中华心血管病杂志刊出科研论文的研究设计及统计学方法分析

Review of research design and statistical methods in Chinese Journal of Cardiology

目的评价中华心血管病杂志刊出科研论文的研究设计和统计方法使用现状。方法对中华心血管病杂志第35卷第12期及第36卷第1至11期中“专题研究”、“临床研究”、“基础研究”及“流行学与人群防治”四个栏目中所有论文的设计类型及统计方法使用情况进行评述。结果最常见的设计类型为：横断面研究（34％）、前瞻性研究（21％）和实验研究（25％）。统计方法使用错误49篇（25％），欠缺29篇（15％），表述有误或不一致23篇（12％）。五类使用频率最高的统计方法分别为：t检验、方差分析、列联表、多元回归及非参数检验，不同方法的使用正确率差异有统计学意义（P〈0．001）。多因素分析正确率较低及重复测量资料未使用相应的统计分析方法。结论2008年度中华心血管病杂志刊出的科研论文研究设计和统计方法使用仍存在一定问题，建议加强流行病与统计学专家审稿。

Objective To evaluate the research design and the use of statistical methods in Chinese Journal of Cardiology. Method Peer through the research design and statistical methods in all of the original papers in Chinese Journal of Cardiology from December ：2007 to November 2008. Result The most frequently used research designs are cross-sectional design （34%）, prospective design （21%）and experimental design（25% ）. In all of the articles,49（25% ） use wrong statistical methods,29（ 15% ） lack some sort of statistic analysis ,23 （12%） have inconsistencies in description of methods. There are significant differences between different statistical methods （ P 〈 0. 001 ）. The correction rates of multifactor analysis were low and repeated measurement datas were not used repeated measurement analysis. Conclusion Many problems exist in Chinese Journal of Cardiology. Better research design and correct use of statistical methods are still needed. More strict review by statistician and epidemiologist is also required to improve the literature qualities.