Vollert Jan, Spivack John, Adamson Blythe, Baron Ralf, Farrar John T, Gilron Ian, Hohenschurz-Schmidt David, Kerns Robert D, Mackey Sean, Markman John D, McDermott Michael P, Parides Michael, Rice Andrew S C, Turk Dennis C, Wasan Ajay D, Dworkin Robert H, Langford Dale J
Department of Clinical and Biomedical Sciences, Faculty of Health and Life Sciences, University of Exeter, Exeter, United Kingdom.
Biostatistics and Bioinformatics Program, Hospital for Special Surgery Research Institute, New York, NY, United States.
Pain. 2025 Sep 3. doi: 10.1097/j.pain.0000000000003798.
Real-world data (RWD) can be defined as routinely collected clinical or administrative data that might be used for research purposes and to generate real-world evidence (RWE). Computerized search and data mining methods, large electronic databases, and the development of novel computational and statistical methods allow for improved access to and analysis of RWD. Although RWD afford the opportunity to generate RWE with potentially improved efficiency and generalizability over prospective clinical studies, it is important to understand and apply best practices when analysing RWD, particularly when the goal is to generate RWE of diagnostic, prognostic, or treatment effectiveness. Real-world evidence can provide evidence complementary to randomized clinical trials (RCTs), especially in scenarios where RCTs are difficult to conduct. Real-world evidence studies need to be carefully designed, the research question clearly defined and addressable with the available RWD source, variables (treatment, outcome, covariates) operationalized, and hypotheses and analyses specified before data access. Sound interpretation of results requires a deep understanding of the benefits and limitations of RWE studies, including often deficient data quality, confounding, and other potential sources of bias. Registered protocols, registered reports as a publishing model, and/or restricted access to data until protocols are in place can be encouraged by journals and enforced by data guardians and will contribute to the emergence of high-quality RWD studies. Here, we summarize guidance documents on generating RWE of treatment effectiveness or comparative effectiveness, discuss the strengths and limitations of RWD and RWE, and provide recommendations for conducting effectiveness RWE studies in the pain field.
真实世界数据（RWD）可定义为常规收集的临床或管理数据，这些数据可用于研究目的并生成真实世界证据（RWE）。计算机化搜索和数据挖掘方法、大型电子数据库以及新型计算和统计方法的发展，使得对RWD的获取和分析得到改进。尽管RWD有机会以可能高于前瞻性临床研究的效率和普遍性生成RWE，但在分析RWD时，尤其是当目标是生成关于诊断、预后或治疗效果的RWE时，理解并应用最佳实践非常重要。真实世界证据可以提供与随机临床试验（RCT）互补的证据，特别是在难以开展RCT的情况下。真实世界证据研究需要精心设计，研究问题要明确界定且可通过可用的RWD来源解决，对变量（治疗、结局、协变量）进行操作化定义，并在获取数据之前明确假设和分析方法。对结果进行合理的解释需要深入理解RWE研究的益处和局限性，包括数据质量往往不足、混杂因素以及其他潜在的偏差来源。期刊可以鼓励采用注册方案、注册报告作为一种出版模式，和/或在方案确定之前限制数据访问，数据监管者可以强制执行这些措施，这将有助于高质量RWD研究的出现。在此，我们总结了关于生成治疗效果或比较效果的RWE的指导文件，讨论了RWD和RWE的优势与局限性，并为在疼痛领域开展疗效RWE研究提供建议。
