Cancer Group, Medical Research Council Clinical Trials Unit, London, UK.
Lancet Oncol. 2010 Jul;11(7):694-700. doi: 10.1016/S1470-2045(10)70041-3. Epub 2010 May 25.
Randomised controlled trials are the gold standard method for developing evidence-based medicine. Good trial design and an awareness of some potential pitfalls are likely to maximise the chances of a successful trial with a conclusion that adds meaningfully to the evidence base. This paper is aimed at people early in their research careers and focuses on some common, usually avoidable, pitfalls in trial design. The areas covered include: assessing the scientific idea; trial design; size and duration of the trial; analysis; and reporting and presentation.
随机对照试验是循证医学发展的金标准方法。良好的试验设计和对一些潜在陷阱的认识，可能会最大限度地提高成功试验的机会，并得出对证据基础有意义的结论。本文面向研究生涯早期的人员，重点介绍试验设计中一些常见且通常可以避免的陷阱。涵盖的领域包括：评估科学理念；试验设计；试验的规模和持续时间；分析；以及报告和呈现。
