O'Connor Annette, Sargeant Jan
Department of Veterinary Diagnostic and Production Animal Medicine, Iowa State University College of Veterinary Medicine, Ames, IA, USA.
Centre for Public Health and Zoonoses, University of Guelph, Guelph, ON, Canada; Department of Population Medicine, Ontario Veterinary College, University of Guelph, Guelph, ON, Canada.
Vet J. 2015 Dec;206(3):261-7. doi: 10.1016/j.tvjl.2015.08.025. Epub 2015 Aug 28.
Reviews of the scientific literature are critically important for synthesizing the state of knowledge and are used extensively in teaching, clinical practice and public policy. Despite the importance of literature reviews, in veterinary science little attention has been paid to the science of research synthesis. In the same manner that diverse study designs address diverse research questions, different approaches to combining scientific literature serve different and valid purposes. However, and again reflective of the underlying primary research, the potential for bias in a review should also be considered when interpreting the results. This article introduces some basic concepts in research synthesis and discusses some of the basic forms of reviews including narrative reviews, systematic reviews and meta-analysis. Also discussed are potential sources of bias and design features that can be incorporated into reviews to either reduce, or at least acknowledge, the potential for bias.
对科学文献的综述对于综合知识状态至关重要，并且在教学、临床实践和公共政策中被广泛使用。尽管文献综述很重要，但在兽医学中，对研究综合的科学却很少受到关注。与不同的研究设计解决不同的研究问题的方式相同，综合科学文献的不同方法服务于不同且合理的目的。然而，同样反映基础的原始研究，在解释结果时也应考虑综述中存在偏差的可能性。本文介绍了研究综合中的一些基本概念，并讨论了一些综述的基本形式，包括叙述性综述、系统综述和荟萃分析。还讨论了偏差的潜在来源以及可以纳入综述以减少或至少承认偏差可能性的设计特征。
