 peer-reviewed articles addressing fraudulent participation and deceptive behaviors in online research. Drawing from both our experiences and the literature, we synthesize actionable strategies to prevent, detect, and manage fraud throughout the research process, including study design (eg, community involvement), advertisement (eg, limited information), recruitment and screening (eg, knowledge-based eligibility questions), assessment (eg, embedding repeated questions), data analysis (eg, removing fraudulent data), and remuneration (eg, clarifying reimbursement conditions). Despite the availability of these strategies, each study design presents unique vulnerabilities, and not all approaches are universally applicable. Researchers are encouraged to assess the specific risks associated with their study design and select strategies that align with their study objectives, balancing the need to minimize fraud, recruit genuine participants, and maintain data integrity.

来源:Arch Phys Med Rehabil,查看原文。版权归原作者所有。

引用

  1. Dhariwal A, Mohammadi S, Golden K, et al. Fraud, Deception, and Subversion: Recommendations for Maintaining Data Integrity. Arch Phys Med Rehabil. 2025. PMID: 40783091. doi:10.1016/j.apmr.2025.07.016

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。