中医药治疗肝癌随机对照试验报告的质量评价

目的调查已发表的中医药治疗肝癌的随机对照试验（RCTs）的方法学质量和报告质量。方法采用计算机和手工检索的方法检索中国2004年1月-2006年6月发表的中医药治疗肝癌的RCTs报告，采用CONSORT声明修订版的评价指标对纳入的所有RCTs报告进行质量评价。结果共检索到46篇符合纳入标准的RCTs报告。其中，描述了随机分配序列的产生方法有9篇（19．57％），无1篇描述随机分配的隐藏和样本量的估算；采用盲法的有3篇（6．52％）；报告了病例纳入、排出标准的分别为18篇（39．13％）、1篇（2．17％）；描述了受试者流程的有3篇（6．52％），但无1篇使用了流程图；描述患者依从性的有5篇（10．87％）；对具体P值进行了描述的有6篇（13．04％），无1篇报告了可信区间；有关不良事件描述的报告有15篇（32．61％）；在讨论部分，只有5篇（10．87％）描述了研究的局限性。结论目前中医药治疗肝癌的随机对照试验的方法学质量和报告质量与目前国际公认的CONSORT标准比较尚有差距。

Quality Assessment of the Report of Randomized Controlled Trials on Treatment of Liver Carcinoma with Traditional Chinese Medicine

OBJECTIVE: To investigate the published randomized controlled trials (RCTs) concerning the treatment of liver carcinoma with TCM, and to assess their methodologic quality and quality of report. METHODS: Related RCT reports published from Jan. 2004 to June 2006 were retrieved by computerized and manual search, and the criteria of CONSORT statement (revised edition) were applied to assess the articles searched. RESULTS: Forty-six articles up to the RCTs inclusion standard were searched, among them, the method for randomized distribution sequence producing was described in 9 articles (19.57%), none mentioned randomization concealment and the estimation of the sample size. The blinded method was adopted in 3 articles (6.52%), the inclusion and exclusion criteria were reported in 18 (39.13%) and 1 (2.17%), respectively. Participant flow was reported in 3 (6.52%), but flow chart was not described in any article. Patients' compliance was reported in 5 (10.87%), and the concrete P value was reported in 6 (13.04%), but with no report involving the confidence interval. Adverse events were described in 15 (32.61%). Only 5 articles (10.87%) discussed the limitations of the present study in their section of discussion. CONCLUSION: The quality of RCTs reports on treatment of liver carcinoma with TCM are still in low quality, not meeting the CONSORT statement.