目的了解我国临床药物治疗试验论著方法学质量问题。方法选取 18种中华医学系列期刊 ,1996～ 1998年刊载的临床药物防治性同期对照研究论著为调查对象。参照国际临床试验报告标准小组和Downs制订的临床试验报告质量评价量表 ,确定调查内容 ,形成调查表。结果2 2 0篇论著中 ,2 9.1%没有交待病例来源 ,采用随机化分组的占 73 .2 % ,使用盲法试验的有16 .8% ;11.8%的论著没有报告病例基线资料 ,2 4.1%的论著对所报告的基线资料未进行均衡性检验 ;防治效果的评价指标最多达 2 7个 ,中位数为 5 .4;17.7%的论著报告了病例的依从性 ,统计推断应用有误的占 2 5 .0 % ;影响研究结论可靠性的主要原因依次为样本例数太少 (占 30 .0 % )、统计分析有误 (占 2 5 .0 % )、基线均衡性差 (占15 .9% )等。结论我国药物防治性论著质量离临床试验系统评价工作的要求尚有较大距离 ,制订并实施临床试验规范 ,对改善我国临床防治性研究论著质量十分必要  Aim  To understand the quality of methodological issues in clinical trials of clinical trials in China.  Methods  Select 18 kinds of Chinese medicine series of journals, 1996 to 1998 published in the same period of clinical drug control study on the investigation object. Reference to the international clinical trial report standard panel and Downs clinical trial report quality evaluation scale to determine the content of the survey to form a questionnaire. Results 220 articles, 9.11% did not explain the source of the cases, using randomized group accounted for 73.2%, using blind test 16.8%; 11.8% of the papers did not report the baseline data, 2 4.1% of the papers on the baseline data reported did not conduct an equilibrium test; prevention and control of the evaluation index up to 27, the median was 5.4; 17.7% of the papers reported the compliance of the case, the statistical inference application The errors accounted for 25.0%; the main factors influencing the reliability of research conclusions were as follows: the number of sample cases was too small (30.0%), the statistical analysis was incorrect (25.0%), and the baseline balance was poor Accounting for 15.9%) and so on.  Conclusions 质 The quality of the articles on drug prevention and treatment in our country is far from the requirements of clinical trial system evaluation. It is very necessary to formulate and implement the clinical trial norms to improve the quality of our clinical research articles