论文部分内容阅读
目的 :评价国内医学文献中有关糖尿病及其并发症 (以下简称糖尿病 )治疗研究随机对照试验(RCT)的质量 ,为改变和提高临床治疗试验的水平提供依据。方法 :选择国内 6种糖尿病核心期刊 :用手工检索调查 1995 - 1999年间发表的糖尿病治疗的RCT数共 30卷 2 5 0期 ,各杂志发表的RCT数分别为 :中华内分泌杂志 18篇 ;中华内科杂志 5篇 ;中国糖尿病杂志 19篇 ;中国新药与临床杂志 4篇 ;中国中西医结合杂志 35篇 ;中国中西医实用临床急救 13篇。发表的RCT年代分布为 :1995 - 1997年 49篇 ,1998- 1999年 45篇。调查内容包括研究对象选择、随机化方法、盲法、组间可比性、样本大小、依从性、统计学处理、疗效评价指标、副反应观察及随访等。结果 :6种杂志 5年发表的糖尿病治疗RCT共计 94篇。调查表明 ,选择研究对象有明确的纳入标准的有 5 2篇占 5 5 3 % ,有排除标准的有 2 2篇占 2 3 4%。仅有 9篇RCT交待了随机分组的方法 ,没有 1篇交待随机方案隐藏 ,盲法试验有 9篇 (9 6 % ) ,其中双盲为 5篇 (5 3% ) ;80 %采用药物对照 ;17篇采用空白对照(18 10 % ) ;2篇采用安慰剂作对照 (2 % )。治疗组平均样本含量为 32 3± 1(11- 6 6 9)例 ;对照组平均样本含量为 2 8 7± 1 0 5 (6 - 70 7)例。 5 7 45 %有组间可比性 ,仅
OBJECTIVE: To evaluate the quality of randomized controlled trials (RCTs) on the treatment of diabetes mellitus and its complications (referred to as diabetes mellitus) in the domestic medical literature for the purpose of changing and improving the level of clinical trials. Methods: Six domestic core journals of diabetes were selected. The number of RCTs published in 1995-1999 and 1995-1999 was searched by hand. The number of RCTs published in each journal were: 18 in Chinese Journal of Endocrinology; Magazine 5; Chinese Journal of Diabetes 19; Chinese new medicine and clinical 4; Chinese Integrative Medicine magazine 35; Chinese and Western medicine practical clinical emergency 13. The distribution of published RCTs was 49 in 1995-1997 and 45 in 1998-1999. The survey included the choice of subjects, randomized methods, blinded methods, comparability between groups, sample size, compliance, statistical analysis, evaluation of efficacy, side effects and follow-up. Results: A total of 94 RCTs were published in five kinds of diabetes treatment published by 6 kinds of magazines. The survey showed that 52 out of 52 selected subjects had a clear inclusion criteria, while 22 out of the 23 criteria excluded the criteria. Only 9 RCTs presented randomized methods, none of which was concealed by randomization. There were 9 (96%) blinded trials, of which 5 were double-blind (53%); 80% Seventeen blank controls (18 10%) and two placebo controls (2%). The average sample content of the treatment group was 32 3 ± 1 (11- 6 6 9), while the average sample content of the control group was 2 8 7 ± 1 0 5 (6 - 70 7). 5 7 45% Comparable between groups, only