我国中医药治疗急性心肌梗死临床试验文献的质量评价

目的 评价我国关于中医药治疗急性心肌梗死（AMI）的临床随机对照试验的质量。方法 （1）手工检索我国83种中医药杂志，其中最早者从1977年开始，所有杂志均检索至2002年最近一期。纳入标准：中医药防治AMI的，并在文中注明了“随机对照”，“随机分组”的临床研究；（2）根据随机对照试验设计质量的评价标准，设计文献评价表，从诊断标准，基线可比性，随机分组方法，隐匿，盲法，意图治疗分析，统计学方法及结果等7个方面进行分析统计，计算出各种情况的百分比，并评价其质量。结果 共检出与心脏疾病相关的文献2501篇，其中与AMI有关的文献102篇，符合纳入标准者42篇，无1篇文献说明了榇一含量的计算方法，分配隐匿和意图治疗分析，仅4篇文献采用了单盲法，占9．52％；有明确诊断标准者共50篇，占95．24％，对基线情况进行说明者27篇，占64．29％；提到随机方法者共8篇，占19．05％，其中随机分配方法正确者8篇；统计学方法错误者6篇，占14．28％,未对统计学方法进行说明者18篇，占42．86％。结论提示中医药治疗AMI的临床试验设计仍存在一些问题：（1）随机方法运用错误；（2）临床试验研究人员对样本量的计算，分配隐匿，意图治疗分析运用不多；（3）统计学方法的运用有待提高和加强；（4）盲法使用率过低等。

Comment on Literatures of Clinical Randomized Controlled Trials of Traditional Chinese Medicine for Acute Myocardial Infarction

Objective ] To assess the quality of clinical randomised controlled trials (RCT) of traditional Chinese medicine (TCM) for acute myocardial infarction (AMI). Methods] Eighty - three journals of TCM from the first issues of 1977 to the latest issues of 2002 were manually searched and reviewed under the following inclusion criteria: clinical trials of TCM for AMI with 'randomized controlled or randomised being mentioned. And a questionnaire was designed according to the criteria of RCT with the items of diagnostic standard, consistent of baseline, method of randomization, concealment in allocation, blind method, intention - to - treat analysis, statistic methods and conclusion. Percentages of each item were figured out to assess the quality. Results] Two thousand five hundred and one trials were associated with heart diseases, among which 102 were associated with AMI and 42 accorded with the inclusion criteria. Calculation method of sample size, concealment in allocation and intention - to - treat analysis after drop - out were mentioned in none of the 42 articles, single - blind method adopted in 4 (9.52%) , diagnostic standard described in 40 (95.24%), baseline consistent illuminated in 27 (64.29%), method of randomization presented in 8 (19.05%) , statistic methods used mistakenly in 6 (14.28%) and statistic methods not described in 18 (42.86%). Conclusion] There still exist some problems in the design of RCT of TCM for AMI: (1) mistaken use of randomization method; (2) less adoption of calculation method of sample size, concealment in allocation and intention - to - treat analysis after drop - out; (3) improper application of statistic methods; (4) lower rate of utilization for blind study.