A method to evaluate scientific evidence in clinical preventive medicine
A method to evaluate scientific evidence in clinical preventive medicine
复制标题
DOI：
10.5124/jkma.2011.54.10.1006
复制
发表时间：
2011-10-01
影响因子：
0.3
通讯作者：
Park, Byung-Joo
中科院分区：
文献类型：
--
作者：
Lee, Joongyub;Park, Byung-Joo
医学文献中科学证据的评价以研究方法论为基础，研究方法论是医学专业人员提高实践水平的基本工具。由于分析性研究通常为提出假设和选择合适的研究设计提供证据，本文旨在综述随机对照试验（RCT）、队列研究和病例对照研究中研究质量的评价方法。对系统性错误（包括选择偏差、绩效偏差、消耗偏差、检测偏差和发表偏差）的批判性评估是评估的基本要素。临床试验需要对序列生成和分配隐藏、暴露测量设盲、随访和意向治疗分析的完整性以及结局评价设盲进行充分描述。对于队列研究，如随机对照试验，需要适当控制混杂变量以防止选择偏倚，并且盲法和随访的完整性也非常重要。病例组与对照组的匹配、访谈者的盲法以及病例和对照的正确定义对于防止病例对照研究中的偏倚是重要的。由于批判性评价的过程取决于报告的质量，因此一直在努力提高医学文献的报告质量。然而，报告避免偏倚所需的所有要素并不能自动保证排除偏倚或保证研究的质量;相反，重要的是研究要素之间的逻辑联系以及要素的最终清晰表达。
Evaluation of scientific evidence in the medical literature is based on the research methodology, which can be the fundamental tool for medical professionals to improve their practice. Since analytic studies usually provide evidence for generating hypotheses and selecting appropriate research designs, this article aims to review the methods for evaluating the study quality in randomized controlled trials (RCTs), cohort studies, and case-control studies. Critical appraisal of systematic errors including selection bias, performance bias, attrition bias, detection bias, and publication bias are the essential elements of the evaluation. Clinical trials need to have an adequate description of sequence generation and allocation concealment, blinding for exposure measurement, completeness of follow-up and intention to treat analysis, and blinding for outcome evaluation. For cohort studies, like RCTs, appropriate control of confounding variables is needed to prevent selection bias, and blinding and completeness of follow-up are also very important. Matching of the case group to the control group, blinding of interviewers, and proper definition of the cases and controls are important to prevent bias in a case-control study. Since the process of critical appraisal depends on the quality of reporting, there have been efforts to improve the reporting quality of the medical literature. However, reporting all of the elements necessary to avoid bias does not automatically guarantee an exclusion of bias or the quality of a study; instead, what is important is the logical connection of the elements of a study and the eventual lucid expression of the elements.