中文期刊发表的中医药系统评价/Meta分析现状调查

目的调查国内中文期刊发表的中医药系统评价/Meta分析文献的质量。方法电子检索1995年1月到2006年12月的CNKI和CBM,筛选中医药的系统评价、Meta分析全文,提取有关中医特色、治疗和对照干预措施的详细资料进行分析评价,并调查文献是否采用QUOROM声明报道结果。结果共纳入文献111篇,其中预防、不良反应、危险因素及先兆症各1篇,理化指标2篇,疗效及安全性评价106篇。涉及42个病种,报道脑血管疾病的文献41篇。干预措施为中医药者25篇,针灸者12篇。对照组:未设对照者2篇;未描述对照组干预措施者15篇;阳性药物对照50篇(其中西药26篇,中药12篇,中药 西药12篇);空白对照14篇;基础对照17篇;假针灸、穴位注射等4篇;安慰剂对照5篇;空白对照14篇及“互为对照”4篇。对照组和部分研究治疗组的干预措施种类均多。每篇文献纳入RCT1～35篇不等,其中24个研究纳入CCT。111篇文献在Cochrane协作网注册者仅14篇。有16篇未对纳入RCT进行质量评价,有22篇仅对纳入RCT进行简单不规范的评价。无一篇文献采用QUOROM声明报道结果。结论应重视提高中医药系统评价的质量。根据中医药特点,选题范围不宜过宽,药物选择不宜过多,注意中药的目标适应病证,应重视对干预措施的评价。系统评价应注意由临床工作者和方法学研究者共同完成,才能做出质量高又具中医特色的评价。

Assessment of Systematic Reviews and Meta-analyses on Traditional Chinese Medicine Published in Chinese Journals

Objective To assess the quality of published systematic reviews and meta-analyses on Traditional Chinese Medicine(TCM)published in Chinese journals.Methods We searched CNKI,CMB from January 1995 to December 2006 and The Cochrane Library(Issue 4,2006)for systematic reviews and meta-analyses on TCM.We extracted details of the interventions used in the treatment and control groups,analyzed the validity of included studies and investigated whether the reports used QUOROM statement or not.Results We identified 111 reports,of which 1 on prevention,1 on adverse events,1 on risk factors and premonitory symptoms,2 on physiochemical parameters,and 106 on effectiveness and safety assessment.In total,42 types of diseases were involved,and 41 reports were related to cerebrovascular diseases.As for the investigated interventions,25 studies assessed TCM and 12 assessed acupuncture.Two had no control intervention design control in the group,15 did not describe the interventions in the control group,50 used active control(26 for western medicine,12 for Chinese medicine,12 for western plus Chinese medicine),14 used blankcontrol,17 used baseline control,4 used sham acupuncture or acupoint injection control etc.,5 used placebo control and 4 used "mutual control".The interventions used in the treatment and control groups varied widely.The number of trials included in the reviews and meta-analyses ranged from 1 to 35,and 24 studies included non-randomized controlled trials.Of the 111 reports,14 were Cochrane reviews,16 did not assess the quality of included randomized trials and a further 22 performed only simple and nonstandard quality assessment of the included trials.None of the reviews or meta-analyses used the QUOROM statement to report their results.Conclusions Because of the unique characteristics of TCM,systematic reviews of TCM should focus on a specific topic,avoid the selection of too many drugs,address the target indications of the test drugs and pay attention on intervention evaluation.High quality systematic reviews of TCM are needed but they will only be produced through the concerted efforts of clinicians,TCM practitioners and methodologists.